UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Gender differences in spouses’ coping with marital tension Daylen, Judith Lynne 1993

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1994-894148.pdf [ 11.93MB ]
Metadata
JSON: 831-1.0088152.json
JSON-LD: 831-1.0088152-ld.json
RDF/XML (Pretty): 831-1.0088152-rdf.xml
RDF/JSON: 831-1.0088152-rdf.json
Turtle: 831-1.0088152-turtle.txt
N-Triples: 831-1.0088152-rdf-ntriples.txt
Original Record: 831-1.0088152-source.json
Full Text
831-1.0088152-fulltext.txt
Citation
831-1.0088152.ris

Full Text

GENDER DIFFERENCES IN SPOUSES' COPING WITH MARITAL TENSION by JUDITH LYNNE DAYLEN B.A., The University of North Carolina, Greensboro, 1982 M.A., The University of British Columbia, 1985 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDffiS (Department of Psychology) We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA December 1993 © Judith Lynne Daylen, 1993  In presentin g thi s thesi s i n partia l fulfilmen t o f th e requirement s fo r a n advance d degree a t th e Universit y o f Britis h Columbia , I freely availabl e fo r referenc e an d study . I  agre e tha t th e Librar y shal l mak e i t  furthe r agre e tha t permissio n fo r extensiv e  copying o f thi s thesi s fo r scholarl y purpose s ma y b e grante d b y th e hea d o f m  y  department o  r  rb  y hi s o r he  r representatives  .I  ti  s understoo  d tha  t copyin  go  publication o f thi s thesi s fo r financia l gai n shal l no t b e allowe d withou t m y w r i t t e n permission.  (Signature)  Department o f //  <^jdLfirByt^  The Universit y o f Britis h Columbi a Vancouver, Canad a  Date 2 ) ^  DE-6 (2/88 )  r,3;/ff-3  ii  ABSTRACT This study provides an examination of gender differences in the coping process of married couples dealing with naturally occurring marital tension. Emphasizing the interpersonal aspects of marital coping, this study examined spouses' reports of their use of confrontation, compromise, and withdrawal strategies and the relationship between spouses' coping reports and their own mood and their partner's mood. The data analyzed in this study were drawn from a community sample of 168 heterosexual, married couples who completed daily structured diaries for a six-week period. The diary questionnaires assessed each spouse's daily reports of marital tension or arguments, appraisals of marital tension, their own coping and their perceptions of their partner's coping (scored using both raw and relative coping values), and daily reports of mood. Analyses of the daily responses across the diary period revealed that wives reported proportionately more days in which marital tension occurred than did their husbands. In addition, wives were significantly more distressed by the occurrence of marital tension than were their husbands. Analyses of a subset of the data (n = 59 couples) in which both spouses within a couple reported marital tension on the same day revealed no gender differences in spouses' appraisal of marital tension or in their coping with marital tension. Significant gender differences emerged, however, in the extent to which spouses' distress was related to their own coping or their partner's coping. The findings from this study suggest that the influence of gender operated at the initial level at which spouses reported the existence of marital tension, but when both spouses agreed that marital tension had occurred, husbands and wives tended to report dealing with marital tension in similar ways. The role of gender also operated in the relationship between coping and spouses' mood. These gender similarities and differences are discussed in relation to the gender-based demand-withdraw patterns reported in the  iii  previous literature. The results of this study suggest that gender differences found in studies with maritally distressed couples who respond to reenacted marital conflict may not generalize to community residing couples dealing with naturally occurring marital stressors in their day-to-day lives.  iv TABLE OF CONTENTS ABSTRACT TABLE OF CONTENTS LIST OF TABLES ACKNOWLEDGMENTS INTRODUCTION Stress and Coping Cognitive appraisal Coping The effectiveness of coping strategies Interpersonal aspects of coping Summary Gender Differences Defining Gender Gender differences in the use of general coping strategies Gender differences in coping with marital tension Overview Gender differences in the general literature Relationship Involvement Interpersonal Stress Overview HYPOTHESES Hypothesis I Hypothesis II Hypothesis III Hypothesis IV Hypothesis V Exploratory Analyses Exploratory Analyses 1 Exploratory Analyses II METHODS Sample Data Subset utilized in Hypotheses II-V Measures Overview of diary Assessment of marital tension Appraisal of Threat Coping Inventory Distress scale Missing data on scale items RESULTS Analyses related to Hypothesis 1 Comparisons between husbands' and wives' reports of marital tension (Hypothesis la)  ii iv vi viii 1 6 7 9 12 16 22 25 25 26 31 47 53 53 60 64 68 68 69 71 73 78 79 79 82 84 85 88 98 98 99 100 101 105 106 107 107 107  V  Comparisons between husbands' and wives' reports of "serious" marital tension (Hypothesis lb) 109 Associations between marital tension and distress (Hypothesis Ic) Ill Summary 115 Analyses related to Hypothesis II - V 115 Analyses related to Hypothesis II 120 Analyses related to Hypothesis III and Exploratory Analysis I 122 Intercorrelations among coping scales 123 Summary 129 The relationship between husbands' and wives' coping (Hypothesis III) 131 Summary 136 The relationship between respondents' own coping and their perception of their spouses' coping (Exploratory Analysis I) ..136 Summary 143 The relationship between respondents' perceptions of their spouses' coping and spouses' coping (Exploratory Analysis I). 144 Overall Summary of Hypothesis III and Exploratory Analysis 1 150 Analyses relating to Hypotheses IV and V and Exploratory Analyses II 153 Bivariate analyses 153 Multiple regression analyses 159 Summary of Hypotheses IV and V and Exploratory Analysis II 171 Overall Summary of Results 173 DISCUSSION 176 Spouses' reports of marital tension 176 Spouses' appraisal of threat 183 Comparisons between husbands' and wives' coping 185 Evidence of reciprocity in couples' coping 186 Explanations for no gender differences in spouses' coping 191 Respondents' perceptions of their spouses' coping 196 Relationship between coping and distress 199 Self Coping 200 Spouses' Coping 206 Perceived Spouses' Coping 213 The Issue of Causality 216 Conclusion and Future Directions 217 FOOTNOTES 222 REFERENCES 224  VI  LIST OF TABLES Table 1. Comparisons between Data Subset and Remaining Diary Sample  91  Table 2. Comparisons between Data Subset and Remaining Diary Sample  92  Table 3. Comparisons between Data Subset and Remaining Diary Sample  94  Table 4. Comparisons between Data Subset and Remaining Diary Sample  95  Table 5. Comparisons between Data Subset and Remaining Diary Sample  96  Table 6. Comparisons between Data Subset and Remaining Diary Sample  97  Table 7. Paired /-tests Comparing the Proportion of Diary Days Husbands and Wives Reported Marital Tension  108  Table 8. Hierarchical Repeated Measures Multiple Regression Analysis Regressing Current Distress on Marital Tension  113  Table 9. Correlation Matrix of Study Variables for Husbands (Raw Score Values)  116  Table 10. Correlation Matrix of Study Variables for Wives (Raw Score Values)  117  Table 11. Correlation Matrix of Study Variables for Husbands (Relative Score Values)  118  Table 12. Correlation Matrix of Study Variables for Wives (Relative Score Values)  119  Table 13. Paired Mests Comparing Husbands' and Wives' Appraisal of Threat  121  Table 14. Intercorrelations Among Self Coping Strategies (Raw Score Values)  124  Table 15. Intercorrelations Among Perceived Spouse's Coping Strategies (Raw Score Values)  126  Table 16. Intercorrelations Among Self Coping Strategies (Relative Score Values)  128  Table 17. Intercorrelations Among Perceived Spouse's Coping Strategies (Relative Score Values)  130  Table 18. Paired Mests Comparing Respondents' Self Coping and Their Spouses' Coping (Raw Score Values)  133  Vll  Table 19. Paired Mests Comparing Respondents' Self Coping and Their Spouses' Coping (Relative Score Values)  135  Table 20. Paired Mests Comparing Self Coping and Perceived Spouses' Coping (Raw Score Values)  138  Table 21. Paired /-tests Comparing Self Coping and Perceived Spouses' Coping (Relative Score Values)  140  Table 22. Paired Mests Comparing Perceived Spouses' Coping and Spouses' Coping (Raw Score Values)  145  Table 23. Paired Mests Comparing Perceived Spouses' Coping and Spouses' Coping (Relative Score Values)  148  Table 24. Comparison of Husbands' and Wives' Correlations Between Current Distress and Coping on Self, Perceived Spouses, and Spouses' Coping Scales  156  Table 25. Hierarchical Multiple Regression Analysis Regressing Current Distress on Confrontation  161  Table 26. Hierarchical Multiple Regression Analysis Regressing Current Distress on Compromise  166  Table 27. Hierarchical Multiple Regression Analysis Regressing Current Distress on Withdrawal  169  viii ACKNOWLEDGMENTS I would like to express my most sincere gratitude to my supervisor, Anita DeLongis, for her support throughout the development of this dissertation. I particularly appreciate her provision of articulate feedback and constructive criticism that was always delivered in a thoughtful and supportive way. I would also like to express my indebtedness to Dr. DeLongis for making available the data used here, along with other resources that have been necessary for the completion of this dissertation. I also would like to gratefully acknowledge the other members of my committee, Tannis MacBeth Williams and Charlotte Johnston, who provided not only helpful suggestions (both substantive and stylistic), but also moral support. Thanks are also extended to Raymond Corteen and Wolfgang Linden for their helpful comments that aided in the final draft of this dissertation. I would also like to thank Chris Davis for providing technical assistance and statistical support. I also would like to acknowledge the library assistance provided by various support persons, but particularly Jennifer McNaught, whose cheerful helpfulness was much appreciated. Finally, I want to thank my partner, John Yuille, who provided immeasurable amounts of technical support, intellectual stimulation, and compassionate understanding that allowed me to be preoccupied with marital tension as a research topic rather than as a lived experience.  1 INTRODUCTION A satisfying marriage, or involvement in a committed relationship, continues to be one of life's top priorities for many North Americans. In a 1970s study of American college students, only 3% of males and 1% of females currently involved in a heterosexual relationship said they planned to never marry (Hill, Rubin, & Peplau, 1976). Most adults in the United States and Canada get married at least once during their lifetimes. A recent Canadian census reported that 74% of those surveyed over the age of 15 had been married at least once (Statistics Canada, 1991). Not only is marriage an important concern for most, but it is also a uniquely intense relationship in most peoples' lives. In a study assessing sources of satisfaction and conflict in long-term relationships (focusing on nine relationships including family, friends, and work associates), marriage emerged as both the greatest source of satisfaction and the greatest source of conflict (Argyle & Furnham, 1983). It is this latter facet of marriage, i.e., marriage as a unique source of conflict, that is the focus of this dissertation. Although some researchers have noted that being married does not necessarily involve being emotionally close or intimate, the marital relationship is likely to be more enduring, to involve more frequent interaction, to span a wider range of activities, and to be subject to a greater number of cultural norms than are most other adult relationships (Huston & Robins, 1982). Furthermore, because marriage involves frequent, diverse, and intense behavioral and psychological interdependence over long periods of time, marriage requires that the two partners negotiate a variety of different interests or goals (Huston & Robins, 1982; Kelley et al., 1983). Given the extent of interdependence inherent in marriage, spouses' interests necessarily compete at times. It is virtually inevitable that both spouses within a relationship will experience conflict, frustration, and anger with one another. Some writers have gone so far as to define marriage as a "continuous confrontation between participants with conflicting-though not always opposing—  2  interests" (Fitzpatrick, 1988, p. 149). Other researchers have called attention to the naturally occurring "ups and downs" and everyday "hassles" or tensions (e.g., marital arguments) that are a feature of every marital relationship, no matter how satisfied or dissatisfied the two partners may be with their relationship (DeLongis & Lehman, 1989; Rusbult, 1987). Marital tension may arise from any number of sources: differences in partners' beliefs, interests, desires, or values; the scarcity of some resource; or, a rivalry in which one person tries to outdo or undo the other (Deutsch, 1969). Another way of conceptualizing marital tension is in terms of incompatible goals; conflict may occur when spouses perceive themselves as having incompatible goals and perceive the other as interfering with the achievement of their own goals (Folger & Poole, 1984). Or, more simply, conflict occurs when one spouse seeks an outcome that he or she believes that the other spouse is, at least initially, unwilling to provide. Although differences leading to conflict between partners in a heterosexual relationship can arise from a multitude of sources, differences are likely to exist if for no other reason than that the spouses are of different genders. After all, relationships between men and women have been characterized as a "battle of the sexes" (Doyle, 1989) or as the "longest war" (Tavris & Wade, 1984). Furthermore, there is evidence to suggest that women and men may differ in how they cope with stress (Miller & Kirsch, 1987; Thoits, 1991), and more specifically, that wives and husbands may differ in how they cope with marital tension (Christensen & Heavey, 1990; Pearlin & Schooler, 1978). If it is the case that heterosexual marriage involves interdependence between partners with gender-based differences and that conflict or tension is an inevitable, ubiquitous phenomenon within marital relationships, the question then becomes how might spouses differ in their management of marital tension. The issue of gender differences in coping with marital tension is the focus of the research being presented here. Before considering these issues more directly, however, it is important to establish what, if any,  3  impact marital tension or conflict is thought to have on spouses and their relationships. Exploration of this issue will establish the importance of studying marital tension and will indicate some of the reasons why marital tension presents a unique opportunity for exploring gender differences in coping. Research within the stress and coping literature relating to major life stressors has indicated that among married couples, conflict or tension with network members (including close significant others, family, and friends) is a potent source of stress. In Mattlin, Wethington, and Kessler's (1990) research based on interviews with 778 couples, stressors related to network members accounted for over 65 % of the events that were reported as being the most stressful event experienced during the previous year. Approximately 23% of these singularly most stressful events related specifically to marital difficulties and interpersonal problems with other family members and friends. When the effects of negative interactions were examined more closely, using this same data set, it was found that reports of negative interactions with one's spouse (compared to negative interactions with friends or relatives) had the most negative impact on mental health for both men and women (Schuster, Kessler, & Aseltine, 1990). Within the stress and coping literature, there also has been increasing evidence that everyday stressors such as marital arguments influence health and psychological wellbeing (e.g., DeLongis, Coyne, Dakof, Folkman, & Lazarus, 1982; Eckenrode, 1984; Kanner, Coyne, Schaefer, & Lazarus, 1981). In fact, recent research by Wagner, Compas, and Howell (1988) has demonstrated that daily hassles are more strongly associated with psychological symptoms than are major life events. For example, in a study of the relationship between daily stressful events and mood, Lewinsohn and Amenson (1978) found that interpersonal events accounted for over 70% of the variance in mood-related events. Similarly, Hamilton and Fagot (1988) reported that negative interpersonal interactions were rated as the most stressful events experienced by respondents during an eight-week reporting period. In addition, there is evidence  4  suggesting that more people (approximately 40% of all clients) seek help in mental health clinics for marital problems than for any other single problem (Veroff, Kulka, & Douvan, 1981). Although marital conflict is often viewed as a stressful experience with negative consequences, the outcome of marital conflict is not necessarily deleterious. Thus, the assumption, once widespread among researchers and clinicians, that conflict and satisfaction in marriage are inversely related appears to be unfounded (Madden & JanoffBulman, 1981). "Constructive marital conflict is NOT an oxymoron," Markman (1991a) declared in the title of his paper in a special issue of Behavioral Assessment that was entitled "Negative communication in marital interaction: A misnomer?" (Vol. 13, 1991). There are, for example, reports of associations between conflict engagement and increased marital satisfaction over time (Gottman & Krokoff, 1989; Smith, Vivian, & O'Leary, 1991). There is also evidence that it is more detrimental if conflict is avoided in marriage rather than engaged in (Pearlin & Schooler, 1978). In fact, it has been suggested that an ability to confront, to reconcile, and to accept differences must be developed between spouses if a relationship is to be an arena of growth rather than one of stagnation or oppression (Weingarten & Leas, 1987). When spouses engage in conflict, they have an opportunity to get to know one another more fully, to gain better access to each other's interpersonal perceptions, and thereby to understand more fully each other's versions of reality. According to Sayers, Baucom, Sher, Weiss, and Heyman (1991), constructive conflict engagement potentially "enhances couples' future satisfaction by increasing the likelihood that the spouses will discuss and solve conflicts in their relationships" (p. 27). Successfully resolved conflicts thus allow spouses to more readily adapt to stress and change within their marriage (Knudson, Sommers, & Golding, 1980). On the other hand, ineffective management of conflict or a prolonged and hurtful process of conflict resolution can be extremely stressful for marital partners and damaging to their relationship (Deutsch, 1969; Schaap, Buunk, & Kerkstra, 1988). Although it is  5  likely that, by definition, virtually all marital conflict is initially perceived as stressful, research indicates that it is not marital conflict per se that is critical in predicting future success of a marriage; how conflicts are handled within the marriage appears to be the more important predictor (Markman, 1991a, b; Storaasli & Markman, 1990). Similarly, the stress and coping literature suggests that the way in which spouses cope with stressors such as marital conflict will affect both the level of stress experienced and the longer term consequences of the stressor (e.g., Billings & Moos, 1981; 1984; Coyne, Aldwin, & Lazarus, 1981; Mitchell, Cronkite, & Moos, 1983). For example, Pearlin and Schooler (1978) demonstrated that it is the specific things that spouses do in dealing with marital strains or stressors that most closely determines the extent of emotional stress that they experience. In fact, in additional analyses of the Pearlin and Schooler (1978) data, Ilfeld (1980) found that the spouses' style of coping was the most predictive factor of marital stress, accounting for 38% of the variance. Thus, it appears that it is not merely the presence of marital tension but how one copes with marital tension or conflict that affects spouses' personal well-being and marital satisfaction. Thus, given that marital tension is a naturally occurring feature of married life with potentially stressful consequences for the spouses and the relationship, and given that there is evidence that the way in which spouses cope with marital tension affects the degree of stress experienced, marital conflict emerges as a phenomenon of particular interest to those interested in the process of stress and coping. In addition, given that research has suggested gender differences in coping and given that marital conflict involves a clash between a woman and a man, the study of husbands' and wives' responses to marital tension affords an opportunity to study gender differences in stress and coping. Furthermore, if investigation is focused on a specific, naturally occurring marital argument between a husband and wife, one is able to assess gender differences in how spouses cope with the same, everyday, interpersonal stressor. Conflict or tension in marriage (e.g., marital arguments) thus provides a unique opportunity to study how  6  husbands and wives, individually and collectively, cope with the same, naturally occurring stressful event. (1) Are husbands and wives differentially affected by marital tension? (2) Do husbands and wives employ different strategies when coping with marital tension? (3) Do husbands and wives appraise marital tension differently, i.e., do they evaluate what is at stake in a marital argument differently? (4) How do the types of coping employed by wives and husbands affect their own emotional distress and that of their spouse? These are the questions that the research presented here addresses. However, before formally introducing the hypotheses that guided this research, the relevant conceptual and empirical literature is reviewed. Firstly, an overview of the stress and coping framework that formed the conceptual basis for this research is presented. In this section, the focus is on issues relevant to an understanding of coping as it pertains to marital conflict without an emphasis on gender differences in coping. This review is followed by a review of research within the stress and coping literature pertaining to gender differences both in general coping responses, and more specifically, in coping with marital conflict. Then, a selected review of the literature on gender differences within the context of intimate heterosexual relationships is presented. Stress and Coping Stress is generally conceived as a "particular arrangement between the person and the environment that is appraised by the person as taxing or exceeding his or her resources and endangering his or her well-being" (Lazarus & Folkman, 1984, p. 19). Coping, as a response to stress, is generally described in terms of cognitive and behavioral efforts to master, manage, tolerate, or reduce the demands that tax or exceed a person's resources (Pearlin & Schooler, 1978; Cohen & Lazarus, 1983). Lazarus and his colleagues (e.g., Lazarus & DeLongis, 1983; Lazarus & Folkman, 1984) have developed a theoretical framework for understanding the stress and coping process that is grounded in empirical work (DeLongis, Coyne, Dakof, Folkman, & Lazarus, 1982; DeLongis, Folkman, &  7  Lazarus, 1988; Folkman & Lazarus, 1980; Folkman & Lazarus, 1985; Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986; Folkman, Lazarus, Gruen, & DeLongis, 1986; Kanner, Coyne, Schaefer, & Lazarus, 1981). Because this model of stress and coping provided a framework for the research being presented here, relevant aspects of this model will be more fully explicated. In particular, the focus here is on the applicability of this model as it pertains to coping with marital tension. Three aspects of the stress process that are considered in more depth in the following section include the following: (1) cognitive appraisal of a stressor; (2) coping strategies employed to deal with a stressful encounter; (3) the relationship between coping and adaptational outcomes, i.e., the effectiveness of particular coping strategies used to deal with stress. Finally, the limitations of this model for understanding marital conflict are addressed. Cognitive appraisal In keeping with the 1960s and 1970s cognitive revolution within psychology that led to a re-emphasis on the role of cognitive factors in explanations of human behavior, Lazarus and his colleagues place particular importance on cognitive factors in their model of stress and coping. In fact, in the Lazarus et al. model of coping, cognitive appraisal of a potentially stressful event is the single most important determinant of how an individual will cope with that event (Lazarus, 1990; Lazarus, DeLongis, Folkman, & Gruen, 1985; Lazarus & Folkman, 1984). According to Lazarus and his colleagues, when a potential stressor is encountered an individual engages in a cognitive appraisal process to determine why and to what extent a given event or situation is stressful for him or her. Given the role of primary appraisal in determining the perceived threat of a situation, it is not merely the occurrence of a particular event or encounter that is important, but its meaning or significance for the person involved (see also Bradbury & Fincham, 1992; Holtzworth-Munroe & Jacobson, 1985, for a discussion of similar issues within an attribution framework). For example, when a couple is having breakfast  8  together and one spouse picks up a newspaper and proceeds to read it throughout the meal, this could be perceived as of no consequence to his or her partner or it could engender a stressful encounter depending on the significance of this event for the other spouse. Reading the newspaper during breakfast could be perceived as an acceptable routine for either or both partners. Alternatively, it could be appraised as a demonstration of disinterest in communicating with a partner who considers communication important and feels ignored by his or her partner. Or, this act could be perceived as an angry gesture related to other marital issues. As noted above, the primary cognitive appraisal that an individual makes involves an evaluation of what, if anything, the person has at stake in a particular situation or what he or she believes is threatened by the stressful encounter. In the preceding example, a spouse's need for support and understanding may be threatened if the other spouse's behavior is perceived as a way to avoid interpersonal contact. Although cognitive appraisal is central to the Lazarus et al. theory of stress and coping, surprisingly little empirical research has specifically explored the appraisal process. In one of the few studies focusing on cognitive appraisal, Folkman, Lazarus, Dunkel-Schetter, DeLongis, and Gruen (1986) developed a measure composed of thirteen items that described various concerns or threats drawn from a review of the relevant literature and from self-reports about perceived stakes in stressful encounters gathered in a previous study of a community sample of 100 adults aged 45-64 (Folkman & Lazarus, 1980). A factor analysis of these items identified two major concerns that are purported to guide individuals' evaluations of potentially stressful situations: threats to self-esteem and threats to a loved one's well-being (Folkman, Lazarus, Dunkle-Schetter, DeLongis, & Gruen, 1986). However, because Lazarus and his colleagues have focused on the general coping process and individual coping rather than on the interpersonal aspects of coping that are particularly germane in marital conflict (see later section, "Interpersonal aspects of coping"), the applicability of Folkman, Lazarus, Dunkle-Schetter, DeLongis, and Gruen's (1986) primary appraisal scale to the specific domain of marital conflict remains unclear.  9  Furthermore, with the exception of one study's finding that wives tend to endorse more concerns about their loved ones and husbands tend to endorse more concerns about a goal at work (Folkman, Lazarus, Gruen, & DeLongis, 1986), gender differences in the primary appraisal process have not been explicated. Given the importance of cognitive appraisal to the coping process in stress theory, spouses' cognitive appraisal of the threats engendered by marital conflict were explored in the present research. In the approach taken here, the Folkman, Lazarus, Dunkle-Schetter, DeLongis, and Gruen (1986) primary appraisal scale was revised slightly to capture the interpersonal concerns that might arise in marital conflict (see Methods section). In addition, research findings in the marital and gender differences literatures were reviewed and hypotheses were developed pertaining to possible gender differences in cognitive appraisal. However, before discussing these issues, other aspects of the Lazarus et al. model of stress and coping will be further explicated. The following section deals with issues related to the use of specific coping strategies. Coping After having appraised a situation as stressful, the individual then must somehow cope with the situation. Coping is what the person actually thinks or does to deal with the stressful situation. Two broad functions of coping have been proposed by Lazarus and his colleagues: problem-focused coping and emotion-focused coping. Problem-focused coping aims to alter the stressful situation itself. This form of coping includes cognitive and behavioral strategies akin to problem solving. These strategies focus on defining the problem, generating alternative solutions, seeking information, weighing the alternatives in terms of costs and benefits, choosing among the possible alternatives, delineating obstacles, and then acting in specific ways to deal with the problem (Lazarus & Folkman, 1984; Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986). The other major function of coping delineated by Lazarus and his colleagues is emotion-focused coping, a  10  form of coping that functions to manage or regulate one's stressful emotions. In Lazarus et al.'s model this is largely a cognitive process by which one alleviates emotional distress by employing particular cognitive strategies such as avoidance, minimization, and positive reappraisal (Lazarus, 1985; Lazarus & Folkman, 1984; Folkman, Lazarus, DunkelSchetter, DeLongis, & Gruen, 1986). There is, however, controversy within the stress and coping literature over the conceptualization of coping and the development of typologies of coping strategies (see Haan, 1982, and Moos & Billings, 1982, for reviews). For example, as noted above, Lazarus and his colleagues distinguish between two coping functions, problem-focused and emotion-focused coping; however, factor analyses of Lazarus et al.'s coping strategies (Ways of Coping inventory, Folkman & Lazarus, 1985) yielded eight factors rather than two (Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986). Other prominent researchers (e.g., Billings & Moos, 1981; Pearlin & Schooler, 1978; Stone & Neale, 1984a) present a typology with three categories: active behavioral strategies (corresponding to altering the problem directly), active cognitive strategies, (corresponding to changing one's way of viewing the problem) and avoidance strategies (corresponding to managing the emotional distress aroused by the problem). In an attempt to transcend these various typologies, a dichotomous classification scheme has been proposed by Roth and Cohen (1986). This classification scheme distinguishes between strategies that are active in nature and oriented toward confronting the problem and strategies that entail an effort to reduce tension by avoiding dealing with the problem (Holahan & Moos, 1987). Roth and Cohen (1986) review formulations of approach-avoidance coping from Freud (1915/1957) to Horowitz (1976,1979), spanning psychodynamic, behavioral, and phenomenological approaches, to present the case for the centrality of these concepts to our understanding of coping with stress. Thus, according to these authors, coping is defined as "cognitive and emotional activity oriented either toward (approach) or away from (avoidance) threat" (Roth & Cohen, 1986, p.813). This scheme  11  can be applied to the three part typology noted above by combining active cognitive and behavioral strategies and viewing these in opposition to avoidance strategies (although see Billings & Moos, 1981, and Stone, 1984a, who would argue against collapsing the distinction between the former two types of coping). This categorization scheme is less applicable, however, to the Ways of Coping inventory developed by Lazarus and his colleagues (Folkman & Lazarus, 1980; Folkman & Lazarus, 1985; Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986); problem-focused strategies are generally approach oriented, whereas emotion-focused strategies include both approach and avoidance strategies. Although Roth and Cohen's (1986) dichotomous classification scheme may not capture all of the complexities of the coping process, the approachavoidance formulation of coping appears to be particularly germane to the study of marital conflict and therefore this organizational scheme is employed in discussing the relevant research findings. These various conceptualizations of coping have expanded our understanding of the types of coping strategies employed in a variety of stressful situations (e.g., relating to home, work, and health). However, the coping inventories developed thus far within the stress and coping literature (e.g., Billings & Moos, 1981; Folkman & Lazarus, 1980, 1985; Pearlin & Schooler, 1978; Stone & Neale, 1984a) do not fully reflect the types of coping employed within the interpersonal context of marital conflict (Bolger, DeLongis, Kessler, & Schilling, 1989; Kessler, Price, & Wortman, 1985). These limitations are further explicated in a following section pertaining to interpersonal aspects of coping. Before elucidating these issues, however, a brief overview is presented of the current perspective within the stress and coping literature pertaining to the effectiveness of coping strategies.  12 The effectiveness of coping strategies A large body of research suggests that coping is a major mediator between stressful events and adaptational outcomes such as psychological symptoms, mood, and somatic health (e.g., Billings & Moos, 1981, 1984; Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986; Folkman, Lazarus, Gruen, & DeLongis, 1986; Menaghan, 1982; Mitchell et al., 1983; Pearlin & Schooler, 1978; Vaillant, 1977). There is little consensus among researchers, however, as to which coping strategies or modes of coping are most effective. Although some theorists (Haan, 1982; Vaillant, 1977) categorize coping strategies (including defensive strategies) a priori according to their presumed effectiveness in reducing stress, many stress and coping researchers argue that coping strategies are not inherently good or bad. Instead, these researchers suggest that different strategies are efficacious in different circumstances and/or at different points in the coping process (Lazarus & Folkman, 1984; Folkman & Lazarus, 1988; Pearlin & Schooler, 1978). In other words, the value of particular coping strategies must be determined in reference to the context in which they are used (Folkman, Lazarus, Gruen, & DeLongis, 1986; Lazarus & Folkman, 1984). Thus, coping efforts used in one domain may be ineffectual or even detrimental in another domain. For example, the use of avoidance or denial strategies can be adaptive in very negative situations that cannot be changed but somehow must be accepted (Collins, Baum, & Singer, 1983; Lazarus, 1985); whereas, in interpersonal situations, where the problem could be changed with attention and appropriate action, avoidance or denial strategies can be associated with negative outcomes and increased distress (Pearlin & Schooler, 1978). Thus, despite a tendency to see active approach strategies as positive and passive avoidance strategies as negative, the efficacy of these strategies depends upon the circumstances within which they are applied (Lazarus, 1985). Allowing for the effects of the circumstances in which coping is taking place, Roth and Cohen (1986) presented a model outlining possible benefits and costs for both  13  approach strategies and avoidance strategies. Approach strategies can be beneficial when they lead to appropriate action, ventilation of affect, and/or assimilation or resolution of a stressful experience; approach strategies can be ineffectual or harmful when they lead to nonproductive worry, increased distress, or the inability to accept an uncontrollable outcome (Roth & Cohen, 1986). On the other hand, avoidance strategies can be beneficial in that they may reduce stress and prevent anxiety from becoming overwhelming. Avoidance strategies may thereby provide the needed time for assimilation of stressful information and the mobilization of efforts to deal with the situation. However, avoidance strategies may be costly in that they may interfere with appropriate action by producing emotional numbness, disruptive avoidance behavior, and/or decreased awareness of the relationship between the stressful event and one's own felt distress (Roth & Cohen, 1986; Lazarus, 1985). In addition, Lazarus (1985) reviewed research evidence suggesting that the effectiveness of these two types of strategies is associated with the time point at which effectiveness is measured. For example, Mullen and Suls (1982) reported consistent findings across several studies that indicated that avoidance strategies were effective when outcome measures were immediate or short-term, whereas approach strategies were effective when outcome measures were long-term. Research has also indicated that the effectiveness of approach strategies depends upon the extent to which these strategies include an aggressive or hostile component. For example, there is evidence that non-aggressive, non-hostile, planful problem-solving is associated with positive emotional states (Folkman & Lazarus, 1988) and satisfactory outcomes (Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986) and is negatively correlated with psychological symptoms (Folkman, Lazarus, Gruen, & DeLongis, 1986). Similarly, negotiation strategies, which involve congenial, problemsolving efforts have been associated with low marital stress (Sabourin, Laporte, & Wright, 1990). Also, Rands, Levinger, and Mellinger (1981) found that compromise was positively associated with marital satisfaction. In keeping with Mullen and Suls1 (1982)  14  finding that approach strategies may be more effective when evaluated in the long-term, Menaghan (1982) found that negotiation strategies were not an immediate reducer of stress among respondents dealing with marital strains, but these strategies were associated with fewer problems at a later point in time. On the other hand, more aggressive confrontive coping strategies have consistently been found to be associated with negative outcomes, e.g., negative emotional states (Folkman & Lazarus, 1988), unsatisfactory outcomes (Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986), psychological distress (Folkman, Lazarus, Gruen, & DeLongis, 1986), low marital satisfaction (Bowman, 1990; Rands et al., 1981), and chronic marital strain (Ilfeld, 1980). In an observational study in which couples' discussions of a marital conflict were coded by trained observers, conflict engagement (including disagreement and criticism) was found to be negatively associated with marital satisfaction at the time of the conflict enactment, but positively associated with marital satisfaction three years later (Gottman & Krokoff, 1989). Regarding research concerning the efficacy of avoidance strategies, Lazarus (1985) reviewed several studies that suggest that denial or avoidance strategies may be effective in helping one deal with uncontrollable, severe, and sudden crises or with incapacitating illnesses, injuries, or surgical procedures. When the focus moves away from lifethreatening or health-related concerns, however, research has often indicated that avoidance strategies are associated with negative outcomes (e.g., Billings & Moos, 1981; Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986; Holahan & Moos, 1987; Kobasa, 1982; although see Long & Gessaroli, 1989, for evidence that avoidance strategies may be perceived to be effective in alleviating work stress, especially higher levels of stress and especially for males). Some research also supports the notion that avoidance strategies can be especially detrimental when employed in interpersonal situations. For example, in research that focuses on spouses coping with marital stress, avoidance strategies have been found to be negatively associated with marital satisfaction (Bowman, 1990; Rands et al., 1981) and positively associated with increased stress  15  (Pearlin & Schooler, 1978). In Gottman and Krokoff's (1989) observational research relating to marital conflict, husbands' withdrawal had no effect on their immediate marital satisfaction, but it was associated with declines in their marital satisfaction over time. Wives' withdrawal was negatively associated with their immediate marital satisfaction, but no relationship was found between wives' withdrawal and their later marital satisfaction. Although this brief review has offered some insights into the effectiveness of various coping strategies for individuals dealing with stressful events, researchers within the mainstream coping field have tended to ignore interpersonal aspects of coping that are particularly germane within the context of close relationships such as marriage (Coyne & DeLongis, 1986; DeLongis & O'Brien, 1990). In particular, coping effectiveness has been considered almost exclusively in terms of its effectiveness for the individual herself or himself, without taking into account its effect on others. There is, however, interdependency and mutual influence in marriage—the actions of each spouse have effects on the other (Coyne, Ellard, & Smith, 1990; Robins, 1990). As noted by Coyne et al. (1990) in a study of heart attack patients, coping strategies that were beneficial to the patient were not always beneficial to their spouses or their relationships. When spouses are dealing with marital conflict, it is particularly likely that the way in which one spouse copes with the conflict will affect the coping of the other spouse. For example, observational research dealing with reenactments or role plays of marital conflict has consistently demonstrated that one spouse's way of dealing with marital conflict affects his or her partner's responses (e.g., Levenson & Gottman, 1983; Margolin & Wampold, 1981; Notarius, Benson, Sloane, Vanzetti, & Hornyak, 1989; Roberts & Krokoff, 1990; Ting-Toomey, 1983). The interdependence of spouses' responding has even been demonstrated at the physiological level. For example, Levenson and Gottman (1983) found that "physiological linkage" between spouses during a marital problem interaction explained 59% of the variance in concurrent marital satisfaction. To explore more fully  16  the ramifications of findings such as these, we turn now to a discussion of the interpersonal aspects of coping. Interpersonal aspects of coping As noted above, the Lazarus et al. model of stress and coping has expanded our knowledge of the stress and coping process; however, it is limited in terms of its applicability to interpersonal conflicts. With the exception of issues pertaining to social support, little attention has been paid in the stress and coping literature to the social context within which stress and coping attempts occur (DeLongis & O'Brien, 1990). Thus, coping strategies pertaining to efforts to seek informational or emotional support from others are included in the Lazarus et al. model (Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986), but other interpersonal factors receive little attention. Inclusion of social support-seeking in models of coping is consistent with the large social support literature that emphasizes the critical role played by close others in the process of adapting to stress. There is evidence, for example, that others can be helpful in attenuating stress responses or buffering individuals from stress (Cohen & Wills, 1985; Vanfossen, 1981), as well as intensifying one's stress responses (Pagel, Erdly, & Becker, 1987; Rook, 1984; Wortman & Lehman, 1985; see also Rook & Pietromonaco, 1987, for a review of both positive and negative effects of social support). However, DeLongis and her colleagues have pointed out that the impact of interpersonal factors on the coping process goes beyond the role of social support (Coyne & DeLongis, 1986; DeLongis & O'Brien, 1990). For example, research has demonstrated that stress of an interpersonal nature has an especially negative impact on mood (Bolger et al., 1989). Furthermore, DeLongis and O'Brien (1990) argue that interpersonal factors are often critical to our understanding of the coping process, particularly within marital dyads; thus, interpersonal factors may greatly influence how a spouse copes with a stressor and how a spouse's wellbeing is affected by the coping process.  17  Several research findings suggest that how one spouse copes with a stressor may affect the other spouse's coping and well-being. For example, in a study that suggested that wives may be particularly responsive to the feelings and behavior of their husbands, Cronkite and Moos (1984) found that wives' depressed mood was related to undesirable events in their lives and to their husbands' alcohol consumption, whereas husbands' depressed mood related only to their own experiences and problems, not to their wives' behavior. In keeping with this finding, other research has also indicated that one spouse's choice of coping strategy may affect the other's well-being. For example, in research with couples in which the husband had suffered a heart attack, Coyne et al. (1990) found that wives' self efficacy after their husbands' heart attack was positively affected by their husbands attempts to actively engage them in constructive problem solving and negatively affected by their husbands "buffering," i.e., hiding of concerns or avoiding disagreements. On the other hand, when wives adopted a buffering coping style it contributed positively to the patients' (their husbands') well-being and self-efficacy, but negatively to wives' own well-being (Coyne et al., 1990). Similarly, in their research with heart attack patients, Stern and Pascale (1979) found that when males coped with their condition by denial, their wives experienced significantly more anxiety than wives whose husbands did not employ denial strategies. In Manne and Zautra's (1989) research with female arthritis patients and their husbands, the response of the husband also appeared to affect the type of coping engaged in by his wife. Positive support from the husband was associated with wives employing more cognitive restructuring and information-seeking efforts, strategies predictive of positive psychological adjustment. On the other hand, criticism by the husband was associated with more wishful thinking on the part of the wife, a strategy predictive of poor psychological adjustment. Thus, spouses' responses were related to the patients' coping behavior and were associated with the psychological adjustment of patients (Manne & Zautra, 1989). Similarly, Repetti (1989) found that wives' emotional support-giving was  18  associated with their husbands' response to work overload. When wives responded with high levels of emotional support, their husbands showed increased social withdrawal and diminished anger responses at home (a process proposed by the author to enhance recovery from work overload). This pattern of husbands' responses was not found with low levels of support from their wives (Repetti, 1989). These research findings, along with the findings from marital observational research, suggest that it is especially important to consider interpersonal factors when studying coping within a marital dyad. As noted above, interpersonal issues are likely to influence not only one's choice of a coping strategy, but also the effectiveness of a particular coping strategy. For example, DeLongis and O'Brien (1990) have suggested that avoidance may be a particularly ineffective strategy in interpersonal situations because it inhibits emotional expressiveness and may thereby interfere with the ability of a spouse to comprehend and respond empathically to her or his partner's experience. Consistent with this notion, Gottman and Krokoff (1989) found that wives' withdrawal was associated with decreased concurrent marital satisfaction for both husbands and wives. It is also possible in some circumstances in which spouses are embroiled in a marital argument that one spouse's withdrawal may serve to initially reduce his or her own emotional distress (Roth & Cohen, 1986), but this withdrawal may also result in increased distress for the spouse's partner who may wish to resolve the situation (Levenson & Gottman, 1985). Roth and Cohen (1986) suggested that approach strategies can be beneficial when they lead to appropriate action or ventilation of affect. Gottman and Krokoff (1989), however, found that conflict engagement by either spouse was negatively associated with concurrent marital satisfaction for both spouses. Consistent with Roth and Cohen's (1986) view, it is also possible that confrontation, an approach strategy defined as an "aggressive effort to alter the situation" (Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986) may initially be beneficial for the individual who ventilates his or her negative affect. However, consistent with Gottman and Krokoff's (1989) findings,  19  we would not expect confrontation to be beneficial when spouses are dealing with marital conflict in which one spouse receives the brunt of the other spouses' ventilation. Regardless of the specific effects that one spouse's coping may have on the other, it is clear that in order to evaluate the effectiveness of spouses' attempts to cope with marital conflict, it is important to consider not only their individual coping responses, but also their partners' coping responses. Although the research being presented here considers the individual's coping efforts, coping responses are also considered within the context of the social unit, in this case the dyad. In this research dyadic coping thus refers to the interpersonal context of coping, i.e., coping that focuses on interpersonal coping strategies employed by both members of a marital couple when they are dealing with marital tension. Because the coping literature has not focused on the interpersonal context of coping, coping inventories, such as the Ways of Coping (Folkman & Lazarus, 1980, 1985; Folkman, Lazarus, Gruen, & DeLongis, 1986), have been developed to study individual coping rather than interpersonal coping. Thus, it is not surprising that some researchers (e.g., Bolger et al., 1989; Kessler, Price, & Wortman, 1985; O'Brien & DeLongis, in press) have noted that standard coping inventories (e.g., Folkman & Lazarus, 1980; Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986; Stone & Neale, 1984a) are not well suited when interpersonal conflict is the topic of interest. When considering marital conflict, it is particularly important that coping strategies capture the interpersonal dynamics of the situation. Thus, researchers in the communication literature have noted that marital partners have three interpersonally-based options when faced with a conflict within their relationship: avoid or withdraw from their spouse, compete or aggress against their spouse, or cooperate or compromise with their spouse (Fitzpatrick, 1988; Sillars, Colletti, Parry, & Rogers, 1982). In the preceding discussion of general coping inventories within the literature, two of these strategies were noted as being well represented in the coping literature: confrontation (an approach  20  strategy that relates to competitive or verbally aggressive attempts to deal with a stressor) and avoidance (a strategy entailing avoidance or withdrawal from a stressor). However, the third strategy noted here, cooperation or compromise, has received virtually no attention within the mainstream coping literature. Given the lack of emphasis in the stress and coping literature on interpersonal aspects of coping, it is perhaps not surprising that cooperation or compromise, interpersonal strategies related to satisfying concerns of others along with concerns of one's own, generally have not been included in studies of the stress and coping process. Even the few inventories that have been developed within the stress and coping literature to deal more specifically with marital coping have not addressed adequately issues related to the use of compromise (Bowman 1990; Pearlin & Schooler, 1978). Pearlin and Schooler (1978) included negotiation items in their questions about coping responses to marital stress (e.g., "How often do you try to find a fair compromise in marriage problems?" "How often do you sit down and talk things out?" p. 20). These questions focused on how respondents usually coped with general marital stressors occurring during the past year. Given that retrospective contamination may be of particular concern when asking about coping with general, as opposed to specific, situations (Folkman & Lazarus, 1980; DeLongis, Hemphill, & Lehman, 1992), Pearlin and Schoolers' (1978) negotiation items did not allow for an assessment of how spouses actually cope with specific marital stressors (see pp. 33-36 for a more complete discussion of this research and its applicability to the issues being discussed here). Similarly, Rands et al. (1981) included compromise strategies in their study of marital conflict resolution, but this research assessed only how respondents perceived that their spouses usually coped with marital conflict, not how they themselves or their spouses actually coped with specific marital conflict. In a more recent study of marital coping, Bowman (1990) developed a Marital Coping Inventory that included five scales (i.e., conflict, introspective self-blame, positive appraisal, self-doubt, and avoidance), but this inventory did not include the interpersonal  21  strategies of cooperation or compromise. Perhaps because her scales were based on the extant stress and coping literature, this interpersonal dimension of coping was not included in her research. Although compromise strategies are included in some scales outside of the mainstream coping literature (e.g., Rusbult, Johnson, & Morrow, 1986) and in the most widely used observational coding system [Marital Interaction Coding System (MICS), Weiss & Summers, 1983], compromise tends to be imbedded in larger categories such as voice and loyalty (Rusbult, Johnson, & Morrow, 1986) or combined with other positively coded strategies such as agreement, assent, and approval (MICS coding in Gottman & Krokoff, 1989) or problem-solving and accepting responsibility (MICS coding in Margolin & Wampold, 1981). To address the limited interpersonal focus in the stress and coping literature, an interpersonal conflict coping inventory was developed for the research presented here based on conceptualizations of interpersonal conflict drawn from the marital, business management, and human communication literatures, where interpersonal conflict is a central issue. Within the marital literature, the Conflict Tactics Scale (Straus, 1979) is widely used in research that focuses on marital and intrafamily conflict. However, this scale, with its three subscales related to reasoning, verbal aggression, and violence, was designed to distinguish varying degrees of verbal and physical aggression/violence, and therefore it was not suited to the issues of interest in this dissertation. Other classification schemes have been developed, however, to capture the tactics, excluding reliance on violence, that make up the coping strategies employed in managing interpersonal conflict. These classification schemes have generally reflected two major dimensions: the directness or disclosiveness of the strategies employed and the competitiveness of these strategies (e.g., Blake & Mouton 1964; Fitzpatrick, 1988; Putnam & Wilson, 1982; Sillars et al., 1982). The dimension of directness or disclosiveness is akin to Roth and Cohen's (1986) distinction between approach and avoidance strategies; approach strategies are direct and relatively disclosive about perceptions and feelings, whereas avoidance  22  strategies are not. The second dimension, competitiveness, refers to the degree to which strategies are intended to satisfy one's own concerns or one's partner's concerns (Pruitt & Rubin, 1986). This dimension brings into focus cooperation and compromise, strategies generally neglected within the stress and coping literature. It should be noted that cooperation and compromise can also be considered approach strategies, as defined by Roth and Cohen (1986) in that they entail movement toward a stressor. However, because Roth and Cohen based their analysis on the extant stress and coping literature, interpersonal aspects of coping, such as cooperation, were not given consideration. It is important for our purposes, however, to note that of the various categorization schemes within the stress and coping literature discussed earlier, the approach-avoidance scheme best captures the dimensions of coping noted here as being particularly relevant to interpersonal coping (e.g., coping with marital tension). Based on the above general conceptualization of interpersonal conflict resolution, Sillars et al. (1982) developed and validated an observational coding procedure that classified statements made by participants during interpersonal conflict into the two dimensions of directness and competitiveness. In this classification scheme three broad classes of conflict behavior were delineated: competition or confrontation, cooperation or compromise, and avoidance or withdrawal. Because these three types of responses appear to capture, both empirically and conceptually, the primary dynamics of interpersonal conflict (Burggraf & Sillars, 1987; Fitzpatrick, 1988; Sillars et. al., 1982; Zietlow & Sillars, 1988), the conflict inventory developed for this research included these three categories of coping responses, labeled here as confrontation, compromise, and withdrawal (see Methods section for a complete description of this inventory). Summary Although the Lazarus et al. model of stress and coping has proven useful in the explication of the coping process for individuals across a range of stressful situations  23  (including home, work, and health), the foregoing overview of this model has demonstrated its limitations when applied to the interpersonal domain of coping with marital tension. By limiting the role of interpersonal factors to that of social support, this model has neglected other important aspects of the social context in the stress and coping process (DeLongis & O'Brien, 1990). Thus, existent measures of coping (Billings & Moos, 1981; Bowman, 1990; Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986; Stone & Neale, 1984a) have not included interpersonally based coping strategies such as compromise or cooperation that have been found to be critical to an understanding of interpersonal conflict (Fitzpatrick, 1988; Pruit & Rubin, 1986; Sillars et al., 1982). In addition, Lazarus et al.'s distinction between two broad functions of coping, problemfocused coping and emotion-focused coping, has been found inadequate for characterizing the dimensions of coping that are most pertinent to coping with interpersonal stress (i.e., marital tension or conflict). Furthermore, the extant coping literature presents an incomplete understanding of the relationship between coping and adaptational outcomes. Although individual coping responses are important in our understanding of the effectiveness of particular coping strategies, research has demonstrated that how one's spouse copes with a stressor may be associated with one's own coping and well-being (Coyne et al., 1990; Cronkite & Moos, 1984; Manne & Zautra, 1989; Stern & Pascale, 1979). Regarding primary cognitive appraisal, very little research has focused on this critical aspect of the coping process and what research there is (e.g., Folkman, Lazarus, Gruen, & DeLongis, 1986; Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986) has included aggregated stressors (i.e., pertaining to work, home, and health). As will be demonstrated in the following section, disaggregated analyses are necessary for our understanding of the dynamics of coping with marital tension (Compas, Forsythe, & Wagner, 1988; Folkman & Lazarus, 1980; Pearlin & Schooler, 1978). In order to understand the primary appraisal process as it pertains to spouses experiencing marital  24  tension or potential conflict, we must look at the significance that spouses place on their relationships. These issues are explored in the following review of the gender differences literature. The research being presented here attempts to address these various limitations. In this research, cognitive appraisal as it relates to marital conflict is conceptualized within a framework that takes into account the significance that spouses place on their relationship, how this might differ according to one's gender, and how this difference might affect the appraisal process. Acknowledging the importance of interpersonal factors, the coping inventory used in this research also includes interpersonally based strategies such as cooperation. In addition, this research considers the effectiveness of coping in light of both the spouse's own coping and his or her partner's coping. Having presented the stress and coping framework that, with modification, provided the theoretical structure for the exploration of coping in my research, I turn now to issues related to gender differences that are also fundamental to the research being presented here. Because the focus in this research is on issues related to gender differences in coping, a review of gender differences within the stress and coping literature is presented. Following a brief review of these research findings, a review of research related more specifically to how wives and husbands differ in their attempts to cope with marital conflict is presented. Then, a selected review of the more general marital and gender differences literature is presented to provide theoretical direction to issues related to gender differences in cognitive appraisal of marital conflict. This review also provides additional support for the hypotheses that are presented in this research related to gender differences in coping with marital conflict. The review of these various literatures relating to gender differences begins with a clarification of the terminology employed in this dissertation to discuss gender differences.  25 Gender Differences Defining Gender Before turning to the relevant research pertaining to gender differences in heterosexual relationships, it is important to clarify some of the terminology used in this dissertation. Specifically, distinctions should be made between use of the words gender and sex when discussing female and male differences. During the 1970s, feminist writers adopted the convention of using the word gender as a psychological and cultural term that refers both to an individual's subjective sense of her or his femaleness or maleness (gender identity), and to society's evaluation of behavior on the basis of perceptions of masculinity and femininity (gender role) (Basow, 1992; Unger, 1979). Sex, on the other hand, is a biological term that establishes maleness or femaleness on the basis of one's sex organs and genes. In the research presented here this convention is followed by using the expression gender differences to emphasize the psychological and social aspects of being male or female rather than the biological aspects. Making this distinction does not imply that biological factors are not important in the development of gender and gender differences. The long-standing controversy concerning the contribution of nature vs. nurture has demonstrated that human nature develops from a complex interaction of biological, psychological, and sociocultural factors. However, because explorations of this interaction are beyond the scope of this research (see Adkins-Regan, 1989, for a discussion of the status of biosocial interactionist models in gender research), the focus here is on the psychological and social features of gender. From this perspective, gender is viewed as a concept created by humans and shaped by historical, cultural, and psychological processes. Thus, gender differences are not inherent or immutable; gender differences are subject to change as cultural and historical pressures change. For example, there is evidence that most of the previously reported gender differences in cognitive functioning have virtually disappeared in the past two decades as females' roles in our society have expanded (Feingold, 1988; Hyde & Linn, 1988).  26  Gender differences in the use of general coping strategies One of the first problems encountered when surveying the literature pertaining to gender differences in coping is that there is surprisingly little in the literature that relates specifically to gender differences (e.g., see reviews by Miller & Kirsch, 1987; Thoits, 1991). Once the few studies that address gender differences are located, a further difficulty arises concerning their comparability. As noted previously, there is considerable controversy within the literature over the conceptualization of coping and the development of typologies of coping strategies (Haan, 1982; Moos & Billings, 1982). Therefore, it is difficult to compare studies because of the variability in both defining and categorizing coping strategies. For example, as noted above, Lazarus and his colleagues distinguish between two coping functions: problem-focused and emotion-focused coping. Other researchers (Stone & Neale, 1984a; Pearlin & Schooler, 1978; Billings & Moos, 1981) present a typology with three categories: active behavioral strategies, active cognitive strategies, and avoidance strategies that can be further broken down into approachavoidance strategies (Roth & Cohen, 1986). Although these latter researchers and Lazarus and his colleagues have developed similar measures of coping, these measures group coping strategies into different typologies that make comparisons difficult. Thus, because of various conceptual differences, a specific strategy may be defined as belonging in one category by one researcher, whereas another researcher conceptualizes it as belonging in another category. For example, Kobasa (1982) included "getting angry" as an avoidance strategy, whereas Folkman, Lazarus, Dunkel-Schetter, DeLongis, and Gruen (1986) conceptualized the same strategy as an active problem-focused (i.e., confrontive) response. Such differences render comparisons across different research paradigms extremely difficult. Keeping in mind the limitations of comparing across research studies, the following review of gender differences in coping is offered. Although Roth and Cohen's  27  (1986) distinction between approach and avoidance strategies is particularly germane to the research being presented here (i.e., marital coping) and thus preferable as a categorization scheme, the research findings relating to gender differences in the use of general coping strategies do not lend themselves to this particular typology. Because of the prominence of Lazarus et al. 's research pertaining to general coping across a wide range of situations, these research findings are presented within the literature in terms of problem-focused vs. emotion-focused strategies (sometimes mistakenly equated with active vs. passive strategies). The following review is thus organized accordingly. In a review of gender differences in coping, Miller and Kirsch (1987) reported equivocal findings regarding gender differences in types of cognitive coping strategies employed to deal with stress. These authors reported, however, that the strongest findings of difference between males and females related to gender differences in the use of problem-focused strategies. Their review concluded that males used problem-focused coping more often than did females, thus suggesting the stereotype of the rational, active problem-solving male and the more indirect, emotionally expressive female. However, a review of the evidence for this conclusion, including some studies not available at the time of Miller and Kirsch's (1987) review, presents a much more ambiguous picture. When men and women were asked in a large interview study how they typically cope with stress, Veroff et al. (1981) found that men reported using more rational, problem-solving strategies than did women. Stone and Neale (1984b) also reported that males were more likely to report taking direct action, whereas females were more likely to report employing indirect (emotion-focused) strategies, such as distraction, relaxation, and reliance on religion. Also, in two studies of medical patients, men reported using more active problem-solving strategies than did women (Viney & Westbrook, 1982; Shanan, DeNour, & Garty, 1976). It is important to note, however, that Shanan et al. (1976) found no gender differences in the reported use of coping strategies by a normal, healthy, and apparently non-distressed control group. Folkman and Lazarus (1980) found gender  28  differences, but only in the reported use of coping strategies related to specific problem areas. No gender differences were found in the males' and females' reports of emotionfocused coping for either work, health, or family problems; however, males were more likely than were females to report problem-focused coping strategies in response to work problems, a finding that Folkman and Lazarus (1980) related to the fact that females usually occupy lower level employment positions where fewer opportunities exist for the use of problem-solving strategies (see further discussion of this issue on p. 54). Consistent with this interpretation, research pertaining only to higher levels of employment in which the employees have some control over their work environment (Long, 1990; Rosario, Shinn, Morch, & Huckabee, 1988) found no gender differences in the reported use of problem-solving strategies. Several other studies have reported no gender differences in respondents' reported use of problem-focused strategies (Billings & Moos, 1981; Folkman, Lazarus, Gruen, & DeLongis, 1986; Folkman, Lazarus, Pimley, & Novacek, 1987; Hamilton & Fagot, 1988; Long, 1990; Rosario et al., 1988; Thoits, 1991). For example, Folkman, Lazarus, Gruen, and DeLongis (1986) and Folkman et al. (1987) found no gender differences in the reported use of confrontive coping or planful problem solving (i.e., problem-focused strategies); the only differences they found related to males reporting the use of more selfcontrol strategies (e.g., keeping their feelings to themselves), and females reporting the use of more positive reappraisal strategies. Billings and Moos (1981) also reported no gender differences in problem-focused coping, although they did find that women reported using significantly more emotion-focused strategies than did men. In a study with 90 undergraduates in which coping responses were coded either as instrumental (e.g., problem-solving) or expressive (e.g., emotion-focused), Hamilton and Fagot (1988) found no gender differences in reported use of either of these two categories of coping responses.  29  In contrast to Folkman and Lazarus's (1980) finding of more reports of problemfocused coping among men dealing with job-related stressors than among women dealing with job-related stressors, more recent studies have found no gender differences in these types of stressors. For example, a comparison of male and female managers revealed no gender differences in the use of active, problem-solving coping (Long, 1990). Female managers, compared to male managers, did report more avoidance and reappraisal coping, however (Long, 1990). Interestingly, in a study comparing the perceived effectiveness of coping for male and female teachers, Long and Gessaroli (1989) found that both males and females perceived problem-focused coping to be effective, but females considered problem solving to be more effective than did males, and males considered avoidance coping to be more effective than did females. In two studies examining coping with job stress by group therapists and child-care workers, and in a third study pertaining to undergraduates coping with multiple role stress, Rosario et al. (1988) found no gender differences in the reported use of problem-focused strategies in these three separate studies. These researchers also found no gender differences in the reported use of emotion-focused strategies in two of their studies; however, in one study they found that male child care workers reported using significantly more emotion-focused strategies than did female child care workers (Rosario et al., 1988). This latter finding was due primarily to males reporting the use of exercise and sports (coded as an emotion-focused strategy) in response to an open-ended question about what they do to relieve workassociated stress. Taken together these research findings offer a very ambiguous picture of gender differences in coping. However, research by Compas et al. (1988) and Folkman and Lazarus (1980) sheds some light on why such a confusing picture might emerge. These researchers provided empirical evidence that coping is problem-specific rather than problem-invariant (see also Pearlin & Schooler, 1978), suggesting that it is meaningless to try to understand gender differences in coping without taking into account the domain  30  (e.g., work, home, health) of the stressful problem with which the individual is coping. In a study focusing on academic and interpersonal stressors, Compas et al. (1988) found moderate consistency in coping with the same type of stressor over time, but low consistency in coping across the two different types of stressors. Thus, coping efforts vary as a function of the type of stressor; that is, coping is situation specific. Folkman and Lazarus (1980) found that across all stressful episodes, females reported more emotion-focused strategies than did males. However, this finding was due to the fact that women reported significantly more health problems, with which they coped by using emotion-focused strategies. As noted above, when episodes were disaggregated and coping was compared within specific problem areas, this difference in the use of emotionfocused strategies disappeared. Furthermore, research (Folkman, Lazarus, Gruen, & DeLongis, 1986; Pearlin & Schooler, 1978) has also found that males compared to females tend to report more stress-related work problems; whereas females, compared to males, tend to report more stress-related family problems. Because there is evidence that these domains pull for different types of coping (Folkman & Lazarus, 1980; Pearlin & Schooler, 1978), gender differences in coping found in studies of events which are aggregated across domains (e.g., work, home, and heath domains) may relate primarily to the fact that males and females are reporting coping strategies that relate to different kinds of events or stressors. To deal with this confound, it is important that investigations of gender differences in coping focus on males and females who are coping with the same stressors (e.g., marital conflict). The above review has thus demonstrated that, on the basis of current research, there is no conclusive evidence for the stereotypical belief that across stressful situations men are more efficacious copers who use more active problem-solving strategies than do women, or that women are less efficacious copers who use more passive or emotionallyfocused strategies than do men. As argued above, this inaccurate characterization of gender differences in coping developed largely due to a focus on studies of aggregated  31  events that don't take into account the domain-specific nature of coping. Furthermore, there are reasons to question the relevance of the problem-focused/emotion-focused typology for understanding either the effectiveness of coping efforts or gender differences in coping, particularly in relation to interpersonal conflict. For example, although problem-focused coping pertains to active approach strategies, emotion-focused coping includes both active and passive, approach and avoidance strategies. This typology thus obscures the approach-avoidance dimension of coping that is important to our understanding of coping effectiveness (Roth & Cohen, 1986) and interpersonal coping (Fitzpatrick, 1988; Sillars et al., 1982). The significance of the approach-avoidance dimension of coping will become increasingly clear in the following review of the literature relating to gender differences in coping with marital tension. Gender differences in coping with marital tension Gender differences in marital interactions, based on spouses' complaints about their partners, have been reported in the marital literature for over fifty years. For example, in one of the earliest studies of marriage, Terman, Buttenweiser, Ferguson, Johnson, and Wilson (1938) reported that husbands complained about their wives' high levels of criticism and complaining, whereas wives complained about their husbands' withdrawal from interactions. More recently, interview studies have echoed these sentiments, noting that wives tend to express their feelings in confrontations with their spouses, whereas husbands tend to withdraw or try to avoid conflict engagement (Komarovsky, 1962, 1976; Rubin, 1976, 1983). From her interviews with spouses, Rubin (1976, 1983) concluded that these gender differences lead some spouses to become "intimate strangers" who "talk at each other, past each other, or through each other, (but) rarely with or to each other" (Rubin, 1976, p. 116). Similarly, in describing how spouses deal with marital conflict, clinicians often report a pattern in which one partner pressures the other to confront a particular problem,  32  while the other partner retreats through withdrawal, avoidance, or passive inaction (Christensen & Heavey, 1990). Various labels have been applied to this pattern including pursuer-distancer (Fogarty, 1976), rejection-intrusion (Napier, 1978), closeness-distance (Jacobson, 1989; Jacobson & Margolin, 1979), and demanding-withdrawn partners (Greenberg & Johnson, 1986; Wile, 1981). According to clinical lore, this pattern is gender linked, with females taking on the demanding role and males the withdrawn role. It is interesting to note that this characterization of gender differences in coping with marital conflict is virtually opposite to that of the passive female and active male presented by Miller and Kirsch (1987) in their examination of more general gender differences in coping with stress. In addition, just as the characterization of gender differences in general coping presented a stereotypical view of men and women that portrayed females more negatively than males, the tenor of the gender distinctions relating to marital conflict often mirrors the negative cultural stereotype of the wife as a scold or shrew constantly nagging her husband, and the husband as a bewildered and resentful Caspar Milquetoast seeking escape from his overbearing wife. The meaning conveyed by these descriptors is also consistent with research indicating that assertive behavior engaged in by females is evaluated more negatively than the identical behavior engaged in by males (Connor, Servin, & Ender, 1978; Gervasio & Crawford, 1989). Within the empirical literature, the findings are equivocal regarding this pattern of the confronting female and withdrawing male partner. Several studies (including unpublished studies cited by other authors) have reported no gender differences in the use of approach and avoidance strategies [e.g., Burggraf & Sillars, 1987; Gottman & Levenson, 1992; Kahn, Coyne, & Margolin, 1985; Buunk & Nijskens, 1980, and Kerkstra, 1985, (cited in Schaap, Buunk, & Kerkstra, 1988); Roberts & Krokoff, 1990]. Other studies have reported data supporting the notion of more confrontation from the female partner and more withdrawal from the male partner (e.g., Bowman, 1990; Christensen & Heavey, 1990; Margolin, Burman, & John, 1989; Rusbult, Johnson, &  33  Morrow, 1986). Other researchers (Pearlin & Schooler, 1978) have suggested a different pattern, with wives demonstrating more avoidance strategies than are demonstrated by their husbands. A closer look at some of these studies will illuminate some of the issues being addressed in this dissertation. Pearlin and Schooler's (1978) seminal research, involving interviews with 2,300 people, was one of the first large scale community research projects that looked specifically at coping within several domains, including marriage (see also, Pearlin, 1975). Pearlin and Schooler (1978) found that wives more often reported using strategies of negotiation, self-assertion (vs. passive forbearance), and selective ignoring when dealing with marital stress; whereas husbands more often reported using self-reliance (vs. advice seeking) and controlled reflectiveness (vs. emotional discharge). Pearlin and Schooler's (1978) data also indicated that the use of avoidance and withdrawal strategies was associated with increased stress within the marriage. Because wives reported using more of these strategies (i.e., selective ignoring) than did husbands, the wives were described in this research as "less efficacious copers" than their male counterparts (Pearlin & Schooler, 1978). Billings and Moos (1981) also reported that wives are less effective copers than are husbands because wives reported greater use of avoidance strategies than did husbands. However, these latter researchers did not specifically include marital stressors in their categories of stressful events (instead, they included events related to illness, death, economic concerns, children, and other interpersonal or non-interpersonal issues); therefore, it can not be determined whether spouses differed in the ways in which they dealt with marital conflict. The applicability of the Pearlin and Schooler (1978) data to the questions being raised by this dissertation is limited, however. Perhaps the greatest limitation of these data to the issues being raised here is that information was gathered from only one member of each household; therefore, there were no data relating to how two spouses within the same marriage might deal with stressful marital events. In addition, Pearlin  34  and Schooler (1978) focused on how spouses typically coped with general sources of marital stress. Thus, they asked their respondents how frequently they had responded in particular ways to general marital strains during the past year (e.g., "How often do you yell or shout to let off steam?" "How often do you tell yourself that marital difficulties are not important?" p. 20). If we are interested in understanding how spouses actually cope with specific marital stressors (i.e., marital arguments), the methodology employed by Pearlin and Schooler (1978) is inconclusive for several reasons. As pointed out by Folkman and Lazarus (1980) there is generally a poor relationship between what individuals say that they typically do and what they actually do in specific situations. By asking respondents how they usually coped, the likelihood is increased that respondents may rely on stereotypes and awareness of how they should have coped or how they ideally would have coped rather than how they actually coped. For example, in studies in which data were gathered from women concerning their mood across their menstrual cycle, retrospective reports indicated that the respondents felt worse in the premenstrual stage than did reports based on prospectively obtained records of the women's mood across their cycle (Kessler, DeLongis, Haskett, & Tal, in press; McFarlane, Martin, & Williams, 1988). Just as stereotypical beliefs about hormonal changes apparently led respondents to report symptoms consistent with a "premenstrual syndrome," stereotyped beliefs about coping may have affected respondents' reports of their coping efforts in Pearlin and Schooler's study. In addition, the use of standard survey methodology such as that employed by Pearlin and Schooler (1978) yields data based on retrospective reports that may be systematically biased (DeLongis et al., 1992; DeLongis & Lehman, 1989). DeLongis et al. (1992) delineated several sources of retrospective contamination, including the following: (a) searching for meaning and making attributions for past events may render retrospective reports unreliable and invalid (Stone, 1973); (b) some experiences are more salient than others and thus remembered more readily than are more mundane or everyday  35  experiences (Hedges, Jandork, & Stone, 1985); (c) because memory is mood-congruent (Lewis & Williams, 1989), one's mood at the time of recall may differentially affect recall (i.e., memories congruent with current mood are more likely to be recalled). Furthermore, the risk of retrospective contamination and systematic error being introduced into self-report data increases with the passage of time between the event and recall (Schuman & Kalton, 1986; Sudman & Bradburn, 1982). Given that recall spanned one year in Pearlin and Schooler's (1978) study, the data may not have accurately reflected how spouses actually coped with naturally occurring marital stressors and strains. To illustrate how retrospective recall bias may have affected Pearlin and Schooler's data, consider their finding that wives reported more "selective ignoring" strategies (leading them to be described as "less efficacious copers") than did husbands. Selective ignoring was measured by asking the respondent the following questions: "How often do you: (1) tell yourself that marital difficulties are not important; (2) try to overlook your spouse's faults and pay attention only to good points; (3) try to ignore difficulties by looking only at good things" (Pearlin & Schooler, 1978, p. 20). Because endorsement of these statements is in keeping with the cultural stereotype of the "good wife" and because retrospective recall focusing on typical coping tends to enhance stereotypical responding, women, more often than men, may have endorsed these items. Furthermore, there is evidence that husbands tend to be less reflective about their relationship, more satisfied with their marriage, and less aware of its problems than are their wives (HoltzworthMunroe & Jacobson, 1985; Margolin, Talovic, & Weinstein, 1983); therefore, the salience of "ignored" problems is likely to differ for husbands and wives. Given that salience affects retrospective recall (Hedges et al., 1985), husbands may be reporting less use of "selective ignoring" because this is not a salient issue for them: husbands may not have had to ignore concerns about which they remained unaware or undisturbed, or they may have forgotten instances of "selective ignoring" because of less reflection about issues considered to be of little consequence to them.  36  Research findings based on both heterosexual dating couples and married couples have indicated a greater use of avoidance strategies by males compared to their female partners, and a greater use of approach or confrontation strategies by females compared to their male partners. In a study of heterosexual dating college students, Rusbult, Johnson, and Morrow (1986) obtained questionnaire data from both dating partners about how they dealt with problems within their relationships. The questionnaires focused on four categories of responses: voice (i.e., actively and constructively trying to improve the situation), loyalty (i.e., passively, but optimistically waiting for things to improve), neglect (i.e., passively allowing the relationship to deteriorate), and exit (i.e., actively leaving or destroying the relationship) (Rusbult, Johnson, & Morrow, 1986). Two of these categories of responses, voice and neglect, corresponded somewhat to the coping strategies of interest in this dissertation. Relating to both confrontation and compromise strategies, voice responses included behaviors such as discussing a problem or seeking a compromise solution; neglect responses included withdrawal or refusing to discuss an issue-responses that are conceptually similar to withdrawal or avoidance strategies. In their research, Rusbult, Johnson, and Morrow (1986) found that females, compared to their male partners, reported engaging in significantly more voice responses, and males reported slightly, but not significantly, greater use of neglect strategies than were reported by females. Regarding the other two response types, there were no gender differences in reports of exit strategies, but females reported significantly more loyalty responses than did males. In a series of three studies using the same four categories of responses (voice, loyalty, neglect, and exit), Rusbult, Zembrodt, and Iwanisjek (1986) obtained data from both undergraduates and other adults residing in the community pertaining to their generalized reaction to dissatisfaction in romantic relationships. Although no gender differences were found for "constructive responses" (i.e., voice and loyalty), males were  37  more likely than females to report exit responses in one study, and males were marginally more likely than females to report neglect responses in two studies. In another series of studies involving six studies, Rusbult, Verette, Whitney, Slovik, and Lipkus (1991) used the same typology of responses to explore accommodation in close relationships. Accommodation was defined as a person's willingness in interpersonal encounters to "enhance tendencies toward constructive reactions (voice and loyalty) and inhibit impulses toward destructive reactions (exit and neglect)," (Rusbult et al., 1991, p. 54). Although consistent gender differences in willingness to accommodate were not found, in three of four cases women were found to exhibit greater accommodation than were men. Given that accommodation was found to be primarily mediated by relationship commitment, women's greater level of accommodation appears to be related to their greater commitment to the relationship (Rusbult et al., 1991). Generalizing the results of Rusbult and her colleagues research to the issues of interest here is problematical, however. Because Rusbult, Johnson, and Morrow (1986) and Rusbult et al. (1991) relied primarily on undergraduate dating couples rather than on married couples, there may be little correspondence between relationship issues and conflict as they are experienced and managed within a dating relationship compared to a marital relationship. In addition, all of these studies (Rusbult, Johnson, & Morrow, 1986; Rusbult, Zembrodt, & Iwanisjek, 1986; Rusbult et al., 1991) dealt with general patterns of problem solving among couples; therefore, participants responded to questions about general tendencies in dealing with relationship problems, including being upset with their partner or disapproving of their partner's behavior. Thus, as noted earlier, it is questionable that their responses were representative of couples' actual behavior in conflict situations within their relationship. The responses of most of the subjects, given the methodology employed and their relative youth (average age of early 20s in most of this research) and relative lack of relationship experience, might resemble stereotypical responses rather than actual responses to interpersonal conflictual situations. Furthermore,  38  the response typology used by Rusbult and her colleagues was very broadly constructed to study the development and deterioration of relationships, and therefore it did not correspond precisely with the issues of interest in this dissertation (i.e., couples dealing with marital conflict via confrontation, compromise, and withdrawal). For example, neglect responses included making critical or insulting remarks, a type of response that may be in keeping with confrontation strategies; and, loyalty responses included choosing to ignore or forget about certain bothersome behaviors, a type of response that may be in keeping with withdrawal strategies. Also, compromise and confrontation strategies were collapsed into voice responses, obscuring important interpersonal differences (i.e., the competitive dimension, Sillars et al., 1982) between these types of strategies. Thus, it is difficult to generalize the findings from these studies to the specific issues being raised in this dissertation. Regarding marital coping, Bowman (1990) asked 368 married persons (some couples) to consider a recurring problem of most concern in their marriage (later categorized as communication, money, children, health, and other) and to indicate how they coped with that problem (e.g., conflict, introspective self-blame, positive appraisal, self-distress, avoidance). Bowman (1990) reported no significant relationship between type of marital problem and particular coping efforts, but she found significant gender differences in overall coping strategies employed. Females reported the use of more conflict (i.e., confrontation) strategies and males reported the use of more avoidance strategies. This research differed from that of Rusbult and her colleagues (Rusbult, Johnson, & Morrow, 1986; Rusbult, Zembrodt, & Iwanisjek, 1986; Rusbult et al., 1991) and Pearlin and Schooler (1978) in that it focused on how the respondents dealt with a specific problem within their relationships, rather than on how they typically coped or coped with a contrived event. However, because Bowman (1990) asked respondents to retrospectively describe their coping efforts with a stressor that may have occurred days, weeks, or months before, we cannot be confident about how well respondents were able to  39  remember and report the nuances of their coping process. Furthermore, respondents were treated in this study as individuals rather than as dyads. That is, only a portion of the respondents' spouses also participated in the research, and all responses were analyzed individually, not within couples. In addition, as noted earlier, dyadic coping strategies such as compromise were not included. Research by Christensen and his colleagues (Christensen, 1987; 1988; Christensen & Heavey, 1990; Christensen & Shenk, 1991) also lends support to the demand-withdraw pattern in marital conflict. For example, in a study of 31 couples with at least one son, each spouse chose one of five parenting changes that they would like their partner to make (Christensen & Heavey, 1990). Each spouse's desired change was discussed within the dyad, videotaped, and rated by naive observers for behaviors related to the demandwithdraw pattern. In addition, each spouse completed a questionnaire in which he or she rated the likelihood of particular communication patterns within their marriage (e.g., How likely is it that "Dad tries to start a discussion while Mom tries to avoid a discussion" or "Mom tries to start a discussion while Dad tries to avoid a discussion"). Although reports by wives, husbands, and observers indicated that a wife-demand/husband-withdraw interaction pattern was significantly more likely to occur than was a husbanddemand/wife-withdraw interaction, this pattern was more likely only when the wife was discussing the issue related to a requested change in her husband. Separate analyses of the demand and withdraw behaviors indicated that both husband and wife were more likely to be demanding when discussing a change that they wanted from their spouse, and both were more likely to withdraw when discussing a change their partner wanted. Overall, men withdrew more than did women, but women were not more demanding overall than were men (Christensen & Heavey, 1990). Although there was some evidence that withdrawal occurred significantly more often with men, the findings relating to demanding behavior appeared to reflect the research situation: a spouse, regardless of gender, was  40  more demanding when the experimental situation involved requesting desired change from the other partner. In another series of studies (reported in Christensen, 1987, 1988; Chnstensen & Shenk, 1991), Christensen and his colleagues assessed communication patterns with questionnaires. The demand-withdraw pattern was assessed by asking couples, individually, to indicate to what extent they dealt with relationship problems in the following way: "Wife nags and demands while husband withdraws, becomes silent, or refuses to discuss the matter further;" the reverse pattern was also presented, with husband and wife in the opposite roles (Christensen, 1988). In a study of 142 married, livingtogether heterosexual couples, Christensen (1988) reported that the spouses indicated that the wife demand/husband withdraw pattern was more likely to occur than the reverse pattern. In addition, husbands and wives tended to agree (.73 intraclass correlation) in their assessments of this pattern (although in two previous studies reported by the author, spouses' assessments did not significantly correlate). Christensen (1988) also reported that a significant correlation was found between those who expressed a desire for more intimacy and those who tended to be demanders, and between those who expressed a desire for more independence and those who tended to be withdrawers. In a similar study comparing nondistressed married couples, distressed couples seeking marital therapy, and divorcing couples, Christensen and Shenk (1991) reported marginal significance for a greater likelihood of a wife-demand/husband-withdraw interaction than a husbanddemand/wife-withdraw interaction pattern. This pattern of results was also found more often with maritally distressed and divorcing couples than with nondistressed couples. These results are difficult to interpret, however, in that they did not assess spouses' responses related to specific marital conflicts experienced within their relationship, but they relied instead on self-reports of spouses' typical responses. Not only were there problems due to retrospective recall bias (e.g., DeLongis et al., 1992), but given the wording of the response choices it was also likely that these findings might have reflected  41  stereotyped perceptions of the spouses. As noted previously, research related to women's menstrual cycles has demonstrated that beliefs about hormonal changes can influence women to retrospectively report symptoms that are consistent with a "premenstrual syndrome," (Kessler et al., in press; McFarlane, et al., 1988). Similarly, in Christensen's (1988) research, stereotypes about the "nagging wife" might have led spouses to endorse these descriptors in a manner consistent with the stereotype. Margolin, Burman, and John (1989) videotaped 73 couples in their homes enacting "typical" conflict situations, staged before an observer. Although the major goal of this study was to compare couples with previously identified marital conflict styles (i.e., physically aggressive, verbally aggressive, withdrawing, nondistressed), the authors also explored main effects for gender. Margolin, Burman, and John (1989) reported that wives, compared to husbands, were rated as expressing significantly more indications of overt hostility, problem solving, and despair; husbands, compared to wives, were rated as being significantly more patronizing, defensive, and uninvolved during the marital discussions. Corresponding to the demand-withdraw pattern, these findings suggested that wives took a more active role in marital conflict discussions, both in terms of problem solving and the expression of disapproval, whereas husbands tended to respond in a condescending, self-defensive manner or to become aloof and uninvolved in marital conflict discussions. However, because the conflict discussions were staged and the reenactments included reconstructions based on both previous single arguments and generalized versions of previous multiple arguments (Margolin, Burman, & John, 1989), the relevance of these findings to naturally occurring marital arguments is unclear. The data, for example, may be compromised by both lack of recall and biased memories. In addition, as noted below, being observed may induce various self-presentation tendencies in people, such as trying to present themselves in a socially desirable fashion or avoiding a negative evaluation by the observer (Baumeister, 1982; Rosenberg, 1965; Tedeschi, 1981). Interpretation of these results is also complicated by the use of a new, global  42  coding system that relied heavily on the judgment of the coders (interrater reliabilities for 46 different behaviors classified in seven categories ranged from .53 to .89; see Huston & Robins, 1982, for a discussion of problems related to coder contamination and other issues with observational research). Other observational studies lend partial support to the notion that wives tend to exhibit more confrontive or demanding behavior than do their husbands or that husbands tend to withdraw from conflict more than do their wives. Using Sillars et al.'s (1982) coding system, Zietlow and Sillars (1988) reported that wives had a more direct, engagement style of conflict (more confrontative and conciliatory codes) than did husbands in a study of couples ranging from 23 to 83 years of age. These spouses did not differ, however, in their coded withdrawal strategies (denial codes). Two studies using MICS coding of a role-played interaction found that wives, compared to husbands, were coded for more complaining and criticizing behaviors (Fichten & Wright, 1983; Margolin & Wampold, 1981). Although these behaviors were characterized in both studies as negative verbal behaviors, they may be interpreted (particularly by the wives' spouses) as confrontational or demanding behaviors. However, in light of additional findings, Margolin and Wampold (1981) interpreted these results somewhat differently. In their study, wives, compared to husbands, also exhibited more behavior that was coded as smile/laugh, and husbands, compared to wives, exhibited more behavior that was coded as excuse and as "not tracking" (this latter code, interpreted as withdrawal, was significant only in univariate analysis, and not in multivariate analysis). Margolin and Wampold (1981) suggested that these findings indicated that wives tend to exhibit greater emotional expressiveness, whereas husbands tend to resort to factual explanations (e.g., excuse) or withdrawal (not tracking their spouses' responses). Another group of studies have found no gender differences in the use of approach and avoidance strategies by spouses. In two unpublished studies within the communication literature reported by Schaap et al. (1988), husbands and wives did not  43  differ in their self reports of communication behavior; destructive communication, intimacy and avoidance were not reported more frequently by females or males (Kerkstra, 1985, and Buunk & Nijskens, 1980, cited in Schaap et al., 1988). In a study primarily interested in the relationship between depression and coping with marital conflict, Kahn et al. (1985) found differences between depressed and non-depressed spouses' reports of their general use of problem-solving, aggression, and withdrawal to deal with marital conflict, but they found no gender differences in spouses' reports of their use of these strategies. In addition, no gender differences were found in two observational studies. Roberts and Krokoff (1990) found no observer-rated gender differences on hostility, withdrawal, or displeasure in 22 couples that they observed resolving a marital problem chosen from a problem inventory. Similarly, Burggraf and Sillars (1987) found no gender differences in observational coding of couples' use of avoidance, confrontative, or conciliatory tactics during conflict resolution tasks in two separate studies with a total of 77 couples. Burggraf and Sillars (1987) also reported an interesting finding pertaining to the effect of context on observational research. They conducted two observational studies, one in the laboratory and another at the participants' homes. Conflict avoidance was more likely in the home context (51 % of coded acts vs. 16% in the lab), whereas analytic, information-laden communication was more likely in the lab context (43% vs. 18% in the home; Burggraf & Sillars, 1987). Thus, it appears that the context of contrived observational research will affect the type of observed behavior. Furthermore, it is possible that these effects may differentially influence males and females, thus obscuring or distorting gender differences. Gottman and his colleagues have offered another explanation for the variability in findings reviewed above. These researchers suggest that when couples are maritally distressed (i.e., marital satisfaction is rated low) husbands withdraw more from their spouses (Gottman & Krokoff, 1989; Gottman & Levenson, 1986, 1988; Levenson &  44  Gottman, 1985). For example, when Gottman and Levenson (1988) reanalyzed Komarovsky's (1976) interview data relating to gender differences in self-disclosure (i.e., husbands were reported to be less disclosing than wives), they found no gender differences in "happy" couples' self-disclosure, but males disclosed less (i.e., withdrew more) in unhappy marriages. In addition, Gottman and Porterfield (1981) found that wives were able to decode husbands' ambiguous messages equally well regardless of level of marital dissatisfaction, but husbands' decoding abilities were much worse in dissatisfied marriages. These same dissatisfied husbands weren't impaired, however, when decoding a female outside the relationship. In a longitudinal study Levenson and Gottman (1985) found that lower levels of husbands' negative and positive affect during a marital problem discussion (self ratings obtained by participants as they viewed a videotape of their marital interaction) were associated with subsequent (measured three years later) declines in marital satisfaction. Levenson and Gottman (1985) interpreted these low levels of affect as indications of emotional withdrawal on the part of the husbands that had negative ramifications for the couple's marital satisfaction. From observations such as these, Levenson and Gottman (1983, 1985) developed the hypothesis that males can't function as well as females in the context of highly negative affect and when conflict becomes intense, males withdraw. According to Gottman, "in a sea of conflict, men sink and women swim" (quoted in Markman, 1991b, p. 422). To support this hypothesis, Gottman and Levenson (1986, 1988) presented a review of data indicating that males generally are more physiologically reactive, and that they become aroused more readily and more slowly return to prestressed levels than do females. In a meta-analysis of sex differences in physiological arousal, Stoney, Davis, and Matthews (1987) also provided support for the notion that men, compared to women, show greater physiological responses to acute behavioral stress. In keeping with these findings, Notarius and Johnson (1982) found that males exhibited greater physiological reactivity (i.e., skin conductance) to females' negative speech than the reverse. They  45  suggested that this heightened physiological response leads husbands, especially those in distressed marriages, to withdraw from conflict with their wives. Similarly, in their longitudinal study of marital satisfaction, Levenson and Gottman (1985) presented data suggesting that males preferred to vent their feelings and then to be left alone, whereas females found it more satisfying to have their negative feelings responded to and found it punishing when their feelings, positive or negative, were ignored. Levenson and Gottman (1985) reported that marital satisfaction declined most over a three year period for those couples in which the husband did not reciprocate their wives' negative affect and the wives did reciprocate the husbands' negative affect. They also reported that the extent of physiological arousal at Time 1 in this study predicted marital satisfaction at Time 2; higher arousal led to greater dissatisfaction (Levenson & Gottman, 1985). Gottman et al.'s findings that wives' confrontation during marital conflict and husbands' withdrawal from marital conflict is more pronounced in distressed marriages than nondistressed marriages is consistent with anecdotal reports from marital therapists who observe wife-demand/husband-withdraw patterns among their distressed clients (e.g., Greenberg & Johnson, 1986; Jacobson, 1989; Wile, 1981). In addition, as noted above, Christensen and Shenk (1991) found the wife-demand/husband-withdraw pattern more often among maritally distressed and divorcing couples than among nondistressed couples. However, conflicting findings in the literature, including results from a recent study by Gottman and Levenson (1992) mitigate against drawing firm conclusions concerning the relationship between these gender differences and distressed marriages. For example, when observational coding was used to record the emotional behaviors of 79 couples engaged in a marital problem, Gottman and Levenson (1992) found no gender differences in coded conflict engagement or withdrawal. These researchers did find, however, that wives showed more anger and whining and less affection than did husbands, but these differences were not related to their being classified as either "regulated" (relatively nondistressed) or "nonregulated" (relatively distressed) spouses. Thus, although  46  nonregulated spouses, compared to regulated spouses, exhibited more negative emotional expressions, lower marital satisfaction, more negative ratings of interactions, and more stubbornness and withdrawal from interactions, there were no significant gender differences between these two groups of spouses. Roberts and Krokoff (1990) also found no differences in husbands' withdrawal related to distressed vs. nondistressed marriages. Thus, the data are not consistent in demonstrating gender differences in demand or withdrawal tactics in marital conflict or in demonstrating a relationship between gender differences and marital satisfaction or distress. Although there appears to be strong evidence to support the notion that men tend to be more reactive physiologically to stressful stimuli than are women (e.g., Gottman & Levenson, 1988; Stoney, Davis, & Matthews, 1987), our understanding of the relationship between physiological measures and marital distress remains incomplete. Different physiological measures appear to be predictive in different studies and they sometimes vary by gender (cf. Levenson & Gottman, 1985; Notarius & Johnson, 1982). For example, in their most recent published research, Gottman and Levenson (1992) used the same set physiological measures and procedures that they had used in their earlier research (Levenson & Gottman, 1983), but their results differed. In this study no significant physiological differences emerged to suggest that husbands were more physiologically aroused during marital problem discussion than their wives or that maritally distressed husbands were more aroused than nondistressed husbands (Gottman & Levenson, 1992). Out of the five physiological measures used with both husbands and wives, the only two differences obtained found that maritally distressed wives had shorter cardiac interbeat intervals and smaller finger pulse amplitudes than did nondistressed wives. In addition, recent research has suggested that under some circumstances wives may become more physiologically aroused (i.e., their immune systems are more negatively affected) by heated discussions of marital problems than is the case for their  47  husbands (Kiecolt-Glaser's findings presented at National Institute of Health symposium, April, 1993, cited in Adler, 1993a). Furthermore, regarding the long-term effects of demand and withdraw behaviors, Gottman and Krokoff (1989) reported data from two longitudinal studies, based on observational data at one time point and questionnaire data three years later, that are difficult to reconcile with longitudinal data reported by Levenson and Gottman (1985). Gottman and Krokoff (1989) reported that observer-rated conflict engagement (e.g., disagreement and anger exchanges) at Time 1 was associated with improved marital satisfaction three years later; whereas Levenson and Gottman (1985) reported that physiological measures indicating arousal during a conflict interaction at Time 1 was associated with declines in marital satisfaction three years later. Thus, their findings differed when different measures were employed: physiological measures of arousal predicted lower marital satisfaction (Levenson & Gottman, 1985); observer ratings of arousal predicted increased marital satisfaction (Gottman & Krokoff, 1989). To further complicate interpretation of these results, in their most recent longitudinal study, Gottman and Levenson (1992) found a nonsignificant relationship between marital satisfaction and physiological measures obtained four years previously. They noted, however, that the recent data were consistent with the Levenson and Gottman (1985) findings in that there was a trend toward an association between increased physiological arousal at Time 1 and decreased marital satisfaction at Time 2 (Gottman & Levenson, 1992). Thus, although Gottman et al.'s data are intriguing and somewhat provocative, the relationship between physiological arousal and later marital satisfaction awaits clarification from further research (Weiss & Heyman, 1990). Overview In summary, it appears that the research findings pertaining to gender differences in coping, both in general and when specifically related to marital conflict, are somewhat  48  inconsistent. Empirical research has supported Lazarus and his colleagues' theoretical position that coping is situationally determined (Compas et al., 1988; Folkman & Lazarus, 1980; Pearlin & Schooler, 1978). Therefore, it is not meaningful to compare males' and females' coping processes without taking context into account. The use of aggregated stressors in much of the research to date helps to explain the lack of consistent findings in the literature related to gender differences in the use of generalized coping strategies. Instead, coping must be considered within a particular role domain, such as the marital role. However, as the preceding review of this more specific literature indicates, even within a particular domain (e.g., marital tension or conflict) evidence for gender differences in coping remains confusing and seemingly contradictory. Within the clinical literature there are reports of a communication pattern in which demands are made by one spouse (usually reported to be the wife) and met with withdrawal by the other spouse (usually reported to be the husband). Although clinical researchers such as Jacobson (1989) have called for empirical research to confirm or disconfirm the existence of this conflict pattern among spouses, empirical attempts to demonstrate this pattern have met with mixed results. There appears to be a trend in the data demonstrating a greater likelihood of wifeconfrontation and husband-withdrawal in marital conflicts, especially among distressed couples; however, methodological issues mitigate against any firm conclusions. Standard survey methodology is limited because of a focus, in some studies, on what respondents report that they usually do to cope, rather than on their reports of what they do to cope with specific problems (e.g., Christensen, 1988; Pearlin & Schooler, 1978; Rusbult et al., 1991). In this type of research, problems related to retrospective recall biases (DeLongis, et al., 1992) are likely to contaminate self-reports of coping that occurred days, weeks, or months ago (e.g., Bowman, 1990; Pearlin & Schooler, 1978). In addition, this methodology often surveys only one member of a couple (e.g., Bowman, 1990; Pearlin & Schooler, 1978); and even when both members of a couple participate, there are no  49 indications that the two spouses are responding to the same event. Because different types of events tend to elicit different types of coping strategies (Compas et al., 1988; Folkman & Lazarus, 1980; Pearlin & Schooler, 1978), comparisons between spouses in which the partners are reporting their coping with different events may be misleading. Observational research related to marital conflict is advantageous in that it focuses on spouses' interactions within the dyad and it allows microanalysis of spouses' communication, including the sequential nature of their responses to one another (Weiss & Heyman, 1990). There are other methodological issues in observational research, however, that may obscure or exaggerate gender differences in spouses' interactions when dealing with marital conflict. Many of the conflict scenarios used in observational research are generated from inventories rather than from the couples' experiences, and all of the "conflict resolutions" are role-played. Contrived, reenacted marital conflict situations, even those that are drawn from couples' past experiences, may not represent spouses' actual conflict coping behavior due both to reliance on retrospective accounts of "typical" conflicts to create conflict scenarios (DeLongis et al., 1992; Folkman & Lazarus, 1980) and to the reactivity of respondents in observational research (Baumeister, 1982; Rosenberg, 1965). For example, Vincent, Friedman, Nugent, and Messerly (1979) found that spouses (both distressed and nondistressed) could "fake good" or "fake bad" in marital interactions when verbally instructed to do so. Furthermore, there is evidence demonstrating that gender serves as a social category that influences judgments, explanations, and expectations of behavior; thus there is the potential for bias in observers' ratings of spouses' performance (Deaux, 1984). In addition, the conflict episodes enacted in observational research are extremely time-limited, typically conducted within 10 or 15 minute time frames which are unlikely to represent the more gradual unfolding of naturally occurring marital conflict. It is also possible that observational studies of marital conflict may be particularly unlikely to accurately reflect certain kinds of avoidance or withdrawal responses. Just by agreeing to participate in such research,  50  spouses implicitly agree to engage in conflict and to not withdraw from the procedure. Observational marital research may thus be biased toward ratings of direct confrontation strategies and may not capture less direct and less obvious distancing or withdrawal tactics (Margolin, John, & O'Brien, 1989). In addition, observational research with its emphasis on overt behavior observed by trained or naive coders fails to capture the spouses' internal states or subjective reality (Margolin, 1987). To address many of the methodological and conceptual issues raised in this literature review, data are needed examining how husbands and wives, considered as a dyad, attempt to either deal with or avoid actual conflictual issues within their relationship as they occur on a day-to-day basis. By collecting self-report data on a daily basis from both members of a marital couple who are recording the occurrence of marital arguments or marital tension and their responses to these specific incidents of marital tension or arguments, attempts could be made to address the issues raised in this dissertation. Collection of data via a daily structured diary methodology such as being advocated here minimizes recall errors due to memory lapse or forgetting of specific details and it reduces the amount of retrospective bias (e.g., forgetting or positively reframing irritating or negative events) because entries are completed shortly after events occur (DeLongis et al., 1992). Furthermore, structured diaries allow for the collection of data reflecting the kind of naturally occurring "ups and downs" that characterize intimate relationships (DeLongis & Lehman, 1989). This type of data collection is important given that everyday marital stressors or hassles are commonly occurring types of stressors that have been found to be more strongly associated with social functioning, somatic health, and psychological well-being than are major life events assessed retrospectively from life event inventories (DeLongis, et al., 1992; Kanner, et al., 1981; Lazarus & DeLongis, 1983; Wagner et al., 1988). The advantage of studying everyday hassles (e.g., marital arguments) may relate primarily to the fact that these are proximal rather than distal variables (DeLongis et al., 1992). As noted by DeLongis et al. (1992) proximal variables  51  (e.g., a recent marital argument) may be better predictors of stress responses than are distal variables (e.g., a major life event experienced during the preceding year) because proximal variables reflect "here-and-now pressures of living as these are sensed or appraised by the individual" (p. 121). The structured diary methodology, with its daily entries, thus allows for the collection of "average" events compared to retrospective recall, which may focus more on salient, peak experiences (Hedges, et al., 1985). In addition, respondents' completion of diary entries anonymously and in the privacy of their own homes reduces the demand characteristics of the task and biases associated with observer reactivity. A survey of the marital conflict literature revealed only two studies that employed a methodology similar to the one advocated here. Huston and his colleagues (Huston, McHale, & Crouter, 1986; Huston, Robins, Atkinson, & McHale, 1987) collected selfreport data from both spouses in 168 couples via nine telephone interviews over a two or three week time period in three phases over two years. However, because these researchers gathered information only about the occurrence and type of conflict spouses experienced rather than information about how the spouses coped with the conflict, their research did not address the issues of concern here. Margolin, John, and O'Brien (1989) also collected diary-like data in telephone contacts; they contacted, on a daily basis, 75 couples who recorded open-ended daily descriptions of their most important marital interactions for fourteen days. Three episodes from each respondent were segmented into acts and coded for aggression, distress, withdrawal, affection, and calm (coding reliability was described as moderate with a kappa value of .56). Margolin, John, and O'Brien's (1989) research focused on sequential patterning of conflictual interactions among couples whose interaction style had previously been classified as either physically aggressive, verbally aggressive, withdrawing, or nondistressed. These researchers' interest was in specific group patterns (e.g., an aggressive event increased the probability of another aggressive event for physically aggressive couples), rather than in response patterns  52  related to gender. Thus, although the methodology employed by Margolin, John, and O'Brien (1989) fits many of the criteria noted above, their research was not designed to address the issues raised by this dissertation. In addition to the methodological issues raised here concerning research pertaining to gender differences in spouses' coping with marital conflict, there appears to be little theoretical direction in the research in regard to the purported gender differences. Gottman and Levenson (1988) have suggested that gender differences in physiological arousal are related to gender differences in how spouses respond to marital conflict; however, this biologically-based theory needs more clarification and more stringent empirical support (O'Leary & Smith, 1991; Weiss & Heyman, 1990). Although intriguing, this particular approach to gender differences in marital conflict was not explored in the research presented here; instead, the focus here was on more psychologically based approaches to these issues. For example, drawing from his clinical observations with marital therapy clients, Jacobson (1989) speculated that wives are more confrontive and husbands are more avoiding in marital conflicts because of underlying gender differences in intimacy preferences: wives actively seek intimacy and husbands withdraw in an attempt to maintain autonomy (i.e., avoid intimacy). Similarly, Christensen (1988) reported a correlation between those who expressed a desire for more intimacy (primarily wives) and those who were "demanders," and between those who expressed a desire for more autonomy (primarily husbands) and those who were "withdrawers." Thus, wives may place more significance on intimate contact in their marriage than do their husbands. Therefore, wives may be more invested in confronting and resolving marital conflicts that threaten this contact than are their husbands. If gender differences exist for women and men in the meaning or significance that they place on relationships and in the ways in which they respond to their partners in relationships, these differences could enhance our understanding of how women and men deal with marital  53  conflict, including how they appraise conflict, cope with conflict, and are affected by conflict. In order to explore these issues in more depth, a review is presented of the more general literature pertaining to differences between males and females in their views of relationships and their ways of relating to one another within intimate relationships such as marriage. A review of this literature provides not only a theoretical basis for approaching gender differences in appraisal of and coping with marital conflict, but it also provides an empirical basis for predictions about differences in how wives and husbands might deal with marital disagreements or conflicts. To address these issues, the following literature review focuses on gender differences in the relevant areas of involvement in intimate heterosexual relationships, relationship reflectiveness, and interpersonal distress. This review will be followed by presentation of the hypotheses being explored in the research presented here. Gender differences in the general literature Relationship Involvement Within the literature pertaining to gender differences within intimate heterosexual relationships, there is conflicting evidence regarding the significance to males and females of family roles such as spouse and parent. The literature is relatively consistent in suggesting the centrality of close, interpersonal relationships to females (Barnett & Baruch, 1987; Surrey, 1985, Veroff etal., 1981). For example, when 197 adult females (aged 25-44) from a probability sample in the New York Metropolitan area were asked which of nine possible roles were most important in terms of how they saw themselves, 35% mentioned family-related roles (e.g., wife, partner, mother), whereas only 7% mentioned work or career-related roles (Kandel, Davies, & Raveis, 1985). In fact, until recently women who were neither wife nor mother were often assumed to be defective, incomplete, unfeminine, or unnatural (Barnett & Baruch, 1987; Basow, 1992; Rossi,  54  1984). Thus, "women's family roles typically have been seen as natural and as crucial for women's well-being" (Barnett & Baruch, 1987, p. 126). The literature is less consistent, however, in relation to males due in part to the fact that the paid employee role is theoretically often assumed to be central to males, and non-workplace roles are often relegated to positions of minor importance (Barnett & Baruch, 1987; Wethington, 1988). The most often cited research has suggested that career roles are central to males (e.g., Vaillant, 1977). Some researchers have, however, reached a different conclusion, suggesting that, in general, married men experience family roles as more psychologically significant or critical to their well-being than they do paid work roles (Pleck, 1985; Veroff, et al., 1981). The discrepancies in these conclusions may be due in part to socioeconomic and age differences in the research samples. For example, Veroff et al. (1981) found an older group of men (retirees) in their study who considered self-actualization or security to be more important than a sense of belonging, and Pleck (1985) noted a subgroup of highly educated men and/or men with high occupational status (more similar to Vaillant's sample) who prized work over family or who were more involved with their work than with their family. A previously noted research finding in the stress and coping literature indicated that men more often have reported stress related threats to their careers and women more often have reported stress related threats to their loved ones (Folkman, Lazarus, Gruen, & DeLongis, 1986). This finding appears to support differences in the significance that family and work have in men's and women's lives. Many of these reported differences, however, may be primarily due to gender differences in the work force. Data from Canada's 1991 national census indicated that approximately 60% of Canadian women, compared to 70% of Canadian men, were currently in the labor force. Women, however, were primarily concentrated in the lowest paid and lowest status jobs (Statistics Canada, 1991). Of the lowest paid workers in Canada, the 1991 national census reported that 72% were women. Furthermore, a larger proportion of women (approximately 30%) reported  55  that they were part-time employees than was the case for men (12%) (Statistics Canada, 1991). Thus, it can be argued that the majority of female-dominated jobs don't involve long-term commitment, requiring instead little advanced training and allowing for flexible hours (Basow, 1992). Therefore, given that women are less likely than are men to have a long-term career commitment that is either financially or personally rewarding, it is perhaps not surprising that research has found that many women place less significance on paid employment and report less work related stress than do men. In order to better understand the significance of close interpersonal relationships such as marriage to both males and females, it is important to examine the extent and type of both partners' involvement or investment in their intimate heterosexual relationships. Some research has indicated, for example, that females tend to be the more involved or committed person in an intimate, heterosexual relationship (Rubin, Peplau, & Hill, 1981; Rusbult, Zembrodt, & Iwaniszek, 1986). Other research has suggested that men and women do not differ substantially in the extent of their commitment to their marriage (Barnett, Marshall, Raudenbush, & Brennan, 1993). Regardless of whether their commitment to their relationships differ, there is evidence suggesting that women and men differ in the meanings or significance that they attribute to their relationships. For example, it has been noted that women tend to focus more on interpersonal needs in their intimate relationships, leading them to expect more companionship and emotional support from their relationship than do their male partners (Argyle & Furnham, 1983; Huston & Ashmore, 1986; Worell, 1988). Similarly, it has been found that men and women report similar degrees of love in their relationships, but they interpret love differently. Using a six-style model of love, it was found that women differed from men in that women reported more pragmatic, euphoric, and companionate aspects to their experiences of love (Hendrick & Hendrick, 1986; Hendrick, Hendrick, Foote, and Slapion-Foote, 1984). There is also evidence that women tend to do more to accommodate the development and maintenance of emotion intimacy in their relationships than do their male  56  partners (Huston, Surra, Fitzgerald, & Catel, 1981). For example, research has found that wives report greater use of cooperative types of coping strategies (e.g., negotiation and compromise) to deal with marital conflict than do their husbands (Pearlin & Schooler, 1978; Zietlow & Sillars, 1988). These findings are consistent with the notion that females, compared to males, tend to be more cooperative in their interactions with others than are males (see Basow, 1992, or Deaux, 1976, for a review). In addition, in a survey of 1,755 adults, Schuster et al. (1990) found that men reported significantly more supportive interactions with their wives than women reported with their husbands. Not only may females tend to be more supportive than are males, but females may also be more effective than are males in their support attempts. Research has indicated, for example, that both females and males tend to report less loneliness and greater intimacy from their interpersonal contacts with females compared to their contacts with males (Wheeler, Reis, & Nezlek, 1983). These findings are in keeping with research indicating that men and women tend to differ in how they define emotional support or nurturance; women tend to focus on the egalitarian, feeling, and giving aspects of support, whereas men tend to emphasize the problem-solving aspects of support (Worell, 1988). Holtzworth-Munroe and Jacobson (1985) reported that wives in both distressed and nondistressed couples made more interpersonal attributions (described in terms of demonstrating "relationship mindfulness") than did their husbands. Although wives were moderately inclined toward making interpersonal attributions regardless of their marital satisfaction, husbands in distressed marriages reported significantly more attributional thoughts than did husbands in nondistressed marriages. The authors interpreted their findings as evidence that husbands tend to invest little time or energy in thinking about their marriage until things start going badly. Wives, on the other hand, tend to be more mindful of the relationship all along, obviating the need for enhanced mindfulness when displeasing or conflictual events occur (Holtzworth-Munroe & Jacobson, 1985).  57  As further evidence of their greater relationship involvement or "mindfulness," wives appear to take more responsibility than do their husbands for de-escalating conflicts (Gottman, 1979; Margolin & Wampold, 1981; Markman, Floyd, & Dickson-Markman, 1982; Krokoff, 1987). For example, Gottman (1979) found that wives in nondistressed marriages tended to "edit" their husbands' negative communication in such a way as to prevent negative reciprocity and the escalation of conflict. Another study suggested that not only do males less readily engage in soothing or conflict de-escalating behaviors, but they may experience considerable difficulty in being able to actually produce such behaviors (Lochman & Allen, 1979). For example, when couples were asked to become more approving during role play of conflict vignettes, females became significantly more verbally and nonverbally approving and less verbally disapproving. When the male partners were asked to become more approving, they only produced a tendency toward being more verbally approving and they actually emitted fewer nonverbal approving behaviors (Lochman & Allen, 1979). Perhaps related to differing definitions of intimacy, some researchers have suggested that males' and females' expressed needs in their intimate relationships may be so fundamentally different that they may not understand the needs of their partners (Caldwell & Peplau, 1982). In one study, for example, husbands reported that their marital satisfaction was more affected by their wives' instrumental activities at home (e.g., housecleaning, child care) than by how affectionate they were, whereas the wives reported that their marital satisfaction was more affected by how affectionate their husbands were (Wills, Weiss, & Patterson, 1974). When researchers requested that the husbands in their study increase pleasurable affectionate behaviors with their wives, these men increased both their affectionate and instrumental behaviors (Wills et al., 1974). When one wife reported no increase in affection after her husband had been instructed (by the researchers) to increase these behaviors, the husband was confronted by the researcher  58  concerning the request. The husband replied that he had, indeed, increased his affectionate behaviors, i.e., he had washed her car (Wills et al., 1974). Perhaps because females generally are monitoring the relationship more closely than males (Holtzworth-Munroe & Jacobson, 1985), research has also indicated that females are more likely to express their dissatisfaction with their relationship and to do so earlier than do males (Hagestad & Smyer, 1982; Margolin, Talovic, & Weinstein, 1983; Rubin et al., 1981). Thus, women are usually the ones who seek marital therapy, and they are the ones who express the most complaints about their marital relationship (Margolin et al., 1983). In addition, when dissatisfied with their relationships, women, compared to men, tend to express more negative affect in conflictual marital interactions (Fichten & Wright, 1983; Margolin & Wampold, 1981; Notarius et al., 1989). Regarding the types of dissatisfaction expressed by wives, a study by Hawkins, Weisber, and Ray (1980) relating to spouses' communication styles found that 171 couples rankordered their preferred styles of marital communication in roughly the same way, with both males and females reporting that they most preferred a "contactful" style (high disclosure, conveying interest in and respect for the other) and least preferred a "controlling" style (closed, rejecting stance toward mutual exploration of meaning) of communication with their spouse. Wives reported, however, that their husbands actually used significantly less contactful communication and significantly more controlling communication than the husbands themselves reported. Wives also said that they preferred that their husbands use significantly more contactful and less controlling communication than the husband himself reported that he preferred. There were no significant differences between the husbands' reports of either the wives' actual communication styles or the husbands' preferences for her styles and the wives' reports themselves. In other words, husbands tended to be relatively content with their wives' communication styles, which mirrored those they preferred (i.e., contactful); wives,  59  however, tended to be dissatisfied with their husbands' communication styles, preferring significantly more contact and less distancing from them (Hawkins et al, 1980). Similarly, in their study of 264 college couples, Davis and Oathout (1987) found that females' reported satisfaction with their relationships was most strongly correlated with the perception that their male partner was a receptive listener or willing discloser; males' reported satisfaction, on the other hand, was unaffected by this perception of their female partners. Rhyne's (1981) research with 2,190 married Canadians also indicated that wives tended to focus more on companionship and verbal interaction in their marriage than did husbands. More specifically, in this study, wives' marital satisfaction, compared to that of husbands, tended to be affected more by their spouses showing an interest in them and in what they had to say to one another and by their spouses' attempts to meet their needs for companionship and understanding (Rhyne, 1981). Consistent with these findings, other research has indicated that a confiding, contactful relationship with one's spouse tends to be particularly important for females. Hobfoll and Lieberman (1987) found, for example, that women high on spousal intimacy were less depressed when adapting to childbirth than were those who were low on spousal intimacy. Although self esteem was also relevant to adaptation in this study, women who were low on self esteem, but high on spousal intimacy, adapted nearly as well as the women who were high on self esteem (Hobfoll & Lieberman, 1987). Underscoring the importance of spousal intimacy to wives' well-being, Brown and Harris (1978) found that, in terms of females' vulnerability to depression, their having a confiding relationship with a parent, sister, or friend did not compensate for lack of a confiding relationship with their spouse. Looking more closely at the types of complaints made by both spouses and how they may differ, Margolin et al. (1983) assessed both distressed and nondistressed spouses' desires for change (either increases or decreases) in their partner's behavior using the Areas of Change Questionnaire [Weiss & Birchler (cited in Margolin et al., 1983)]. These researchers found significant gender differences on 15 of the 34 assessed change  60  items: only two of these items, relating to sexual relations and having meals served on time, were endorsed more frequently by husbands; the remaining 13 items, spanning areas such as communication, companionship, appreciation, finances, housework, and childrearing, were endorsed more frequently by wives (Margolin et al., 1983). Of the five most often endorsed items, there was overlap between wives and husbands on three items: "express emotions clearly," "show appreciation for things I do well," and "argue with me." The remaining two items endorsed more frequently by husbands related to attending to their sexual needs; the other two items endorsed more frequently by wives were "start interesting conversations with me" and "give me attention when I need it." These results suggest, as noted above, that husbands generally tend to be more content with their wives' behavior than the reverse. However, it also appears that discontent among both wives and husbands may often relate to desire for different kinds of contact with their spouses; wives tend to seek more communication, more emotional support, and more companionship from their husbands (Argyle & Furnham, 1983; Hawkins et al., 1980; Margolin et al., 1983; Worell, 1988), whereas husbands tend to request more sexual contact with their wives (Margolin et al., 1983). Clinicians have noted that once couples engage in therapy, wives' requests for change often mirror Margolin et al.'s (1983) findings: wives express more dissatisfaction and they tend to request more emotional expressiveness, affection, and nurturance from their husbands; whereas husbands tend to complain about their wives demands and typically express a desire to maintain the status quo (Jacobson, 1983). Interpersonal Stress Research has demonstrated that women are more likely than men to manifest certain psychological disorders, such as depression and anxiety (Belle, 1980; Myers et al., 1984; Nolen-Hoeksema, 1987; Robins et al., 1984; Weissman & Klerman, 1977). The relationship between gender and mental health becomes more complex, however, when we consider the association between mental health and marriage. Research suggests that  61  marriage is positively related to psychological well-being for both males and females (Broman, 1991; Crohan, Antonucci, Adelmann, & Coleman, 1989; Glenn & Weaver, 1988; Gove & Zeiss, 1987), but marriage appears to be more advantageous for men than for women (Aneshensel, Frerich, & Clark, 1981; Gove, 1972; Gove, Hughes, & Style, 1983; Radloff & Rae, 1979). Thus, married women tend to be more prone to psychological disturbance such as depression when compared to married men; however, when single men and women or divorced men and women are compared, women show slightly fewer signs of psychological disturbance than do their male counterparts (Gove, 1972). The psychological burden of marriage for women may relate to the greater demands placed on them in regard to such tasks as household care and child-rearing. In particular, research has shown that compared to other groups of women, those with children under six years of age are at greater risk for depression (Barnett & Baruch, 1985, Gore & Mangione, 1983; Pearlin, 1975; Radloff & Rae, 1979). Thus, married women's greater susceptibility to depression may relate to a greater number of stressors in married women's lives and/or to specific types of stressors in women's lives such as those associated with caregiving (Belle, 1982). Regarding the possibility of more stress in the lives of women, research has indicated that although females report significantly higher rates of psychological distress than do males, they don't tend to differ in the overall number of undesirable life events that they report (Al-Issa, 1982; Hamilton & Fagot, 1988; Ilfeld, 1980; Wethington, McLeod, & Kessler, 1987). Men and women do differ, however, in the types of stressors they report. In particular, it appears that women experience greater exposure to network events, including the death of loved ones and interpersonal conflict (Kessler & McLeod, 1984; Schuster et al., 1990). Compared to males, females appear to have a larger circle of significant or intimate others and greater access to the interpersonal lives of others, and therefore they tend to experience greater numbers of a particularly potent kind of stressor, i.e., network or interpersonal events. Having been socialized to be "relationship  62  caretakers," women's investments in their close interpersonal relationships may lead to greater experiences of distress as a result (Rusbult, Zembrodt & Iwanisjek, 1986). For example, Pearlin (1975) reported that certain types of events are differentially stressful for men and women; in his research, males were found to be more depressed by work strains and females were more depressed by marital problems. Although Pearlin's finding may be confounded by the extent to which men and women were involved in these two different roles, Long's (1990) research found gender differences in the perceived stressfulness of interpersonal stressors in the work environment of male and female managers. The female managers not only perceived interpersonal conflicts as more stressful than did the male managers, but the females also perceived their work environment as more interpersonally supportive than did the male managers (Long, 1990). These findings suggest, as interpreted by Long (1990), that women's sensitivity to interpersonal concerns may have resulted in both interpersonal conflicts and support being more salient for the female managers than for the male managers. Whiffen and Gotlib (1989) presented data indicating that marital distress reported by husbands was associated with reports of more depressive symptoms and more life stress for both husbands and wives. Marital distress reported by wives, however, was related to decreased functioning (e.g., more depressive symptoms and more life stress) for wives only; their husbands' reports of their own functioning did not differ from those of husbands whose wives reported no marital distress (Whiffen & Gotlib, 1989). This research suggests that men, compared to women, may express their negative feelings in ways that are more obvious and detrimental to their spouses, and/or that women tend to be more sensitive than men to their partner's feelings and expressions of dissatisfaction. These findings are also consistent with the notion that not only do women, as compared with men, tend to care more about what happens to others in their lives, but also that they pay a price for their caring in terms of their own psychological well-being (Belle, 1982; Kessler & McLeod, 1984).  63  The foregoing is consistent with other research, such as the empathy literature (Davis & Oathoat, 1987), in suggesting that females are especially sensitive to issues relating to their intimate others. Perhaps related to an empathic perspective, females' satisfaction with their relationships, compared to that of males, is more closely related not only to receiving emotional support, but also to giving emotional support (Argyle & Furnham, 1983; Worell, 1988). Although the research noted above suggests that women tend to be particularly vulnerable to threats or losses within their interpersonal networks, it would be a mistake to confuse females' relationship orientation with simplistic notions of dependency. Similarly, it is inappropriate to view males, with their more independent orientation as being solely independent or autonomous. There is evidence to suggest, for example, that males fall in love more quickly and more easily than do females (Cate, Huston, & Nesselroade, 1985; Rubin et al., 1981); that women, more often than men, initiate the breakup of a relationship or marriage (Gray & Silver, 1990; Rubin et al., 1981); that unmarried men experience more emotional and physical distress than do unmarried women (e.g., Gove, 1972); and that men tend to have a more difficult time with both divorce and widowhood than do women (Price & McKenry, 1988; Stroebe & Stroebe, 1983). Several factors noted in the previous literature review could explain these seemingly anomalous findings. For example, given that females tend to be more reflective or mindful about their relationships, they may also tend to be more "realistic" and less romantic about their relationships than are males (Rubin et al., 1981). Also, by being attentive to the relationship, females tend to be more aware of potential problems and more invested in resolving problems as they occur; therefore, they are more likely to be the first to signal a breakdown of the relationship and they may also be the first to suggest an end to the relationship (Margolin et al., 1983). Males, on the other hand, tend to be less reflective about their relationships and they are less likely to mobilize efforts to address problems before the relationship has already begun to break-up (Holtzworth-  64  Munroe & Jacobson, 1985). In addition, males may be more vulnerable to the loss of a close, heterosexual relationship because of the extent to which they rely on their female partner or wife to provide intimate, emotional support. For instance, males tend to have fewer close friendships with other males after marriage (Tschann, 1988). Thus, not only may men tend to rely exclusively on their wives as confidants, but they may leave it up to their wives to develop and maintain other social contacts for the couple as well (Bock & Webber, 1972). Furthermore, because both females and males report less loneliness and greater intimacy from their interpersonal contacts with females compared to males (Wheeler, Reis, & Nezlek, 1983), ex-wives or widows may have greater access to potentially supportive others than do ex-husbands or widowers. In addition, women may feel more comfortable than men establishing and using social supports when they need help (Hobfoll, 1986) and there is evidence suggesting that women may gain more benefit from supportive interactions than do men (Schuster et al., 1990; Wethington & Kessler, 1986). Also, females may be less stressed than males by the loss of the relationship because females may tend to have less to lose: females report less satisfaction with their heterosexual relationships than do males, particularly in terms of meeting their emotional needs (Atkinson, 1980; Rhyne, 1981). Therefore, they may not only lose less in terms of emotional support than would their male partners, but females may more readily fill the gap with emotional support from their larger network of intimate others (e.g., Kessler & McLeod, 1984). Overview To summarize the findings from the foregoing literature review, there appear to be differences in the quality of males' and females' relationship involvement, their responsiveness to others, their communication styles within relationships, and their experiences of interpersonal distress. The preceding review demonstrated, for example, that females' reported satisfaction with their relationships tends to be more strongly  65  associated with greater intimacy and expressiveness from their partner than does males' reported relationship satisfaction (Rhyne, 1981; Wills et al., 1974). Females also tend to do more to accommodate both the development and maintenance of relationships than do their male partners: they monitor their relationships more closely and intervene to both confront issues and to de-escalate conflicts in the relationship (Gottman, 1979; Holtzworth-Munroe & Jacobson, 1985; Krokoff, 1987). Research also suggests that given their greater sensitivity to interpersonal concerns, females tend to be more often distressed by negative events that occur to significant others and by threats to their on-going connections with others than are males (Kessler & McLeod, 1984; Whiffen & Gotlib, 1989). The findings presented in the preceding review lend support to the argument that interpersonal relationships have different significance and meaning for females and males. Adopting the terminology used by Bakan (1966), the different significance placed on intimate relationships by men and women may be referred to as differences in tendencies related to agency and communion. Agency and communion have been posited as two central inclinations, associated with gender, that affect the development of one's selfconcept and relationship to others (Bakan, 1966; Gilligan, 1982; Miller, 1976; Markus & Kitayama, 1991; McAdams, 1985; Wiggins, 1991). More specifically, agency shows itself in "strivings for mastery and power which enhance and protect ...differentiation" (Wiggins, 1991, p. 89). According to Bakan (1966), agency manifests itself in "selfprotection, self-assertion, and self-expansion...in the formation of separations...(and) in isolation, alienation, and aloneness" (p. 15). Communion, on the other hand, refers to the condition of "being part of a larger social or spiritual entity" and it shows itself in "strivings for intimacy, union, and solidarity with that larger entity" (Wiggins, 1991, p. 89). Communion manifests itself in "the sense of being at one with other organisms...in the lack of separations...(and) in contact, openness, and union." (Bakan, 1966, p. 15). Agency is thus synonymous with concepts such as instrumentality, autonomy, and  66  individuation, whereas communion represents concepts such as intimacy, affiliation, and interdependency. Gender differences in the significance of interdependency (i.e., communion) and autonomy (i.e., agency) in the development of the self have been the focus of recent feminist scholarship (e.g., Gilligan, 1982; Goldberger, Clinchy, Belenky, & Tarule, 1987; Jordon & Surrey, 1989; Miller, 1976). A self-in-relation model of female development has been proposed by these scholars to describe female development (Jordon & Surrey, 1989, Miller, 1984, Surrey, 1985). According to this model, the self, for females, tends to be organized around "being able to make and then to maintain affiliation and relationships" (Miller, 1976, p. 83). Females' development is hypothesized to differ from males' self development, which is more likely to be based on strivings for autonomy or separateness rather than connection. Other theories based on the premise that female identity tends to be relationship-focused have also been developed, including a theory of morality organized around the notions of responsibility and care (Brown & Gilligan, 1992; Gilligan, 1982) and a theory of epistemology focused on contextual ways of knowing rather than abstract critical thought (Belenky, Clinchy, Goldberger, & Tarule, 1986). Similarly, researchers interested in theories of self-schema have suggested that women are more likely than men to include valued others as an integral part of their self schemas. This type of self-schema, referred to as a collectivist, ensembled, or connected schema of the self, is in contrast to an individualist, independent, or autonomous schema of the self in which others are represented as distinct from the self rather than as part of the self (Josephs, Markus, Tafarodi, 1992; Markus & Kitayama, 1991; Markus & Oyserman, 1989; Sampson, 1989; Triandis, 1989). These theorists contend that women are relatively more likely to develop connected self-schema, whereas men (especially those raised in Western culture, e.g., North America and Europe) are relatively more likely to develop autonomous self-schema (Josephs, et al., 1992). Thus, according to the theoretical position taken by these social scientists and by feminist scholars, women's perspectives on  67  themselves and their relationships tend to reflect an interpersonal focus that differs from their male counterparts who tend to be more focused on the attainment of independence rather than connection. Although this theoretical position has been criticized for an overemphasis on essential and dichotomous gender differences (Hare-Mustin & Maracek, 1988), empirical findings such as those previously reviewed are consistent with the notion that women and men tend to develop self concepts that differ along the dimensions of connectedness and separateness (see Markus & Oyserman, 1989, for a review of gender differences in some aspects of thinking that are relevant to this issue). In addition, in a series of three studies, Josephs et al. (1992) recently provided research evidence suggesting that men's self esteem derives, at least in part, from separation and independence, whereas women's self esteem is more related to connection and interdependence. Gilligan and her colleagues (Brown & Gilligan, 1992; Gilligan, 1982; Pollack & Gilligan, 1982) and Goldberger, Clinchy, Belenky, and Tarule (1987) have also provided some empirical support, gleaned primarily from extensive structured interviews, for their theories of female development based on connectedness with others. Their research has been criticized, however, for its lack of appropriate comparisons to data gathered from males and for its lack of consideration of other potential factors related to gender (Sher, 1987; Walker, 1984; 1986; Walker, de Vries, & Trevethan, 1987). It was hypothesized in the research being presented here that the gender differences enumerated above relating to the meaning or significance of interpersonal relationships might be reflected in the ways that wives and husbands deal with interpersonal issues within their marriage. In the following section in which the hypotheses that guided my research are presented, the theoretical and empirical relevance of the previously reviewed literature is discussed as it pertained to the development of these hypotheses.  HYPOTHESES On the basis of the preceding reviews of the literature related to stress and coping, gender differences in coping with marital stress, and gender differences in men's and women's investments in their intimate heterosexual relationships (e.g., marriage), five hypotheses relating to how husbands and wives might differ in their ways of dealing with marital conflict or tension were developed and tested. The hypotheses explored in this research postulated that wives and husbands differed in how they initially experienced marital tension, cognitively appraised marital tension, reported coping with marital tension, and responded to marital tension. Hypothesis I Regarding Hypothesis I, the literature review indicated that wives tend to monitor their relationships more closely and therefore are more likely to both notice and call attention to potential conflicts (Gottman, 1979; Holtzworth-Munroe & Jacobson, 1985; Huston & Ashmore, 1986). Furthermore, evidence was presented that wives generally tend to be more dissatisfied with their relationships and more acutely aware of the problems in their relationships than are their husbands (Margolin et al., 1983; Rhyne, 1981). In addition, if wives tend to prefer more sharing of emotions and intimacy in their marriage than do their husbands, then wives may also be more likely to perceive conflicts relating to these issues than are their husbands. For example, some research has suggested that wives are more likely to perceive the absence of intimate, emotionally satisfying behavior than are their husbands, who are less likely to perceive the absence of such behavior either in themselves or in their wives (Hawkins et al., 1980). Thus, wives may be more likely to perceive conflicting interests or needs in their relationship and they may be more likely to experience a conflict than are their husbands. Furthermore, if wives tend to be more "relationship mindful" and place more emphasis and importance on their intimate relationships than do their husbands, it is likely that marital conflict would not  only be perceived more often by wives, but it is likely to also be experienced as more distressing for wives than for their husbands. Although wives tend to be less distressed than are their husbands by the aftermath of divorce or widowhood (Price & McKenry, 1988; Stroebe & Stroebe, 1983), the previous literature review relating to the experience of interpersonal stress was consistent with the notion that within an ongoing relationship, wives are more likely to become distressed by events (e.g., marital arguments) that disrupt or threaten intimacy or connectedness with their marital partner. In addition, evidence from other contexts (i.e., the work environment) has indicated that interpersonal conflicts may be particularly stressful for wives (Long, 1990). Thus, it was expected in this research that wives would experience more distress associated with tension or conflicts with their spouse than would their husbands. Given these expectations the following hypothesis and its specific derivatives were advanced: Hypothesis I: It was expected that a greater association between marital tension and distress would be found for wives than for their husbands: (a) Wives, compared to husbands, were expected to report a greater proportion of diary days in which marital tension occurred. (b) Wives, compared to husbands, were expected to report a greater proportion of diary days in which marital tension was reported to be the most serious tension experienced during a particular day. (c) It was expected that the association between marital tension and distressed mood would be greater for wives than for husbands. Hypothesis II According to the Lazarus et al. model of coping, cognitive appraisal of a potentially stressful event is the single most important determinant of how an individual will cope with that event (Lazarus, 1990; Lazarus & Folkman, 1984). However, as noted previously, very little research has been conducted related specifically to cognitive appraisal (i.e., appraisal of threat) and there appears to be no research focusing on the  70  appraisal of marital conflict. There is a good deal of literature related to attribution processes in close interpersonal relationships, including marriage (see Bradbury & Fincham, 1990, 1992, for a review), but this research pertains to explanations of relationship events that focus on attributions of blame or responsibility along the dimensions of locus, stability, control, and globality (Bradbury & Fincham, 1990, 1992; Fincham & Bradbury, 1987) rather than on appraisal of the significance of a particular event or encounter and whether that event might pose a threat to one's well-being. One study has found that when asked to report concerns, wives more often reported about their loved ones and husbands more often reported about a goal at work (Folkman, Lazarus, Gruen, & DeLongis, 1986). This finding of a gender difference in cognitive appraisal does little, however, to inform us about how spouses are likely to appraise marital encounters that engender tension in their marriage. In order to gain insight into some of the ways wives and husbands might differ in their appraisals of potentially stressful marital events or encounters, the relevant gender differences literature was reviewed. It emerged that relationship issues such as emotional connectedness tend to be of more significance in marital relationships for females than for males (Hawkins et al., 1980; Margolin et al., 1983; Rhyne, 1981). That is, research has suggested that wives, compared to husbands, tend to be more concerned with being understood and emotionally responded to by their spouses (Davis & Oathout; 1987; Huston et al., 1981; Levenson & Gottman, 1985; Wills etal., 1974). It was thus hypothesized that when appraising marital tension, wives, with their greater relationship focus, would appraise marital tension as threatening to a greater extent than would their husbands. The following hypothesis was thus advanced. Hypothesis II. It was expected that when appraising the perceived threat in marital tension, wives would report higher levels of perceived threat than would their husbands.  71 Hypothesis III As noted in the previous review of gender differences in coping, there is considerable confusion in the literature regarding ways in which males and females might differ in their use of particular coping strategies. Within the literature pertaining to how husbands and wives cope with marital conflict, clinicians have anecdotally reported a demand-withdrawal communication pattern among many couples seeking counseling (e.g., Greenberg & Johnson, 1986; Jacobson, 1989; Wile, 1981). Although there has also been some empirical research reporting such a pattern among marital couples dealing with marital conflict, with wives seen as demanding and husbands seen as withdrawing, (Bowman, 1990; Christensen & Shenk, 1991; Gottman & Levenson, 1988), other research has found no gender differences in spouses' use of these tactics to deal with marital tension (Burggraf & Sillars, 1987; Gottman & Levenson, 1992; Roberts & Krokoff, 1990). In addition, as noted in the previous literature review, methodological concerns mitigate against drawing firm conclusions from research to date regarding naturally occurring marital conflict. The current study attempted to address a number of these methodological concerns. Given the equivocal findings in the literature regarding gender differences in spouses' coping with marital conflict, additional research was reviewed that might be applicable to our understanding of these issues. Evidence emerged from this literature indicating that (a) women tend to be more involved in and committed to their intimate heterosexual relationships and more willing to accommodate the development and maintenance of their relationships than are their male partners (Huston et al., 1981; Markman et al., 1982; Rubin et al., 1981; ); (b) being more attentive to their relationships, wives tend to be more aware of potential problems and more invested in confronting and resolving problems as they occur than are husbands (Gottman, 1979; Holtzworth-Munroe & Jacobson, 1985); (c) wives, compared to their husbands, tend to be less satisfied with their marriage and desire more changes in their partners (Margolin et  72  al., 1983). These findings thus suggest that wives may be particularly likely to raise issues of concern with their husbands and then to report confronting their husbands with their requests (or demands) for change or resolution of the problem. 1 These findings are also consistent with the expectation that wives would be more likely than their husbands to report compromise to deal with marital conflict and less likely than their husbands to report withdrawal from marital conflict. These two latter issues are discussed in more detail below. The findings noted above suggesting that wives, compared to husbands, tend to be more attentive to and involved in their relationships and more willing to accommodate to their relationships are consistent with research indicating that wives report greater use of compromise-like strategies to deal with marital conflict than do their husbands (Pearlin & Schooler, 1978; Zietlow & Sillars, 1988). In addition, because compromise involves balancing one's own concerns with another's, reported use of this conflict strategy is consistent with other findings in the literature that indicate that females are more likely than males to take into account the needs of others (Davis & Oathoat, 1987; Kessler & McLeod, 1984; Wethington et al., 1987). Some research has also suggested that females tend to be more cooperative in their interactions with others than are males (see Basow, 1992, or Deaux, 1976, for a review). Given that compromise entails cooperation, this finding would also support the hypothesis that wives might tend to report compromise more than would their husbands to deal with marital conflict. The findings noted above pertaining to wives being more involved or committed to their relationships and therefore more likely to pursue issues of concern within the relationship (Holtzworth-Munroe & Jacobson, 1985; Huston et al., 1981; Margolin et al., 1983; Markman et al., 1982) are also consistent with the notion that wives might report less withdrawal from marital conflict than would their husbands. Given evidence suggesting that males tend to be less reflective about their intimate relationships and more avoidant of relationship issues until things start going badly in the relationship  (Holtzworth-Munroe & Jacobson, 1985), husbands may therefore be more inclined to report withdrawing from potentially stressful marital conflict. Levenson and Gottman (1983, 1985) and Gottman and Levenson (1988) suggested that males will withdraw from conflict because of their greater susceptibility, compared to females, to noxious arousal. A greater tendency on the part of husbands to report withdrawing from marital conflict is also in keeping with the previously reviewed literature that suggested that males, compared to females, prefer to distance themselves from intimate contact (with the exception of sexual contact) and from the concerns of the other (Jacobson, 1989, Wethington et al., 1987). On the basis of the foregoing, the following general hypothesis and its derivatives were developed: Hypothesis III. It was expected that husbands and wives would differ in the types of coping that they reported using when dealing with marital tension or arguments. (a) Wives were expected to report the use of more confrontation than their husbands would report. (b) Wives were expected to report the use of more compromise than their husbands would report. (c) Husbands were expected to report the use of more withdrawal than their wives would report. Hypothesis IV As noted previously, research suggests that coping is a major mediator between stressful encounters (e.g., marital tension) and adaptational outcomes such as psychological distress (e.g., Billings & Moos, 1984; Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986; Folkman, Lazarus, Gruen, & DeLongis, 1986; Pearlin & Schooler, 1978). Consistent with this literature, Hypothesis IV postulated that respondents' reported use of particular coping strategies (confrontation, compromise, and  74  withdrawal) to deal with marital tension would be associated with either increases or decreases in their reported distress. Expectations regarding the relationship between each of these reported types of coping and spouses' distress are presented below. Regarding confrontation, this type of coping has been defined by Folkman, Lazarus, Dunkel-Schetter, DeLongis, and Gruen (1986) as an "aggressive effort to alter the situation." As noted in the Introduction section, this aggressiveness dimension may relate to the adaptational outcomes associated with the use of particular coping strategies. For example, although Roth and Cohen (1986) suggested that confrontation strategies may serve to reduce distress by providing an outlet for the ventilation of emotions, empirical research has consistently indicated negative outcomes associated with the use of aggressive or competitive confrontation strategies, including negative emotional states (Folkman & Lazarus, 1988), psychological distress (Folkman, Lazarus, Gruen, & DeLongis, 1986), low marital satisfaction (Bowman, 1990), and chronic marital strain (Ilfeld, 1980). In the research being reported here it was expected that, regardless of gender, spouses' reports of employing confrontation strategies would be associated with reports of increased psychological distress (i.e., distressed mood) at the time of, or shortly after, the conflict. Regarding compromise, there has been little research directly applicable to an understanding of the immediate impact on mood of compromise strategies used to deal with marital stress. Rands et al. (1981) found that compromise was positively associated with marital satisfaction. Sabourin et al. (1990) also found that use of negotiation strategies was associated with less marital distress (reported here in terms of greater marital satisfaction). In a follow-up of the Pearlin and Schooler (1978) study in which a second interview was conducted with respondents four years after the original interview, Menaghan (1982) found that the original reports of negotiation had no effect on initial reports of marital distress (i.e., how respondents felt when they thought about their relationship), but negotiation was associated with a reduction in marital problems later on (four years later). Unfortunately none of these studies specifically evaluated the  75  immediate impact of negotiation or compromise on respondents' psychological distress or mood. In addition, because the latter two studies obtained global coping ratings related to how respondents usually coped rather than how they coped with a recent, specific marital stressor, the relevance of these findings to our understanding of the impact of reported compromise strategies on mood when spouses have reported specific marital tension within the past 24 hours remains unclear. If compromise is conceptualized as a relatively congenial, reasoned approach strategy, the relationship between reports of compromise and mood might resemble that between reports of planful problem solving, a non-aggressive, reasoned approach strategy that has been studied more extensively, and mood. For example, there has been evidence that planful problem solving is associated with positive emotional states (Folkman & Lazarus, 1988) and satisfactory outcomes (Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986) and is negatively correlated with psychological symptoms (Folkman, Lazarus, Gruen, & DeLongis, 1986). In addition, because compromise strategies focus primarily on congenial attempts to find a solution that is fair to both spouses, it is likely that the reported use of these strategies would be associated with decreased psychological distress for both husbands and wives. Although some researchers in the stress and coping literature have suggested that the use of avoidance or denial-like strategies (e.g., withdrawal) might be associated with short-term beneficial effects, such as reduction of psychological distress (Lazarus, 1985; Mullen & Suls, 1982; Roth & Cohen, 1986), research has also suggested that avoidance strategies (e.g., selective ignoring) might be associated with a negative impact on an individual's mood when these strategies are employed within the context of dealing with marital stress (Pearlin & Schooler, 1978). Applying these seemingly contradictory findings to the issues being raised in this dissertation presented several difficulties. Research focusing on the short-term benefits of avoidance was applicable to interest in the short-term effects of avoidance. Because the benefits of avoidance have largely been  76  found in studies relating to uncontrollable, severe, and sudden crises, or incapacitating illnesses, injuries, or surgical procedures (Lazarus, 1985; Mullen & Suls, 1982), however, these findings might not be relevant to the study of naturally occurring marital tension. Similarly, although Pearlin and Schooler's (1978) data suggesting a relationship between the use of avoidance strategies and psychological distress pertains to the general issue of interest in this dissertation, the mood measurements did not relate to specific incidents of marital conflict and they covered a time-frame of as much as one year. Findings in the coping literature were, thus, equivocal concerning the relationship between the reported use of withdrawal strategies to deal with marital tension and immediate measures of a spouse's mood. Findings from observational research relating to marital conflict were also only marginally relevant to the issue being explored here. For example, Gottman and Krokoff (1989) found that husbands' withdrawal was not significantly associated with their concurrent ratings of marital satisfaction, but it was associated with declines in their marital satisfaction ratings over time. Wives' withdrawal was negatively associated with their concurrent marital satisfaction, but it was not related to their later marital satisfaction. Although these results do not relate specifically to the relationship between withdrawal and mood, they are suggestive in terms of their findings of gender differences. Drawing from the previously reviewed gender differences literature, further indications are found that withdrawal strategies may be associated to mood differently for wives and husbands. The findings of a number of studies (e.g., Holtzworth-Munroe & Jacobson, 1985; Huston et al., 1981; Rhyne, 1981) suggested that men, compared to women, tend to be less reflective about their marital relationships and less emotionally involved in their relationships. Withdrawal from marital conflict may be consistent with this position and may therefore provide emotional relief from potentially stressful emotional involvement. In addition, Gottman and his colleagues' findings (Levenson & Gottman, 1983; 1985; Gottman & Levenson, 1988; see also Stoney et al., 1987)  77  suggested that males are particularly reactive to emotionally laden stimuli. Therefore, husbands' withdrawal from marital conflict might serve the positive emotional function of reducing intolerable levels of distress at the time of the conflict (Gottman & Levenson, 1988; Levenson & Gottman, 1983, 1985). Also, some research has found that men are more likely, under some circumstances, to perceive avoidance strategies as effective in reducing stress than are women (Long & Gessaroli, 1989). On the basis of these arguments, it was expected that, for husbands, the reported use of withdrawal strategies would be associated with decreased psychological distress. Regarding the relationship between the reported use of withdrawal strategies and wives' mood, information gleaned from the previous reviews of the gender differences literature was consistent with the prediction that withdrawal might be more distressing for wives than for their husbands. In particular, it has been proposed in the gender differences literature that women, whose self concept is more likely to be based on connectedness to others, tend to be more relationship-focused than are men (Jordon & Surrey, 1989; Markus & Kitayama, 1991; Markus & Oyserman, 1989; Miller, 1976). In keeping with this theoretical position, research has indicated that emotional connectedness tends to be of more significance in relationships for wives than for husbands, and that wives tend to be more concerned about being understood and emotionally responded to by their husbands than the reverse (Hawkins et al., 1980; Margolin et al., 1983; Rhyne, 1981). This is in keeping with Gottman and Krokoff s (1989) finding that wives' withdrawal, in contrast to that of their husbands, was associated with decreased marital satisfaction at the time of the staged interaction. If a similar dynamic operated in the research presented here, wives might have experienced increased distress in association with their reports of withdrawal from a tense marital interaction. Given these findings, it is likely that wives would tend to become distressed by events, such as withdrawal, that disrupt or threaten intimacy or connectedness with their marital partner. Therefore, it was expected that wives would not only report withdrawal from marital conflict less often than  78  would their husbands (see Hypothesis III), but also that when wives did report employing withdrawal strategies, these reports would be associated with increased distressed mood. To summarize the arguments presented above, the following general hypothesis and its derivatives were put forward to be tested: Hypothesis IV. It was predicted that the type of coping (i.e., confrontation, compromise, or withdrawal) reported by respondents to deal with a specific incidence of marital tension would be differentially related to that respondent's mood. (a) For both husbands and wives, it was expected that reports of their own confrontation would be associated with increased psychological distress. (b) For both husbands and wives, it was expected that reports of their own compromise would be associated with decreased psychological distress. (c) 1. It was expected that husbands' reports of their own withdrawal would be associated with a decrease in their psychological distress. 2. It was expected that wives' reports of their own withdrawal would be associated with an increase in their psychological distress.  Hypothesis V Whereas Hypothesis IV dealt with the association between a respondent's coping and his or her mood, Hypothesis V pertained to the association between a respondent's coping and his or her spouse's mood. In keeping with previously reviewed literature suggesting that wives are more relationship-focused than are their husbands, it was expected that a greater association between respondents' coping and their spouses' mood would be found in regard to wives' mood than husbands' mood. Findings from the empathy literature suggesting that females are more likely than males to experience distress in the presence of others who are experiencing distress (Davis & Oathoat, 1987) are consistent with this expectation. In addition, in keeping with this expectation, Cronkite and Moos (1984) found that wives' depressed mood was associated with their husbands' behavior, whereas husbands' depressed mood related only to their own  79  behavior. Similarly, Whiffen and Gotlib's (1989) findings suggested that wives tend to be more sensitive to their partners' feelings and expressions of dissatisfaction than are their husbands. Based on these findings and on previously reviewed evidence suggesting that wives tend to be more relationship-focused, the following hypothesis was developed: Hypothesis V. It was expected that wives' mood would be more strongly related to husbands' coping than husbands' mood would be related to wives' coping. Exploratory Analyses Exploratory Analyses I In addition to analysis of the above hypotheses pertaining to respondents' reports of their own coping responses (see Hypothesis III), exploratory analyses were conducted regarding spouses' reported perceptions of their partner's coping responses. Because in the stress and coping literature interpersonal factors have generally been subsumed under the construct of social support (DeLongis & O'Brien, 1990), spouses' perceptions of one another's coping have not been assessed in the current coping literature. Within the marital literature, however, interspouse perception has received considerable attention. During the 1970s as research on marital interaction grew, the "insiders' perspective" (Olson, 1978), i.e., spouses' observations of their partners, was utilized as an assessment tool to determine the efficacy of behavioral intervention in marital therapy, presupposing that spouses provided accurate, reliable data about their partners (Jacobson, 1977, 1979; Margolin & Weiss, 1978). Empirical research has indicated, however, discrepancies between spouses' reports of behavior and between spouses' and observers' reports of behavior (Floyd & Markman, 1983; Margolin, 1978; Robinson & Price, 1980). For example, Floyd and Markman (1983) found that spouses perceived their partners' behavior as positive or negative regardless of the "objective" valence of the behavior. Other research that has found higher rates of agreement between members of a couple than between spouses and outside observers has led to the suggestion that couples have a  80  "private message system" by which they interpret one another's messages as they perceive that they are intended. Thus, as "insiders," spouses' interpretations of one another's behavior necessarily differ from those made by "outsiders" who are not privy to the spouses' shared history (Gottman & Porterfield, 1981; Margolin, Hattem, John, & Yost, 1985). Spouses see and interpret their partners' behaviors through perceptual biases and filters colored by their own needs and desires and their shared history (Bradbury & Fincham, 1992; Fincham, Bradbury, & Beach, 1990; Fincham, Bradbury, & Scott, 1990). Although research has consistently found differences between spouses' self reported behavior and their partners' reports of the spouses' behavior (Christensen, 1988; Floyd & Markman, 1983; Guthrie & Noller, 1988; Jacobson & Moore, 1981; Margolin, 1978; Margolin et al. 1985; Robinson & Price, 1980), a wide range of differences have been reported. For example, using the Spouse Observation Checklist of specific behavioral acts (SOC: Patterson, 1976), Jacobson and Moore (1981) found mean correlations between spouses of .38 for distressed couples and .47 for nondistressed couples, with an average agreement across both groups of spouses of 47.8%. Using a revised version of the SOC to create scales relating to expression of anger and social withdrawal, Repetti (1989) found mean correlations between spouses of .47 for anger and .49 for withdrawal. The range of agreement between spouses who were asked to identify the frequency of various interaction patterns and the roles played by the spouse in these interactions was particularly striking in three studies reported by Christensen (1988). Across the three studies, each using different presentations of the interaction patterns, agreement on the frequencies of the patterns varied from . 18 to .55, whereas agreement on the roles played in the interactions ranged from .32 to .73 (Christensen, 1988). Given the wide range of agreement between spouses in these various studies and the lack of previous research comparing spouses' perceptions of one another's coping with naturally occurring marital stressors, no specific hypothesis was proposed regarding the  81  extent of agreement expected between spouses' reports of one another's coping in this research. However, because findings based on the SOC (Jacobson & Moore, 1981; Repetti, 1989) pertained to specific, circumscribed behavioral acts rather than the more interpretive, cognitively-based processes captured by the coping strategies reported in this research, it was expected that lower rates of agreement might be found among spouses in the research being reported here than the rates reported based on the SOC. Exploratory analyses were also conducted to examine not only ways in which respondents' perceptions of their spouses' coping differed from the spouses' reports of their own coping, but also ways in which respondents' own coping differed from their perceptions of their spouses' coping. Although this issue has not been examined within the coping literature, some research has indicated that spouses tend to perceive their own behavior during marital conflict more favorably than that of their partners (Guthrie & Noller, 1988; Sillars & Scott, 1983). In the present study it was expected that respondents might report more congenial coping for themselves (e.g., compromise) and more aggressive coping (e.g., confrontation) for their spouses. Although specific hypotheses were not advanced, exploratory analyses were conducted to examine the relationship between respondents' reports of their own coping and their perceptions of their spouses' coping. Regarding gender differences in husbands' and wives' perceptions of one another's coping or in respondents' comparisons between their own coping and their perceived spouses' coping, there is little evidence to suggest that significant gender differences exist. Females, compared to males, have consistently been found to be more effective senders of interpersonal communication (especially in relation to nonverbal cues) than have been males (Gallagher & Shuntich, 1981; Lochman & Allen, 1981; Margolin et al., 1985; Noller, 1980, 1986; Noller & Venardos, 1986). Research has been less consistent, however, in regard to gender differences in males'and females' reception and decoding of communication. Although some research has indicated that females are more effective  82  receivers and decoders of interpersonal communication (Buck, 1984; Hall, 1978), other research has found no gender differences when looking at decoding of marital communication (Guthrie & Noller, 1988; Margolin et al., 1985; Noller, 1986; Noller & Venardos, 1986). Given the inconsistencies in this research and the lack of research specifically comparing spouses' perceptions of one another's coping, no specific hypotheses were proposed regarding gender differences in spouses' perceptions. Given the interest in gender differences in this research, the possibility of gender differences in spouses' perceptions of one another's coping was explored. Exploratory Analyses II Although no specific hypotheses were proposed regarding gender differences in the relationship between respondents' reported use of specific coping strategies and their spouses' mood (Hypothesis V), exploratory analyses were conducted to determine if gender differences existed in the relationship between particular coping strategies reported by respondents and their spouses' mood. For example, although it was possible (as predicted) that reports of confrontation strategies by either spouse might be associated with increased distress for the other, it was also possible that husbands might perceive the use of these strategies by their wives as particularly threatening. Therefore, increases in husbands' distress might be more closely associated with the reported use of these strategies by their spouses than would be the case for wives' distress. Or, if males are more physiologically reactive to emotion as suggested by Gottman and Levenson (1988) and Stoney et al. (1987), given the emotionally charged nature of confrontation, husbands' increased distress might be more closely associated with their spouses' reports of confrontation than would be the case for wives' distress. Similarly, the reported use of withdrawal strategies by either husbands or wives might be associated with increased distress for their spouses, but wives might find the use of these strategies by their husbands to be particularly threatening given their relationship focus. Therefore, wives'  83  increased distress might be more closely associated with the reports of these strategies by their spouses than would be the case for their husbands. Although these issues were too speculative for the proposal of specific hypotheses, they were approached in an exploratory fashion in this research. In addition, the relationship between respondents' perceptions of their spouses' coping and respondents' distress was also explored to determine if respondents' distress was related to their perceptions of their spouses' coping as opposed to their spouses' reports of their own coping. Lazarus (1990) has suggested that stress and emotion depend more on a person's inferential interpretations of events than on the actual events themselves. Consistent with this cognitive perspective, research has found that subjective appraisals of events are better predictors of emotional reactions than are objective measures (Repetti, 1987; Solomon, Mikulinver, & Hobfoll, 1987). Although these findings suggest that respondents' perceptions of their spouses' coping might be more related to respondents' distress than would be spouses' reports of their own coping, no specific hypotheses were advanced given the paucity of relevant research pertaining to the marital context. However, examination of the similarities or discrepancies in the relationship between respondents' distress and both their spouses' reports of coping and respondents' perceptions of their spouses' coping could clarify our understanding of dyadic coping.  METHODS Data for this study were part of a larger data base collected by Anita DeLongis and Ronald Kessler. 2 These researchers gathered data from couples who completed daily structured diaries for six weeks concerning the types and influences of daily stressors in their lives. The data set included information on each spouse's demands (family demands and other demands) and overloads at home or work, transportation problems, financial problems, and interpersonal conflicts with either one's spouse, child, or others. The data set included information about participants' coping processes when dealing with interpersonal problems and conflicts, but these coping data have not been previously analyzed. Instead, previously reported analyses based on this data set have examined individual data pertaining to the effects of daily stress on mood, the contagion of stress across multiple roles (e.g., work and home) (Bolger, DeLongis, Kessler, & Schilling, 1989; Bolger, DeLongis, Kessler, & Wethington, 1989a; Bolger, DeLongis, Kessler, & Wethington, 1989b), and the relationship between neuroticism and distress (Bolger & Schilling, 1991). In the present study, only coping data relating to marital stressors were analyzed. That is, the analyses in this research utilized only data pertaining to reported interpersonal tension between spouses, i.e., the information recorded by husbands and wives about marital tension (e.g., arguments) as it occurred during the 42 days of the study. Analyses related to Hypothesis I of this study utilized data obtained from 168 couples relating to their reports of interpersonal tension with their spouses. The remaining analyses (pertaining to Hypotheses II-V) were limited to a smaller subset of cases in which both spouses within a couple reported marital tension on the same diary day. Of these spouses, only those couples in which both spouses also reported that the marital tension experienced that day was the most serious interpersonal tension of that particular day were included in  85  these analyses. This subsample of couples (n = 59 couples) and the sample from which these couples were drawn is discussed in more detail below. Sample The couples in the diary study were drawn from a larger sample that participated in a community survey of stress and coping conducted in the spring and summer of 1985 (Mattlin, Wethington, & Kessler, 1990; McGonagle & Kessler, 1990; Schuster, Kessler, & Aseltin, 1990). In the original 1985 survey study, the target population consisted of non-black, married couples living in non-institutionalized housing in the Detroit Metropolitan Area in which at least one of the spouses was between the ages of 18 and 65. On the basis of census data indicating that 90% of non-black married couples in this area lived outside the central city of Detroit, the central city was excluded from the sampling frame. A total of 2,579 housing units were sampled in a multistage stratified cluster design. Selections were made so that each married couple in the population had an equal probability of selection. Of these, 1,332 housing units were found to be eligible for inclusion in the survey. Respondents were eliminated who were not fluent in English, were senile, or had a physical illness that made participation difficult. Interviews were obtained with 977 (73%) of the 1,332 eligible housing units; 311 (23%) refused and 44 (2%) could not be interviewed for other reasons (e.g., failure to make contact or schedule an interview after numerous visits to house). This step in the data collection yielded a screening response rate of 98% and a household interview response rate of 73% (Mattlin et al., 1990). Survey interviews took place in respondents' homes. When an interview was completed successfully with one spouse, an attempt was made to interview the other spouse. Of the 977 eligible couples, both spouses were interviewed in 778 (80%) cases; in 162 (17%) cases the second spouse declined to participate and in an additional 37 (3%) cases, the second spouse could not be interviewed for other reasons (primarily hospitalization). The response rate for the second spouse was  86  80%; the couple response rate was 58%. Interviewer reports indicated that nonresponse by the second spouse may have been systematically related to marital problems (Wethington, 1988). That is, when couples were having serious marital problems, one spouse was often not available for an interview because he/she (usually the husband) had in effect moved out of the house, or was too angry or disengaged to participate. Thus, it is likely that marriages in the process of breaking down were underrepresented in this sample. The couple response rate is comparable to that found in other surveys of married couples (e.g., Hiller & Philliber, 1985). From this sample of 778 couples from the Detroit Metropolitan Area, contact was attempted with 489 (63%) couples one year after their initial interview to solicit their participation in a diary study of stress and coping. The remaining 289 couples were not contacted because they had earlier been approached about participating in another, unrelated study. The researchers were successful in tracing and recruiting both spouses in thirty four percent (i.e., 168 couples) of the 489 couples with whom contact was attempted. The low response rate to the diary study was likely due, in part, to the long period of time that elapsed between the end of the initial interview and contact to solicit participation in the diary study (Bolger et al., 1989). Because of the low response rate, comparison between the participants in the larger sample and those in the diary study was particularly important. When the diary sample was compared to the larger sample no systematic differences in background variables, such as age, education, hours worked, number of children, and family income were found (see Bolger et al., 1989a, b, for more details). Descriptive information received from the participants in the diary study revealed that the average family income (from 1984) was approximately $43,000. The relatively high income is not only due to the profile of the target population (i.e., AfricanAmericans were excluded, as well as the elderly), but also to the relatively high personal incomes found in the Detroit metropolitan area (Mattlin et al., 1990). In the diary  87  sample, the couples had an average of 2.5 children; husbands were significantly older than wives (43.3 vs. 40.5 years) and husbands had insignificantly higher levels of education (13.7 vs. 13.4 years; Bolger et al., 1989a). Approximately 90% of the husbands were employed outside the home compared to 61 % of the wives. Regarding their religious affiliation, approximately half of the sample reported that they were Protestant, 40% reported being Catholic, and 10% reported other affiliations. All respondents in the diary study were asked to complete a short structured diary on a daily basis for 42 consecutive days (six weeks). Diary booklets containing forms for seven days were mailed to participants each week and upon completion were returned by mail to the investigators. To ensure confidentiality, booklets were mailed separately to each spouse and each booklet was equipped with a set of seven adhesive tabs so that each day's entry could be sealed after completion. To monitor when diary entries were actually completed, respondents were asked to indicate the date they completed each entry, if different from the date of the recorded events. Eighty percent of the time, respondents reported filling out the diary entry on the day of the events; 95% of the time they reported having completed the diary entry within one day of the target date. Respondents reported data on a total of 12,348 diary days, with an average of 37 diary days per respondent. Although respondents were not paid for their participation, a $5.00 gift was enclosed with the first diary booklet. In the present research, data were included from spouses who reported marital tension on any of the days during the 42-day diary period. Data analyses pertaining to Hypothesis la and lb were based on responses from all of the 336 spouses (168 couples), focusing on their reports of marital tension during any of the 42 diary days. As described more fully below (see Results section), the analysis pertaining to Hypothesis Ic utilized data from 273 spouses who provided complete diary data on at least 24 of the 42 diary days, yielding a total n of 8,736 diary days. In addition, several analyses (related to  88  Hypotheses II-V) were based on responses from a subset (118 respondents) of the 336 respondents, described below. Data Subset utilized in Hypotheses II-V As noted previously, research has demonstrated that different situations call for different types of coping (Compas et al., 1988; Folkman & Lazarus, 1980). Thus, in order to determine if husbands and wives differed in their coping efforts when dealing with marital tension, it was important that both members of a couple reported about their coping regarding the same marital stressor. In addition, in order to compare spouses' perceptions of one another's coping and the association between one spouse's coping and the other spouse's mood, it was necessary that both members of a couple reported about the same marital stressor. In order to address these issues, data were needed from those diary days in which both members of a couple (i.e., husband and wife) reported the occurrence of marital tension.3 Furthermore, because coping data were collected only for those incidents that were judged by participants to be the most serious interpersonal tension experienced during that particular day (i.e., referred to here as "serious" marital tension), data were needed for those diary days in which both spouses of a couple reported "serious" marital tension on the same day.  A subset of the data was thus used in this  research to explore Hypotheses II-V that contained only those couples who reported "serious" marital tension on the same diary day at least once during the 42-day diary period (n = 59 couples). Because the couples in this data subset differed in the number of diary days (total of 96 days) during which both spouses in a couple reported that marital tension was the most serious interpersonal tension experienced during a given day (Af = 1.6, range = 1-7 diary days), it was inappropriate to combine or retain all of the relevant diary days for each couple. Therefore, one diary day indicating "serious" marital tension reported by both spouses in a couple was chosen for each of these couples for inclusion in the data  89  subset. The first diary day during the 42 days that both spouses agreed that marital tension was the most serious tension of that day was chosen for inclusion in the data subset. This inclusion rule was applied because both spouses in 66% (n = 39) of these couples reported "serious" marital tension on the same day on only one diary day. The only exception to this rule was in the cases in which the first incident of agreed upon "serious" marital tension was the first day of the 42-day study period. Because mood ratings were used in some analyses from the diary day prior to the day on which "serious" marital tension was reported, the first diary day of the study period would not allow for analysis of this variable. Therefore, in cases in which the first incident of "serious" marital tension was the first day of the study, when possible, the second incident of "serious" marital tension reported by both spouses was chosen for inclusion in the data subset. This data subset consisted of 118 diary days representing one diary entry from both spouses of 59 couples (118 respondents). Three of these couples met the criteria for inclusion in this data subset only on day one of the diary study; the first diary day indicating "serious" marital tension for the remaining 56 couples was chosen from diary days between day 2 and day 40 of the 42-day diary period. Because the data subset (n = 59 couples) represented only 35 % of the overall diary sample (n = 168 couples), comparisons were made between the data subset and the other couples in the diary sample (n = 109) on two sets of variables. One group of these variables (Tables 1-2) pertained to the extent to which marital tension was reported and included measures of spouses' responses to interpersonal tension. The second group of variables (Tables 3-6) pertained to spouses' appraisals of threat and measures of coping. Mean scores obtained on these variables by the husbands in the data subset were compared via independent Mests to the mean scores obtained by the husbands in the remaining diary sample. Wives in the two samples were compared in the same manner. For the first set of comparisons (Tables 1 and 2), data from husbands and wives in the data subset were compared via Mests to data from the remaining husbands and wives  90  on the following variables: mean proportion of reports of marital tension (based on the total number of diary days completed by each spouse); mean proportion of reports of "serious" marital tension (based on the total number of diary days completed by each spouse); mean scores on daily mood ratings (referred to as distress); mean stressfulness rating on days in which interpersonal tension was reported; mean scores on daily ratings of physical symptoms. Because two of these measures, a one-item stressfulness rating and a checklist of nine physical symptoms, were not included in other analyses in this study, a brief description of each of these measures is given here. For the stressfulness rating, respondents were asked "How stressful was this situation for you?" and they responded on a five-point scale, ranging from "not at all" to "extremely." Spouses' ratings of the stressfulness of the most serious tension that they experienced on a particular day were summed across the 42 diary days and a mean stressfulness rating was obtained for use in this analysis. The symptom checklist, completed on each diary day, included the following items: hay fever or allergies, cold or flu, headache, stomach pain or cramps, other joint or muscle pain, chest pain, nausea or upset stomach, loss of appetite, diarrhea. Each item was scored as 0 or 1 and summed to generate a daily physical symptoms score. A mean daily physical symptoms score from the 42 diary days was obtained for use in this analysis. Table 1 presents mean scores for husbands and Table 2 presents mean scores for wives on these variables derived from spouses' scores across the 42 days of the diary period. Given the selection criteria (i.e., reporting of marital tension by both spouses in a couple) for the data subset, it was perhaps not surprising that both husbands (M = .09, SD = .08) and wives (M = .11, SD = .10) in the data subset reported that marital tension occurred on a significantly greater proportion of diary days than did the other husbands (M = .03, SD = .04) and wives (M = .04, SD = .06) in the remaining diary sample [husbands / (166) = 6.36, p < .001; wives t (166) = 5.03, p < .001]. Husbands (M = .08, SD = .07) and wives (M = .10, SD = .08) in the data subset also reported a  Table 1 Comparisons between Data Subset and Remaining Diary Sample Study Variables (Set 1)  Subset = 59)  i(n  Others (n = 109)  M  SD  M SD  Proportion of diary days with marital tension  .09  .08  .03  .04  Proportion of diary days with "serious" marital tension  .08  .07  .02  .04  Distress across study  1.32  .30  1.27  .37  Stressfulness rating for serious interpersonal tension  3.12  .65  3.09  .73  Physical symptoms rating  .55  .53  .60  .65  Variables  ***p < .001  Table 2 Comparisons between Data Subset and Remaining Diary Sample Study Variables (Set 1) Wives Subset (n = 59)  i  Others (n = 109)  M  SD  M SD  Proportion of diary days with marital tension  .11  .10  .04  .06  5.03  Proportion of diary days with "serious" marital tension  .10  .08  .03  .05  5.41  Distress across study  1.30  .26  1.29  .38  .31  Stressfulness rating for serious interpersonal tension  3.38  .59  3.38 .7  3  .04  .60  .46  .52  -.68  Variables  Physical symptoms rating ***p < .001  .66  93  significantly greater proportion of diary days on which marital tension was considered to be the most serious interpersonal tension experienced on that day than did the other husbands (M = .02, SD = .04) and wives (M = .03, SD = .05) in the remaining diary sample [husbands t (166) = 6.63; p < .001; wives t (166) = 5.41, p < .001)]. As indicated in Tables 1 and 2, there were no significant differences between husbands and wives in the data subset and husbands and wives in the larger diary sample on mean scores of mood, ratings of stressfulness of interpersonal tension, or reports of physical symptoms. The second set of comparisons between the husbands and wives in the data subset and the husbands and wives in the remaining diary sample included variables specific to respondents' coping with marital tension. The comparison reported above indicated that the data subset reported a significantly higher proportion of "serious" marital tension than did the remaining diary sample. Because respondents' coping responses regarding how they dealt with marital tension were solicited only for those incidents identified as the most serious interpersonal tension experienced during a particular day ("serious" marital tension), marital coping responses were given by only a small proportion of those in the larger diary sample. Thus, the comparisons reported in Tables 3-6 include coping variables for 59 husbands and 59 wives in the data subset and for 38 husbands and 60 wives in the larger diary sample. The variables included in these comparisons are: mean scores on appraisal of threat associated with "serious" marital tension; mean scores on respondents' reports of their own coping (self coping) on the three separate coping scales (confrontation, compromise, and withdrawal); mean scores on respondents' reports of their perceptions of their spouses' coping (perceived spouses' coping) on the three coping scales. Coping measures were scored using both raw scores and relative scores in the research being reported here (see discussion in section below pertaining to coping measures). Therefore, both types of scoring were given in the comparisons of coping  Table 3 Comparisons between Data Subset and Remaining Diary Sample Raw Score Coping Variables Husbands Subset (n=59) Variables M  Others (n = 38)  SD  M  SD  Self confrontation  2.37  .64  2.07  .60  2.30  Self compromise  2.61  .71  2.35  .74  1.76  Self withdrawal  2.34  .62  2.31  .70  .24  Perceived spouses confrontation  2.55  .62  2.43  .77  .86  Perceived spouses' compromise  2.18  .83  1.99  .71  1.17  Perceived spouses' withdrawal  2.03  .63  2.11  .66  -.58  Threat appraisal  2.01  .66  1.93  .54  .62  *p < .05  95 Table 4  Comparisons between Data Subset and Remaining Diary Sample Relative Score Coping Variables  Husbands  Subset (n=59)  Others («=38)  M  SD  M  SD  t  Self confrontation  .32  .08  .32  .10  .30  Self compromise  .35  .07  .35  .08  .47  Self withdrawal  .32  .08  .34  .08  -1.34  Perceived spouses' confrontation  .39  .10  .38  .12  .16  Perceived spouses' compromise  .31  .07  .30  .07  .71  Perceived spouses' withdrawal  .30  .07  .32  .07  -1.59  Variables  Table 5  Comparisons between Data Subset and Remaining Diary Sample Raw Score Coping Variables  Wives  Others («== 60)  Subset («==59) M  SD  M  SD  Self confrontation  2.51  .63  2.46  .76  Self compromise  2.59  .55  2.45  .61  Self withdrawal  2.38  .58  2.29  .60  Perceived spouses' confrontation  2.61  .69  2.56  .74  Perceived spouses' compromise  2.13  .76  1.93  .77  Perceived spouses' withdrawal  2.13  .71  2.03  .68  Threat appraisal  2.14  .64  2.11  .68  Variables  Table 6 Comparisons between Data Subset and Remaining Diary Sample Relative Score Coping Variables Wives Others (n=60)  Subset (n=59) Variables  M  SD  M  SD  t  Self confrontation  .34  .07  .35  .11  -.54  Self compromise  34  .05  .34  .07  .35  Self withdrawal  32  .08  .32  .08  .28  Perceived spouses' confrontation  39  .10  .41  .14  -.91  Perceived spouses' compromise  .31  .08  .30  .08  .71  Perceived spouses' withdrawal  .31  .08  .31  .07  -.10  98  measures presented here. Table 3 presents comparisons of raw score coping for husbands in the two diary samples; Table 4 presents comparisons of relative score coping for husbands in the two diary samples. Table 5 presents comparisons of raw score coping for wives in the two samples and Table 6 presents comparisons of relative score coping for wives in the two samples. The data reported in Table 3 indicated that the husbands in the data subset reported that they used significantly more (raw score) confrontation (M = 2.37, SD — .64) to deal with marital distress than did the husbands in the remaining diary sample [M = 2.07, SD = .60; t (95) = 2.30, p < .05]. No significant differences were found between the two groups' raw scores pertaining to the remaining coping measures: husbands' reports of their own use of compromise, husbands' reports of their own use of withdrawal, their reports of their perceptions of their spouses' coping on the three coping strategies, or their reports of appraisal of threat related to marital tension. In addition, as indicated in Table 4, there were no significant differences in the two groups of husbands on any of the coping variables when relative coping scores were used.5 The data pertaining to comparisons of wives' coping with marital tension in the two groups, using both raw coping scores and relative coping scores (Tables 5 and 6), indicated no significant differences on any coping variables between wives in the data subset and the wives in the remaining diary sample. Measures Overview of diary The diary booklet contained a structured daily questionnaire that inquired about various aspects of spouses' experiences related to interpersonal tension. This research pertained to diary data collected from respondents who indicated that they had experienced tension with their spouses. In addition, respondents were asked to indicate which of their daily experiences of interpersonal tension they considered to be the most serious tension experienced that day. Those respondents who indicated the occurrence of "serious"  99  marital tension on a given day, then replied to a series of forced choice questions pertaining to how they had appraised their "serious" marital argument or tension with their spouse that day and how they had responded to (i.e., coped with) this "serious" tension and how they perceived that their spouses coped with the marital tension. Mood ratings related to psychological distress were obtained at the end of each day. Each of these measures is described in detail below. Assessment of marital tension Each respondent was asked to indicate in a structured questionnaire whether or not any interpersonal tension or arguments had occurred on a given day and with whom the tension or argument had occurred. Specifically, the structured questionnaire introduced the issue to respondents in the following way: "We would like to know about any tension or arguments you had with any of these people during the past 24 hours. Please check each box that applies, and circle the one box that indicates the person you had the most serious problem with today." Respondents then indicated on a checklist the person(s) with whom tension had occurred (the checklist included thirteen items denoting respondents' spouses, their children, other family members, friends, co-workers, and acquaintances). The question about interpersonal tension posed in the diary inquired about the occurrence of tension or arguments rather than about the occurrence of conflict in order to avoid the presupposition that the encounter involved conflict. Use of the more neutral terms, tension and arguments, allowed an assessment of a wider range of events, including those that might have been appraised as being less stressful than those events that connote conflict. Furthermore, use of the term conflict might prejudice respondents' reports of their coping attempts; that is, the notion of conflict may assume engagement or confrontation and may be antithetical to withdrawal or avoidance of conflict. After indicating whether or not interpersonal tension had occurred on a particular day and with whom the tension had occurred (more than one person could be indicated),  100  respondents indicated with whom they had experienced their most serious interpersonal tension that day. In this research the focus was only on respondents' reports of "serious" tension with their spouses; "serious" tension with others (e.g., co-workers, friends, children) was not considered in the research being presented here. Appraisal of Threat The questionnaire included a revised version of a measure of primary appraisal developed by Folkman, Lazarus, Dunkel-Schetter, DeLongis, and Gruen (1986). Two factors, threats to self esteem and threats to a loved one's well-being, along with four additional items (not achieving an important goal at your job or in your work; harm to your own health, safety, or physical well-being; a strain on your financial resources; and losing respect for someone else) emerged from Folkman, Lazarus, Dunkel-Schetter, DeLongis, and Gruen's (1986) scale development of a primary appraisal measure. In the diary employed in this research, a revised version of this inventory was used. Eight threat items were included in the diary: things not running as smoothly as you would like; not accomplishing something you set out to do; not getting the support or understanding you want; losing another's respect or love; losing your self-respect; losing money, possessions or other resources; harm to your physical health or safety; and something bad happening to someone you care about. The eight threat inventory items were introduced in the diary with the following statement: "Here are some difficulties which can be created by a tension or disagreement. How much was each of these an issue for you in your most serious tension or disagreement (during the past 24 hours)?" This statement was followed by a list of eight specific items with responses given on a 4-point scale ranging from "not at all" to "a lot". To determine the internal reliability of the diary's appraisal of threat scale for assessing marital tension, data were analyzed from all respondents who reported "serious" marital tension at some point during the six-week study period. For respondents who  101  reported more than one day on which "serious" marital tension occurred, threat appraisal data from one of the days was randomly chosen (n=214) for inclusion in the analysis.^ As a result of this analysis, three items were dropped from the appraisal of threat scale due to their low item-total correlations and low rates of endorsement. The excluded items had total-item correlations ranging from .27 to .35. The excluded items included the following: losing money, possessions or other resources; harm to your physical health or safety; something bad happening to someone you care about. Analysis of the remaining five-item appraisal of threat scale indicated acceptable internal reliability (Cronbach's alpha =.78). A threat score was derived by summing the spouses' responses (1-4) across the five items and calculating the mean threat score. Coping Inventory The specific coping items used in the diary's 15-item coping inventory were developed to capture three dimensions of interpersonal coping: confrontation, compromise, and withdrawal. Some of the interpersonal conflict items were drawn from that Ways of Coping inventory developed previously by Lazarus and his colleagues (Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986). Confrontation is parallel to the Confrontive coping scale, which includes aggressive efforts to the alter the situation (e.g., "stand your ground and fight for what you want") (Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986). However, because the Ways of Coping inventory was developed without an interpersonal focus, dyadic items were developed specifically for the confrontation scale in this research (e.g., "try to tell the other person why they were wrong"). The following confrontation items were included in the diary's confrontation scale: stand your ground and fight for what you wanted; lose your temper; try to get your own way; try to tell the other person why they were wrong; confront the other person.  102  Compromise, on the other hand, represented a complete departure from the Folkman, Lazarus, Dunkel-Schetter, DeLongis, and Gruen (1986) inventory, which included no strategies that focused on this interpersonal dimension of coping. Drawing from the marital, business management, and human communication literatures noted earlier (Blake & Mouton, 1964; Fitzpatrick, 1988; Putnam & Wilson, 1982; Sillars et al., 1982), the following compromise coping strategies were developed for inclusion in the compromise scale: try to meet the other person half-way; try to find a solution that was fair to both of you; try to appreciate the other person's point of view; try to use reason to settle things; give in and let the other person have their way. Conceptually, withdrawal is similar to the Distancing, Self-controlling, and Escape-Avoidance scales of the Ways of Coping inventory (Folkman, Lazarus, DunkelSchetter, DeLongis, & Gruen, 1986). Three items were thus drawn from these scales. In addition, two dyadic items relevant to withdrawal from interpersonal conflict were also developed specifically for the coping inventory used here (e.g., decide that your differences were not worth arguing about, avoid talking about the problem). The following withdrawal items were thus included in the diary's withdrawal scale: decide that your differences were not worth arguing about; try to do what was necessary to avoid tension; try to keep your feelings from interfering with settling things; try not to take things too seriously; avoid talking about the problem. The diary booklet included a 15-item coping inventory of strategies that described how respondents dealt with the most serious interpersonal conflict that they experienced each day. This inventory was introduced with the statement: "On each line below, circle the number that best describes how you and the other person involved handled this tension or disagreement. How much did you..." This was followed by 15 items, each describing a coping strategy, which was then responded to on a 4-point scale ranging from "not at all" to "a lot." These 15 statements were followed by a second list of 15 similarly worded items and response options that were introduced by the phrase, "How much did the other  103  person..." Thus, each respondent was instructed to indicate both their own use of the 15 coping strategies (self coping) and their perceptions of their spouses' use of the 15 coping strategies (perceived spouses' coping). Internal reliabilities for these three coping scales (i.e., confrontation, compromise, withdrawal) as they specifically pertained to marital arguments were assessed in the same manner as that of the appraisal of threats scale. Coping data were drawn from all respondents who reported marital tension that was considered to be the most serious tension experienced during a particular day at some point during the 6-week study period. For respondents who reported more than one day on which a "serious" marital tension occurred, coping data from one of the days was randomly chosen for use in the analysis (n=2l4). Due to low item-total correlations (.40 and .28, respectively) one item (i.e., give in and let the other person have their way) was dropped from the compromise scale and one item (i.e., avoid talking about the problem) was dropped from the withdrawal scale, for both the self coping scales and the perceived spouses' coping scales. With these two revisions, acceptable internal reliabilities for the three coping scales (both self coping and perceived spouses' coping) were obtained. For the revised scales relating to self coping, the following scale alphas were obtained: for the confrontation scale, Cronbach's alpha = .85; for the compromise scale, Cronbach's alpha = .82; for the withdrawal scale, Cronbach's alpha = .74. For the revised scales relating to perceived spouses' coping, the following scale alphas were obtained: for the confrontation scale, Cronbach's alpha = .84; for the compromise scale, Cronbach's alpha = .89; for the withdrawal scale, Cronbach's alpha = .76.  Scoring Two conceptually and mathematically distinct scoring procedures (raw score method and relative score method) were used to score the three coping scales (compromise, confrontation, withdrawal). The first scoring method followed the raw  104  score procedure most often used in the coping literature (e.g., Billings & Moos, 1984; Folkman & Lazarus, 1980; Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986; Stone & Neale, 1984a). Given the predominance of raw score results in the general coping literature, use of this method allowed comparison of the present results to other studies in the coping literature. In addition, raw score coping is informative in that it reflects the extent to which respondents use a particular coping strategy without reference to their total coping effort. Raw score coping scores were generated for each coping scale by summing the spouses' responses (1-4) across all items in each of the three coping scales and calculating a mean coping score for each of the scales. Coping scale scores derived by this method are referred to in this research as raw scores and reflect the extent to which a respondent reported using, or perceived their partner to use, a given form of coping. In addition to the raw score coping score, all coping scales were scored using the relative scoring procedure advocated by Vitaliano, Maiuro, Russo, and Becker (1987). Whereas the raw score method is based on the frequency of particular types of coping without reference to total coping effort, the relative score method is based on the percentage or proportion of coping effort for a particular type of coping relative to the total coping effort expended across all other types of coping. Vitaliano et al. (1987) point out that raw scores do not take into account respondents' total coping efforts across all scales. The relationships between specific coping scales and overall strategies are thus ignored with raw scores, but these relationships are taken into account with the use of relative scores. Given that marital coping takes place within a dyad where the interplay among various coping strategies may be especially important, relative scores may be particularly relevant to our understanding of spouses' coping with marital conflict. Within the marital context, the relative use of a particular strategy may be more informative than the absolute frequency of that particular strategy. For example, when members of a couple are embroiled in a marital argument, the fact that one spouse uses confrontation strategies an equivalent of four times may be of less consequence than whether the use of  105  confrontation represented 10% or 90% of his or her total coping effort (i.e., the use of confrontation compared to the total use of confrontation, compromise, and withdrawal strategies). Furthermore, as noted by Vitaliano et al. (1987), the relative score method has the advantage of using both idiographic and nomethetic norms. That is, the research participant serves as his or her own standard of reference (one type of his or her coping is compared to all of his or her coping effort) and the participant's relative coping is also compared to other research participants' relative coping (Vitaliano et al., 1987). As noted by Vitaliano et al. (1987), by combining the use of ideographic and nomothetic methods, "relative scores may be better suited for empirically evaluating the dynamic interplay of coping efforts and distress and, as such, they can be especially useful in studying the transactional model of stress" (p. 16). The relative score method was thus employed in this research, along with the more standard raw score method. Relative scores were derived by dividing the mean raw score for each coping scale by the sum of the mean raw scores of the three coping scales. For example, the relative confrontation score was derived by dividing the mean raw confrontation score (obtained as described above) by the sum of the mean raw confrontation score, the mean raw compromise score, and the mean raw withdrawal score. A corresponding procedure was used to calculate relative scores for compromise and withdrawal. Both raw scores and relative scores are reported for all hypothesis testing analyses that involve the coping scales. Distress scale The diary included 15 mood items from the Affects Balance Scale (Derogatis, 1975). These consisted of five items each from three negative mood subscales: anxiety, hostility, and depression. Given the high intercorrelations among the three subscales in the present data (ranging from .81 to .89), the three subscales were combined to produce a single scale, referred to here as the distress scale. This scale included the following mood items: nervous, timid, tense, anxious, afraid, irritable, resentful, angry, enraged, bitter,  106  sad, hopeless, worthless, miserable, unhappy. On the basis of their emotional state over the previous 24 hours, respondents were asked to rate each of the 15 items on a 4-point scale ranging from "not at all" to "a lot." Following the procedure noted above (n = 214), the distress scale was found to have high internal reliability (Cronbach's alpha = .89). A distress score was calculated by summing the respondent's responses (1-4) on the 15 items and obtaining the mean. Missing data on scale items There were very few missing data on the appraisal scale, the three coping scales, or the distress scale for the data subset with which analyses were conducted. Complete data were available from 91% to 96% of respondents, depending on the scale being analyzed. Only 1.5% to 3% of respondents were dropped from the analyses due to missing data, and scale scores were derived from partial data for only 2% to 7% of the respondents, depending on the scale being analyzed.  107 RESULTS The data analyses are organized by hypotheses. Thus, results are considered below as they relate to each of the five sets of hypotheses.7 Analyses related to Hypothesis I Comparisons between husbands' and wives' reports of marital tension (Hypothesis Ial The first set of hypotheses were tested using diary data from all of the 168 couples. When responses from all 336 respondents across the entire diary period (a total of 6,031 diary days for husbands and 6,317 diary days for wives) were analyzed, support was received for the prediction that wives would report proportionately more days on which marital tension occurred than would their husbands. Thirty-six percent (60) of husbands reported no diary days on which marital tension occurred; the remaining 64% (108) of husbands indicated marital tension occurred on a total of 295 diary days, or 4.89% of the 168 husbands' total diary days. Twenty-nine percent (48) of wives reported no diary days in which marital tension occurred; whereas 71% (120) of wives indicated marital tension occurred on a total of 401 diary days, or 6.35% of the 168 wives' total diary days. For husbands, the proportion of diary days in which marital tension was reported (Table 7) ranged from 0% to 31% (Af = .05, SD = .06). For wives, the proportion of diary days in which marital tension was reported (Table 7) ranged from 0% to 54% (M = .07, SD = .08). As predicted (Hypothesis la), paired Mests revealed that wives reported a significantly greater proportion of diary days in which marital tension occurred than did their husbands, t (167) = -2.70, p < .01. This difference in husbands' and wives' reporting of marital tension can also be viewed in terms of their agreement with one another's reports of marital tension. Independently, husbands and wives reported the occurrence of marital tension on a total of 696 diary days. Both spouses agreed that marital tension occurred on the same day on 121 (17%) of their combined reports of marital tension. Examination of their levels of  108  Table 7 Paired r-tests Comparing the Proportion of Diary Days Husbands and Wives Reported Marital Tension Husbands (n=168) M SD  Wives (n = 168) SD  M  Proportion of diary days .05 marital tension reported  .06  .07  .08  -2.70**  Proportion of diary days "serious" marital tension reported  .05  .06  .07  -2.84**  .04  Note. Proportion of diary days for husbands based on total of 6,031 diary days completed during 42-day diary period. Proportion of diary days for wives based on total of 6,317 diary days completed during 42-day diary period.  **p < .01  109  agreement by gender indicates that when husbands reported the occurrence of marital tension (295 diary days), their wives agreed on 41 % of those days that marital tension existed between them. When wives reported the occurrence of marital tension (401 diary days), their husbands agreed on 30% of those days that marital tension existed between them. Although data relevant to respondents' reports of occurrence of other types of arguments or tension during the diary period was not a focus of this research, it is interesting to note that marital tension was reported proportionately more frequently than any other type of interpersonal tension for both husbands and wives in the diary study. The proportion of diary days in which tension with their children was reported was the second most frequently reported interpersonal tension for both husbands (M = .03, SD = .06) and wives (M — .05, SD = .07). Paired Mests revealed that the proportion of diary days in which marital tension was reported was significantly greater than the proportion of diary days in which respondents reported tension with their children, t (167) = 3.77, p < .001, for husbands; t (167) = 2.31, p < .05, for wives. Comparisons between husbands' and wives' reports of "serious" marital tension (Hypothesis lb) The expectation that wives, compared to husbands, would indicate proportionately more diary days in which marital tension was considered to be the most serious interpersonal tension experienced during a particular day was also supported by the data. Considering all respondents (168 couples) across the entire diary period, 60% (101) of the husbands reported, on at least one occasion, that marital tension was the most serious interpersonal tension experienced during the day. For husbands, "serious" marital tension was reported on a total of 261 days, or 4.33% of the 168 husbands' total diary days. Sixty seven percent (113) of the wives reported, on at least one occasion, that marital tension was the most serious interpersonal tension experienced during that day. For  110  wives, "serious" marital tension was reported on a total of 340 days, or 5.38% of the 168 wives' total diary days. The proportion of diary days on which husbands reported marital tension to be the most serious tension occurring on a particular day (Table 7) ranged from 0% to 31 % of the diary days (M = .04, SD = .05). The proportion of diary days on which wives reported marital tension to be the most serious tension occurring on a particular day (Table 7) ranged from 0% to 40% of the diary days (M = .06, SD = .07). As predicted (Hypothesis lb), paired Mests revealed that wives reported a significantly greater proportion of incidents in which marital tension was considered to be the most serious interpersonal tension than was reported by their husbands, t (167) = -2.84, p < .01. Given the finding noted above that wives reported significantly more incidents of marital arguments per se, it is possible that the finding indicating that wives reported more incidents of "serious" marital tension was a function of their having reported more occurrences of marital tension to begin with. In fact, when husbands and wives reports of "serious" marital tension were compared to the overall number of diary days in which marital tension was reported by husbands and wives, the proportions were virtually identical. Of all of their reported incidents of marital tension (295), husbands rated 88% (261) of them as the most serious interpersonal tension experienced that day; similarly, wives rated 85% (340) of their overall reported incidents marital tension (401) as the most serious interpersonal tension experienced that day. The difference found in husbands' and wives' reporting of "serious" marital tension can also be viewed in terms of their degree of agreement with one another's reports of "serious" marital tension. Independently, husbands and wives reported the occurrence of marital tension that they considered to be the most serious interpersonal tension experienced that day on a total of 601 diary days. Both spouses agreed that "serious" marital tension occurred on the same day on 96 (16%) of their combined reports of "serious" marital tension. Examination of their level of agreement by gender indicates  Ill  that when husbands reported "serious" marital tension (261 diary days), their wives agreed on 37% of those days that "serious" marital tension occurred. When wives reported "serious" marital tension (340 diary days), their husbands agreed on 28% of those days that "serious" marital tension occurred. In keeping with the results reported above, the proportion of diary days in which marital tension was judged to the most serious interpersonal tension experienced during a particular day exceeded the proportion of diary days in which other types of interpersonal tension were reported to be the most serious interpersonal tension experienced during a particular day. The proportion of diary days in which tension with respondents' children was considered to be the most serious tension of the day was the second most often reported type of "serious" interpersonal tension experienced by both husbands (M = .02, SD = .04) and wives (M = .04, SD = .05). Paired Mests revealed that respondents reported proportionately more "serious" marital tension than they reported "serious" tension with their children, t (167) = 4.23, p < .001 for husbands; t (167) = 3.43, p = .001 for wives. Associations between marital tension and distress (Hypothesis Ic) The association between marital tension and mood (i.e., current distress) was examined using hierarchical repeated measures multiple regression (see Edwards, 1979). Because this analysis disallowed any missing data on any of the 42 diary days (i.e., none of a respondent's data were retained if any data were missing on one day) and only 31 % of respondents (31 husbands and 72 wives) provided complete data on all 42 diary days, it was necessary to devise a procedure for dealing with missing data. Fifty-six of the 336 respondents did not complete diary days 33-42. Therefore, in order to minimize data loss due to respondents dropping out toward the end of the diary study, only the first 32 diary days of the 42-day diary period were included in the analysis. In addition, only those respondents who provided complete data for 3/4 of the 32 days (i.e., 24 days) were  112  included in the analysis. This produced a subsample of 273 respondents: 126 husbands who responded to a total of 4,032 diary days and 147 wives who responded to a total of 4,704 diary days. Of these 273 respondents, 190 respondents provided complete data on all 32 days and 83 respondents had partial or complete missing data on 1-8 days (the average was 3.2 days of partial or complete missing data). Missing data were filled by replacing missing data with data from a randomly chosen day that had complete data from that particular respondents' diary days. Using this sample of 273 respondents (126 husbands and 147 wives), the following hierarchical repeated measures multiple regression analyses were conducted. The dependent variable in the analyses was respondents' distress score on those days on which marital tension was reported (referred to as current distress). Following Edwards (1979), predictor variables included the following: a set of 31 dummy coded variables for time point, mean distress across the 31 time points (subject variable), the distress score on the previous day (referred to as prior distress), and a dummy-coded variable for the presence of marital tension. Given the nonindependence between husbands and wives, analyses were run separately for husbands and wives, and statistical differences between husbands and wives were examined via comparison of beta weights (Pedhazur, 1982). Table 8 summarizes the results of this hierarchical repeated measures multiple regression analysis (Edwards, 1979). The time points (representing diary days 2-32) included in the analysis were entered in step 1 as a control variable. These 31 variables combined to account for a small, but significant proportion of the variance in the equations for husbands (R^ = .01, p < .001) and for wives (R^ = .004, p < .01). Not surprisingly, this analysis indicated that when entered in the equation as a control variable in step 2, the subject variable, respondents' mean distress across the study, accounted for a large and significant proportion of the variance in both husbands' (R^ change = .46, fi =.68, p < .001) and wives' (R^ change = .45, fi = .67, p < .001) current distress. Respondents' mean distress accounted for 46% of the variance in husbands' current  113  Table 8 Hierarchical Repeated Measures Multiple Regression Analysis Regressing Current Distress on Marital Tension  Husbands (n= 126) Step  Predictors  &  Wives 147)  (71 =  Beta  R2  Beta  z  1  Time points  Ol***  Notea  .004**  Noteb  2  Mean distress over study  4^***  .68***  45***  67***  .16  3  Prior distress  Q^***  .26***  .03***  .22***  1.83  4  Marital tension  .03***  .19***  .06***  .25***  3.03*  Note. For husbands, significance tests were based on 30, 3749 df for step 1; 125, 3749 dfTor step 2; and 1, 3749 dfTor steps 3 and 4. For wives, significance tests were based on 30, 4374 df for step 1; 146, 4374 df for step 2; and 1, 4374 dfTor steps 3 and 4. a  Betas for the thirty-one time points ranged from -.03 to .04. No individual beta was significant.  ^Betas for the thirty-one time points ranged from -.02 to .03. No individual beta was significant. *p < .05 **p < .01 ***p < .001  114 distress (p < .001) in addition to the 1 % of the variance accounted for by the time point variables entered in step 1. Respondents mean distress accounted for 45 % of the variance in wives' current distress (p < .001). Entered as a control variable in step 3, the respondents' prior distress was also significantly associated with current distress (R2 change = .04, fi =.26, p < .001 for husbands; R2 change = .03, fi =.22, p < .001 for wives). Respondents' prior distress accounted for 4% of the variance in husbands' current distress (p < .001), above and beyond the 47% of the variance explained by the variables in steps 1 and 2. Respondents' prior distress accounted for 3% of the variance in wives' current distress (p < .001), above and beyond the 45% of the variance explained by the variables in steps 1 and 2. Entered as step 4, marital tension emerged as a significant variable in accounting for the variance of respondents' current distress (R2 change = .03, fi = .19, p < .001 for husbands; R2 change = .06, fi = .25, p < .001 for wives). Marital tension accounted for an additional 3% of the variance in husbands' current distress (p < .001), above and beyond the 51 % of the variance explained by the variables in steps 1-3. Marital tension accounted for an additional 6% of the variance in wives' current distress (p < .001), above and beyond the 48% of the variance explained by the variables in steps 1-3. To compare gender differences in these effects, betas were converted to Fisher z scores and compared.8 No significant difference was found (z = 1.83, p < .10) when husbands and wives were compared regarding the extent to which prior distress explained the variance in current distress (4% of the variance for husbands, p < .001; 3% of the variance for wives, p < .001). A significant gender difference was found, however, regarding the extent to which marital tension explained the variance in respondents' current distress. In support of Hypothesis Ic, marital tension explained significantly more of the variance in current distress for wives (6% of the variance, p < .001) than was the case for husbands (3% of the variance, p < .001; z = 3.03, p < .01).  115 Summary The above analyses provided support for the prediction that wives would report a significantly greater proportion of diary days in which marital tension or arguments occurred than would their husbands. Although perhaps related to the higher proportion of days in which marital tension was reported, wives, compared to husbands, also reported a significantly greater proportion of diary days in which marital tension was considered to be the most serious interpersonal tension experienced during a particular day. It was also found that increases in marital tension were significantly related to increases in daily distress for both husbands and wives. However, in keeping with the research hypothesis, the relationship between marital tension and current distress was significantly greater for wives than for their husbands. Analyses related to Hypothesis II - V In contrast to the analyses reported above which were based on data received from the 336 respondents (or a subsample of 273 respondents used in the hierarchical repeated measures multiple regression analysis) across 42 diary days (or a subsample of 32 diary days used in the repeated measures multiple regression analysis), the following analyses (hypotheses II-V) were based on one diary day from each of 118 respondents (59 couples). This data subset includes data from the first diary day (unless this was day 1 of the diary period) in which both spouses reported on the same day that marital tension was the most serious interpersonal tension experienced that particular day (this sample is described in detail in a previous section). Although the analyses of each of the remaining four hypotheses are discussed in separate sections below, the reader is directed to Tables 9, 10, 11 and 12 for a summary of the intercorrelations of the variables included in these analyses. Tables 9 and 10 present correlation matrices with all coping measures scored by the raw score method; Table 9 presents a correlation matrix for husbands and Table 10 presents a correlation matrix for wives. Tables 11 and 12 present correlation matrices  116  Table 9 Correlation Matrix of Study Variables for Husbands Raw Score Method (n=59) a  ••p < .01  ***p < .001  a Note: Due to missing data n varied for various comparisons; minimum n = 51.  117 Table 10 Correlation Matrix of Study Variables for Wives Raw Score Method (n=59) a  r  <*»  £  OC O  \t * - s a8  k  s  1 is  2  ©  ©  k  is  i  bi  £j is  s  \s  "s s  2  s  8  3  a  5  B  53  3  k k  s  fe s  LA  s  2  5  2  S3  fe k  i  fo  k  ©  8  a  3  s  t*  8  k  b  k  ©  *  ' ."  *.  -  3  -  O  s  k  SO  1-  o  3  5  -  &  H  5  rif Ilf PIPi|H1'r'1II m "  |  g  ©  'K  If  5  a,  B.  k  w  U\  '8  is  ON  k  2  1  1 i  2,  k>  fe  u>  8  8.  1  to  8  s b -  1  <_A  ON  -J  s OC  S  & -J# • CO  So  W  VO  **  *  k  1  s  »•  %o  k h  **p < .01 ***p < .001 a Note: Due to missing data n varied for various comparisons; minimum n = 51.  «©  3  118 Table 11 Correlation Matrix of Study Variables for Husbands Relative Score Method (n=59) a  **p < .01 ***p < .001 a Note: Due to missing data n varied for various comparisons; minimum n = 51.  119 Table 12 Correlation Matrix of Study Variables for Wives Relative Score Method (n=59) a  **p < .01 ***p < .001 a Note: Due to missing data n varied for various comparisons; minimum n — 51.  120  with all coping measures scored by the relative score method; Table 11 presents a correlation matrix for husbands and Table 12 presents a correlation matrix for wives. Analyses related to Hypothesis II A paired Mest was conducted to examine gender differences in husbands' and wives' reports of perceived threat when dealing with marital tension. The prediction that wives, compared to husbands, would report a higher level of perceived threat when appraising the extent to which they felt threat by the existence of marital tension that day was not supported (Table 13). A paired Mest comparing husbands' (M = 2.04, SD = .72) and wives' (M = 2.06, SD = .72) threat scores on a day in which both spouses reported marital tension to be the most serious interpersonal tension experienced during that day revealed no significant difference between husbands' and wives' appraisals of threat, t (52) = -.09, p > .10. For both husbands and wives, higher threat scores were associated with higher distress scores on the same day. As indicated in the correlation matrices (Tables 9-12), higher levels of threat were associated with increased current distress for both husbands (r = .62, p < .001) and wives (r — .41, p < .01). When these correlation coefficients were converted to Fisher z scores and compared, no significant gender differences were found in the relationship between threat and current distress for husbands and wives (z = 1.47, p > .10). In summary, in contrast to prediction, no differences were found between husbands' and wives' reported appraisals of threat related to marital tension. Thus, husbands and wives did not differ in the extent to which they perceived marital tension as threatening. Although there was a significant correlation between spouses' appraisals of threat and their reported distress, no gender differences were found in these associations. For both husbands and wives, higher appraisals of threat were associated with greater distress.  121 Table 13 Paired /-test Comparing Husbands' and Wives' Appraisal of Threat Threat Scores M SD Husbands  2.04  t .72  vs Wives  -.09 2.06  .72  122  Analyses related to Hypothesis III and Exploratory Analysis I Next, husbands' and wives' reported use of coping strategies when dealing with marital tension was examined. As noted in the Methods section, the coping scales were scored in two ways, using the raw score method and the relative score method. The raw score method scores each of the three coping scales individually, without reference to the total amount of coping effort expended across the three types of coping. The relative score method scores each of the three coping scales relative to the total amount of coping effort expended across the three types of coping (Vitaliano et al., 1987). Results obtained using both types of scoring are reported here. In addition, the following results pertain to the three coping strategies (confrontation, compromise, withdrawal) as reported by the respondents in terms of their own coping (self coping), their perceptions of their spouses' coping (perceived spouses' coping), and their spouses' reports of their own coping (spouses' coping). Results pertaining to the relationships between these differently reported coping strategies (i.e., self, perceived spouses', and spouses' coping) are examined separately below. However, before presenting these results, intercorrelations among the coping scales, confrontation, compromise, and withdrawal (both self coping and perceived spouses' coping) are presented. As noted previously (see footnote 7), a .01 alpha level was set for determining the statistical significance of the correlations in order to reduce the likelihood of a Type I error. The power of these analyses was estimated. Using a population correlation of .30 (a medum effect size per Cohen, 1992) and a sample size of 59, calculation of power estimates indicated that setting an alpha level of .01 resulted in .41 power. Using these same research parameters, setting the alpha level at .05 would result in .65 power (Cohen, 1969; 1992). Given the relatively low power at the .01 alpha level and the probability of committing a Type II error, trends toward significance (p < .05) in the bivariate analyses  123 are reported below. However, given that these trends do not represent statistically significant results as defined in this research, they should be interpreted cautiously. Intercorrelations among coping scales Intercorrelations among self coping scales (raw score method) Intercorrelations among the three raw score self coping scales were examined separately for husbands and wives in the data subset. Similar patterns emerged for husbands and wives in regard to the associations found among the three self coping scales (summarized in Table 14). Using raw score coping values, no significant relationship was found between confrontation and compromise for either husbands (r = .19, p > .10) or wives (r = -.05, p > .10). Similarly, no significant relationship was found between confrontation and withdrawal for husbands (r = -.21, p > .10) or for wives (r = -.26, p > .01). However, raw score compromise was positively and significantly related to raw score withdrawal for wives (r = .44, p < .001). Although raw score compromise was not associated with raw score withdrawal for husbands at the .01 level of significance, there was a trend toward significance in the association between these two variables (r = .33, p < .05). Thus, similar patterns emerged for husbands and wives in the intercorrelations between the three raw score measures of self coping. For both husbands and wives, no significant relationship was found between confrontation and compromise or between confrontation and withdrawal. A significant positive relationship was found between compromise and withdrawal for wives; a trend toward a significant positive relationship was found between compromise and withdrawal for husbands.  Table 14 Intercorrelations Among Self Coping Strategies Raw Score Values Husbands Confrontation Confrontation  Compromise . 19  Compromise  Withdrawal -.21 .33  Wives Confrontation Confrontation Compromise  Compromise -.05  Withdrawal -.26 A A $£$&%•  Note. Missing data resulted in the n ranging from 55 to 57. ***p < .001  125  Intercorrelations among perceived spouses' coping scales (raw score method) Intercorrelations among perceived spouses' raw score coping scales were examined in order to determine if these relationships differed from those obtained with the self coping scales. A somewhat different pattern of correlations was found when raw score perceptions of spouses' coping were examined (Table 15). Using raw score values, no significant relationship was found between perceived spouses' confrontation and perceived spouses' compromise for either husbands (r = -.05, p > .10) or wives (r = -.18, p > .10). For wives, a significant negative relationship was found between perceived spouses' confrontation and perceived spouses' withdrawal (r = -.36, p < .01). Although not significantly different from their wives, no significant relationship was found between these variables for husbands (r = -.23, p > .01). Perceived spouses' compromise was positively and significantly related to perceived spouses' withdrawal for both husbands (r = .62, p < .001) and wives (r = .48, p < .001). For husbands, the correlation obtained between perceived spouses' compromise and perceived spouses' withdrawal (r = .62, p < .001) was significantly different from the correlation found between self compromise and self withdrawal (r = .33, p < .01; z = 1.98, p < .05); there was no significant difference between the corresponding correlations for wives. Thus, regarding correlations among husbands' perceived spouses' coping scales scored with the raw score method, two of the correlations were not significant, but the third intercorrelation, between compromise and withdrawal, was significantly positive. For wives, when raw score values were used, there was no significant relationship between perceived spouses' confrontation and compromise, but there was a significant negative relationship between perceived spouses' confrontation and withdrawal and a significant positive relationship between perceived spouses' compromise and withdrawal.  Table 15 Intercorrelations Among Perceived Spouse's Coping Strategies Raw Score Values Husbands Confrontation Confrontation  Compromise -.05  Withdrawal -.23 .62***  Compromise  Wives Confrontation Confrontation Compromise  Note, n = 56. **-p  < .01 ***p < .001  Compromise -.18  Withdrawal -.36** .48***  127 Intercorrelations among self coping scales (relative score method) Intercorrelations among the three relative score self coping scales were also examined separately for husbands and wives in the data subset (Table 16)." Using relative score values, a significant, negative relationship was found between self confrontation and self compromise for both husbands (r = -.40, p < .01) and wives (r = -.61, p < .001). Similarly, a significant, negative relationship was found between self confrontation and self withdrawal for husbands (r = -.62, p < .001) and wives (r = -.58, p < .001). Relative score self compromise was also negatively and significantly related to relative score self withdrawal for husbands (r = -.35, p < .01), but not for wives {r = -.03, p > .10). Thus, for husbands, significant, negative relationships were found among the three relative score coping strategies pertaining to husbands' own coping attempts. Similar relationships were found for the wives, with the exception of no significant relationship being found between self compromise and self withdrawal. For both husbands and wives, the intercorrelations among the three relative score self coping scales differed significantly from the intercorrelations obtained via the raw score method. The z score values comparing the relative and raw score correlations varied from 2.00 (p < .05) to 3.30 (p < .001). Intercorrelations among perceived spouses' coping scales (relative score method) As noted above, intercorrelations among perceived spouses' raw score coping scales were examined in order to determine if these relationships differed from those obtained with the self coping scales. A pattern of results somewhat similar to those obtained for relative score self coping strategies was found when correlations were examined among relative score coping scales pertaining to husbands' and wives' perceptions of their spouses' coping (Table 17). Using relative score coping values, a significant, negative relationship was found between perceived spouses' confrontation and  Table 16 Intercorrelations Among Self Coping Strategies Relative Score Values Husbands Confrontation Confrontation  Compromise -.40**  Withdrawal -.62*** -.35**  Compromise  Wives Confrontation Confrontation Compromise  Compromise -.61***  Withdrawal -.58*** -.03  Note. Missing data resulted in the n ranging from 55 to 57. **p  < .01 ***p < .001  129 perceived spouses' compromise for both husbands (r = -.76, p < .001) and wives (r = .70, p < .001). A significant, negative relationship was also found between perceived spouses' confrontation and perceived spouses' withdrawal for both husbands (r = -.71, p < .001) and wives (r = -.70, p < .001). In contrast to the results found using relative scores for the self coping scales, no significant relationship was found between perceived spouses' relative score compromise and perceived spouses' relative score withdrawal for either husbands (r = .07, p > .10) or wives (r = -.02, p > .10). As was the case with raw score values, there was a significant difference for husbands between the relative score correlation obtained between perceived spouses' compromise and perceived spouses' withdrawal (r = .07, p > .10) and the relative score correlation found between self compromise and self withdrawal (r = -.35, p < .01; z = -2.24, p < .05), but no significant difference was found for the corresponding correlations for wives. In summary, the pattern of intercorrelations among perceived spouses' relative score coping was similar for husbands and wives. Using relative score coping values, for both husbands and wives, significant, negative relationships were found between perceived spouses' confrontation and perceived spouses' compromise and between perceived spouses' confrontation and perceived spouses' withdrawal, whereas no significant relationship was found between perceived spouses' compromise and withdrawal. The intercorrelations found between the perceived coping scales based upon raw scores differed significantly from those obtained using relative scores. The z scores varied from 2.45 (p < .05) to 4.80 (p < .001). Summary The raw score and relative score analyses produced different patterns of significance in the intercorrelations between coping strategies for both the self coping measures and the perceived spouses' coping measures. Using relative scores, a negative relationship was generally found among the coping measures. Using raw scores, one  Table 17  Intercorrelations Among Perceived Spouse's Coping Strategies Relative Score Values  Husbands Confrontation Confrontation  Compromise  Withdrawal  -.76***  Compromise  .07  Wives Confrontation Confrontation Compromise  Note, n = 56. ***p < .001  Compromise -.70***  Withdrawal _7Q#** -.02  131  significant negative relationship was found (between perceived spouses' confrontation and withdrawal for wives) and significant positive relationships were found between compromise and withdrawal scales. In addition, for husbands (using both raw and relative scores) the relationship between withdrawal and compromise was significantly different depending on whether the strategies pertained to husbands' own coping or to their perceptions of their wives' coping. The relationship between husbands' and wives' coping (Hypothesis III) In the first set of the following comparisons, mean differences in husbands' and wives' reported use of the same type of coping (confrontation, compromise, and withdrawal) were compared via paired Mests. These comparisons indicated the presence or absence of gender differences in spouses' use of the same type of coping strategy. This set of comparisons was then complemented by a second set of comparisons consisting of correlations between husbands' and wives' use of the same type of coping strategy. These comparisons indicated whether or not there was a relationship between husbands' and wives' use of the same type of coping strategy. Finally, correlations were analyzed pertaining to possible relationships between husbands' and wives' reported use of different types of coping (e.g., confrontation compared to withdrawal). Comparisons between self coping and spouses' coping (raw score method) Analyses were conducted to examine the relationships between respondents' own coping (self coping) and their spouses' coping. Contrary to prediction (Hypothesis III), paired r-tests conducted to analyze gender differences in husbands' and wives' reported use of specific raw score coping strategies revealed no significant gender differences (Table 18). Specifically, these analyses of the data subset, using raw score coping values, revealed no significant difference between husbands' (M = 2.44, SD = .73) and wives' (M = 2.62, SD = .71) reported use of confrontation strategies, t (53) = -1.42, p > .10.  132 There was also no difference between husbands' (M = 2.54, SD = .79) and wives' (M = 2.61, SD = .66) reported use of compromise strategies, t (52) = -.47,/? > .10. And, there was no difference between husbands' (M = 2.29, SD = .74) and wives' (M = 2.40, SD = .79) reported use of withdrawal strategies, t (54) =-.99, p > .10. Thus, the extent to which husbands reported using the three different types of coping strategies, scored without reference to overall coping efforts, did not differ significantly from the extent to which their spouses' (wives) reported using these three types of coping. The correlation coefficients presented in Tables 9 and 10, comparing raw score self coping and raw score spouses' coping, also indicated that no significant relationship was found between husbands' and wives' reported use of confrontation strategies (r = .15, p > .10). Similarly, no significant relationship was found between husbands' and wives' reported use of compromise strategies (r = .07, p > .10). A moderate positive correlation was found, however, between husbands' and wives' reported use of withdrawal strategies (r = .39, p < .01). Thus, respondents' reports of their own withdrawal when dealing with marital tension were positively associated with their spouses' reports of withdrawal to deal with marital tension. In addition, Tables 9 and 10 indicated that no significant relationships were found between husbands' and wives' reported use of different types of raw score coping. Of particular interest in this research, it was found that wives' reported use of confrontation was not significantly associated with their husbands' reported use of withdrawal (r = -.02, p > .10). Similarly, husbands' reported use of confrontation was not significantly associated with their wives' reported use of withdrawal (r = -.16, p > .10). Comparisons between self coping and spouses' coping (relative score method) Data analyses were also conducted to examine the relationships between respondents' own coping (self coping) and their spouses' coping when coping was scored using relative coping scores. Consistent with the results obtained with the raw score  Table 18  Paired f-tests Comparing Respondents' Self Coping and Their Spouses' Coping Raw Score Values  Husbands vs Wives  Coping Husbands' confrontation vs. Wives' confrontation  Husbands' compromise vs. Wives' compromise  Husbands' withdrawal vs. Wives' withdrawal  M  SD  2.44  .73  2.62  .71  2.54  .79  2.61  .66  2.29  .74  2.40  .79  -1.42  -.47  -.99  134  method, but contrary to prediction (Hypothesis III), paired Mests revealed no significant gender differences (Table 19). Similar to the raw score results, analyses of the relative score coping variables revealed no significant difference between husbands' (M = .34, SD = .10) and wives' (M = .35, SD = .10) reported use of confrontation, t (53) = -.50, p > .10. There was also no difference between husbands' (M = .34, SD = .07) and wives' (M = .34, SD = .06) reported use of compromise, t (52) = .39, p > .10. And, there was no difference between husbands' (M = .32, SD = .09) and wives' (M = .33, SD = .12) reported use of withdrawal, t (54) =-.48, p > .10. Thus, husbands' reported relative use of the three types of coping strategies did not significantly differ from their spouses' (wives') reported relative use of these coping strategies. The correlation coefficients presented in Tables 11 and 12, comparing relative score self coping and spouses' relative score coping, also indicated that no significant relationship was found between husbands' and wives' reported use of coping strategies. Consistent with the raw score correlations, when relative score coping values were used, no significant relationship was found between husbands' and wives' reported use of compromise (r = -.04, p > .10). Although no significant relationship was found between husbands' and wives' reported use of relative score confrontation (r = .28, p < .05) or between their reported use of withdrawal (r = .31, p < .05), a trend towards significance was found for both of these relationships. This latter relationship is consistent with the positive relationship found between husbands' and wives' raw score withdrawal. In addition, Tables 11 and 12 also indicated that no significant relationships were found between husbands' and wives' reported use of different types of relative score coping. Of particular interest in this research is the relationship between confrontation and withdrawal. The results indicated that with relative score coping values, wives' reported use of confrontation was not significantly associated with their husbands' reported use of withdrawal (r = -.21, p > .10). Similarly, husbands' reported use of  Table 19 Paired t-tests Comparing Respondents' Self Coping and Their Spouses' Coping Relative Score Values  Husbands vs Wives M  SD  Husbands' confrontation vs. Wives' confrontation  .34  .10  .35  .10  Husbands' compromise vs. Wives' compromise  .34  .07  .34  .06  .32  .09  .33  .12  Coping  Husbands' withdrawal vs. Wives' withdrawal  .50  .39  .48  136  confrontation was not significantly associated with their wives' reported use of withdrawal (r =-.18, p > .10). Summary In summary, neither the raw score nor the relative score analyses provided support for Hypothesis III. There were no gender differences in spouses' reported use of the same type of coping strategies. Also, using both raw scores and relative scores, no significant correlations were found between husbands' and wives' reports of confrontation or between their reports of compromise. However, using raw scores, respondents' reported use of withdrawal was positively correlated with their spouses' reported use of withdrawal. Although not significant, a similar trend was found for reports of withdrawal in the relative score analysis. In addition, no significant correlations were found between respondents' and their spouses' reported use of different coping strategies, with either raw score or relative score coping. The relationship between respondents' own coping and their perception of their spouses' coping (Exploratory Analysis I) In keeping with the preceding organization of results, two sets of comparisons between respondents' reported self coping and their perceptions of their spouses' coping are presented below. The first set of comparisons relates to mean differences between respondents' own coping reports and their perceptions of their spouses' coping. The second set of comparisons pertains to correlations between respondents' own coping reports and their perceptions of their spouses' coping. Comparisons between self coping and perceived spouses' coping (raw score method') To explore the relationship between respondents' own reported coping (self coping) and their perceptions of their spouses' coping (perceived spouses' coping), a series of paired Mests were conducted separately for husbands and wives. The results of these  137  exploratory analyses, using raw score coping values, are displayed in Table 20. These results revealed that husbands tended to report that they used less confrontation (M = 2.40, SD = .75) than they ascribed to their wives [M = 2.65, SD = .74, t (55) = -2.26, p < .05]. There was no significant difference, however, in wives' reports of their own use of confrontation (M = 2.59, SD = .73) compared to their perceptions of their spouses' confrontation [M = 2.72, SD = .72, t (55) = -1.01,/? > .10]. Regarding compromise, husbands tended to report that they used significantly more compromise themselves (M = 2.54, SD = .79) than they perceived that their wives used [M = 2.23, SD = .91, t (54) = 2.63, p < .01]. Similarly, wives tended to report that they used significantly more compromise themselves (M = 2.57, SD = .68) than they perceived that their husbands used [M = 2.16, SD = .84, t (54) = 3.35, p < .001]. In addition, husbands tended to report that they used significantly more withdrawal themselves (M = 2.31, SD = .75) than they perceived that their wives used [M = 1.96, SD = .70, t (55) = 3.60, p < .001]. Wives also tended to report that they used significantly more withdrawal themselves (M = 2.43, SD = .76) than they perceived that their husbands used [M = 2.13, SD = .72, t (55) = 2.85,p <.01]. The raw score correlation coefficients presented in Table 9 revealed a significant positive relationship for husbands between their own confrontation and their perceived spouses' confrontation (r = .38, p < .01). There was no significant relationship, however, for wives (Table 10) between self confrontation and perceived spouses' confrontation (r = .17, p > .10). Using raw score coping values, there was a significant positive correlation between self compromise and perceived spouses' compromise for husbands (r = .50, p < .001), but there was only a trend toward significance in this relationship for wives (r = .30, p < .05). Self withdrawal and perceived spouses' withdrawal were significantly, positively correlated for both husbands (r = .50, p < .001) and wives (r = .41, p < .01). In addition, no significant relationships were found between self withdrawal and perceived spouses' confrontation (r = -.11, p > .10, for  Table 20  Paired Mests Comparing Self Coping and Perceived Spouses' Coping Raw Score Values  Wives  Husbands Coping Self confrontation vs. Perceived spouses' confrontation  Self compromise vs. Perceived spouses' compromise  Self withdrawal vs. Perceived spouses' withdrawal  M  SD  2.40  .75  t  M  SD  2.59  .73  2.72  .72  2.57  .68  -2.26* 2.65  .74  2.54  .79 2.63**  2.23  .91  2.16  .84  2.31  .75  2.43  .76  1.96  .70  2.13  .72  3.60***  *p < .05 **p < .01 ***p < .001  139 husbands; r = .05, p > .10, for wives) or between self confrontation and perceived spouses' withdrawal (r = .05, p > .10, for husbands; r = -.12, p > .10, for wives) for either husbands or wives. When each of the correlations obtained for husbands and wives were compared, no significant gender differences emerged. In summary, examination of the raw score coping correlations revealed that, for husbands, there was a significant, positive association between their own reported use of the three coping scales and their perceptions of their spouses' use of the same three coping scales. Although there were no significant gender differences in these relationships, for wives there was no significant relationship between their own reported use of confrontation and their perception of their spouses' confrontation. There was also no significant relationship between wives' own reported use of compromise and their perception of their spouses' compromise, although there was a trend toward significance in this relationship. There was a significant, positive association between wives' reported use of withdrawal and their perceptions of their spouses' use of withdrawal. Comparisons between self coping and perceived spouses' coping (relative score method) The preceding exploratory analyses comparing respondents' own coping reports with their perceptions of their spouses' coping were also conducted using relative score coping values (Table 21). Examination of the relative score results from the series of paired Mests shown in Table 21 revealed that, similar to the results found with the raw score method, husbands tended to report that they themselves used less confrontation (M = .34, SD = .10) than they ascribed to their wives [M = .40, SD = .11, t (55) = -3.91, p < .001]. The relative score coping results also indicated that wives reported significantly less use of confrontation themselves (M = .35, SD = .10) compared to their perceptions of their spouses' confrontation [M = .40, SD = .11, t (55) = -2.62, p < .05]. As was found with the raw score method, comparison of relative coping scores indicated that both husbands and wives perceived that their spouses used relatively less  Table 21 Paired /-tests Comparing Self Coping and Perceived Spouses' Coping Relative Score Values  Husbands M  SD  Self confrontation vs. Perceived spouses' confrontation  .34  .10  .40  Self compromise vs. Perceived spouses' compromise  Coping  Self withdrawal vs. Perceived spouses' withdrawal  Wives M  SD  .35  .10  .11  .40  .11  .35  .07  .34  .06  .32  .08  .30  .08  .32  .09  .32  .08  .30  .08  t  -3 91***  2.44*  2.85** .29  *p < .05 **p < .01. ***p < .001  .07  141  compromise than the respondents reported regarding themselves. Specifically, husbands reported that they used significantly more compromise themselves (M = .35, SD = .07) than they perceived that their wives used [M = .32, SD = .08, t (54) = 2.44, p < .05]. Similarly, wives reported that they used significantly more compromise themselves (M = .34, SD = .06) than they perceived that their husbands used [M = .30, SD = .08, t (54) = 2.63, p < .01]. Husbands also tended to report that they used significantly more relative score withdrawal themselves (M = .32, SD = .09) than they perceived that their wives used [M = .29, SD = .07, t (55) = 2.85, p < .01]. On the other hand, no significant relationship was found between wives' reports of their own relative score withdrawal (M = .32, SD = .08) and their perceptions of their husbands' relative score withdrawal [M = .30, SD = .08, t (55) = 1.40, p > .10]. Examination of the relative score correlation coefficients presented in Tables 11 and 12 revealed a similar relationship regarding confrontation to that found in the raw score correlation matrix. There was a significant positive relationship for husbands (Table 11) between their own reported confrontation and their perceived spouses' confrontation (r = .38, p < .01), but no significant relationship was found (Table 12) between these variables for wives (r = . 16, p > . 10). Contrary to the results found using the raw score method, with relative scores only a trend toward significance was found in the relationship between self compromise and perceived spouses' compromise for husbands (r = .33, p < .05) and no significant relationship was found between these variables for wives (r = .07, p > .10). However, relative score self withdrawal and perceived spouses' relative score withdrawal were significantly positively correlated for husbands (r = .36, p < .01). Only a trend toward significance was found in the relationship between wives' reported self withdrawal and their perceived spouses' withdrawal (r = .29, p < .05). In addition, no significant relationships were found between self withdrawal and perceived spouses' confrontation (r = -.24, p > .01, for husbands; r = -.09, p > .10, for wives) for either husbands or wives. Although no significant relationships were found  142  between self confrontation and perceived spouses' withdrawal (r = .24,/? > .01, for husbands; r = -.28, p < .05, for wives) for either husbands or wives, there was a trend toward a significant negative relationship between wives' confrontation and their perception of their husbands' withdrawal. Comparisons between each of the relative score correlations found for husbands and for wives revealed one significant gender difference. There was a significant gender difference between husbands' and wives' own reported confrontation and their perceptions of their spouses' withdrawal (z = 2.74, p < .01). Although neither of the individual correlations is significantly different (using .01 alpha level) from zero (r = .24, p > .01, for husbands; r = -.28, p < .05, for wives), the two correlations are significantly different from one another (p < .01). The difference between these correlations suggests that increases in wives' relative reports of confrontation are more likely to be associated with decreases in their perceptions of their spouses' relative use of withdrawal than would be the parallel case for husbands. The reverse interpretation is also possible. That is, increases in husbands' relative reports of confrontation are more likely to be associated with increases in their perceptions of their spouses' relative use of withdrawal than would be the parallel case for wives. In summary, examination of the relative score coping correlations revealed that for husbands there was a significant, positive association between their own reported relative use of confrontation and withdrawal and their perceptions of their spouses' relative use of these two strategies, but there was only a marginally significant association between husbands' own reported relative use of compromise and their perceptions of their spouses' relative use of compromise. Although there were no significant gender differences in these relationships, no significant relationships were found between wives' own reported relative use of any of the three coping strategies and their perceptions of their spouses' relative use of these same coping strategies. There were no significant relationships between husbands' and wives' reported use of either confrontation or withdrawal and their perceptions of their spouses' use of different coping strategies, i.e., either withdrawal or  143  confrontation. There was, however, a significant gender difference suggesting that increases in wives' reported confrontation, compared to that of their husbands, was more likely to be associated with decreases in wives' perceptions of their spouses' withdrawal, or that increases in husbands' reported confrontation, compared to that of their wives, was more likely to associated with increases in husbands' perceptions of their spouses' withdrawal. Summary When self coping and perceived coping were compared, the raw score and relative score methods yielded different patterns of significance. For both scoring methods, husbands perceived their wives as more confrontational and less withdrawing and less compromising than they viewed themselves. Both scoring measures indicated that wives viewed themselves as compromising more than their husbands. However, for the wives' data, the two measures were mixed with respect to withdrawal and confrontation. The raw scores suggested that wives viewed themselves as withdrawing more than they perceived their husbands as withdrawing, but there was no significant difference in the measures pertaining to confrontation. The relative scores suggested, however, that wives perceived themselves as confronting significantly less often than their husbands, with no differences in terms of the measures pertaining to withdrawal. Both types of scoring produced results in the same direction, but the significance levels differed. The correlational analyses for both raw and relative scores indicated that husbands tended to perceive their spouses as coping in a similar way as they did themselves. Although not significantly different from their husbands, fewer significant relationships were found between wives' perceptions of their spouses' coping and wives' reports of their own coping. There were also no significant correlations, using both scoring methods, for either husbands or wives, between reports of self confrontation and perceived spouses' withdrawal and between reports of self withdrawal and perceived spouses' confrontation.  144  The relationship between respondents' perceptions of their spouses' coping and spouses' coping (Exploratory Analysis I) Consistent with the preceding organization of results, two sets of comparisons between respondents' perceptions of their spouses' coping and their spouses' reports of their own coping are presented below. The first set of comparisons relates to mean differences between these variables. The second set of comparisons pertains to correlations between respondents' perceptions of their spouses' coping and their spouses' reports of their own coping. Comparisons between perceived spouses' coping and spouses' coping (raw score method) Exploratory analyses were conducted separately for husbands and wives to examine the relationship between respondents' perceptions of their spouses' coping and their spouses' reports of their own coping. The relationships between husbands' and wives' perceptions of their spouses' coping and their spouses' reports of their own coping, using raw score coping values, are summarized in Table 22. The findings of this series of paired r-tests indicated that wives perceived that their spouses used more confrontation (M = 2.70, SD = .73) than was reflected in their spouses' own coping reports [M — 2.44, SD = .73, t (53) = 2.37, p < .05]. There was no significant difference, however, between husbands' perceptions of their spouses' use of raw score confrontation (M = 2.66, SD = .74) and their spouses' reports of their own use of raw score confrontation [M = 2.62, SD = .71, t (53) = .37, p > .10]. Regarding raw score compromise, these analyses revealed that husbands tended to perceive that their spouses compromised less (M = 2.25, SD = .91) than their spouses reported for themselves [M = 2.60, SD = .66, t (52) = -2.56, p < .05]. Similarly, it was found that wives tended to perceive that their  Table 22 Paired Mests Comparing Perceived Spouses' Coping and Spouses' Coping Raw Score Values Wives  Husbands Coping Perceived spouses' confrontation vs. Spouses' confrontation Perceived spouses' compromise vs. Spouses' compromise Perceived spouses' withdrawal vs. Spouses' withdrawal  M SD 2.66  .74 .37  M  SD  2.70  .73  2.62  .71  2.44  .73  2.25  .91  2.18  .85 -2  -2.56* 2.60  .66  2.54  .79  1.96  .70  2.13  .73  2.40  .79  -3.71***  *p < .05 **p<.01 ***p < .001  2.30  -1 .75  146 spouses compromised less (M = 2.18, SD = .85) than their spouses reported for themselves [M = 2.54, SD = .79, t (52) = -2.88, p < .01]. It was also found that wives' perceptions of their spouses' use of raw score withdrawal (M = 2.13, SD = .73) did not differ significantly from the spouses' reports [M = 2.30, SD = .73, t (53) = -1.39, p > .10]. Husbands, on the other hand, tended to perceive that their spouses used significantly less raw score withdrawal (Af = 1.96, SD = .70) than their spouses reported themselves [Af = 2.40, SD = .78, t (53) = -3.71, p < .001]. Examination of the raw score correlation coefficients in Tables 9 and 10 revealed, as was suggested, relatively low rates of agreement between respondents' perceptions of their spouses' coping and spouses' reports of their own coping. For husbands, none of the three correlations (Table 9) comparing perceived spouses' coping to spouses' coping reached the .01 level of significance, but there was a trend toward significance (p < .05) in two of these correlations: perceived spouses' confrontation and spouses' confrontation (r = .31, p < .05); perceived spouses' compromise and spouses' compromise (r = .21, p > .10); perceived spouses' withdrawal and spouses' withdrawal (r = .30, p < .05). For wives (Table 10), a significant, positive relationship was found between perceived spouses' confrontation and spouses' confrontation (r = .39, p < .01) and between perceived spouses' compromise and spouses' compromise (r =.37, p < .01). There was a trend toward a positive relationship between wives' perceived spouses' withdrawal and spouses' withdrawal (r = .33, p < .05). There were no significant differences between the correlations obtained for husbands and wives. Thus, there were no significant gender differences in the rates of agreement between respondents' perceptions of their spouses' coping and their spouses' reports of their own coping.  147  Comparisons between perceived spouses' coping and spouses' coping (relative score method) The preceding exploratory analyses comparing respondents' perceptions of their spouses' coping with their spouses' reports of their own coping were also conducted using relative score coping values. The relationships found between husbands' and wives' perceptions of their spouses' relative score coping and their spouses' reports of their own relative score coping are summarized in Table 23. Similar to the results found using raw scores, when relative coping values were used, a series of paired Mests indicated that wives perceived that their spouses used more confrontation (M = .39, SD = .12) than was reflected in their spouses' reports of confrontation [M = .34, SD = .10, t (53) = 3.30, p < .01]. However, in contrast to the raw score findings, husbands' also perceived that their spouses used more relative score confrontation (M = .40, SD = .11) than was reported by the spouses themselves [M = .35, SD = .10, t (53) = 2.87, p < .01]. Regarding relative score compromise, paired Mests revealed that wives perceived that their spouses compromised less (M = .30, SD = .08) than their spouses reported for themselves [M = .34, SD = ..07, t (52) = -3.19, p < .01]. The findings pertaining to the compromise variables did not reach significance for the husbands; husbands' perceptions of their spouses' compromise (M = .32, SD = .08) was not significantly different from their spouses' reports of their own use of compromise [M = .34, SD = .06, t (52) = -1.72, p > .05]. Consistent with the raw score results, it was found that wives' perceptions of their spouses' use of relative score withdrawal (M = .30, SD = .08) did not differ significantly from the spouses' reports themselves [M = .32, SD = .09, t (53) = -1.19, p > .10]. Husbands, however, tended to perceive that their spouses used significantly less relative score withdrawal (M = .29, SD = .07) than their spouses reported themselves [M = .33, SD = .12, t (53) = -2.45,/? < .05].  Table 23  Paired f-tests Comparing Perceived Spouses' Coping and Spouses' Coping Relative Score Values  Husbands M  SD  Perceived spouses' confrontation vs. Spouses' confrontation  .40  .11  .35  Perceived spouses' compromise vs. Spouses' compromise  Coping  Perceived spouses' withdrawal vs. Spouses' withdrawal  Wives M  SD  .39  .12  .10  .34  .10  .32  .08  .30  .08  .34  .06  .34  .07  .29  .07  .30  .08  .33  .12  .32  .09  t  2.87**  -1.72  -2.45*  *p < .05 **p<.01 ***p < .001  149  Examination of the relative score coping correlation coefficients in Tables 11 and 12 also revealed low rates of agreement between perceived spouses' relative score coping and spouses' reports of their own relative score coping similar to those found using raw score coping values. For husbands, one of the three correlations (Table 11) comparing respondents' perceived spouses' relative score coping to spouses' relative score coping reached significance. Higher relative scores for husbands' perceptions of their spouses' confrontation were associated with higher relative scores for their spouses' reports of their own confrontation (r = .36, p < .01). The relationships between husbands' perceived spouses' compromise and their spouses' compromise (r = .13, p >  .10), and between  husbands' perceived spouses' withdrawal and their spouses' withdrawal (r = .23, p > .10) did not reach significance. Similarly, for wives (Table 12), higher relative scores for perceived spouses' confrontation were significantly associated with higher relative scores for their spouses' confrontation (r = .40, p < .01). The relationships between wives' perceived spouses' compromise and their spouses' compromise (r =.27, p < .05) and between wives' perceived spouses' withdrawal and their spouses' withdrawal (r = .30, p < .05) did not reach significance at the .01 alpha level, although there was a trend toward a significant positive relationship in both of these comparisons. In summary, when coping was considered relative to total coping effort, for both husbands and wives, respondents' perceptions of their spouses' confrontation was positively associated with their spouses' reported confrontation. Although no significant relationships were found for either husbands or wives between respondents' perceptions of their spouses' compromise and withdrawal, and their spouses' reported use of compromise and withdrawal, there was a trend toward significance for both of these relationships in the wives' data. There were, however, no gender differences between these relationships for husbands and wives.  150 Overall Summary of Hypothesis III and Exploratory Analysis I Contrary to prediction no gender differences were found in husbands' and wives' reports of their own coping when dealing with marital tension that was reported to be the most serious source of interpersonal stress during a particular day. Thus, husbands and wives did not differ from one another in their reported use of confrontation, compromise, and withdrawal strategies. Regarding the relationship between respondents' own reports of coping and their perceptions of their spouses' coping, exploratory analyses (paired Mests) revealed that husbands perceived that their wives used more confrontation, and less compromise and withdrawal than did the husbands themselves. These results were obtained both when coping was considered in terms of individual strategies without reference to total coping effort (i.e., raw score coping) and when coping was considered in terms of its relative use, compared to total coping effort expended (i.e., relative score coping). The pattern of results for the wives, although not significantly different from the pattern found for husbands, was not as straight-forward. For example, it was found that wives perceived that their husbands reported proportionately more confrontation (relative to their total coping effort) than wives themselves reported. This relationship was not found, however, when confrontation was considered without reference to total coping effort. When withdrawal was considered without reference to total coping effort, wives reported using more withdrawal strategies themselves than they perceived their spouses using; however, this was not the case when withdrawal was considered relative to total coping effort. Regarding compromise, the findings for the wives were similar to that found for husbands; wives reported using more compromise strategies (both with and without reference to total coping effort) themselves than they perceived that their spouses used. Regarding correlations between respondents' own coping and their perceptions of their spouses' coping, the following exploratory results were found. For husbands, a  151  significant, positive association was found between the extent to which they reported use of the three coping strategies and their perceptions of the extent to which their spouses used the same three coping strategies when coping strategies were considered without reference to total coping effort. When coping was considered relative to the total coping effort expended, the association between husbands' own reported use of both confrontation and withdrawal and their perceptions of their spouses' use of these two strategies remained significant and positive; however, there was only a trend toward significance in the relationship between husbands' own reported relative use of compromise and their perceptions of their spouses' relative use of compromise. For wives, when coping was considered without reference to total coping effort, no significant relationship was found between wives own reported use of confrontation and compromise and their perceptions of their spouses' use of these two coping strategies, although there was a trend toward significance in the relationship between wives' compromise and their perceived spouses' compromise. A significant, positive association was found between wives' reported use of withdrawal and their perceptions of their spouses' use of withdrawal. However, when coping was considered relative to total coping effort, no significant relationship was found between wives' own reported relative use of any of the three coping strategies and their perceptions of their spouses' relative use of these same coping strategies, although a trend toward significance was found between wives' reported withdrawal and their perceptions of their spouses' withdrawal. Even though the patterns among the correlations for husbands and wives appeared to differ somewhat, no significant gender differences were found. No significant correlations were found, using either raw scores or relative scores, in regard to the relationships between husbands' and wives' reported use of confrontation and their perceptions of their spouses' use of withdrawal or between husbands' and wives' reported use of withdrawal and their perceptions of their spouses' use of confrontation. However, there was a significant gender difference between the relative score correlations  152  suggesting that husbands and wives differed in the association between their own reports of confrontation and their perceptions of their spouses' withdrawal. Results from exploratory analyses (paired Mests) pertaining to the relationship between respondents' perceived spouses' coping and the spouses' own reports of their coping were in keeping with the expectation of low levels of agreement between spouses regarding their perceptions of one another's coping. These analyses revealed that wives reported that their husbands used more confrontation (both raw score and relative score) than husbands reported themselves. Husbands also perceived that their wives used more confrontation than the wives reported using themselves when confrontation was considered relative to total coping effort. This finding was not significant, however, when confrontation was considered without reference to total coping effort. These analyses also indicated that husbands and wives both reported that their spouses employed less compromise than spouses reported themselves, but this finding did not reach significance for husbands when compromise was considered relative to total coping effort. Data analyses also revealed that husbands reported that their wives used fewer withdrawal strategies than wives reported using themselves and this finding was statistically significant when withdrawal was considered both with and without reference to total coping effort. An exploratory comparison of correlations between respondents' perceptions of their spouses' coping and their spouses' own reports of their coping also indicated relatively low levels of agreement between spouses. No significant relationships were found between husbands' perceptions of their wives' coping and their wives' reports of their own coping when coping was considered without reference to total coping effort; however, there was trend toward significance in two of these relationships (pertaining to confrontation and to withdrawal). When coping was considered relative to total coping effort, a significant, positive relationship emerged between husbands' perception of their wives' confrontation and wives' reports of their own confrontation, but no other significant relationships (or trends toward significance) were found in the husbands' data.  153  For wives, when coping was considered without reference to total coping effort, significant positive relationships were found between wives' perceptions of their husbands' confrontation and their husbands' own reports of confrontation and between wives' perceptions of their husbands' compromise and their husbands' own reports of compromise. In addition, a trend toward significance was found between wives' perceptions of their husbands' withdrawal and their husbands' own reports of withdrawal. When coping was considered relative to total coping effort, a similar pattern emerged. Only the relationship between wives' perceptions of their husbands' confrontation and their husbands' reports of their own confrontation obtained significance, but there were trends toward significance in regard to the other two relationships pertaining to compromise and withdrawal. No significant gender differences were found in the relationships between husbands' and wives' perceptions of their spouses' coping and their spouses' reports of their own coping. Analyses relating to Hypotheses IV and V and Exploratory Analyses II Bivariate analyses Analyses were conducted to examine the relationship between respondents' distress (current distress) and their reported use of the three types of coping (i.e., confrontation, compromise, withdrawal) to deal with marital tension. The relationship between respondents' current distress and their reports of the three coping strategies was considered in terms of the types of coping reports. Coping scales were analyzed in terms of reports by respondents concerning their own coping efforts (self coping), in terms of reports by respondents concerning their perceptions of their spouses' coping (perceived spouses' coping) and in terms of reports by respondents' spouses concerning the spouses' own coping (spouses' coping). In addition, both raw coping scores (scores derived for each individual coping scale, without reference to total coping effort expended) and relative coping scores (scores derived taking into account total coping effort expended across all  154  three coping scales) were obtained for all coping measures. Bivariate correlations, for both husbands and wives, between current distress and the three types of coping strategies reported as self coping, perceived spouses' coping, and spouses' coping can be found in Tables 9-12. For the readers' convenience, the correlations relevant to the present discussion of results, along with pertinent z score comparisons, have also been summarized in Table 24. Given that trends toward significance (p < .05) are reported when relevant, alpha levels of .05 have been indicated in Table 24. Correlations between respondents' self coping and respondents' current distress (Hypothesis IV) Regarding the association between respondents' reports of their own coping and respondents' current distress, the following bivariate relationships were found. Contrary to prediction (Hypothesis IVa) no significant relationship (using .01 alpha level) was found between respondents' reports of their own confrontation and current distress for either husbands (raw score r = .17, p > .10; relative score r = .33, p < .05) or wives (raw score r = .15, p > .10; relative score r = .24, p > .01). There was, however, a trend toward significance in the relationship between husbands' reports of their own confrontation and their current distress using relative scores. Similarly, contrary to prediction (Hypothesis IVb) no significant relationship (using .01 alpha level) was found between wives' reports of their own compromise and their current distress (raw score r = -.27, p < .05; relative score r = -.28, p < .05), although there was a trend toward significance in this relationship using both scoring methods that was consistent with prediction. Contrary to prediction (Hypothesis IVb), there was also no significant relationship between husbands' reports of their compromise and their current distress (raw score r = -.16, p > .10; relative score r = -.04, p > .10). When withdrawal was considered without reference to total coping effort, consistent with prediction (Hypothesis  155  IVcl), husbands' reports of withdrawal were significantly negatively related to husbands' current distress (raw score r = -.37, p < .01). Although a similar trend was found in the data when withdrawal was considered relative to total coping effort, the correlation did not reach the .01 level of significance (relative score r = -.30, p < .05). Contrary to prediction (Hypothesis IVc2), for wives, no significant relationship was found between their own reports of withdrawal and current distress, whether or not withdrawal was considered relative to total coping effort (raw score r = -.21, p > .10; relative score r = .29, p < .05). However, it is interesting to note that the two types of scoring produced significantly different results (z = -2.55, p < .01) in regard to this relationship. There was a trend toward a positive relationship (p < .05) between wives' reports of withdrawal and wives' current distress when withdrawal was considered relative to all other coping efforts. This is in contrast to a non-significant (p > .10), but negatively signed correlation between these variables when withdrawal was considered without reference to total coping effort. When correlation coefficients were converted to Fisher z scores and compared, a significant gender difference was found pertaining to the relationship between respondents' reports of withdrawal and their current distress. When withdrawal was considered relative to total coping effort, the relationship between withdrawal and distress differed significantly for husbands and wives (z = -3.14, p < .001). As reported above, there was a trend in the relative score data toward husbands experiencing less distress when they withdrew from marital tension (relative score r = -.30, p < .05) and a trend toward wives experiencing more distress when they withdrew from marital tension (relative score r = .29, p > .01). While neither of these correlation coefficients is significantly different from zero, the two correlations are significantly different from each other (p < .001). Consistent with prediction, it thus appears that increases in husbands' reported  Table 24 Comparison of Husband's and Wives' Correlations Between Current Distress and Coping on Self. Perceived Spouses' and Spouses' Coping Scales Husbands  Wives  z  Self confrontation Self compromise Self withdrawal  .17 -.16 -.37**  .15 -.27* -.21  .10 -.61 -.97  Perceived spouses' confrontation Perceived spouses' compromise Perceived spouses' withdrawal  .21 -.11 -.26  .30* -.03 .06  -.51 -.41 -1.62  Spouses' confrontation Spouses' compromise Spouses' withdrawal  .28* -.24 -.21  -.10 -.02 -.24  1.98* -1.10 .15  Husbands  Wives  Self confrontation Self compromise Self withdrawal  .33* -.04 -.30*  .24 -.28* .29*  .51 1.23 -3 14***  Perceived spouses' confrontation Perceived spouses' compromise Perceived spouses' withdrawal  .30* -.13 -.29*  .30* -.03 .06  0 -.51 -1.84  Spouses' confrontation Spouses' compromise Spouses' withdrawal  .31* -.28* .08  .02 .16 -.19  1.54 -2.27* 1.39  Raw Score Coping  Relative Score Coping  *p < .05 **p < .01 ***p < .001  z  157 relative withdrawal were significantly more likely than increases in wives' reported relative withdrawal to be related to decreases in that respondents' current distress. Correlations between respondents' spouses' coping and respondents' current distress (Hypothesis V and Exploratory Analysis ID Although no hypotheses were advanced regarding specific associations between respondents' current distress and their spouses' coping reports, exploratory analyses were conducted to determine if particular strategies reportedly employed by respondents had gender-related effects on their spouses' mood. For example, it was suggested that increases in husbands' distress might be more closely related to their spouses' reported confrontation than would be the case for wives' distress. In addition, it was suggested that increases in wives' distress might be more closely related to their spouses' reported withdrawal than would be the case for husbands' distress. None of the bivariate relationships between respondents' distress and their spouses' reported coping reached the .01 level of significance (see Table 24). There were, however, trends toward significance in the husbands' data. There was a trend toward husbands' distress being positively related to their wives' reported confrontation using both raw scores (r = .28, p < .05) and relative scores (r = .31, p < .05). In addition, there was a trend toward a negative relationship between husbands' distress and their wives' reported compromise when relative scores were used (r = -.28, p < .05). In Hypothesis V, it was predicted that wives' distress would be more related to their husbands' coping than husbands' distress would be related to their wives' coping. The bivariate data did not support this hypothesis. Instead, gender differences emerged suggesting that husbands' distress was more closely related to their wives' coping than wives' distress was related to their husbands' coping. When the correlations for women and men between the extent to which their spouses reported using confrontation were compared with their own distress, a significant gender difference emerged (raw score z =  158 1.98, p < .05). As noted above, using the .01 alpha level there was no significant relationship (although there was a trend toward significance) between husbands' distress and wives' reported confrontation (raw score r = .28, p < .05). There was no significant relationship between wives' distress and husbands' reported confrontation (r = -. 10, p > . 10). While neither of these correlation coefficients is significantly different from zero, the two correlations are significantly different from each other (p < .05). Thus, increases in husbands' distress were significantly more likely than increases in wives' distress to be related to increases in their spouses' reported confrontation. In addition, when the correlations for women and men between their spouses' relative reported use of compromise and their own distress were compared, a significant gender difference emerged (relative score z = -2.27, p < .05). Using a .01 alpha level, no significant relationship was found between husbands' distress and their wives' relative reported compromise (relative score r = -.28, p < .05), although there was a trend toward a significant negative relationship. There was also no significant relationship between wives' distress and their husbands' relative reported compromise (r = .16, p > . 10). While neither of these correlation coefficients is significantly different from zero, the two correlations are significantly different from each other (p < .05). It thus appears that decreases in husbands' distress were significantly more likely than decreases in wives' distress to be related to increases in their spouses' reported compromise. Correlations between respondents' perceived spouses' coping and respondents' current distress (Exploratory Analyses II) Exploratory analyses concerning the association between respondents' current distress and their perceptions of their spouses' coping revealed that no significant relationships were found (see Table 24). Using both raw score and relative score coping values, there was not a significant relationship (with an alpha level of .01) between respondents' perceptions of their spouses' confrontation and their own current distress for  159 either husbands (raw score r = .21, p > .10; relative score r = .30, p < .05) or wives (raw score r = .30, p < .05; relative score r = .30, p < .05). However, there was a trend toward significant positive relationships (p < .05) in these analyses using relative scores for husbands and using both scoring methods for wives. No significant relationship nor trend toward significance was found between respondents' perceived spouses' compromise and their own current distress for either husbands (raw score r = -.11, p > .10; relative score r = -.13, p > .10) or wives (raw score r = -.03, p > .10; relative score r = -.03, p > .10). In addition, no significant relationship (using .01 alpha level) was found between respondents' perceptions of their spouses' withdrawal and their own current distress for either husbands (raw score r = -.26, p > .01; relative score r = -.29, p < .05) or wives (raw score r = .06, p > .10; relative score r = .06, p > .10). There was, however, a trend toward a negative relationship between husbands' perceptions of their wives' withdrawal and husbands' distress when relative scores were used. Consistent with the lack of significant findings among these variables, no significant gender differences were found when correlation coefficients were converted to Fisher z scores and compared. Multiple regression analyses Hierarchical multiple regression analyses were also employed to examine the above relationships. A series of hierarchical multiple regression analyses was performed separately for husbands and wives for each type of coping (confrontation, compromise, and withdrawal) reported by respondents concerning their own coping (self coping), their perceptions of their spouses' coping (perceived spouses' coping) and their spouses' own coping (spouses' coping). Each analysis was scored both as raw score coping and relative score coping. These analyses resulted in three sets of equations for both husbands and wives, with each set including both raw score and relative score coping and with each set  160  pertaining to either confrontation (Table 25), compromise (Table 26), or withdrawal (Table 27). The dependent variable in these analyses was the respondents' distress score on the day (current distress) in which marital tension was reported as being the most serious interpersonal stressor of the day. The predictor variable, respondents' distress score on the preceding day (prior distress), was entered in the first step of the regressions as a control variable in each of the three sets of equations. The respondents' coping score (either raw score or relative score) was entered in step 2; the respondents' spouses' coping score (either raw score or relative score) was entered in step 3; and the respondents' perceived spouses' coping score (either raw score or relative score) was entered in the fourth step. 10 These regression analyses were conducted using the previously described data subset of 59 couples. Nine couples were dropped from these analyses due to missing data; thus, 50 couples were included in these analyses. Due to the nonindependence in the husbands' and wives' data, the regressions were performed separately for husbands (n = 50) and wives (n = 50). Betas were then converted to Fisher z scores and compared in order to determine gender differences (Pedhazur, 1982). The results of this series of hierarchical multiple regression equations are presented below, organized by type of coping being analyzed (i.e., confrontation, compromise, withdrawal). Confrontation strategies As indicated in each of the regression equations (see Tables 25-27), the control variable, prior distress, did not contribute significantly to the variance in respondents' current distress. Husbands' prior distress (ii = .21) accounted for 5% of the variance (p > .10) in husbands' current distress. Wives' prior distress (J3 = .07) accounted for less than 1 % of the variance (p > .10) in wives' current distress.  Table 25 Hierarchical Multiple Regression Analysis Regressing Current Distress on Confrontation n = 50 Husbands Steps  Predictors  Wives  Beta  R2 Beta change  .05 .09 .09 .04  .21 .30* .30* .23  .01 .03 .01 .11  .07 .18 -.11 .37*  .68 .63 2.04* -.78  .05 .17 .10 .00  .21 .42** .32* -.01  .01 .01 .01 .03  .07 .10 .12 .18  .68 1.70 1.02 -.92  change  Raw Score Analysis 1 2 3 4  Prior distress Self confrontation Spouses' confrontation Perceived spouses' confrontation  Relative Score Analysis 1 2 3 4  Prior distress Self confrontation Spouses' confrontation Perceived spouses' confrontation  *p < .05. **p < .01.  162  Concerning the predictive value of raw score confrontation as related to husbands' current distress, Table 25 reveals that husbands' own reported confrontation (R = .30), entered in step 2 of the equation, accounted for 9% of the variance in husbands' current distress (p < .05) above and beyond the 5% of variance accounted for by husbands' prior distress. That is, husbands' reported confrontation accounted for almost twice as much of the variance in husbands' current distress than that which was accounted for by husbands' prior distress. Increases in the extent to which husbands reported using confrontation were thus associated with increases in husbands' distress. In the third step of this equation, entry of husbands' spouses' confrontation (fl = .30) accounted for an additional 9% of the variance (p < .05), over and above the 14% of the variance already accounted for by prior distress and self confrontation, entered in steps 1 and 2. That is, increases in the extent to which wives reported that they were confrontational were associated with increases in husbands' distress even after accounting for the variance due to prior distress and self confrontation. Although not significant, husbands' perceived spouses' confrontation (6 = .23), entered in step 4, accounted for an additional 4% of the variance (p > .10) in husbands' current distress. Given that 23% of the variance in current distress was accounted for in steps 1-3, entry of perceived spouses' confrontation in the fourth step was a particularly stringent test of the predictive value of this variable. As noted in footnote 10, steps 3 and 4 were entered in the reverse order to test the possibility that spouses' coping might be associated with distress via perceived spouses' coping. A pattern of results supporting this possibility occurred only in the raw score equation pertaining to the relationship between husbands' distress and confrontation. Entry of husbands' raw score perceived spouses' confrontation (6 = .30) in step 3 accounted for a significant 8% of the variance (p < .05) in husbands' current distress; whereas husbands' raw score spouses' confrontation, entered in step 4, was not significant (6 = .23, p > .10), accounting for an additional 4% of the variance.  163  A pattern of results similar to that noted above (with steps 3 and 4 not reversed) was obtained when confrontation was considered relative to respondents' total coping effort (relative score confrontation). Using relative confrontation scores, husbands' own confrontation (B = .42) accounted for 17% of the variance (p < .01) in husbands' current distress. This represented more than three times the variance in husbands' current distress than that which was accounted for by husbands' prior distress. Husbands' spouses' relative score confrontation (B = .32, p < .05) contributed an additional 10% of the variance, over and above the 22% already accounted for by the variables in steps 1 and 2. Thus, increases in either husbands' relative confrontation or wives' relative confrontation were associated with increases in husbands' distress. With 32% of the variance accounted for in steps 1-3, husbands' perceived spouses' confrontation entered in step 4 (B = -.01, p > . 10) explained no significant additional variance in husbands' current distress. As noted above, this pattern of results was not affected by entering steps 3 and 4 in the reversed order. Regarding the relationship between raw score confrontation and wives' current distress, Table 25 indicates that only wives' perceptions of their husbands' confrontation (B = .37, p < .05), entered in step 4, accounted for a significant proportion of the variance in wives' current distress (11 % of the variance, above and beyond the 5 % accounted for by the variables in steps 1-3). That is, increases in the extent to which wives perceived their husbands as confrontational were associated with increases in wives' distress. Wives' own reported confrontation (B = .18, p > .10) accounted for only 3% of the variance over and above the 1 % accounted for by wives' prior distress; and wives' spouses' confrontation (8 = -.11, p > .10) accounted for only an additional 1% of the variance when entered in step 3. Although similar results were obtained in steps 1-3 of the regression equation when confrontation was considered relative to respondents' total coping efforts (relative score confrontation), wives' perceived spouses' confrontation (B =  164 AS, p > .10), entered in step 4, was not significant in this set of equations (see Table 25). When betas were converted to Fisher z scores and compared, a significant gender difference emerged in the relationship between spouses' raw score confrontation and respondents' current distress (z = 2.04, p < .05). Consistent with the suggestion that husbands might be particularly vulnerable to their wives' confrontation, these results indicated that husbands' distress was significantly more related to the extent to which their spouses reported being confrontational than was the case for wives' distress. Increases in the extent to which wives reported using confrontation were associated with increases in their husbands' current distress (6 = .30, p < .05), whereas the extent to which husbands reported using confrontation was not significantly related to their wives' current distress (6 = -.11, p > .10). In summary, findings from this series of regression analyses were consistent with the prediction that husbands' reported use of confrontation strategies would be associated with their increased current distress. Contrary to prediction, however, a parallel finding was not obtained for wives. Instead, these results indicated that wives' reported use of confrontation was not associated with their current distress. Consistent with exploratory suggestions noted in regard to the relationship between husbands' distress and their spouses' confrontation, these results also indicated that husbands' distress was more positively related to their spouses' confrontation than was wives' distress. Although a significant relationship was found between wives' perceptions of their husbands confrontation (considered without reference to their total coping efforts) and wives' distress, no relationship was found between spouses' confrontation and wives' distress. Thus, the findings related to confrontation do not support the prediction that wives' distress would be more closely related to their spouses' coping than would be husbands' distress.  165  Compromise strategies The results of the regression analyses relating to the relationship between compromise and respondents' current distress (Table 26) indicated no significant effects for the compromise variables entered in these equations. For husbands, when compromise was considered without reference to total coping efforts (raw score compromise), the three compromise variables [self compromise (15 = -.13, p > .10), spouses' compromise (6 = -.20, p > .10), and perceived spouses' compromise (B = .08, p > .10)] did not add significantly to the prediction of husbands' current distress over and above that accounted for by prior distress. Only 6% of the variance in husbands' current distress was accounted for by the three compromise variables combined. A similar pattern was found when compromise was considered relative to total coping efforts (relative score compromise). Although spouses' compromise approached significance (B = -.27, p < .10) in this equation, accounting for 7% of the variance in husbands' current distress, husbands' own compromise (B = -.04, p > .10) and husbands' perceived spouses' compromise (B = -.06, p > .10) accounted for no significant additional variance in husbands' current distress. Thus, contrary to prediction, for husbands neither their own reported use of compromise, their wives' reports of using compromise, or husbands' perceptions of their wives' use of compromise was significantly related to husbands' distress on days in which they reported experiencing marital tension. Similarly, for wives, the compromise variables did not add significantly to the prediction of wives' current distress. In the set of equations in which compromise was considered without reference to total coping efforts (raw score compromise), wives' own coping (B = -.19, p > .10), their spouses' coping (6 = .01, p > .10), and their perceived spouses' coping (B = . 10, p > .10) explained only 4% of the variance in wives' current distress over and above the 1 % accounted for by prior distress (see Table 26).  Table 26 Hierarchical Multiple Regression Analysis Regressing Current Distress on Compromise n = 50 Husbands Predictors Beta  Steps  Wives R2  change  Beta change  z  Raw Score Analysis 1 2 3 4  Prior distress Self compromise Spouses' compromise Perceived spouses' compromise  .05 .02 .04 .01  .21 -.13 -.20 .08  .01 .03 .00 .01  .07 -.19 .01 .10  .68 .29 -1.02 -.10  .05 .00 .07 .00  .21 -.04 -.27 -.06  .01 .05 .02 .04  .07 -.23 .15 -.21  .68 .92 -2.09* .73  Relative Score Analysis 1 2 3 4  Prior distress Self compromise Spouses' compromise Perceived spouses' compromise  * p < .05.  167  A similar pattern of results was obtained when compromise was considered relative to total coping effort (relative score compromise). In this set of equations, the three relative score compromise variables did not contribute significantly to the variance in wives' current distress. Wives' own coping (8 = -.23, p > .10), their spouses' coping (6 = .15, p > .10), and their perceived spouses' coping (6 = -.21, p > .10) combined to explain 11 % of the variance, but no one variable was significant. Thus, contrary to prediction, for wives neither their own reported use of compromise, their husbands' reports of using compromise, nor wives' perceptions of their husbands' use of compromise was significantly related to wives' distress on days in which they reported experiencing marital tension. A significant gender difference was found when betas were converted to Fisher z scores and compared. When spouses' compromise was considered relative to total coping efforts, a significant gender difference emerged in the relationship between spouses' reported compromise and respondents' current distress (z = -2.09, p < .05). There was no significant relationship between husbands' distress and their wives' reported relative score compromise (8 = -.27, p > .05). In addition, no significant relationship was found between wives' distress and their husbands' reported relative score compromise (8 = .15, p >. 10). While neither of these betas is significantly different from zero, the two betas are significantly different from each other (p < .05). It thus appears that husbands were more likely than were their wives to report less distress when their spouses reported using compromise to deal with marital tension. In summary, contrary to prediction, neither husbands' nor wives' current distress was significantly accounted for by their own reported use of compromise, their spouses' reported use of compromise, or their perceptions of their spouses' use of compromise. A significant gender difference was found, however, regarding the relationship between respondents' current distress and their spouses' reported relative use of compromise. These results suggested that it was more likely that husbands' distress was associated with  168  their wives' reported relative use of compromise than it was likely that wives' distress was associated with their husbands' reported relative use of compromise. Contrary to the prediction that wives' distress would be associated more closely with their husbands' coping than husbands' distress would be associated with their wives' coping, the opposite relationship was found in these data pertaining to the relationship between spouses' relative compromise and respondents' distress when dealing with marital tension.  Withdrawal strategies Regarding the predictive power of the raw score withdrawal variables in relation to husbands' current distress, Table 27 indicates that, consistent with prediction, husbands' own reported withdrawal (13 = -.40) accounted for a significant proportion of the variance in husbands' current distress, even after controlling for prior distress. Entered in step 2 of the regression equation, husbands' own raw score withdrawal accounted for a significant 16% of the variance in distress (p < .01) over and above the 5% accounted for by prior distress. That is, increases in the extent to which husbands withdrew were significantly associated with decreases in their distress. Spouses' reported raw score withdrawal (B = .11, p > . 10) and perceived spouses' raw score withdrawal (6 = . 13, p > . 10) did not significantly contribute to the variance in current distress, accounting for only 2% combined additional variance over and above the 21 % of variance accounted for by husbands' own reported withdrawal and prior distress. A similar pattern of results was obtained when withdrawal was scored relative to total coping efforts (relative score withdrawal). As indicated in Table 27, when relative coping scores were used, husbands' own reported withdrawal (B = -.45) accounted for a significant 21 % of the variance in distress (p < .001) in addition to the 5% accounted for by prior distress. That is, increases in husbands' reported use of withdrawal were associated with decreases in their distress. Spouses' reported withdrawal (6 = -.19p >.10) and perceived spouses' withdrawal (B = -.05, p >. 10) did not significantly contribute to the variance in current  Table 27 Hierarchical Multiple Regression Analysis Regressing Current Distress on Withdrawal n = 50 Husbands Steps  Predictors Beta  Wives R2  change  Beta change  z  Raw Score Analysis 1 2 3 4  Prior distress Self withdrawal Spouses' withdrawal Perceived spouses' withdrawal  .05 .16 .01 .01  .05 .21 .03 .00  .21  -.11 .13  .01 .00 .08 .04  .07 -.07 -.33* .24  .68 -1.73 1.11 -.53  .21 _ ^c*** -.19 -.05  .01 .00 .06 .00  .07 -.08 -.26 .06  .68 -1.94 .34 -.53  . 4Q* *  Relative Score Analysis 1 2 3 4  Prior distress Self withdrawal Spouses' withdrawal Perceived spouses' withdrawal  *p < .05 **p < .01 ***p < .001  170  distress, accounting for only 3% combined additional variance over and above the 26% of variance accounted for by husbands' own reported withdrawal and prior distress. Turning now to the predictive power of raw score withdrawal in relation to wives' current distress, Table 27 indicates that wives' own reported withdrawal (B = -.07, p > . 10) did not explain any additional variance beyond the 1 % of variance accounted for by prior distress. Spouses' raw score withdrawal (fi = -.33, p < .05), entered in step 3, was significant, however, accounting for an additional 8% of the variance in wives' current distress. That is, increases in the extent to which their husbands withdrew were associated with decreases in wives' distress. Although not significant, perceived spouses' raw score withdrawal (6 = .24, p >. 10) accounted for an additional 4% of the variance, above and beyond the 9% accounted for by variables in steps 1-3. A similar pattern of results was found when withdrawal was scored relative to total coping efforts (relative score withdrawal). When relative coping scores were used, wives' own reported withdrawal (B = -.08, p > .10) accounted for no additional variance in wives' distress. Although not significant, spouses' relative score withdrawal (6 = -.26, p > .05) accounted for 6% of the variance in addition to the 1 % accounted for by prior distress. Perceived spouses' relative withdrawal (6 = .06, p > .10), entered in step 4, accounted for no additional variance in wives' current distress. A marginally significant gender difference was also found when betas were converted to Fisher z scores and compared. Comparisons between husbands' and wives' self withdrawal, scored relative to total coping efforts, yielded a marginally significant result (z = -1.94, p < .053). The same comparison between husbands' and wives' own withdrawal, scored without reference to total coping efforts, indicated only a slight trend toward significance (z = -1.73,/? < .10 > .05). Thus, these two findings suggest that husbands' distress was more likely than wives' distress to be related to reports of self withdrawal from marital tension. More specifically, the data indicated that decreases in husbands' distress were associated with husbands' reported withdrawal from marital  171 conflict (raw score fi = -.40, p < .01; relative score B = -.45, p < .001), whereas wives' distress was not related to their own reports of withdrawal from marital conflict (raw score 15 = -.07, p > .10; relative score R = -.08, p > .10). No significant gender differences were found in the relationships between spouses' withdrawal and respondents' distress; thus, no support was provided for the prediction that wives' distress would be more closely related to their spouses' coping than would husbands' distress. In summary, these results indicated that, as predicted, husbands' own reported use of withdrawal to deal with marital tension was associated with decreased current distress. When withdrawal was scored relative to total coping efforts, this relationship was marginally, significantly different from the relationship found in the wives' data. Perceived spouses' withdrawal and spouses' reported withdrawal were not significantly related to husbands' current distress. These findings also indicated that, contrary to prediction, there was no relationship between wives' own reported withdrawal and their current distress. A significant relationship between spouses' reported withdrawal, scored without reference to total coping efforts, and wives' current distress was found. Contrary to expectation, however, these results indicated that, for wives, increased withdrawal reported by their husbands was associated with decreased, not increased, distress. Also, as noted above, no support was found for the prediction that wives' distress would be more closely related to their spouses' reported coping than would be their husbands' distress. Summary of Hypotheses IV and V and Exploratory Analysis II Results from the bivariate analyses and the multivariate analyses differed somewhat in the extent to which they provided support for hypotheses related to expected associations between respondents' current distress and their reported use of specific coping strategies. Some of these differences were due, in part, to the more conservative alpha level set for the correlational analyses given the large number, overall, of correlation  172 coefficients reported in this research. Interpretation of trends toward significance (p < .05), however, addressed some of these discrepancies. For example, although no significant correlations (p < .01) were found between husbands' current distress and their reported use of confrontation, there was a trend toward a significant positive relationship between husbands' distress and their reported confrontation (p < .05) when confrontation was considered relative to total coping effort. This finding is in keeping with the results of the hierarchical multiple regression analyses (using both raw scores and relative scores) that provided support for the prediction that increases in husbands' reported use of confrontation to deal with marital tension would be associated with increases in husbands' distress. Consistent with prediction, results of both bivariate and multivariate analyses indicated that greater reported use of withdrawal by husbands was significantly associated with a decrease in husbands' distress. Contrary to prediction, however, both bivariate and multivariate analyses found no significant relationship between husbands' distress and their reported use of compromise strategies. In addition, contrary to prediction, both bivariate and multivariate analyses found no significant relationship between wives' distress and their reported use of any of the three coping strategies: wives' distress was not related to their reported use of confrontation, compromise, or withdrawal. Although there was a trend toward significance in the bivariate analyses pertaining to the relationship between wives' distress and their reported use of compromise (p < .05, for both raw score and relative score correlations) and between wives' distress and their reported use of withdrawal (p < .05, for relative score correlation), the lack of significant relationships in any of the multivariate analyses suggests caution in interpreting these trends in the bivariates. The results of these analyses also did not provide support for the prediction that wives' distress would be more closely related to their spouses' coping than would be husbands' distress. Instead, when gender differences were found in the relationship  173  between spouses' reported coping and respondents' distress, they indicated that husbands' distress was more closely related to their spouses' reported coping than was the case for wives' distress. Both bivariate and multivariate analyses indicated significant gender differences in the relationship between respondents' distress and reports of spouses' confrontation and of spouses' compromise that suggested that husbands' distress was more likely than wives' distress to be related to their spouses' coping reports. Regarding exploratory analyses of the relationship between respondents' distress and their perceptions of their spouses' coping, a significant relationship was found between wives' distress and their perceptions of their spouses' confrontation, considered without reference to total coping effort. Interestingly, wives' distress was significantly related to their perceptions of their spouses' confrontation, but not to their spouses' own reports of confrontation. No significant relationship was found between wives' distress and their perceptions of their spouses' use of compromise or withdrawal. Furthermore, husbands' perceptions of their spouses' coping (confrontation, compromise, or withdrawal) were generally not significantly related to husbands' distress. An exception to this was found, however, when raw score perceived spouses' coping was entered in step 3, rather than step 4, of the regression equation. In this case, increases in perceived spouses' confrontation (i.e., husbands' perceptions of their wives' confrontation) were associated with increases in husbands' distress. Overall Summary of Results The results from these data analyses have demonstrated both similarities and differences in husbands' and wives' reports of coping with marital tension. Both husbands and wives in this study acknowledged incidents of marital tension in their relationship and both husbands and wives reported increased distress that was associated with this marital tension. Consistent with prediction, however, results indicated that wives reported proportionately more instances of marital tension and their levels of distress were  174  significantly higher when such tension occurred. When both husbands and wives acknowledged the presence of marital tension or conflict and when both spouses considered this tension to be the most serious interpersonal tension experienced that day, results indicated that the coping process engaged in by husbands and wives was remarkably similar. Thus, contrary to prediction, spouses tended to appraise the situation similarly in terms of the extent to which they felt threatened by the conflict. Also, contrary to prediction, results indicated that husbands and wives in this study tended to engage in similar coping strategies; there were no significant differences in their reported use of confrontation, compromise, or withdrawal. Although spouses did not differ in their reported self coping, exploratory data analyses showed that husbands and wives perceived that their spouses coped differently than they did themselves. When paired Mests were conducted with coping strategies considered without reference to total coping effort, both husbands and wives reported using more compromise and withdrawal themselves than they perceived that their spouses used, and husbands perceived their wives as using more confrontation than the husbands reportedly used themselves. When coping was considered relative to total coping effort, both husbands and wives reported using more compromise and less confrontation themselves than they perceived that their spouses used, and husbands reported using more withdrawal than they perceived that their wives used. No gender differences were found in the relationships between husbands' and wives' own coping reports and their perceptions of their spouses' coping. Exploratory analyses also indicated that husbands and wives tended to perceive that their spouses coped differently than the spouses themselves reported. When coping was considered without reference to total coping efforts, paired Mests indicated that husbands' perceptions of their wives' confrontation were not significantly different from the wives' reports of their own confrontation; however, husbands perceived that their wives compromised and withdrew significantly less than the wives themselves reported. On the  175  other hand, wives' perceptions of their husbands' withdrawal were not significantly different from their husbands' own reports; however, wives perceived that their husbands compromised less and confronted more than the husbands themselves reported. The same pattern was found for wives when coping was scored relative to total coping effort. The pattern was somewhat different for husbands, however, when relative coping scores were used. Husbands' perceptions of their wives' relative use of compromise was not significantly different from their wives' reports; however, husbands perceived that their wives used relatively more confrontation and less withdrawal than was reported by wives themselves. There were no gender differences, however, in the relationships between husbands' and wives' perceptions of their spouses' coping and the spouses' reports of their own coping. Although spouses endorsed similar coping strategies, results indicated that spouses' reports of coping were differentially related to their current mood. Consistent with prediction, husbands' own reported use of confrontation was associated with their own increased distress, but, contrary to prediction, wives' own reported use of confrontation was not associated with their own distress. Consistent with prediction, husbands' own reported use of withdrawal was associated with their own decreased distress, but, contrary to prediction wives' own reported use of withdrawal was not significantly related with their distress. Contrary to prediction, the reported use of compromise was not significantly related to either husbands' or wives' distress. Also, contrary to prediction, gender differences related to the relationship between spouses' coping and respondents' distress indicated that husbands' distress was more closely related to their spouses' coping (i.e., confrontation and compromise) than was wives' distress.  176 DISCUSSION The organization of the following discussion corresponds to the preceding presentation of the results. Discussion begins with issues related to gender differences in the reports of marital tension and the relationship of those reports to respondents' mood. This is followed by a discussion of the findings related to gender differences in appraisal and respondents' reported use of coping strategies. Differences in respondents' perceptions of their spouses' coping are discussed next, followed by a discussion of the relationship between various coping strategies and respondents' mood depending on whether the coping reports pertained to respondents' own coping, respondents' perceptions of their spouses' coping, or spouses' reports of their own coping. This discussion closes with a summary of the study's findings and suggestions for future research. Spouses' reports of marital tension Consistent with prediction, the results from this research indicate gender differences in spouses' reports of marital tension. During the diary period, wives reported that marital tension occurred more often than did their husbands. Wives also considered marital tension to be the most serious interpersonal tension on proportionately more diary days than did their husbands. Although it was assumed that husbands and wives were both experiencing the "same" marital events, it appears that wives perceived some events as stressful that were not perceived as stressful by their husbands. This finding is in keeping with the cognitively-based model of stress and coping that postulates that there are no objective ways to define stress in terms of the environmental conditions (Lazarus, 1990; Lazarus, DeLongis, Folkman, & Gruen, 1985; Lazarus & Folkman, 1984). Instead of attempting to understand stress on the basis of objective criteria, stress must be considered in terms of the relationship between the environment and the person (Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986; Folkman, Lazarus, Gruen, & DeLongis, 1986; Lazarus, 1990; Lazarus & Folkman, 1984). Within this contextual  177  framework, how a person interprets an event is critical to their experience of stress. Furthermore, a person's interpretation of an event is largely based on the meaning or significance attributed to the event on the basis of the person's needs, beliefs, and commitments (Bradbury & Fincham, 1990; Holtzworth-Munroe & Jacobson, 1985; Lazarus, 1990). Applying this model to the current research provides a framework for interpreting the gender differences found in spouses' reports of marital tension. Consider, for example, the literature review presented in the introduction of this dissertation where evidence was presented suggesting that females and males tend to differ in the meaning that they attribute to interpersonal events (e.g., Belenky et al., 1986; Gilligan, 1982; Josephs et al., 1992; Markus & Oyserman, 1989). If, as the literature suggests, females' sense of self or self-schema is more likely than that of males to be based on connectedness or interdependence (Josephs et al., 1992; Miller, 1976; Jordon & Surrey, 1989; Markus & Oyserman, 1989), it follows that females are more likely to generate meaning or attribute significance to events based on interdependence rather than on independence. This postulated gender difference is consistent with the present research finding suggesting that wives were more sensitive to marital stressors than were their husbands. In addition, although it is not possible to ascertain the accuracy of wives' perceptions of greater tension in their marriages, this finding is consistent with previous research indicating that wives tend to monitor their relationships more closely and therefore are more likely to perceive and call attention to potential conflicts (Holtzworth-Munroe & Jacobson, 1985; Krokoff, 1984). In addition, wives' reporting of proportionately more incidents of marital tension is consistent with research that has indicated that wives generally tend to be more dissatisfied with their relationships and more acutely aware of the problems in their relationships than are their husbands (Margolin et al., 1983; Rhyne, 1981). That husbands and wives tended to disagree in their evaluations of conflict in their marriage was also demonstrated by the fact that both spouses agreed less than 20% of the  178  time that marital tension occurred on the same day. Discrepancies such as these between spouses have resulted in marriage being described as being like two relationships — hers and his (Bernard, 1972). That is, a single objective reality does not exist in a relationship; instead, there are the two spouses' different perspectives, each colored by their individual histories, needs, desires, hopes, and expectations (Bradbury & Fincham, 1990, 1992; Robins, 1990). A manifestation of the different perspectives that spouses can have when dealing with marital conflict was documented in longitudinal research conducted by Huston and his colleagues (Huston, McHale, & Crouter, 1986; Huston, Robins, Atkinson, & McHale, 1987). These researchers attempted to obtain descriptions from individual spouses about their daily experiences of conflict, but this effort was abandoned because "the two spouses' reports were often so different that the episode was barely recognizable as the same one" (Huston et al., 1987, p. 59). Although husbands and wives differed in the extent to which they reported marital tension, both husbands and wives reported relatively few incidents of marital tension during the diary period. Approximately one third of the spouses (36% of the husbands and 29% of the wives) reported no incidents of marital tension during the diary period. The remaining spouses reported marital tension on an average of approximately two (56%) of the 42 diary days. This finding appears to be in contrast with statements from the literature noted in the introduction of this dissertation suggesting that marriage represents "continuous conflict" (Fitzpatrick, 1988) or that everyday marital hassles or daily "ups and downs" are an inevitable aspect of married life (DeLongis & Lehman, 1989; Rusbult, 1987). Given the small number of incidents of marital tension reported by this study's sample of spouses, it does not appear that these spouses would characterize their marriages as being in a state of constant conflict. There may be daily, or very frequent, fluctuations in the ways in which most marital couples interact. However, it may well be the hallmark of healthy marriages that the ups and downs are there, but that they are not appraised as constituting marital tension. This apparent discrepancy between descriptors of marriage in  179  the literature and relatively well functioning community spouses' reports of marital tension in their marriage should be explored more directly in future research. If it is assumed that husbands and wives differed in their perceptions of marital tension, another question arises pertaining to whether or not one gender's reports of marital tension more accurately reflected the couples' interaction than did the other gender's reports. It could be the case that wives more accurately reported marital tension than did their husbands, who might be less attentive to interpersonal issues (HoltzworthMunroe & Jacob son, 1985; Margolin et al., 1983). Alternatively, it could be the case that husbands more accurately reported marital tension than did their wives, whose heightened interpersonal sensitivity may have led to exaggerations in their perceptions of marital tension (Jacobson, 1989; Nolen-Hoeksema, 1987). Whether or not it is possible to objectively assess and measure environmental events appraised by individuals to be stress producing is a matter of debate. For example, recently there has been a debate in the stress and coping literature between Dohrenwend and his colleagues and Lazarus and his colleagues (Dohrenwend, Dohrenwend, Dodson, & Shrout, 1984; Dohrenwend & Shrout, 1985; Lazarus, 1990; Lazarus et al., 1985) concerning this issue. Although objective measurement of behaviors or events that trigger stress appraisals may be a worthwhile goal, Lazarus (1990) has argued that no existing measures are totally free from subjective bias. Thus, particularly in situations in which the stressful event is interpersonal in nature, as in the case of spouses dealing with marital tension, it may be the individuals' perceptions of the stressful event rather than the "accuracy" of their perceptions that are of interest in understanding the coping process (Fincham, Bradbury, & Beach, 1990; Floyd & Markman, 1983; Gottman & Porterfield, 1981; Lazarus et al., 1985; Margolin, 1987). Therefore, although spouses' perceptions of the occurrence of marital tension differed, in the present study one spouse's perception cannot be considered more valid than the perception of the other spouse.  180  The relationship between reports of marital tension and distress In addition to their reports of more incidents of marital tension, wives' daily distress was more closely related to their experiences of marital tension than was the case for husbands. This finding is consistent with the literature which has generally found higher reported symptoms of distress (e.g., depression and anxiety) among women when compared to men (e.g., Belle, 1980; Myers et al., 1984; Nolen-Hoeksema, 1987; Robins et al., 1984; Weissman & Klerman, 1977), especially when married women are compared to married men (e.g., Aneshensel, Frerich, & Clark, 1981; Gove, 1972; Gove, Hughes, & Style, 1983; Radloff & Rae, 1979). Kessler and McLeod (1984) have suggested that this gender difference is largely due to women's greater exposure to network events or interpersonal stressors. However, in the present research, as noted above, even when husbands and wives were presumably exposed to the same marital events, wives reported proportionately more marital stressors and these stressors predicted a significantly greater degree of wives' distress. It was expected in this research that the greater association between wives' distress and their reports of marital tension, compared to the corresponding relationship for husbands, might be related to wives feeling more threatened than did their husbands by the occurrence of marital tension. As discussed more fully below, however, wives did not rate marital tension as more threatening than did their husbands. Gender differences in the relationship between spouses' distress and marital tension can be interpreted in the context of other research findings. For example, Holtzworth-Munroe and Jacobson (1985) interpreted their findings as indicating that husbands, unlike their wives, tend to invest little time or energy thinking about their marriage until things really start going badly (Holtzworth-Munroe & Jacobson, 1985). Thus, husbands may tend to pay less attention than do their wives to the day-to-day fluctuations in their marriage. Husbands may, therefore, experience less distress than do their wives from the day-to-day hassles or tensions that occur in their marriage. Similar to the process noted by Holtzworth-Munroe  181  and Jacobson (1985), husbands may not become distressed by marital tension until things develop to the point that husbands feel threatened by the tension occurring in their marriage. An alternative interpretation for the gender difference in spouses' responses to marital tension is suggested by the explanation for gender differences in unipolar depression offered by Nolen-Hoeksema (1987). According to Nolen-Hoeksema (1987), women become more distressed by events in their lives because of their tendency to amplify their negative moods by ruminating about their state of mind or about the events thought to cause their distress. Men, on the other hand, tend to respond more adaptively to distressful events by engaging in more active behaviors that distract themselves from the source of their distress. It is possible that a similar process occuned in the present study. That is, women's higher levels of distress related to marital tension may be due to their greater tendency, compared to their husbands, to ruminate about marital events, particularly those perceived as stressful. Husbands, on the other hand, may be less distressed by marital tension because they don't tend to dwell on the negative aspects of marital events to the same extent as do their wives. These gender differences in response sets are discussed in more detail later in respect to this study's findings of gender differences in spouses' distress as a function of their coping. Within the present context, Nolen-Hoeksema's (1987) theoretical perspective does offer a possible explanation for why wives' reports of marital tension were more related to their distress than was the case for their husbands. Given that the gender differences found here relating both to spouses' reports of marital tension and to the relationship found between marital tension and distress were based on self-reports, the issue of potential reporting biases should be addressed. For example, the issue has been raised in the literature concerning the possibility that gender differences in distress may be an artifact of reporting differences between males and females (e.g., Hammen & Padesky, 1977; Phillips & Segal, 1969). In response to  182  criticisms of biases in depressive mood scales, Gove and his colleagues (Clancy & Gove, 1974; Gove & Geerken, 1976; Gove, McCorkel, Fain, & Hughes, 1976) investigated the influence of response biases on these scales and found no gender differences in ratings of the undesirability of the mood items or in need for social approval. Other researchers have also demonstrated no gender differences in participants' willingness to disclose symptoms to others (Amenson & Lewisohn, 1981; King & Buchwald, 1982; Tousignant, Brosseau, & Tremblay, 1987). Although these findings arguing against dismissal of gender differences in distress on the basis of gender based response biases may also pertain to the current research findings, the possibility that response bias affected these results cannot be dismissed. It is possible, for example, that wives overreported incidents of marital tension and/or husbands underreported incidents of marital tension. If this reporting pattern occurred, the gender differences found in reports of marital tension would have been influenced by these reporting biases. It is also possible that when women experienced marital tension, they overreported their distress. Alternatively, husbands may have been distressed when they experienced marital tension, but they may not have reported their distress. The plausibility of this explanation in respect to gender differences in the association between marital tension and distress is questionable given that the underreporting or overreporting of general distress by spouses would have to be contingent on their reports of marital tension. Since spouses did not differ in their mean levels of distress across the diary period, it seems unlikely that their reports of distress would become specifically biased when they were related to marital tension. Spouses' appraisal of threat Contrary to prediction, there were no gender differences in the levels of perceived threat when husbands and wives experienced marital tension. Husbands and wives did not differ significantly in the degree to which their threat appraisals were associated with their  183  reports of marital tension. For both husbands and wives, increases in the extent to which they felt threatened by incidents of marital tension were related to increases in their current distress. However, there was no gender difference in the relationship between threat appraisals and distress. It thus appears that when husbands and wives agreed about the occurrence of marital tension, they tended to similarly appraise the extent to which they felt threatened by the situation. This finding of no difference in spouses' threat appraisals may relate primarily to the fact that these appraisals were made only after husbands and wives had acknowledged that marital tension was a concern. As noted in the previous section, in this research the influence of gender operated at the level of initial reports of marital tension, i.e., wives, compared to their husbands, reported proportionately more instances of marital tension. After perceived marital tension was acknowledged to have occurred by both husbands and wives, however, it appears that gender was no longer related to the appraisal process. That is, the results indicated that once a marital stressor was identified by both spouses as the most serious interpersonal tension experienced on a given day, both spouses felt similarly threatened by the conflict in their relationship. This interpretation is also consistent with the observation that although wives reported substantially more incidents of marital tension, husbands and wives both identified an equivalent percentage of their overall reports of marital tension as being the most serious interpersonal tension reported that particular day. The role of gender thus appears to occur in spouses' initial perceptions of an event (i.e., whether or not an event constitutes "marital tension"), rather than in their appraisals of threat once the event has been identified as marital tension. It is also possible that differences in husbands' and wives' appraisals exist, but that these differences were not captured by the measure of cognitive appraisal used in this study. As described in the Methods section, a revised version of the standard primary appraisal scale developed by Lazarus and his colleagues was used in this study (Folkman  184  & Lazarus, 1980; Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986). Because the Lazarus et al. model of stress and coping was designed to apply to the coping process across situations, including domains such as work, family, and health, the primary appraisal measure was also designed to apply across situations. Thus, previous research using this measure has included a wide range of stressful situations (Folkman & Lazarus, 1980; Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986; Folkman, Lazarus, Gruen, & DeLongis 1986). Gender differences in appraisal have been reported in the extant coping literature using this measure, but these differences related primarily to issues of concern across domains other than marriage (i.e., work and family concerns). Such a broad based measure may not have been sufficiently sensitive to the issues of concern for spouses dealing with marital conflict and therefore it may not have captured the presence of more subtle gender differences in appraisal. Use of an appraisal measure more specific to marital concerns would be less applicable to other contexts and therefore results based upon such a measure would be less applicable to coping in other contexts. When one weighs the advantages and disadvantages of specificity vs. breadth, however, the advantage of a specific measure of marital appraisal may outweigh the disadvantage of such a measure. Development of such a measure should therefore be explored in future research. Another possible explanation for the lack of gender differences in these findings relates to characteristics of this study's sample and design. It is possible that threat appraisals do not differ within a community sample of husbands and wives who are dealing with everyday marital tension. Perhaps gender differences in threat appraisal only emerge when couples are dealing with particularly severe or chronic marital stressors. Although speculative, it is possible, given the pattern of the findings reported here, that more severe conflict could result in gender differences opposite to those predicted in this study. For example, findings in this research demonstrated that husbands' distress increased as their wives' confrontation increased. Given that there was a significant  185  correlation between husbands' appraisals of threat and their wives' relative use of confrontation strategies, husbands could become progressively more threatened as conflict escalates if their wives also become progressively more confrontive relative to their total coping effort. Although speculative, the possibility of gender differences in appraisal emerging in highly conflictual marital interactions suggests that gender differences in threat appraisal of more severe marital stressors should be explored in future research. In addition, this speculation underscores the possibility that the present research findings are more readily generalizable to relatively nondistressed couples dealing with naturally occurring, day-to-day tension in their marriage. These results may be more limited in their applicability to either clinical samples or to couples dealing with particularly severe or chronic marital conflict. Comparisons between husbands' and wives' coping Although it was predicted that husbands and wives would differ in their use of particular types of coping to deal with marital tension, the results indicated no gender differences in spouses' reported coping. Instead, the findings of this research suggest that when both spouses acknowledged the presence of marital tension and attempted to cope with that tension or conflict, the coping process was more homogeneous than much of the literature has suggested. Husbands and wives did not differ from one another in their endorsements of confrontation, compromise, and withdrawal strategies. Evidence of reciprocity in couples' coping Although the type of data collected in this study did not allow for analyses of the sequential patterning of coping, one might speculate that given the similarity of husbands' and wives' reports of coping, reciprocity of coping occurred when these couples were dealing with marital tension. That is, it is possible that one spouse's confrontation elicited confrontation from his or her partner. Similarly, compromise may have elicited  186  compromise, and withdrawal may have elicited withdrawal (cf., Burggraf & Sillars, 1987; Kahn, Coyne, & Margolin, 1985). If, on the other hand, the interpretation of coping reciprocity in this study were based on the relationships found in the data between husbands' and wives' reports of their own use of the various types of coping, we would be led to the conclusion that reciprocity was generally not operating in this study. For both husbands and wives (using both raw scores and relative scores), there was no significant relationship between either their own confrontation and their spouses' confrontation or their own compromise and their spouses' compromise. There was, however, a significant positive relationship between the extent to which respondents and their spouses reported withdrawal (raw score withdrawal). There was also a trend in the relative score data suggesting that increases in a spouse's reported use of withdrawal were associated with increases in their partner's reported use of withdrawal. The preceding interpretation, based on a comparison between husbands' and wives' reports of their own coping, indicates very limited evidence for reciprocity in spouses' coping except perhaps in relation to reported use of withdrawal strategies. Spouses' perceptions of one another's coping may be of more relevance to reciprocity of coping than were the actual behaviors of spouses. That is, it might be expected that spouses would respond to their perceptions of their partners' coping rather than to their partners' reported coping (Fincham Bradbury, & Beach, 1990; Margolin, 1987; Repetti, 1987; Sillars & Scott, 1983; Solomon et al., 1987). When the relationships between respondents' use of the three types of coping strategies and their perceptions of their spouses' use of the same coping strategies were examined, a different interpretation than that given above emerged regarding the possibility of reciprocity in coping. For husbands (using both raw scores and relative scores), there were positive relationships between their reports of their own use of both confrontation and withdrawal and their perceptions of their spouses' use of these two strategies. There was also a positive relationship between the extent to which husbands said that they compromised and the extent to which they  187  perceived that their wives compromised. Similarly, there was a trend toward a positive relationship between husbands' reports of their relative use of compromise and their perceptions of their wives' relative use of compromise. Thus, it appears that, for husbands, their reported use of confrontation, compromise, and withdrawal were related to their perceptions that their wives were also using these same three strategies. As suggested above, these relationships between husbands' coping and their perceptions of their wives' coping may represent perceived reciprocity in coping. That is, husbands may respond to their perceptions of their wives' use of specific coping strategies by using the same type of coping themselves (Burggraf & Sillars, 1987; Roberts & Krokoff, 1990). Another possible interpretation is that husbands' perceptions of their wives' coping were based on husbands' own coping. That is, when husbands were confrontive they perceived their wives as also being confrontive, or when husbands were compromising or withdrawing they also perceived their wives as either compromising or withdrawing. This interpretation is consistent with research that has found a strong tendency for spouses to use their own feelings as a reference for predicting their partners' feelings (Sillars, Pike, Jones, & Murphy, 1984). These two alternative interpretations are equally plausible explanations of the findings in the present study. More time points for each respondent and/or observations of the sequential patterns of couples' interactions during marital conflict are needed to definitively distinguish between these alternative explanations. Although there were no significant differences between the correlations found for husbands and those found for wives, a somewhat different pattern of results emerged when wives' reports of their own coping were compared to their perceptions of their husbands' coping. Considering both raw scores and relative scores, there were no significant relationships between wives' reported use of either confrontation or compromise and their perceptions of their husbands' use of the same two coping strategies. Although there was a trend toward a positive relationship between the extent to which wives reported  188  compromise and the extent to which they perceived their husbands using compromise, the lack of a similar trend in the relative score data suggests caution in interpreting the significance of this relationship. The only significant relationship found between wives' reported coping and their perceptions of their husbands' coping pertained to withdrawal. A significant relationship was found in the raw score data and a trend toward significance was found in the relative data suggesting that wives' reports of their own withdrawal were associated with their perceptions of their husbands' use of withdrawal. It is especially interesting to note that withdrawal was the only type of coping for which significant correlations were found, for both husbands and wives, between their reports of their own coping and their spouses' coping, and between their reports of their own coping and their perceptions of their spouses' coping. Although the literature has suggested that withdrawal may be met with confrontation (see discussion below), it appears more likely in these data that one spouse's withdrawal was met by the other spouse's withdrawal. As noted in the earlier literature review, researchers and clinicians have suggested that withdrawal (typically from the husbands) is often met with confrontation (typically from the wife), especially in distressed marriages (e.g., Christensen, 1988; Gottman & Levenson, 1988; Greenberg & Johnson, 1986; Jacobson, 1989). Examination of the correlations between withdrawal and confrontation (both with and without reference to total coping effort) in the present research revealed, however, no relationship between either husbands' or wives' reports of their own withdrawal and their spouses' reports of confrontation. In addition, no significant relationships were found between respondents' own reports of withdrawal and their perceptions of their spouses' confrontation or between respondents' reports of their own confrontation and their perceptions of their spouses' withdrawal. It thus appears that in this study one spouse's withdrawal was not met with his or her partner's confrontation. Rather than risk escalation of the conflict by reengaging (e.g., confronting) their partner, these spouses may have chosen to withdraw or disengage from the marital argument or conflict. In fact, among the couples in this study,  189  it is possible that withdrawal by one spouse signaled not only disengagement but the end or resolution of the conflict. Thus, withdrawal by one spouse would have been particularly conducive to withdrawal by the other spouse. Before leaving this discussion concerning the relationship between one spouse's coping and his or her partner's coping, another relevant finding should be noted. As mentioned above, no significant relationships were found between respondents' reports of confrontation and either their spouses' reported withdrawal or the respondents' perceptions of their spouses' withdrawal. There was a trend, however, toward a negative relationship in the relative score data between wives' reported confrontation and their perceptions of their husbands' withdrawal. Given that this was only a trend in the data (.05 > p > .01) and no significant relationships or trends between these variables were found in the raw score data, this finding is of limited interpretative value. It is interesting to note, however, that there was a significant gender difference related to this finding. Although this finding must be interpreted cautiously, it suggests the possibility that for wives, compared to their husbands, increases in their own proportional use of confrontation were more likely to be associated with decreases in their perceptions of their husbands' proportional use of withdrawal; whereas for husbands, compared to their wives, increases in their own proportional use of confrontation were more likely to be associated with increases in their perceptions of their wives' proportional use of withdrawal. This finding is thus consistent with other evidence from this study suggesting that there was not a pattern in this study of wives' confrontation being associated with husbands' withdrawal. As noted previously, the foregoing discussion regarding the possibility of reciprocity in spouses' coping is speculative. The data in this study did not allow for definitive analysis of the sequential nature of the coping process. In assessing the viability of the speculations presented above, it is important to note that these interpretations have largely been based on relatively conservative estimates of the statistical significance of the relevant correlation coefficients. Given the large number of correlation coefficients  190  generated in this study, concerns were raised about the possibility of Type I errors at the more standard .05 significance level. Therefore a more stringent criterion was applied (p < .01) to evaluate the statistical significance of the correlations between variables in this study (Cohen, 1992). Given that an increase in the alpha level leads to an increase in the possibility of Type II errors, more subtle, reliable differences may have been missed. As noted previously, in an attempt to address this concern, trends toward significance have been noted. For example, in the preceding discussion correlation values of moderate size (r = .27 - .33) were noted as trends toward significance. Correlations of this magnitude are generally significant in the larger stress and coping literature (e.g., DeLongis, Folkman, & Lazarus, 1988; Folkman, Lazarus, Gruen, & DeLongis, 1986), but they were not statistically significant in this study using an alpha level of .01. Given the possibility of committing a Type II error at the .01 level, the interpretations given here have taken into account trends toward significance (p < .05) especially when they are consistent with other significant findings in this study. Explanations for no gender differences in spouses' coping The finding in this research of no gender differences in spouses' reports of their own coping is in contrast to the gender differences reported in much of the marital conflict research reviewed previously (e.g., Bowman, 1990; Christensen, 1988; Gottman & Levenson, 1988; Pearlin & Schooler, 1978). There are several plausible explanations for the differences between the findings in this study and those of other research. For example, it is possible that the null findings noted above were due to the relatively small sample size that may have resulted in insufficient statistical power. On the other hand, these findings may accurately reflect how a community sample of spouses cope with everyday, naturally occurring marital tension. The belief that husbands and wives differ in how they cope with marital conflict has been based on qualitative interviews with couples (Komarovsky, 1962, 1976; Rubin, 1976, 1983);  191  clinical anecdotes (e.g., Greenberg & Johnson, 1986; Jacobson, 1989; Wile, 1981); observational coding of reenacted conflicts (Christensen & Heavey, 1990; Margolin, Burman, & John, 1989); reports of typical coping with non-specific relationship stressors (Christensen, 1988; Christensen & Shenk, 1991; Rusbult et al., 1986, 1991); and retrospective reports of serious marital conflict from respondents, not paired with their spouses (Bowman, 1990). Generalizing from these studies to cases where both spouses are engaged in mutually acknowledged, naturally occurring day-to-day marital arguments in which both have an equivalent investment (i.e., their threat assessments are similar) may be unjustified. When relatively "normal," heterosexual, North American couples deal with everyday marital stressors within the privacy of their homes, their coping process may be more accurately reflected in the data of this study than in studies based on contrived events or retrospective, general reports of spouses' coping. Thus, within the context of their everyday marital life, husbands and wives may be more similar than they are different in the ways in which they cope with day-to-day marital tension. It is also possible that the gender based demand-withdrawal pattern often reported in the literature may be found primarily with more seriously distressed marital couples (Christensen & Shenk, 1991; Gottman & Levenson, 1988; Levenson & Gottman, 1983). It is possible, for example, that among seriously distressed couples the discrepancies found in this research between husbands and wives in their reports of the occurrence of marital tension could become more exaggerated and entrenched and a source of conflict itself. Thus, wives might increasingly call attention to perceived marital tension which their husbands do not acknowledge and which they increasingly try to avoid or ignore. By the time their problems have escalated to the point of seeking marital therapy, the couple could be entrenched in a pattern of "demand-withdraw" conflict behavior. A related issue is that gender differences in coping with marital conflict may relate to responses to severe or chronic marital stressors. As noted above, research reporting gender differences in spousal coping generally has been based on spouses' responses to  192  continuing, highly rated areas of disagreement on problem inventories (e.g., Gottman & Krokoff, 1989; Levenson & Gottman, 1983, 1985) or on spouses' retrospective accounts of highly conflictual or chronic marital problems (e.g., Bowman, 1990; Margolin, Burman, & John, 1989). Because data in the present study pertained to everyday, naturally occurring marital tension, it is likely that these spouses were coping with relatively minor stressors. It may be the case that gender differences in marital coping, such as those found in the demand-withdrawal pattern, reflect how couples deal with severe and/or chronic marital stressors, but not the everyday problems that occur in marriage. This reinforces the caution against generalizing from research based on clinical samples or on couples dealing with chronic or severe marital stressors to the context of average couples dealing with day-to-day hassles in their marriage. It may also be the case that some of the reports of husbands' greater withdrawal, such as those noted by clinicians (Fogarty, 1976; Jacobson, 1989; Napier, 1978; Wile, 1981) may not reflect husbands' behavior when they actually acknowledge marital tension and engage in a marital argument. Instead, these reports of husbands' greater withdrawal, often relayed by wives seeking marital therapy, may reflect husbands' initial lack of acknowledgment of an episode as constituting marital tension in situations in which their wives perceive marital tension. This type of withdrawal would not generally be captured by standard coping inventories given that these inventories pertain to conscious coping attempts made after an event is appraised as being stressful. If spouses differ in their perceptions of the existence of marital tension, as demonstrated in this research, the spouse who does not as frequently acknowledge marital tension (i.e., the husband) may be perceived by the other spouse (i.e., the wife) as avoidant or withdrawn. It is easy to imagine how this pattern could be reflected in clinicians' offices, with wives complaining that their husbands do not engage with them to deal with marital tension and husbands complaining that their wives bring up problems that are inconsequential or non-existent  193  (Jacobson, 1989). Thus, she could be seen as demanding and he could be seen as withdrawing. This scenario calls attention to some of the complexities of measuring a concept such as withdrawal or avoidance. One issue relates to the stage in the coping process at which coping is measured. Withdrawal measured at the initial appraisal stage, i.e., at the point at which the respondent is assessing whether an event is to be considered as a stressor, could produce different results than withdrawal measured at a later stage, i.e., after an event has been appraised as constituting marital tension. Withdrawal measured at the earlier time point reflects lack of engagement, a process perhaps akin to minimization or denial; whereas withdrawal measured at the later time point reflects disengagement, a process possible only after an event has been appraised as a stressor (e.g., marital tension). Consistent with standard coping inventories, the coping scales in this study solicited coping efforts only after marital events or issues were appraised as representing marital tension. Thus, disengagement rather than lack of engagement was measured by the withdrawal scale employed in this study. The type of withdrawal measured in this study may therefore differ from the type of withdrawal often measured in observational marital research. The more cognitively-based, conciliatory withdrawal strategies examined in this study may not have tapped the dimension of withdrawal relating to lack of engagement or lack of emotional responding that is described in the observational marital literature (Gottman & Krokoff, 1989; Margolin, Burman, & John, 1989; Margolin & Wampold, 1981). Measurement of lack of engagement or denial-like processes via self reports is fraught with difficulties, not the least of which is the inability of respondents to directly report a process about which they may be unaware or which they do not readily acknowledge (Haan, 1982, 1985; Horowitz, 1986; Lazarus, 1985). This problem is also not easily solved by relying on observational research which may be prone to genderbased biases (Deaux, 1984) and may be less effective in coding less observable distancing  194  and withdrawal tactics (Margolin, John, & O'Brien, 1989). In addition, there may be difficulties achieving reliability in coding lack of engagement. For example, in a recent study Bradbury and Fincham (1992) dropped this category of codes from their observational data because the coders could not reliably agree on the degree to which spouses were disengaged vs. engaged in the interaction. It is possible, however, that future self report research such as that reported here could expand our present knowledge of the lack of engagement process within marital couples by designing more specific questions about each spouse's cognitive processes during the initial appraisal stage. In addition, by obtaining more detailed information from each spouse concerning their perceptions of their partner's appraisal and coping process, especially as it relates to engagement or lack of engagement in marital conflict, we could expand our current understanding of these issues. As the foregoing suggests, differences between the findings of this study and other research may relate to differences in the conceptualizations and measurements of coping strategies such as withdrawal. In particular, the coping scales used in this study may reflect multidimensional rather than unidimensional forms of coping. For example, the intercorrelations found between the withdrawal and compromise scales suggest that a conciliatory form of withdrawal may have been measured in this research. A fifth withdrawal item, "avoid talking about the problem," was dropped from the withdrawal scale because of low intercorrelations with other scale items. This withdrawal strategy may have reflected a less conciliatory and more hostile dimension of withdrawal. For example, avoiding talking about a marital problem might signal an antagonistic, noncaring form of withdrawal or a more negatively valenced passive-aggressive avoidance strategy. Explication of conceptually meaningful and distinctively different dimensions of withdrawal should thus be explored in future marital coping research. It is possible that a more explicit conceptualization of withdrawal could clarify some of the inconsistencies in  195  the literature regarding gender differences in husbands' and wives' withdrawal from marital conflict. Given that the coping data were derived from self reports, it is also possible that bias or error in husbands or wives reports of coping could have obscured gender differences in coping. For example, one could interpret the low relationships found between respondents' perceptions of their spouses' coping and their spouses' reports of their own coping (see discussion below) as evidence of errors in coping reports. However, the finding that these relationships did not differ by gender mitigates against the possibility that there were systematic gender biases in spouses' coping reports. Respondents' perceptions of their spouses' coping As expected, results from this study indicated that respondents' perceptions of their spouses' coping tended to differ from their spouses' reports of their own coping. For example, in two of the three comparisons (paired Mests), using both types of scoring, there were significant differences between both husbands' and wives' reports of their spouses' coping and their spouses' reports of their own coping. In addition, although significant relationships or trends toward significant relationships were found between respondents' perceptions of their spouses' coping and their spouses' reports of their own coping, these correlations were in the low to moderate range (r ranged between .13 and .36 for husbands and between .27 and .40 for wives). Perhaps because respondents in this research were reporting primarily about covert cognitive processes or ambiguous behaviors that could be interpreted in a variety of ways, the correlations found between respondents' perceptions of their spouses' coping and their spouses' reports of their own coping were somewhat lower than those correlations (averaging between .47 and .49) reported in other research pertaining to spouses' reports of their partners' overt behavior (Jacobson & Moore, 1981; Repetti, 1989). Research has also suggested that perceptual biases are strongest in emotionally expressive and stressful conflicts (see Sillars & Scott,  196  1983, for a review). Thus, interspouse agreement rates may be lower in this research in which spouses were dealing with naturally occurring, "real" tension than was found in research such as that reported by Christensen (1988) (r ranged from .32 to .73) in which conflicts were described in a more detached, generalized fashion. Consistent with other research (Fichten & Wright, 1983; Guthrie & Noller, 1988; Margolin et al., 1985; Noller & Venardos, 1986), no gender differences were found in the relationships between respondents' perceptions of their spouses' coping and the spouses' own reports. Although this might lead to the assumption that husbands and wives are equally perceptive (or non-perceptive) in regard to one another's coping, one cannot make this assumption from these data given that the relative accuracy of husbands' and wives' reports of their own coping cannot be assessed. That is, because there is no objective standard by which to measure coping reports, it cannot be determined whether spouses' own coping reports or their partners' perceptions of their coping are more veridical. What we can determine from these data, however, is that the "match" between respondents' perceptions of their spouses' coping and spouses' reports of their own coping is sufficiently low as to suggest that reliance on respondents' reports of their spouses' coping would lead to a different picture of the coping process than would reliance on spouses' reports of their own coping efforts. For example, in this study, wives' perceptions of their husbands' withdrawal was the only type of coping report that was not significantly different from the spouses' report of their own coping when both types of scoring are considered. Given that marital tension or conflict assumes disagreement, it is not surprising that both husbands and wives in this study perceived that their spouses responded differently and perhaps less favorably than respondents reported that they themselves responded or than their spouses reported that they themselves responded. Although the significance of the results differed slightly according to whether raw scores or relative scores were employed, the following general pattern emerged. No significant differences  197  were found between husbands' and wives' reports of their own use of the three types of coping, however, husbands and wives generally tended to perceive that their spouses were more confrontational, less compromising, and less withdrawing than were the respondents themselves. Husbands also generally tended to perceive that their wives were more confrontational, less compromising, and less withdrawing than their wives themselves had reported. Wives similarly tended to perceive that their husbands were more confrontational and less compromising than their husbands themselves had reported, but wives did not perceive their husbands' withdrawal as being significantly different from that reported by the husbands themselves. Given that the confrontation strategies used in this research have been defined as aggressive attempts to alter a situation (Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986) and given the conciliatory nature of the compromise and withdrawal scales used in this research, it appears that both husbands and wives tended to perceive themselves more in the role of peacemaker than they perceived their spouses in this role. Considered together, these comparisons between respondents' perceptions of their spouses' coping (confrontation, compromise, and withdrawal) and respondents' own reported use of the three coping strategies could thus be considered to be consistent with other research indicating that spouses tend to perceive their own behavior during marital conflict more favorably than that of their partners (Guthrie & Noller, 1988; Sillars & Scott, 1983). The lack of gender differences in these findings also suggests that this bias toward viewing one's own behavior during marital conflict more positively than one's spouse's behavior exists for both husbands and wives. The foregoing discussion has indicated that respondents perceived that their spouses dealt with marital tension differently (possibly less favorably) than did the respondents themselves. In addition to the issues noted above that lend support to this conclusion, another research finding is relevant to an understanding of the differences between respondents' views of their own coping compared to their perceptions of their spouses' coping. For husbands, the associations (i.e., correlations) found between the  198  withdrawal and compromise scales differed significantly depending on whether the scales referred to husbands' reports of their own coping or to their perceptions of their wives' coping. There was no significant relationship between the reported extent of husbands' own use of compromise and their own use of withdrawal. However, there was a significant positive relationship between the extent to which husbands perceived that their wives used compromise and withdrawal strategies. Thus, husbands may have viewed withdrawal as a more conciliatory strategy when they perceived the extent to which their wives withdrew than when they (i.e., husbands) reported the extent to which they themselves withdrew. A different pattern of results emerged, however, when the total amount of coping effort was taken into account. There was a significant negative relationship between husbands' reports of their own relative use of withdrawal and compromise, but there was no relationship between husbands' perception of their spouses' relative use of withdrawal and compromise. In respect to the present discussion, these results indicate that husbands apparently conceptualized withdrawal and compromise somewhat differently depending on whether they were reporting their own coping or their perceptions of their wives' coping. In addition, these results indicate that raw scores and relative scores conveyed significantly different information about the relationships between coping scales. For example, the extent to which husbands reported compromise was unrelated to the extent to which they reported withdrawal (raw score method). However, as the proportion of husbands' reported use of compromise increased, compared to their total coping effort, the proportion of their reported withdrawal decreased (relative score method). Similarly, using the raw score method, the extent to which wives reported compromise was unrelated to the extent to which they reported confrontation. However, using the relative score method, it emerged that as the proportion of wives' reported use of compromise increased, compared to their total coping effort, the proportion of their reported confrontation decreased. Although the mathematical dependence of the relative scores suggests caution  199  in interpreting these intercorrelations, comparisons between the two types of scoring are of interest and together they give a fuller picture of the relationship between coping scales than would either result alone. Relationship between coping and distress During episodes in which husbands and wives both appraised the situation as one involving marital tension, they did not differ in their reported use of coping strategies. They did differ, however, in the relationship between their coping and their mood. Gender differences in the relationships between coping and distress varied as a function of the type of coping (confrontation, compromise, or withdrawal) reported by either the respondents themselves or their spouses. To facilitate the discussion of the relationship of the three coping scales to respondents' distress, the study's findings are discussed below as they related to the respondents' reports of their own coping, respondents' perceptions of their spouses' coping, and their spouses' reports of their own coping. Self Coping Consistent with prediction, husbands' reports of their own coping explained a significant proportion of the variance in their mood. Multiple regression analyses indicated that increases in the extent to which husbands reported using confrontation were significantly associated with increases in husbands' distress. Similarly, these analyses indicated that as the proportion of husbands' reported confrontation relative to their total coping effort increased, their distress increased. It was also notable that husbands' reports of confrontation predicted substantially more of the variance in husbands' current distress (twice as much in the raw score analyses and three times as much in the relative score analyses) than was predicted by husbands' distress on the previous day. These results are consistent with findings in the stress and coping literature indicating that respondents' confrontation is generally associated with negative outcomes such as psychological distress  200  (Folkman & Lazarus, 1988; Folkman, Lazarus, Gruen, & DeLongis, 1986). Although the correlational nature of these data make it difficult to tease apart causal relations, by controlling for prior levels of distress in the regression analyses, these analyses indicated that the previous day's distress did not account for the relationship found between husbands' reported use of confrontation and husbands' increased subsequent distress. Thus, it appears that change in distress level between two days was, to some degree, predicted by the extent to which husbands reported confronting their wives during marital arguments. Although a significant relationship between husbands' reported confrontation and distress was found in the multiple regression analysis using both raw and relative scoring, the relationships between these variables were not significant in the bivariate analyses. The relationship between husbands' relative score confrontation and distress (r = .33, p < .05) was not significant at the .01 alpha level, but there was a trend toward a positive relationship between these variables that is in keeping with the findings of the multivariate analysis. However, the low correlation found between husbands' raw score confrontation and distress (r = .17, p > .10) was not in keeping with the findings of the raw score regression analysis. Comparison of the raw score correlation coefficients and regression betas suggests that husbands' prior level of distress may have operated as a suppressor variable (Pedhazur, 1982). That is, when the variance in current distress due to prior distress was removed in this regression analysis, husbands' reported confrontation predicted a significant amount of the variability in their current distress. Regarding the relationship between husbands' reported withdrawal and distress, multiple regression analyses and bivariate analyses indicated that the extent to which husbands reportedly withdrew from marital tension (raw score withdrawal) was significantly associated with decreases in their distress. Similarly, the multiple regression analyses indicated that as the proportion of husbands' reported withdrawal relative to their total coping effort increased, their distress decreased. Although this latter relationship was  201  not significant at the .01 level in the bivariate analysis, there was a trend toward a significant negative relationship (p < .05) between these variables. The results from these analyses are consistent with research suggesting that males may be particularly physiologically reactive to stress (Gottman & Levenson, 1988; Stoney et al., 1987) and therefore they tend to experience relief when the source of tension is removed or avoided (Gottman & Levenson, 1988; Levenson & Gottman, 1983, 1985). In addition, these results are consistent with evidence presented by NolenHoeksema (1987) suggesting that males experience less depression than females because males relieve their levels of distress by actively seeking to distract or remove themselves from stressful situations. The finding in this study that husbands and wives did not differ in either the frequency or the proportion with which they reported using withdrawal strategies does not appear to be consistent with Nolen-Hoeksema (1987). However, the finding that husbands' reports of withdrawal were associated with decreases in their distress, whereas wives' reports of withdrawal were not associated with changes in their distress level, is in keeping with Nolen-Hoeksema's (1987) suggestion that males' attempts to distract themselves from stress-provoking events result in reduced distress for males compared to females. As noted above, the correlational nature of these data place restrictions on drawing causal inferences. However, given that prior levels of distress were controlled in the regression analyses, these analyses indicated that the previous day's distress did not account for the relationship between husbands' reported use of withdrawal and their decreased subsequent distress. Thus, it appears that change in distress level between two days was, in part, predicted by the extent (and the proportion of their total coping effort) to which husbands reported withdrawing from their wives during marital arguments. Although husbands' reports of confrontation and withdrawal were both significantly associated with their own distress, contrary to prediction husbands' reported compromise was not significantly related to their distress in either the multivariate or  202  bivariate analyses. Although there was a trend toward a significant negative relationship in the bivariate analyses between wives' reported compromise and their distress, wives' reported compromise also was not significantly related to their distress in the bivariate or multivariate analyses. Thus, even though it was predicted that compromise would be associated with decreased distress for both husbands and wives, the results indicated no significant relationship between reports of compromise and their distress for either spouse. That is, reports of compromise were not associated with either alleviations or exacerbations of the distress experienced by husbands and wives on days in which they reported marital tension. Compromise (e.g., trying to meet the other person half-way) calls for collaboration in order to be successful (Sillars et al., 1982; Zietlow & Sillars, 1988). Given that there was no significant correlation between husbands' and wives' reported use of compromise in this study, this lack of reciprocation between spouses may explain why compromise did not have the expected positive impact on spouses' mood. As noted above, wives' reported use of compromise was not significantly related to their distress, but the bivariate analyses indicated a trend toward the predicted relationship. That is, wives' reported compromise was negatively correlated with wives' distress at the .05 alpha level (raw score r = -.27; relative score r = -.28, p < .05). This relationship did not reach significance in these data; however, a similar relationship was found in the bivariates using both types of scoring and the direction of the multivariate betas was also consistent with a negative relationship between compromise and distress. It is possible, therefore, that with a larger sample the predicted relationship (i.e., that increases in wives' compromise would be associated with decreases in their distress) might obtain significance. Future research should explore this possibility. Regarding the relationships between wives' reports of other forms of coping and their distress, in contrast to their husbands and contrary to prediction, wives' reported use of the other two types of coping also did not predict their mood. The multiple regression analyses and the bivariate analyses both revealed that wives' reported use of both  203  confrontation and withdrawal to deal with marital tension was not related to either increases or decreases in their reported distress. These data thus suggest that once wives became distressed by marital tension, their own attempts to deal with the marital problem were not significantly related to their distress level. Although these findings were counter to expectations, there are several plausible interpretations of these results. It is possible, for example, that although wives endorsed coping strategies at a rate and a proportion equivalent to their husbands, wives' coping may have differed in the intentions and meanings that wives ascribed to their own actions or in the cognitive process in which wives were engaged as they performed their coping behaviors. Thus, consistent with evidence presented by Nolen-Hoeksema (1987), it is possible that wives' initial distress upon appraisal of marital tension led to an increased tendency to ruminate about the tension in a way that interfered with the efficacy of their coping process. Alternatively, wives may have interpreted the meaning in their use of the various coping strategies in a way that corresponded to their affective responses. Consider, for example, the finding that wives' reported confrontation was not associated with increased distress. Empirical research in the stress and coping literature has indicated that confrontation is generally associated with increased distress (Folkman & Lazarus, 1988; Folkman, Lazarus, Gruen, & DeLongis, 1986), and therefore it was hypothesized in this study that wives, like their husbands, would experience increased distress when they reported employing confrontation strategies. It is possible, however, that confrontation may have been interpreted and experienced in a less negative way by wives in the marital context. Research reviewed in the introduction suggested, for example, that wives tend to be more involved in and committed to their intimate relationships and therefore they are often more willing to accommodate the development and maintenance of their relationships than are their male partners (Huston et al., 1981; Markman et al., 1982; Rubin et al., 1981). In addition, it was noted that wives more often express dissatisfaction in their marriage and seek change in their relationships (Hagestad & Smyer, 1982; Margolin et al., 1983).  204  Wives may therefore perceive their confrontation as a necessary means toward resolving relationship problems and achieving desired changes in their relationship. Thus, wives' reports of their use of confrontation may not be associated with their distress. Given that the literature has suggested that wives, compared to husbands, tend to be more concerned about the resolution of marital problems (Gottman, 1979; Margolin et al., 1983), it was expected that wives would experience less relief related to their reported withdrawal from marital conflict than would their husbands. Although there were no significant effects related to wives' withdrawal, the expectation of a gender difference in the relationship between withdrawal and distress received some support in this study. In both the bivariate analysis and the multiple regression analysis (using the relative score method), a significant (or marginally significant) gender difference emerged to indicate that for husbands their reported proportional use of withdrawal was more likely to be associated with decreases in their distress than was the case for wives. A significant gender difference was not found, however, when comparisons were made between the extent to which husbands' and wives' reports of withdrawal, without reference to their total coping effort (raw score method), were related to their reported distress. Thus, there are limitations to the interpretation that husbands, compared to their wives, experienced significantly more relief in association with their withdrawal from marital conflict. It is also possible that wives' reports of confrontation and withdrawal might be related to their distress when marital conflict is particularly severe or chronic. In these type of incidents, confrontation might become more pronounced or prolonged and thereby more distressing. Furthermore, it is possible in such circumstances that withdrawal might have a different meaning or consequence for wives. In severe conflicts withdrawal might be associated with higher levels of frustration and/or less possibility for resolution of the conflict. In these circumstances, wives' withdrawal might be associated with increased distress. Alternatively, if severe stressors result in increases in wives' withdrawal relative to their total coping effort (see following discussion), wives' levels of distress might  205  increase. Although this interpretation is based on speculation rather than data, it suggests the possibility of a different pattern of results with more distressed spouses and/or in situations involving severe or chronic marital distress. Given this possibility, caution should be exercised in generalizing from clinical samples or from spouses who are dealing with chronic or severe marital problems to non-clinical samples of spouses dealing with less severe, everyday stressors, such as was the case in this study. Future research should explore the possibility that wives' attempts to cope with marital conflict might have significant effects on their distress within clinical samples or when couples are dealing with more severe marital distress. The results noted above also indicate once again that use of the raw and relative scoring methods results in different perspectives on the coping data. A closer look at these data illustrates the degree to which the two scoring methods inform our understanding of the results found in this study. The correlation between wives' distress and their reports of their relative use of withdrawal, taking into account their total coping effort (relative score withdrawal), was not significant. However, there was a trend toward a positive relationship in these data suggesting the possibility that as wives' reports of their relative use of withdrawal increased, their distress might also tend to increase. When withdrawal was considered only in terms of the extent to which withdrawal itself was reported (raw score withdrawal), wives' reported withdrawal was not only not associated with distress, but the sign of the correlation was negative and the correlation coefficient was significantly different from that obtained using the relative scores. Taken together these results suggest that the relationship between wives' distress and their own reported use of withdrawal differed according to whether or not withdrawal was considered relative to total coping effort. The extent to which wives reported withdrawing from marital conflict clearly was not associated with increased distress. However, as wives' reported withdrawal became proportionately larger in relation to their combined coping effort (i.e., their total reported use of all three coping strategies), there was a significant shift toward a  206  tendency for wives to feel more distressed. Given that neither of these correlation coefficients was significant on its own, the relevance of these relationships should not be overinterpreted. The significant difference between the two correlation coefficients does present the intriguing possibility that wives experienced more or less distress depending on the relationship between their use of withdrawal and their total coping effort. These findings reinforce the advantage of using both types of coping scores in future research pertaining to marital coping. Spouses' Coping Although it was predicted that wives' distress, given wives' greater relationship focus, would be more closely related to their spouses' coping than would their husbands' distress, this expectation was not supported by the results. Instead, significant gender differences emerged suggesting that husbands' distress was more closely related to their spouses' coping (i.e., confrontation and compromise) than was wives' distress. These results are discussed below, along with findings pertaining to the relationship between spouses' withdrawal and respondents' distress. In the multiple regression analyses, increases in the extent of wives' reports of confrontation were significantly associated with increases in their husbands' distress. Similarly, the multiple regression analyses indicated that the proportion of wives' reported confrontation relative to their total coping effort also was significantly associated with increases in their husbands' distress. According to these analyses, the relationship between reports of wives' confrontation and husbands' distress remained significant even after controlling for both the previous day's level of distress and the effects of husbands' own reported use of confrontation. These analyses thus indicate that wives' reported confrontation predicted increases in their husbands' distress over and above that predicted by husbands' prior distress and husbands' own reports of confrontation. These results,  207  therefore, are consistent with the notion that increases in husbands' distress were, in part, predicted by their wives' reported use of confrontation during marital arguments. Although there was a trend toward similar relationships between husbands' distress and their wives' reported confrontation in the bivariate analyses, the correlation coefficients did not reach significance at the .01 level. In this case the discrepancy between the bivariate and multivariate analyses does not appear to be due to suppressor effects. Instead given the trend toward significance in the bivariates (p < .05), the lack of significance is likely related to the relatively low power in these analyses (e.g., small sample size and conservative alpha level due to the large number of bivariate comparisons). There were no significant relationships or trends toward significance in either the bivariate or multivariate analyses between husbands' reports of confrontation (either raw score or relative score) and wives' mood. Taken together, these results indicate that husbands' distress increased as the extent of their wives' reported confrontation increased or as the proportion of their wives' reported confrontation to their total coping effort increased. Wives' distress, on the other hand, was not significantly related to their husbands' reported confrontation. When these relationships in the husbands' and wives' data were compared, a significant gender difference emerged indicating that husbands' distress was more likely to increase in relation to their spouses' reported confrontation than was the case for wives' distress. This finding is consistent with the notion that husbands may feel more threatened by their wives' use of confrontation (Jacobson, 1989). If males have a stronger physiological reaction to emotional stimuli than do females (Gottman & Levenson, 1988; Stoney et al., 1987), the association between husbands' increased distress and their wives' reports of increased confrontation may represent a heightened response to the emotionality or aggressiveness of their wives' confrontation. In addition, there is evidence to suggest that assertive women, by not adhering to the sex role stereotype of congeniality or  208  passivity, may be viewed more negatively by others than are assertive men (Basow, 1992; Gervasio & Crawford, 1989; Kelly et al., 1980). Assertiveness also may be perceived differently in different contexts. For example, in the business management context, use of aggressive influence strategies (frequently attributed to males) has been described as "direct and assertive" (Howard, Blumstein, & Schwartz, 1986), whereas in the marital conflict literature, use of aggressive confrontive strategies (frequently attributed to females) has been described as "nagging and demanding" (Christensen, 1988). A negative bias toward assertiveness in females could heighten for husbands the negativity of their wives' confrontation, and given husbands' greater reactivity to emotionality, this could result in increased distress in association with spouses' confrontation for husbands compared to their wives. Regardless of the interpretation of why husbands' distress increased in association with their wives' confrontation, this finding demonstrates that spouses' emotional response to marital conflict was associated with their partners' coping. This finding is in keeping with recent evidence suggesting that the effects and efficacy of a given coping strategy depend heavily upon the response of involved others (Coyne & DeLongis, 1986; DeLongis & O'Brien, 1990; Kahn et al., 1985). Without taking into account the effects of one spouse's coping on the other spouse, we would have an incomplete picture of the efficacy of coping strategies. Future research on marital coping should therefore include analyses of these interspousal effects. In contrast to their husbands, wives' distress was not related to their spouses' reports of confrontation. The finding that wives' distress was significantly less related to their spouses' reported confrontation than was their husbands' distress is suggestive of Gottman's observation that "in a sea of conflict, men sink and women swim" (quoted in Markman, 1991b, p.422). As discussed previously, support was not found for the expectation that wives would report being more confrontational than would their husbands. However, the finding that wives' distress, in contrast to their husbands'  209  distress, was not significantly associated with either their own or their spouses' confrontation suggests that wives deal with confrontation about marital issues differently than do their husbands. This difference may relate to wives having a relatively positive perception of confrontation. As noted above, if wives are invested in resolving relationship problems and if they perceive confrontation as a necessary means for achieving this goal, they may experience confrontation as less stressful than do their husbands. On the other hand, the finding that wives' distress was significantly related to their perceptions of their husbands' confrontation (discussed below) mitigates against this interpretation. An alternative explanation is that husbands and wives may actually behave differently when they are confronting their spouse. Folkman, Lazarus, Dunkel-Schetter, DeLongis, and Gruen (1986) described the confrontive coping scale, from which most of the confrontation items used in this research were drawn, as "aggressive efforts to alter the situation" (Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986, p. 995). Although aggressiveness is implied in these strategies (e.g., stand your ground and fight for what you want, confront the other person), the degree of hostility employed could vary. For example, one can "stand their ground and fight for what they want" in a firm, assertive, non-hostile way or in a belligerent, angry, hostile way. Distinguishing the degree of hostility perceived, delivered, and received via these strategies could perhaps clarify some of the differences found in the effective use of these strategies. Despite the lack of significant findings pertaining to the relationship between compromise and respondents' distress, a significant gender difference was found regarding the association between respondents' spouses' reports of compromise (considered relative to spouses' total coping effort) and respondents' distress. This difference suggested a greater likelihood that husbands, compared to wives, would experience less distress in association with their spouses' reports of increased use of compromise. Why husbands would be more positively affected by their spouses' reported compromise than would  210  wives is not readily apparent. For example, there was no difference in the extent to which spouses reported compromise, and both husbands and wives perceived that their spouses compromised less than they themselves. Consistent with findings in the literature that suggest that females, compared to males, are more empathic and more effective in their support attempts (Davis & Oathoat, 1987; Schuster et al., 1990; Wheeler et al., 1983), along with findings that suggest that males, compared to females, tend to be more competitive and less cooperative (Ahlgren & Johnson, 1979; Basow, 1992; Doyle, 1989), it is possible that wives are more likely than their husbands to compromise in a supportive or distress relieving fashion. Thus, there may be gender differences in the messages or intentions conveyed by husbands and wives when they compromised. For example, compromise delivered or perceived as a begrudging response might have a different effect than compromise delivered more readily and positively. A slightly different, more speculative, but related interpretation is that the difference between husbands and wives lies not in their ways of delivering compromise, but in the different meanings that they attach to their spouses' compromise. Given that compromise was apparently not reciprocal in this study, i.e., spouses' reports of compromise were not associated with their partners' reports of compromise, when a spouse compromised this could have been interpreted by his or her partner as a victory of sorts. If husbands tend to be more competitively oriented, this might lead them to respond more positively to their spouses' compromise. Regarding the relationship between spouses' reports of withdrawal and their partners' distress, no significant gender differences were found in the analyses. For husbands, there were no significant relationships between their wives' reported withdrawal and husbands' distress in either the bivariate or the multiple regression analyses. The regression analyses did indicate, however, that the extent to which husbands reportedly withdrew from marital conflict (raw score withdrawal) was significantly related to their wives' distress. According to these analyses, husbands' reported withdrawal predicted  211  decreases in their wives' distress over and above that predicted by wives' prior distress and wives' reports of their own withdrawal. These results, therefore, are consistent with the notion that decreases in wives' distress were, in part, a function of their husbands' reported use of withdrawal during marital arguments. In interpreting this finding, it is important to note that neither wives' distress nor husbands' distress was related to their spouses' reported withdrawal in the regression analyses in which withdrawal was considered relative to the spouses' total coping effort (relative score withdrawal). Wives' distress did not decrease (or increase) in association with increases in their husbands' reports of withdrawal considered relative to their total coping effort. That is, wives experienced no relief when their husbands' withdrawal represented a greater proportion of their coping attempts. In addition, it should be noted that none of these relationships was significant in the bivariate analyses. It is possible that, similar to the previously discussed analyses relating to confrontation and distress, a suppressor variable was affecting the raw score bivariate results (Pedhazur, 1982). Comparison of the correlation coefficients and the regression betas in the raw score analysis suggests that wives' withdrawal may have operated as a suppressor variable. Thus, it appears that when the variance in wives' current distress due to wives' reported withdrawal was removed in the raw score regression analysis, spouses' reported withdrawal predicted a significant amount of the variance in wives' current distress. The significant negative relationship found between wives' distress and the extent to which their husbands reported withdrawing from marital conflict is counter to the suggestion that due to wives' relationship investment and their desire to confront and deal with marital concerns, wives would become more distressed as their husbands withdrew from marital conflict. Instead, this finding indicates that wives experienced less, not more, distress in association with increases in the extent to which their husbands reported withdrawing. Returning to an issue raised previously, it is possible that different types or dimensions of withdrawal (e.g., conciliatory withdrawal and hostile withdrawal) might  212  affect the relationship between spouses' withdrawal and their partners' mood. For example, if the withdrawal strategies used by husbands were somewhat conciliatory in tone, their use might have a more ameliorative effect on their wives' distress than if the withdrawal strategies were more hostile in tone. Spouses' use of hostile or angry withdrawal strategies (e.g., either passive-aggressive avoidance techniques or angry refusals to discuss the conflictual issue) might result in decreased personal distress, but increased distress for their partners. On the other hand, wives may experience less distress in relation to the extent to which their husbands withdraw if the husbands' withdrawal signals resolution of the conflict, which might particularly be the case with relatively minor marital tension. As noted above, it was found that none of the wives' own reports of coping was significantly related to their distress. It was suggested that it may be the case for wives that the marital problem itself is what affects their distress and nothing short of resolution of the problem will serve to reduce their distress. The finding that only their husbands' withdrawal was associated with decreases in wives' distress is consistent with this interpretation. Perceived Spouses' Coping Although no specific hypotheses were advanced concerning the relationship between perceived spouses' coping and respondents' distress in general, it was expected that respondents' perceptions of their spouses' coping might be related to respondents' levels of distress. As noted previously, there is considerable evidence that individuals' perceptions affect their behavior and emotional responses toward others (Fincham, Bradbury, & Beach, 1990; Margolin, 1987; Repetti, 1987; Sillars & Scott, 1983; Solomon et al., 1987). In particular, spouses' interactions often appear to be colored by their perceptions of their partners' behavior more so than by "objective" evaluations of their behavior (Bradbury & Fincham, 1990, 1992; Robins, 1990). Given that spouses' reports of their coping were related to their partners' distress, as discussed above,  213  significant relationships between respondents' distress and their perceptions of their spouses' coping were also expected. In particular, given that some of the analyses indicated that spouses' reports of confrontation were associated with husbands' distress and that spouses' reports of withdrawal were associated with wives' distress, it would be expected that perceived spouses' coping pertaining to confrontation and withdrawal would be related to either husbands' or wives' distress. However, none of the bivariate analyses relating to the relationship between perceived spouses' coping and respondents' distress was significant. As was noted previously in regard to correlations between spouses' coping scores, the statistical significance of these correlation coefficients was based on conservative criteria (p < .01) and therefore trends toward significance (p < .05) are discussed when the pattern of results is consistent with other research findings. There was a trend toward a positive relationship between perceived spouses' confrontation and respondents' distress in both the raw score and relative score bivariates for wives (r = .30, p < .05) and in the relative score bivariates for husbands (r = .30, p < .05). Given that some support for a positive relationship between perceived spouses' confrontation and respondents' distress was also found in some of the multivariate analyses (see discussion below), one's confidence in the reliability of these relationships is strengthened. However, a trend toward a significant negative relationship found in the relative score bivariate analysis for husbands between perceived spouses' withdrawal and husbands' distress (r = -.29, p < .05) was not found in the raw score bivariate analysis or in either of the multiple regression analyses (see discussion below). The multiple regression analyses that tested the relationship between respondents' distress and their perceptions of their spouses' coping, after having controlled for prior distress, respondents' coping, and spouses' coping, indicated that perceived spouses' coping explained a significant proportion of the variance in only one regression equation. This equation pertained to the raw score relationship between wives' distress and their  214  perceptions of their husbands' confrontation. According to this analysis, increases in wives' perceptions of their husbands' confrontation predicted increases in wives' distress over and above that predicted by wives' prior distress, wives' own reported confrontation, and husbands' reports of their own confrontation. These results, therefore, are consistent with the notion that increases in wives' distress were, in part, a function of their perceptions of their husbands' use of confrontation during marital arguments. Given that no significant relationship was found between wives' distress and their husbands' own reports of confrontation, this pattern of results indicates that wives were more affected by their perceptions of their husbands' confrontation than by their husbands' reports of their own confrontational behavior. This is in keeping with the notion that one's emotional responses are more associated with one's perceptions of others' behavior than with others' actual behavior (Fincham, Bradbury, & Beach, 1990; Lazarus, 1990; Margolin, 1987; Repetti, 1987; Solomon et al., 1987). The analysis reported above represents a particularly stringent test of the predictive power of perceived spouses' coping. The effect of perceived spouses' coping in this analysis was over and above the effects of prior distress, respondents' own coping, and spouses' coping. Given that in theory spouses' coping affects distress via perceptions (Bradbury & Fincham, 1990; Fincham, Bradbury, & Beach, 1990), the regression analyses were also conducted reversing steps three and four. These analyses were conducted for each type of coping, using both raw scores and relative scores. Significant effects pertaining to the relationship between perceived spouses' coping and distress were found in one of these analyses. In this analysis, husbands' perceptions of their wives' confrontation (using raw scores) accounted for a significant proportion of the variance in husbands' distress, over and above the variance accounted for by husbands' prior distress and husbands' own confrontation. The wives' raw score reports of their own confrontation, however, did not contribute additional predictive power regarding the variance in husbands' distress (as had been the case when this variable was entered in step  215  three of the equation). When the two separate raw score analyses of husbands' data (i.e., the two orderings of perceived spouses' confrontation and spouses' confrontation) are interpreted together, the results indicate that wives' reported coping was not significantly related to husbands' distress, once husbands' perceptions of their wives' coping were taken into account. Future research involving path analysis of the relationships among these variables should be conducted to explore the possibility that the influence of spouses' confrontation on distress is via respondents' perceptions of their spouses' coping. Given that neither self compromise nor spouses' compromise was significantly related to either husbands' or wives' mood, it is perhaps not surprising that no significant relationship was found between perceived spouses' compromise and spouses' mood in any of the bivariate or multivariate analyses. This set of strategies as employed by spouses in this study apparently had little effect on either spouses' mood. There were also no significant relationships between perceived spouses' withdrawal and spouses' mood. Given that the extent to which husbands withdrew was a significant predictor of wives' mood, it was expected that wives' perceptions of the extent to which their husbands withdrew would also predict wives' distress, but no such relationship was found. One explanation of how respondents' mood could be related to their spouses' behavior but not to their perceptions of their spouses' behavior relates to the salience that different behaviors might have for spouses when they report their own coping and their perceptions of their partners' coping. For example, a husband may report having avoided tension in the relationship because he left the house in order to avoid an angry exchange with his wife. His wife, on the other hand, may not have been aware of this behavior as a form of coping and she may therefore not have coded this withdrawal strategy even though she may have benefited (i.e., experienced decreased distress) from the behavior. Instead, the most salient behavior for her may have been an angry remark her husband made before leaving the house, which she coded as confrontation. This scenario could result in the wife feeling relief associated with her husbands' report of withdrawal, whereas her  216  perceptions of his coping, which differed from his own coping reports, may not have been associated with her distress. The Issue of Causality In the preceding discussion pertaining to the relationship between coping and distress, much of the focus was on the potential effect that various types of coping may have on respondents' mood. Because the analyses conducted in this study were correlational in nature, caution must be exercised in interpreting causality. These concerns have been reduced to some extent in this study by the use of hierarchical multiple regression analyses. These analyses statistically removed the effect of an obvious third variable, respondents' prior day's mood rating. Although this feature of this study puts the discussion of the relationship between coping and distress on firmer ground, the obvious limitations of non-experimental methods remain. Conclusion and Future Directions This dissertation has focused on the exploration of gender differences in a community sample of spouses dealing with everyday, naturally occurring marital tension. Given that the importance of gender for coping with marital conflict has been widely cited in both clinical and research settings and given that gender is "one of the earliest and most central components of the self-concept, (serving) as an organizing principle through which many experiences and perceptions of self and other are filtered" (Spence, 1984, p.64), it was hypothesized in this research that gender differences would emerge in the context of spouses dealing with marital tension. The findings indicate that gender differences do exist within the domain of naturally occurring marital tension in the extent to which husbands and wives report marital tension and experience distress as a result of marital tension. Although differences exist in how spouses perceive one another's coping, no gender differences were found in respondents' perceptions of their spouses' coping.  217  Furthermore, no gender differences were found in spouses' reports of either their appraisal of marital tension or their reported use of specific coping strategies to deal with marital tension. Regarding the ways in which husbands' and wives' reports of their own coping and their spouses' reports of coping relate to their mood, gender differences emerged indicating that both respondents' own coping reports and their spouses' coping reports were more predictive of husbands' mood than of wives' mood. By exploring how relatively nondistressed couples deal with day-to-day, naturally occurring marital tension, this research has provided information about gender differences in dealing with marital conflict that was not previously available. This research has provided an initial exploration of gender differences related to how a community sample of couples deal with everyday marital stressors in the real world. By studying real world marital tension, this research has examined couples' reports of coping with naturally occurring marital tension rather than laboratory generated, reenacted incidents of marital tension. This study therefore avoids the types of distortion that are likely to occur when couples are role playing marital tension instead of reporting on their actual experiences of marital tension. In addition, by studying a sample of community residing couples rather than a clinical sample of maritally distressed couples, this research allows an exploration of how relatively well functioning couples deal with stressors within their marriage. Gathering data from these couples provides information about more functional ways of marital coping that may be useful for clinicians working with distressed couples who cope less effectively. This focus on both naturally occurring marital stressors and community residing couples has been lacking in previous research. Although the data from this study represent an important first step toward our understanding of how community couples deal with marital tension, the couples in this study represent a primarily white, English-speaking, Judeo-Christian, heterosexual sample. This sample therefore does not accurately reflect the increasing cultural diversity of North American society. Spouses from other cultural backgrounds may have different  218  attitudes toward marriage or toward the role of husbands and wives in marriage that affect their coping with marital tension along gender lines. Research conducted by Oggins, Veroff, and Lever (1993) comparing perceptions of marital interactions by Black American and White American newlyweds found differences between these two racial groups. For example, these researchers found a gender difference in spouses' reports of self disclosure (husbands reported less disclosure than their wives) among the white couples, but not among the black couples (Oggins et al., 1993). Other cultural differences such as attitudes toward women within the Muslim culture may also result in gender differences in spouses' coping that are different from those found in this research with couples from a Judeo-Christian background. In order to determine the generalizability of the findings of the research reported in this dissertation to couples from other ethnic and cultural backgrounds, future research should therefore include identifiable cultural subgroups such as Black Americans, Asian Americans, Hispanics, and Muslims. To explore the possibility that a different pattern of gender differences might emerge with more maritally distressed couples, as a next step, the methodology employed in this research (i.e., collection of daily, structured diary data pertaining to naturally occurring marital tension) should be applied to a clinical sample of distressed couples. In addition, questions could be added to the diary to distinguish the severity and chronicity of daily marital stressors. Comparisons could then be made between the responses of spouses from a clinical sample and a community sample relating to how they have dealt with naturally occurring marital stressors of varying degrees of severity. Although recruitment of a sufficiently large clinical sample poses difficulties, this problem may be offset to some extent by the likelihood that these spouses would report more days on which a marital argument occurred. In the current study, the relatively small number of days in which both members of a couple reported marital tension disallowed analysis of spouses' coping across several incidents of marital tension. With a clinical sample, this  219  type of analysis, which allows for the study of coping within couples over time, would be more feasible. The findings from the present study also suggest possible refinements of the measures of appraisal and coping used in this research that may enhance future explorations of gender differences in spousal coping. For example, development of an appraisal measure specific to the concerns of husbands' and wives' evaluations of marital tension might reveal gender differences in threat appraisals that did not emerge in this research. Use of a measure specific to the marital context would result in a loss of comparability to appraisals in other contexts, but this loss would be offset somewhat by the ability to explore threats to marital well-being in greater depth. Results from this study also suggest refinements to the coping inventory employed in this research. In particular, the dimension of hostility may be relevant to spouses' coping with marital tension. By distinguishing in future research the degree of hostility perceived, delivered, and received by spouses when they employ various coping strategies, gender differences in the effects of respondents' own coping and their spouses' coping may be further clarified. In addition, this study demonstrated the potential importance of measuring spouses' withdrawal from marital conflict at an earlier point in the stress and coping process. Although difficult to reliably measure, lack of engagement in marital conflict could be approximated in future research by obtaining more specific information from spouses about their own and their partners' initial appraisal process, especially as it pertains to their perceptions of the occurrence of marital tension. Measurement of lack of engagement in marital conflict would perhaps make the research findings pertaining to withdrawal more comparable to those reported in observational research in the marital literature. Data from the present study also indicated that when wives experienced distress on days in which they reported marital tension, their distress was generally not significantly associated with their attempts to cope with marital tension. There was a trend in the data  220  that suggested that wives' reported use of compromise may be associated with decreased distress; however, further research is needed to explore this possibility. The only significant reduction in wives' distress was related to one measure of their husbands' reported withdrawal. It is therefore possible that once wives become distressed as a result of marital tension, their distress is relieved only when the tension is resolved. Future research could examine this possibility by assessing the degree to which spouses consider an argument to be resolved and the extent to which they are satisfied with the resolution. These data may reveal gender differences in spouses' distress as a function of conflict resolution. Due to the theoretical relevance of spouses' perceptions and cognitions relating to the coping process, self-report data were solicited from spouses and their partners in this research. Although use of the diary methodology mitigated against some of the biases inherent in retrospective self reports (DeLongis et al., 1990), self-report data remain suspect without converging evidence from other sources such as observational measures. As discussed previously, however, observations of marital conflict are of questionable validity. Combining daily diary descriptions of spouses' coping with observational coding of spouses' reenactments of recent diary entries of marital conflict would allow a more complete assessment of the dynamics involved in dyadic coping. The results of this study have demonstrated the value of measuring coping both in terms of the frequency of endorsement of various types of coping (raw score coping) and the endorsement of one type of coping relative to the person's total coping effort (relative score coping). Although reporting two sets of results can be somewhat cumbersome, the two types of scoring yield different, but complementary perspectives on the coping process. Given that the two measures together provide useful information not available with one measure alone, it is recommended that future research of the coping process within marital dyads include both raw score and relative score measures of coping.  221  Finally, it is recommended that an emphasis be placed on pursuing naturalistic field studies in future research pertaining to gender differences in marital conflict. Many of the gender differences reported in the literature have been found using restricted samples (e.g., clinical population or undergraduate students) who are responding to contrived conflicts. It may be the case that certain patterns of gender differences emerge as a function of these contrived situations. Although experimental research can provide valuable information regarding couples' interactions during marital conflict, the validity of the findings from this type of research cannot be determined without reference to the behavior of couples within a naturalistic context. It is therefore important that future research attempt to examine gender differences within an ecologically valid context. The present study has demonstrated that naturalistic, ecologically valid research holds promise for the explication of gender differences in spouses coping with marital conflict. Additional research focusing on naturalistic marital conflict is needed, however, to clarify the parameters within which real-world gender differences do or do not emerge.  FOOTNOTES 1 Although there were also findings within the literature indicating that males tend to be more aggressive and confrontational in their interpersonal interactions than are women (e.g., Basow, 1992; Doyle, 1989; Eagly & Steffen, 1986), these studies focused on physical aggression, an aspect of confrontation that was not addressed in the present study. Physical aggression or violence as a response to marital conflict is clearly a significant issue (see, for example, Straus, 1980; Straus & Gelles, 1986; Straus, Gelles, & Steinmetz, 1980, for an in-depth discussion of this issue), but exploration of this issue was beyond the scope of this study and its focus on non-violent coping strategies. ^The collection of these data was funded by the National Institute of Mental Health (1R01-MH42714). 3 Although it seems fair to assume that when both spouses reported marital tension on the same day, they were referring to the same marital tension, it is important to note that this is an assumption that has been made rather than a fact that can be verified. ^It is important to note that the expression "serious" marital tension is being used here to denote marital tension indicated by spouses to be the most serious tension experienced during a particular day. "Serious" marital tension in this context refers only to tension judged to be more serious than any other interpersonal tension experienced during that particular day. This should not be equated with our usual understanding of the term, "serious," that encompasses a broader context of meaning. To avoid attribution of meaning to the word, "serious," that is not appropriate in the current context, "serious" as used to describe marital tension in this research will be set off by apostrophes to indicate its distinctive connotation. ^The relative scoring applied only to the coping scales and their interrelationships. The relative score method was not applicable to the appraisal of threat measure which represented only one scale with five items and therefore this measure is not presented in the tables that pertain to relative score coping. ^The more standard procedure of randomly choosing a diary day from those meeting inclusion criteria was employed in these analyses. Unlike the data subset (n = 59) in which 66% of respondents who met inclusion criteria met this criteria only on one diary day, 67% of respondents in the present analysis met inclusion criteria on two or more diary days. Thus, random selection of diary days for inclusion in this analysis was deemed appropriate. ^Two tailed significance tests were employed for all data analyses and, for most analyses, the alpha level was set at .05. The alpha level for the correlations presented in Tables 9 through 12 was set at .01 in order to reduce the possibility of Type I errors. Trends toward significance relating to these correlations (p < .05) were also noted when appropriate. It should also be noted that given the directional nature of several of the  223  hypotheses, employment of two-tailed analyses represents a conservative approach to the interpretation of significance. ^Because of the nonindependence in husbands and wives data, differences between husbands and wives in beta-weights were not examined by creating interaction terms of gender by the study variables. Instead, equations were run separately and tests for the differences between the betas were examined (Pedhazur, 1982). ^Given that relative coping scores are based on proportions, these three relative coping scores are mathematically dependent. For example, if there were only two relative coping scales, the scales would be perfectly, negatively correlated. In the present case, there were three coping scales, each of which were reported by both husbands and wives approximately one third of the time (see Table 19). Although the intercorrelations among the three relative coping scales in this study are meaningful, these intercorrelations should be interpreted cautiously. test for the possibility that spouses' coping might be related to respondents' distress via perceived spouses' coping, the regression analyses were also conducted reversing steps 3 and 4. The results of these analyses were significantly different only in the raw score equation pertaining to husbands' confrontation. In the third step of this equation, entry of husbands' perceived spouses' confrontation (6 = .30) accounted for 8% of the variance (p < .05), over and above the 14% of the variance already accounted for by prior distress and self confrontation, entered in steps 1 and 2. Husbands' spouses' coping (i.e., wives' coping), entered in step 4 was not significant (R = .23, p > .10), accounting for an additional 4% of the variance.  REFERENCES Adkins-Regan, E. (1989). Genesis of gender (Review of Masculinity/femininity: Basic perspectives}. Contemporary Psychology, 34, 243-245. Adler, T. (1993a, July). Gender and stress: Men and women affected by stress, but differently. APA Monitor, p. 8-9. Ahlgren, A., & Johnson, W. (1979). Sex differences in cooperative and competitive attitudes from the second through the twelfth grades. Developmental Psychology, 15, 45-49. Aldwin, C. M., & Revenson, T. A. (1987). Does coping help? A reexamination of the relation between coping and mental health. Journal of Personality and Social Psychology, 53, 337-348. Al-Issa, I. (Ed.) (1982). Gender andpsychopathology. New York: Academic Press. Amenson, C. S., & Lewinsohn, P. M. (1981). An investigation into the observed sex differences in prevalence of unipolar depression. Journal of Abnormal Psychology, 90, 1-13. Aneshensel, C. S., Frerich, R. R., & Clark, V. A. (1981). Family roles and sex differences in depression. Journal of Health and Social Behavior, 22, 379-393. Argyle, M., & Furnham, A. (1983). Sources of satisfaction and conflict in long-term relationships, Journal of Marriage and the Family, 45, 481-493. Aries, E. J., & Johnson, F. L. (1983). Close friendship in adulthood: Conversational content between same-sex friends. Sex Roles, 9, 1183-1196. Atkinson, T. (1980). Public perceptions of the quality of life. In H. J. Adler & D. A. Brusegard (Eds.), Perspectives Canada 111 (pp. 275-292). Ottawa: Statistics Canada Bakan, D. (1966). The duality of human existence: Isolation and communion in Western man. Boston: Beacon. Barnett, R. C , & Baruch, G. K. (1987). Social roles, gender, and psychological distress. In R. C. Barnett, L. Biener, & G. K. Baruch (Eds.). Gender and stress (pp. 122143). New York: The Free Press. Barnett, R. C , & Baruch, G. K. (1985). Women's involvement in multiple roles and psychological distress. Journal of Personality and Social Psychology, 49, 135-145.  225 Barnett, R. C , Marshall, N. L., Raudenbush, S. W., & Brennan, R. T. (1993). Gender and the relationship between job experiences and psychological distress: A study of dual-earner couples. Journal of Personality and Social Psychology, 64, 794-806. Basow, S. A. (1992). Gender: Stereotypes and roles (3rd ed.). Pacific Grove, CA: Brooks/Cole. Baumeister, R. (1982). A self-presentational view of social phenomena. Psychological Bulletin, 91, 3-26. Belenky, M. F., Clinchy, B. M., Goldberger, N. R., & Tarule, J. M. (1986). Women's ways of knowing: The development of self, voice, and mind. New York: Basic Books. Belle, D. (1980). Who uses mental health facilities? In M. Guttentag, S. Salasin, & D. Belle (Eds.), The mental health of women. New York: Academic Press. Belle, D. (1982). The stress of caring: Women as providers of social support. In L. Goldberger & S. Breznitz (Eds.), Handbook of stress: Theoretical and clinical aspects (pp. 496-505). New York: Free Press. Bernard, J. (1972). The future of marriage. New York: Bantam Books. Billings, A. G., & Moos, R. H. (1981). The role of coping responses and social resources in attenuating the impact of stressful life events. Journal of Behavioral Medicine, 4, 139-157. Billings, A. G., & Moos, R. H. (1984). Coping, stress, and social resources among adults with unipolar depression. Journal of Personality and Social Psychology, 46, 877-891. Blake, R. R., & Mouton, J. S. (1964). The managerial grid. Houston: Gulf. Bock, E. W., & Webber, I. L. (1972). Suicide among the elderly: Isolating widowhood and mitigating alternatives. Journal of Marriage and the Family, 34, 24-31. Bolger, N., DeLongis, A., Kessler, R. C , & Schilling, E.A. (1989). Effects of daily stress on negative mood, Journal of Personality and Social Psychology, 57, 808-818. Bolger, N., DeLongis, A., Kessler, R. C , & Wethington, E. (1989a). The contagion of stress across multiple roles, Journal of Marriage and the Family, 51, 175-183. Bolger, N., DeLongis, A., Kessler, R. C , & Wethington, E. (1989b). The microstructure of daily role-related stress in married couples. In J. Eckenrode & S. Gore (Eds.), Crossing the boundaries: The transmission of stress between work and family (pp. 95-115). New York: Plenum.  226 Bolger, N., & Schilling, E. A. (1991). Personality and the problems of everyday life: The role of neuroticism in exposure and reactivity to daily stressors. Journal of Personality, 59, 355-386. Bowman, M. L. (1990). Coping efforts and marital satisfaction: Measuring marital coping and its correlates. Journal of Marriage and the Family, 52, A63-A7A. Bradbury, T. N., & Fincham, F. D. (1990). Attributions in marriage: Review and critique. Psychological Bulletin, 107, 3-33. Bradbury, T. N., & Fincham, F. D. (1992). Attributions and behavior in marital interaction. Journal of Personality and Social Psychology, 63, 613-628. Broman, C. L. (1991). Gender, work-family roles, and psychological well-being of Blacks. Journal of Marriage and the Family, 63, 509-520. Brown, G. W., & Harris, T. (1978). Social origins of depression. New York: Free Press. Brown, L. M., & Gilligan, C. (1992). Meeting at the crossroads: Women's psychology and girls' development. Cambridge, MA: Harvard University Press. Buck, R. (1984). The communication of emotion. New York: Guilford. Burggraf, C. S., & Sillars, A. L. (1987). A critical examination of sex differences in marital communication. Communication Monographs, 54, 216-29A. Caldwell, M. A., & Peplau, L. A. (1982). Sex differences in same-sex friendship. Sex Roles, 8, 721-732. Cate, R. M., Huston, T. L., & Nesselroade, J. R. (1985) Premarital relationships: Toward the identification of alternative pathways to marriage. Journal of Social and Clinical Psychology; 125, 715-721. Christensen, A. (1987). Detection of conflict patterns in couples. In K. Hahlweg & M. J. Goldstein (Eds.), Understanding major mental disorder: The contribution of family interaction research (pp. 250-265). New York: Family Process Press. Christensen, A. (1988). Dysfunctional interaction patterns in couples. In P. Noller & M. A. Fitzpatrick (Eds.), Perspectives on marital interaction (pp. 31-52). Clevedon, England: Multilingual Matters. Christensen, A., & Heavey, C. L. (1990). Gender and social structure in the demand/withdrawal pattern of marital conflict. Journal of Personality and Social Psychology, 59, 73-81.  227  Christensen, A., & Shenk, J. L. (1991). Communication, conflict, and psychological distance in nondistressed, clinic, and divorcing couples. Journal of Consulting and Clinical Psychology, 59, 458-463. Clancy, K., & Gove, W. (1974). Sex differences in mental illness: An analysis of response bias in self-reports. American Journal of Sociology, 80, 205-216. Cohen, F., & Lazarus, R. S. (1983). Coping and adaptation and health and illness. In D. Mechanic (Ed.), Handbook of health, health care, and the health professions. New York: Free Press. Cohen, J. (1969). Statistical power analysis for the behavioral sciences. San Diego, CA: Academic Press. Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155-164. Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98, 310-357. Collins, D. L., Baum, A., & Singer, J. E. (1983). Coping with chronic stress at Three Mile Island: Psychological and biochemical evidence. Health Psychology, 2, 149166. Compas, B. E., Forsythe, C. J., & Wagner, B. M. (1988). Consistency and variability in causal attributions and coping with stress. Cognitive Therapy and Research, 12, 305320. Connor, J. M., Servin, L. A., & Ender, R. A. (1978). Responses of boys and girls to aggressive, assertive, and passive behaviors of male and female characters. Journal of Genetic Psychology, 133, 59-69. Coyne, J. C , Aldwin, C , & Lazarus, R. S. (1981). Depression and coping in stressful episodes. Journal of Abnormal Psychology, 90, 439-447. Coyne, J. C , & DeLongis, A. (1986). Going beyond social support: The role of social relationships in adaptation. Journal of Consulting and Clinical Psychology, 54, 454460. Coyne, J. C , Ellard, J. H., & Smith, D. A. (1990). Unsupportive relationships, interdependence, and unhelpful exchanges. In I. G. Sarason, B. R. Sarason, & G. Pierce (Eds.), Social support: An interactional view. New York: Wiley. Crohan, S. E., Antonucci, T. C , Adelmann, P. K., & Coleman, L. M. (1989). Job characteristics and well-being at midlife: Ethnic and gender comparisons. Psychology of Women Quarterly, 13, 223-235.  228  Cronkite, R. C , & Moos, R. H. (1984). The role of predisposing and moderating factors in the stress-illness relationship. Journal of Health and Social Behavior, 25, 372-393. Davis, M. H., & Oathout, H. A. (1987). Maintenance of satisfaction in romantic relationships: Empathy and relational competence. Journal of Personality and Social Psychology, 53, 397-410. Deaux, K. (1976). The behavior of women and men. Monterey, CA: Brooks/Cole. Deaux, K. (1984). From individual differences to social categories: Analysis of a decade's research on gender. American Psychologist, 39, 105-116. Deaux, K., & Major, B. (1987). Putting gender into context: An interactive model of gender-related behavior. Psychological Review, 94, 369-389. DeLongis, A., Coyne, J. C , Dakof, G., Folkman, S., and Lazarus, R. S. (1982). Relationship of daily hassles, uplifts, and major life events to health status. Health Psychology, 1, 119-136. DeLongis, A., Folkman, S., & Lazarus, R. S. (1988). The impact of daily stress on health and mood: Psychological and social resources as mediators. Journal of Personality and Social Psychology, 54, 486-495. DeLongis, A., Hemphill, K. J., & Lehman, D. R. (1992). A structured diary methodology for the study of daily events. In F. B. Bryant, J. Edwards, L. Heath, E. J. Posavac, & R. S. Tindale (Eds.), Methodological issues in applied social psychology (pp. 83-109). New York: Plenum. DeLongis, A., & Lehman, D. (1989). The usefulness of a structured diary approach in studying marital relationships. Journal of Family Psychology, 2, 337-343. DeLongis, A., & O'Brien, T. (1990). An interpersonal framework for stress and coping: An application to the families of Alzheimer's patients. In M. A. P. Stephens, J. H. Crowther, S. E. Hobfoll, & D. L. Tennenbaum (Eds.), Stress and coping in later life families (221-239). Washington, D.C: Hemisphere Publishers. DeLongis, A., Tal, M., & Kessler, R. C. (1987). Daily coping with marital conflict. Paper presented at the Annual Convention of the American Psychological Association, New York. Dohrenwend, B. S., Dohrenwend, B. P., Dodson, M., & Shrout, P. E. (1984). Symptoms, hassles, social supports and life events: The problem of confounded measures. Journal of Abnormal Psychology, 93, 222-230.  Dohrenwend, B, P., & Shrout, P. E. (1985). "Hassles in the conceptualization and measurement of life stress variables. American Psychologist, 40, 780-785. Derogatis, L. R. (1975). Affects Balance Scale. Baltimore, MD: Clinical Psychometrics Research Unit. Deutsch, M. (1969). Conflicts: Productive and destructive. Journal of Social Issues, 25, 7-41. Doyle, J.A. (1989). The male experience (2nd ed.). Dubuque, IA: Wm. C. Brown. Eagly, A. H. (1983). Gender and social influence: A social psychological analysis. American Psychologist, 38, 971-981. Eagly, A. H. (1987). Sex differences in social behavior: A social-role interpretation. Hillsdale, NJ: Erlbaum. Eccles, J. S., & Jacobs, J. E. (1986). Social forces shape math attitudes and performance. Signs, 11, 367-380. Eckenrode, J. (1984). The impact of chronic and acute stressors on daily reports of mood. Journal of Personality and Social Psychology, 93, 907-918. Edwards, A. L. (1979). A repeated-measure design. In A. L. Edwards (Ed.), Multiple regression and the analysis of variance and covariance (pp. 117-131). San Francisco: W. H. Freeman. Feingold, A. (1988). Cognitive gender differences are disappearing. American Psychologist, 43, 95-103. Fichten, C. S., & Wright, J. (1983). Problem-solving skills in happy and distressed couples: Effects of videotape and verbal feedback. Journal of Clinical Psychology, 39, 340-352. Fincham, F. D., & Bradbury, T. N. (1987). The impact of attributions in marriage: A longitudinal analysis. Journal of Personality and Social Psychology, 53, 510-517. Fincham, F. D., Bradbury, T. N., & Beach, S. R. H. (1990). To arrive where we began: A reappraisal of cognition in marriage and in marital therapy. Journal of Family Psychology, 4, 167-184. Fincham, F. D., Bradbury, T. N., & Scott, C. (1990). Cognition in marriage. In F. D. Fincham & T. N. Bradbury (Eds.), The psychology of marriage: Basic issues and applications (pp. 118-149). New York: Guilford.  230  Fitzpatrick, M. A. (1988). Between husbands and wives: Communication in marriage. Newbury Park, CA.: Sage Publications Floyd, F. J., & Markman, H. J. (1983). Observational biases in spouse observation: Toward a cognitive/behavioral model of marriage. Journal of Consulting and Clinical Psychology, 51, 450-457. Fogarty, T. F. (1976). Marital crisis. In P. J. Guerin (Ed.), Family therapy: Theory and practice (pp. 325-334). New York: Gardner Press. Folger, J. P., & Poole, M. S. (1984). Working through conflict. Glenview, IL: Scott Foresman. Folkman, S., & Lazarus, R. S. (1980). An analysis of coping in a middle-aged community sample. Journal of Health and Social Behavior, 21, 219-239. Folkman, S., & Lazarus, R. S. (1985). If it changes it must be a process: Study of emotion and coping during three stages of a college examination. Journal of Personality and Social Psychology, 48, 150-170. Folkman, S., & Lazarus, R. S. (1988). Coping as a mediator of emotions. Journal of Personality and Social Psychology, 54, 466-475. Folkman, S., Lazarus, R. S., Gruen, R. J., & DeLongis, A. (1986). Appraisal, coping, health status, and psychological symptoms. Journal of Personality and Social Psychology, 50, 571-579. Folkman, S., Lazarus, R. S., Dunkel-Schetter, C , DeLongis, A., & Gruen, R. J. (1986). Dynamics of a stressful encounter: Cognitive appraisal, coping, and encounter outcomes. Journal of Personality and Social Psychology, 50, 992-1003. Folkman, S., Lazarus, R. S., Pimley, S., & Novacek, J. (1987). Age differences in stress and coping processes. Psychology and Aging, 2, 171-184. Freud, S. (1915/1957). Instincts and their vicissitudes. In J. Strachey (Ed.), Standard edition of the complete psychological works ofSigmund Freud (pp. 111-142). London: Hogarth. (Original work published 1915). Gallagher, D., & Shuntich, R. J. (1981). Encoding and decoding of nonverbal behavior through facial expressions. Journal of Research in Personality, 15, 241-252. Gervasio, A. H., & Crawford, M. (1989). Social evalua