Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Christian beliefs, moral commitment, and marital stability Brandt, Robyn 1997-03-10

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
831-ubc_1997-0149.pdf [ 4.2MB ]
Metadata
JSON: 831-1.0099194.json
JSON-LD: 831-1.0099194-ld.json
RDF/XML (Pretty): 831-1.0099194-rdf.xml
RDF/JSON: 831-1.0099194-rdf.json
Turtle: 831-1.0099194-turtle.txt
N-Triples: 831-1.0099194-rdf-ntriples.txt
Original Record: 831-1.0099194-source.json
Full Text
831-1.0099194-fulltext.txt
Citation
831-1.0099194.ris

Full Text

CHRISTIAN BELIEFS, MORAL COMMITMENT, AND MARITAL STABILITY by ROBYN BRANDT B.A., The University of British Columbia, 1994 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in THE FACULTY OF GRADUATE STUDIES (School of Family and Nutritional Sciences) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA March 1997 © Robyn Brandt, 1997 > / In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Dopartmont of r^a, mi \y -j Ni U~\-f \' [\ QirvocQ ^e^cjrsC^i ^ The University of British Columbia CjF*Kr\«Vy S-U\c(.\^ J Vancouver, Canada Date DE-6 (2/88) Abstract The relationship between moral commitment and marital stability is examined within Johnson's (1991) commitment framework using two religious populations. Beliefs and internal factors are emphasized over external constraints and barriers. Johnson's model of commitment is discussed and modified to incorporate beliefs and Johnson's construct of moral commitment assumes central importance in the study. Moral commitment is hypothesized to directly effect marital stability, and moderate the effects of personal and structural commitment on stability. Self-administered questionnaires are distributed to two Brethren churches (fundamentalist) and two United churches (liberal). These Protestant churches are matched by geographic location to control for ethnic and economic differences. Sixty-three participants return the 10-page survey. Bivariate and multivariate analyses are performed on the variables in the commitment model using logistic and multiple regression. The results indicate that both beliefs and moral commitment are significant factors in the study of marital stability. Moral commitment was related to both structural and personal commitment. Despite there being no direct effect between moral commitment and marital stability, moral commitment is related to the other factors of personal and structural commitment which are in turn related to marital stability. Moral commitment did not moderate the effects of personal and structural commitment on marital stability. The results also suggest that gender is an important control variable in commitment theory, with respect to religious populations. Religious beliefs were highly correlated with moral commitment. The age distribution of the participants and small sample size were among the limitations that prevent generalizability of the results to other religious ii populations. Limited variation in the marital stability scores may have restricted the number and strength of significant findings. It is suggested that future research include both religious and non-religious groups in the study of beliefs as they related to marital stability. iii Table of Contents Abstract ii Table of Contents v List of Figures viiList of Tables ix Introduction 1 Literature Review 5 Predecessors of the Current Model of Commitment 5 Exchange and Attribution Theories 5 The Current Model of Commitment 7 Structural Commitment 8 Personal Commitment 10 Moral CommitmentRevised Model of Commitment 12 Fundamentalist Christianity and the Revised Model of Commitmenl3 Moral Commitment 15 Structural Commitment 20 Personal Commitment 4 Summary of Assumptions in Revised Commitment Model 25 Hypotheses 27 Method 31 Sample and Data Collection 3Measures • 32 Dependent Variable: Marital Instability 32 Independent Variable: Personal Commitment . . . 34 Relational and Religious Identity 35 Attraction to the Relationship 6 Attraction to the Spouse 38 Independent Variable: Structural Commitment 40 Investment 4Attention to Alternative Attractions 41 Influence of Social Reaction 42 Difficulty of Termination 3 Moderating Variable: Moral Commitment 44 Value of Stability 4Person-specific Obligation 45 Independent Variable: Christian Fundamentalism 46 Control Variables 47 Results 48 Description of Sample 4Univariate Distributions 9 Dependent Variable 4Independent Variables 51 Structural Commitment 5Personal Commitment 2 Moral Commitment 53 v Christian Fundamentalism 53 Control VariablesHypothesis Testing 54 Bivariate AnalysisMultivariate Analysis 68 Summary of Significant Relationships 69 Discussion 71 Hypothesis Testing 73 Commitment Model 9 Moral Commitment 7Structural Commitment 80 Personal CommitmentImplications for Research ' 83 Theoretical Implications 8Future Research 4 Limitations • 84 Summary 86 References 8 Appendix A: Definitions 93 Appendix B: Questionnaire 4 Appendix C: Church Bulletin 11Appendix D: Questionnaire Cover Letter 115 Appendix E: Table of Missing Values 116 vi Appendix F: Descriptive Statistics for Control Variables 128 Appendix G: Descriptive Statistics for Dependent and Independent Variables 130 List of Figures Figure 1. Revised Theory of Commitment . .' 28 Figure 2. The Effect of Moral and Personal Commitment on Marital Stability 29 Figure 3. The Effect of Moral and Structural Commitment on Marital Stability 30 Figure 4. Commitment Model 72 Figure 5. Potential Revision for the Model of Commitment 82 viii List of Tables Table 1. Summary of Results from Hypothesisl 56 Table 2. Summary of Results from Hypothesis 3 8 Table 3. Summary of Results from Hypothesis 4 60 Table 4. Summary of Results from Hypothesis 5 4 Table 5. Summary of Results from Hypothesis 6 7 Table 6. Summary of Significant Relationships 70 Table 7. Exclusion of Questions Based on Missing Values 116 Tables 8 & 9. Descriptive Statistics for Males 117 Tables 10 & 11. Descriptive Statistics for Females 118 Tables 12, 13 & 14. Descriptive Statistics for Dependent and Independent Variables .119 ix Introduction The study of marital stability has a long history in the social sciences. Almost twenty years ago Lewis and Spanier were already compiling and analyzing an extensive body of research dating from the 1930's in their examination of marital quality and stability (1979). At the time of their review, one of the basic models of stability explained marital outcomes through attractions, constraints or barriers, and alternatives (e.g., Levinger, 1965). Lewis and Spanier's literature review culminated in a 'propositional inventory' (1979) that included these established variables. The authors organized these, and other variables, into a model containing four main categories or stages: premarital dispositions, marriage, threshold, and postmarriage. Within each of the first three components of Lewis and Spanier's (1979) model there are several sub-factors. Communication and adjustment, for example, are two factors that contribute to marital quality (Lewis & Spanier, 1979, p. 272). In the threshold stage alternatives and external pressures are two of the factors that determine whether or not the couple terminate or continue their relationship. Another variable in this stage is 'religious doctrine' (Lewis & Spanier, 1979, p.272). Lewis and Spanier's conception of this religious factor within their model is typical of its treatment in the stability research, as it is based on a literature review. This conception limits the measurement of religion to church attendance and denominational affiliation (e.g., pages 272, 275, 278, & 283), and presents it as a barrier or constraint that keeps marriages together. Ten years later, Larson and Goltz (1989) are still using these limited conceptions in their examination of religion and the marital relationship. 1 Larson and Goltz use religious homogamy and affiliation as well as church attendance to examine the relationship between religion and marital commitment (1989, p. 388). Using data from a random sample of 428 households in the Edmonton area, the researchers tested these religious variables against two components of marital commitment: personal and structural commitment. These two types of commitment are outlined by Johnson in an unpublished manuscript (1978, cited in Larson & Goltz, 1989). Larson and Goltz propose that "Leaving a marital relationship is much more difficult within the religious community ... because of the obvious constraints against ending marriage" (1989, p. 388). Neither religious affiliation nor homogamy were related to personal commitment in this study, but there was a significant relationship between structural commitment and church affiliation. Both personal and structural commitment were found to be related to church attendance. Despite their significant findings, Larson and Goltz concluded that "The theories and existing measures of commitment ... do not adequately connect with the meaning of commitment in religious covenant" (1989, p. 397). Measures of attendance, homogamy, and affiliation are simple descriptors of a complex experience. However, it is not only the limited scope of religious measures that needs to be addressed in the literature. The conception of religion as an external constraint is also too simplistic and restrictive. The work of Lewis and Spanier and of Larson and Goltz are examples of a common conception in the literature of religion as an external, constraining factor, or barrier. Researchers need to expand this narrow perspective to include an examination of how the belief systems within individuals effect marital stability. The effect of religiosity on marital stability is established: religiosity is associated with increased marital stability (e.g., Chan & Heaton, 1989; Maneker & Rankin, 1993). It is the cause of this effect that has not been 2 explored sufficiently. At least one aspect of the religious experience of individuals has been largely ignored, creating an imbalance in the literature which favors external over internal dimensions in explaining marital stability. The internal aspects of the religious experience need to be acknowledged as well. One purpose of this study, therefore, is to correct this imbalance in the literature by approaching the study of marital stability from the standpoint of the subjects' belief systems. To accomplish this, it is necessary to find a model capable of accommodating belief systems in its predictions. Johnson's model of commitment (1991) provides a theory that encompasses a measure of beliefs. However, the component of his model that addresses this aspect of stability is at the periphery of an exchange-based theory. Johnson proposes that personal commitment (largely a satisfaction measure), structural commitment (measures of external barriers), and moral commitment (beliefs surrounding marriage), predict marital stability. This model is a refinement of the one cited by Larson and Goltz (1989). Johnson has written extensively about commitment. Johnson, like Lewis and Spanier, performed an exhaustive review of the literature in this area and presented a propositional inventory based on his assessment of the existent research (1985). This qualitative meta-theoretical analysis was followed by a discussion (Johnson, 1991) of commitment as it is conceptualized in Johnson's model, and in other related models. The second purpose of this study, therefore, is to modify Johnson's current model of commitment so that beliefs hold a more central position in the theory. A third purpose arose from this need to modify Johnson's model. The variable within Johnson's model that addresses beliefs (moral commitment) is precisely the construct which one author has criticized as superfluous to the theory. Rusbult, in her commentary on Johnson's commitment 3 theory, called for further research and clarification of moral commitment and its usefulness in the understanding of marital stability. She emphasizes the need to develop defensible operational definitions of moral commitment, and empirically demonstrate the conditions under which and the persons for whom this factor contributes to desire to maintain a relationship, [and to] demonstrate that this force contributes to our understanding of commitment ... beyond what we know based on ... investments, satisfaction, and alternatives. (1991, p. 165) Accordingly, the third purpose of this study is to respond to this criticism by examining whether, a) beliefs are important variables in the prediction of stability, and b) whether the moral commitment construct is a useful tool in measuring these beliefs. In order to respond to Rusbult's challenge, and to investigate the contribution of beliefs (as measured by moral commitment) to stability, a religious sample was chosen for this study. Two distinct Protestant Christian groups (Brethren and United) were included in this study. A Christian sample was chosen because this belief system includes specific doctrine on marital stability. In addition, culture and ethnicity do not complicate the study of this group, as would be the case with an Eastern religious system, for example. It has been well established that religiosity has a positive effect on marital stability (e.g., Chan & Heaton, 1989; Maneker & Rankin, 1993). However, whether or not there is a causal basis to this correlation has not yet been completely established. It has been hypothesized that social constraints; social, emotional, and spiritual supports; and practical barriers, among other factors, are responsible for the lower divorce rate of the religious. These explanations tend to view marital stability as created by forces external to the individual and 4 give little credence to the nature and importance of individual beliefs and morality. This thesis attempts to offer an alternative explanation for the stability of Christian marriages, one that balances the view that predominantly external factors affect marriage with a model that accounts for internal beliefs. The following literature review describes, first, the theories and concepts that lead to Johnson's model of commitment, and second, the current state of this model. A revised model is then presented, as well as the hypotheses to be tested with this new model of commitment. Literature Review Predecessors of the Current Model of Commitment Exchange and Attribution Theories The current theory of marital stability includes several concepts first outlined in social exchange theory. The variables from this perspective that are salient to this theory include attraction to the relationship, barriers to dissolution, and alternative attractions (Lewis and Spanier, 1979; Johnson, 1985). Lewis and Spanier (1979) discuss the variables that they feel are applicable in the examination of marital stability. They outline four stages in the process towards divorce. The first, premarital predispositions, is made up of the following factors: personality factors, attitudes and values, social factors, circumstantial factors, marital expectations, and social maturity level. Marital quality is the salient variable during the second stage; marriage. The sub-factors within marriage are: adjustment, satisfaction, happiness, conflict and role strain, communication, and integration. 5 The third stage is the threshold stage. /Threshold factors determine whether the marriage will survive or not. The following are constructs that are predicted to affect stability: marital expectations, commitment and obligations, tolerance, religious doctrine, external pressures and social stigma, divorce law and legal aid, and real and perceived alternatives. The author's fourth stage, post marriage, is made up of separation and divorce, if the threshold variables determine it. Lewis and Spanier (1979, p. 285) describe the effect of the threshold variables as follows: "the threshold variables ... operate as forces which allow some couples to pass over the threshold and separate (and subsequently divorce) while not allowing others to pass." The exchange typology outlined by Lewis and Spanier (1979) relates alternative attractions, attractions, tensions, and external pressures to marital stability in a quadratic form. The authors label tensions and attractions as intradyadic variables (factors which influence satisfaction), and alternative attractions and external pressures as extradyadic factors (influencing stability). Lewis and Spanier state that the following variables influence marital stability within their extradyadic framework (1979, p. 287): strict divorce laws, strong social stigma, strict adherence to or influence from restrictive religious doctrine, low evaluation of nonmarital alternatives, high degree of commitment to marriage, and high tolerance of marital conflict and tension. They conclude by saying that, "the strength of the alternative attractions for the individual outside the marriage are balanced against the external pressures to determine ... marital stability" (1979, p. 287). 6 According to Adam and Sprenkle, (1990, p. 132) "Alternative attractions and external pressures ... serve as contingencies mediating the relationship between marital quality and stability." Adam and Sprenkle suggest that a more realistic model than that of Lewis and Spanier would include attribution and self-perception theories that are better able to accommodate the complexities of the decision to divorce. Adam and Sprenkle state that If a person remains in a relationship when more attractive alternatives are available, it could be because the person 'has to' (structural commitment) or 'ought to' (moral commitment) remain. But if these factors are not present, one could conclude, the person 'wanted to' (personal commitment) remain. This is an internal attribution (1990, p. 135). These authors propose that commitment increases when couples stay together despite attractive alternatives and few barriers, and that this outcome is not conceptualized within the social exchange framework. They conclude that the relationship between commitment and stability is a circular reciprocal one. The Current Model of Commitment Johnson modified Lewis and Spanier's model, most notably adding moral commitment to the theory. He developed a model that relates three types of commitment to marital stability. The following is a brief summary of the factors in this model (Johnson, 1991): 7 Structural Commitment irretrievable investments social reaction difficulty of termination attraction of alternatives Personal Commitment attitude toward the relationship attitude toward the partner relational identity Moral Commitment value of consistency value of the stability of relationships person-specific obligation Structural commitment. Johnson proposes that these three main factors affect marital stability, but that structural commitment has the greatest impact. Structural commitment prevents people from leaving a dissatisfying relationship as it represents barriers to dissolution. It is the variable responsible for people feeling that they 'have to' remain in an unattractive relationship. Johnson classifies these barriers as externally generated and constraining, as opposed to internal variables subject to individual choice (1991, p. 122). These factors are related to 8 Lewis and Spanier's ideas about external barriers. The first factor, investments, is defined by Johnson (1991, p. 123) as irretrievable "to the extent that they were made in anticipation of a future return that has not been realized, and the expenditure was not returnable." High levels of irretrievable investments in a relationship might prevent those involved from leaving it, even if the other types of commitment (personal and moral) are low. Johnson stresses that this factor focuses on the future-oriented nature of investment, and on the types of investments that cannot be regained should one leave the relationship. The second factor, social reaction, involves others' feelings about dissolution, and the pragmatic consequences of a breakup on other people (e.g., children or friends) should the relationship dissolve. Johnson relates the feelings that people in one's social network may have towards one's relationship dissolving to the concepts within his moral commitment construct: the value of consistency, person-specific obligations, and values regarding the stability of relationships (1991, p. 123). The difficulty of termination constitutes the third component of structural commitment. This factor includes the costs of ending the relationship, in terms of monetary and other resources (1991, p. 124). The degree of difficulty in ending a relationship varies depending on, for example, the length of the relationship, the level of interdependence in the dyad, and whether the relationship is legally recognized. The fourth category in Johnson's structural commitment construct is availability of acceptable alternatives. These alternatives extend beyond the replacement of the current marriage to economic consequences, social ties, etc. (Johnson, 1991, p. 124). The consequences of ending a relationship depend on factors such as the age and gender of the 9 person, their occupational and financial status, whether they have children, etc.. These variables will affect not only their future romantic relationships, but also whether their financial situation will improve should the relationship end, whether their social network expands or contracts, as well as numerous other consequences. Personal commitment. Johnson's personal commitment construct is related to satisfaction. This variable measures to what degree the subject 'wants to' stay in the relationship, and is an internal, choice-based factor. Johnson includes three factors in this broad concept. These are: attitude toward the relationship, attitude toward the partner, and relational identity. Johnson distinguishes between attitudes toward the partner and the relationship using the example of an abusive relationship: "the wife abuse literature clearly indicates that it is possible for one to deeply love a person with whom one has a thoroughly unattractive relationship" (1991, p. 120). Johnson describes relational identity as "the extent to which one's involvement in a relationship is incorporated into one's self-concept" (1991, p. 120). Moral commitment. Moral commitment, the third variable in Johnson's commitment model, is an internalized form of self-constraint that makes individuals feel that they 'ought to' act in certain ways in certain situations. Johnson describes moral commitment as containing the following factors (1991, p. 122): a belief in the value of consistency,... values regarding the stability of particular types of relationships, ... and a sense of personal 10 obligation to the particular person with whom one is involved in the relationship. Johnson describes the value of consistency factor as encompassing norms regarding people behaving in predictable ways (1991, p. 121). This concept stems from dissonance theory, which examines the consequences of inconsistent behavior. Johnson's second factor, the value of stability (1991), deals with people's values and beliefs surrounding different types of relationships and their dissolution. For example, it is more acceptable in North American society to end a dating relationship than a marriage (Johnson, 1991, p. 121). The value of stability in Johnson's model is pertinent not for studying the values of one's social network, but rather for examining those values that are internalized by individuals regarding their own relationships. In the current literature, this factor has generally been measured by attitudes toward divorce (Johnson, 1991, p. 121). The marital relationship is not the only relationship that this factor can encompass, however. Johnson views it as applicable to other types of relationships (e.g., friendships, dating partners, etc.). Johnson's third factor within moral commitment, person-specific obligation, is also relevant in terms of internalized values, rather than those found in one's social network (1991, p. 122). The feelings of obligation felt by individuals stem from contracts between them and other individuals. These contracts are not necessarily legally binding, or even articulated (Johnson, 1991, p. 122). 11 Revised Model of Commitment This paper is based on Johnson's theory, although several modifications have once more been made so that the belief systems that impact on Johnson's moral commitment variable can be tested against the other two variables: structural and personal commitment. In the current arrangement of the model, the factors within the two most developed concepts have confusing demarcations of internal and external components. Johnson describes all the factors in personal commitment as 'internal choices', while those in structural commitment are 'external constraints'. However, within these external, structural factors, there appears to be different degrees ofsubjectivity and control. For example, both the availability of acceptable alternatives and the perception of irretrievable investments are seen as equally external to the individual. To reflect the emphasis in this study of internal belief systems and cognitions, Johnson's structural commitment terminology has been altered slightly. Rather than calling the component that measures alternatives the 'availability of acceptable alternatives,' this factor is labelled 'attention to alternative attractions.' This term implies a cognitive evaluation of alternative resources, rather than the passive external presence or absence of these choices. Two people may have, objectively, equal opportunities outside their present marriage. However, subjectively, they may evaluate these alternatives quite differently depending on factors within the marriage, for example. Their satisfaction with the relationship may influence whether they even assess their alternatives. Johnson's (1991) personal commitment terminology has also been altered slightly for this study. He includes 'attitudes' toward the partner and relationship in his model. Because these variables are consistent with satisfaction constructs, they are described in this study as 12 'attractions.' In addition, because this study is an examination of marital stability, and deals exclusively with this specific dyad, the variable relating to attraction to the partner is restricted to 'attraction to the spouse.' The changes made to the model, in terms of altering its focus from external to internal, result from the assumption that beliefs are indeed important in predicting marital stability. The examination of this assumption is the basis of this study. As outlined in the introduction, Christian fundamentalism is the belief system that is used in this study to determine the usefulness of both the beliefs and the construct that measures them (moral commitment). The following is a description of this belief system. The specific components of the commitment model are then examined in relation to these beliefs. Fundamentalist Christianity and the revised model of commitment. The beliefs presented in fundamentalist Christianity are very specific with respect to marital stability. These beliefs are distinct from the attitudes which apply to them. An attitude is an evaluation (e.g., "I like it"), whereas a belief is an existential statement that is true or false (Fishbein, 1967). This thesis is primarily interested in beliefs (e.g., "I believe that God exists" or "I believe that marriage is ordained by God"), as they affect the value the subjects place on stability in marriage, and their feelings of obligation to their spouse. Fundamentalist Christian beliefs are hypothesized to impact moral commitment in ways that affect the rest of the commitment and stability framework. For fundamentalist Christians, these beliefs are internalizations of those outlined in the Bible. One of the most direct statements regarding Christian marriage in the Bible is found in Matthew 19:4. Jesus is quoted as saying of marriage; 13 Haven't you read ... that at the beginning the Creator 'made them male and female ... For this reason a man will leave his father and mother and be united to his wife, and the two will become one flesh'? So they are no longer two, but one. Therefore what God has joined together, let man not separate. (New International Version, also found in Genesis 2:24, Ephesians 5:31, and Mark 10:9) Christian ideas about marriage come from the Bible, but they are also discussed and presented in other Christian literature (sermons, workshops, etc.). In his book on preparation for marriage, McRae (1980) states that there are three elements to the "biblical concept of marriage ... monogamy, absolute fidelity, and permanency" (p. 80). He describes marriage as "a relationship between two people — only two people — for life" (1980, p. 80). According to the Bible, God intended the marital union to outstrip in importance all other human relationships or ties (e.g., the bond between parents and children). The two individuals in the marriage make up a new entity in combination with one another. God commanded that the marriage bond remain faithful ("You shall not commit adultery," Exodus 20:14), or divine judgment would result (Hebrews 13:4). If a subject internalizes the beliefs contained in the Bible, it is proposed that these beliefs will significantly affect what Johnson (1985, 1991) describes as moral commitment. Christianity promotes a view of lifelong commitment to marriage and one's partner. This is in opposition to the secular view that marriages are behavioral contracts that should be dissolved if one of the partners fails to live up to the expectations of the exchange bargain. It is hypothesized that if fundamentalist Christians live by the principles and conceptions 14 about marriage presented in the Bible, moral commitment to their unions will be high, and consequently their marriages stable. It is suggested that the greater longevity of Christian marriages can be attributed to a belief in the sacred, God-created nature of marriage held by fundamentalist believers, and to the effect that these beliefs have on the level of commitment in their relationships. This thesis combines the idea that beliefs affect stability within the commitment framework by relating beliefs to moral commitment. Although many religions see marriage as important, only Christians believe that the union is created and established by God. The uniqueness of this conception of marriage is that the stability of the union is not based on a contract between two people, or even between the couple and God, but rather on the believers' desire to be obedient to the will of God in maintaining the permanence of the union established during creation. For Christians, commitment to their marriage is irrevocably linked with their desire to obey God. Moral commitment. Moral commitment is the concept that has been modified to the greatest extent for this study. Its position within the model has also been modified, but the concept will be addressed here first. The definition of moral commitment is the first modification. Johnson sees moral commitment as half-way between the external forces of structural commitment and the internal choices of personal commitment. He defines it as a feeling of self-constraint. This investigation sees moral commitment as stemming entirely from the internal beliefs of the individual. 15 What is unique in this discussion of marital commitment is that it proposes that the belief system held by the individual has consequences for moral commitment that have not been addressed in previous studies. Fundamentalist Christianity, with its specific discussions of the importance of permanent unions, allows us a unique opportunity to study religiosity and marital stability from a new perspective. Fundamentalist couples are not only committed to each other, or to the institution of marriage, they are committed to submitting themselves to a union that they believe was chosen for them by God. Christians not only define themselves in terms of their relationship to their spouse, but also in terms of their relationship with their God. They are not simply husbands and wives, they are Christians, and there are certain expectations that are associated with this self-perception (these ideas regarding Christian fundamentalism reflect those discussed in a meeting between the author and George Tabert, a member of a Brethren church not sampled in this study). This type of commitment is not acknowledged in the general social science literature since religiosity is usually seen as a barrier to dissolution which is external in nature. This thesis attempts to discern the importance of this type of commitment (i.e., a social psychological internal belief held by the individual), to the prediction of marital stability. Alternative definitions of commitment were examined in the literature, to be used in the revised construct of moral commitment. Commitment, in the secular literature, is generally defined as a decision to pursue a course of action to its completion, although there are almost as many definitions of commitment as there are studies examining it. In a society in which the idea of permanent marital relationships is in decline, the concept of commitment is an important one. Glenn (1991) discusses an increasingly common view of marriage, saying that marriage partners now feel that they are permanently available, 16 should something better come along, or should the love or satisfaction decline in their relationship. Personal commitment is essential for stability in unions where the belief in permanence is low. Rosenblatt defines commitment as "an avowed or inferred intent of a person to maintain a relationship" (1977, italics in original, p. 73-74). He also distinguishes between "Commitment as personal dedication and ... commitment as conformity to external pressures" (1977, p. 73). This distinction is similar to that made here between the secular view of the impact of religiosity on stability (external constraints), and the suggestion that internal belief systems are responsible for the differences. Researchers have attributed the increased rate of marital stability among religious populations to external pressures, but it is hypothesized here that it is the personal dedication of the Christian to the sacred nature of the marriage bond that is responsible. The distinction made here that differs from the current conceptions of religiosity and stability is that religion is not an external constraint (e.g. social pressure), but rather a belief system that is an internal, social psychological variable. Swensen and Trahaug refer to instrumental and intrinsic commitment in their discussion of marital stability (comparable to structural and personal commitment, respectively). In an instrumental marriage, the commitment is to the institution, whereas in an intrinsic marriage, the "bond is to that other person rather than to the functions the other person serves" (Swensen and Trahaug, 1985, p. 939). Swensen and Trahaug (1985, p. 944) found that "an increasing personal commitment between a couple produces a more active, satisfying relationship." They found that marriages with intrinsic commitment were both more stable and more satisfying. 17 Mary Lund, in her barrier model of marital stability, defines commitment as "an attitude about continuing a relationship that is strengthened by a person's own acts of investing time, efforts and resources in that relationship" (1985, p. 4). She separates commitment into three components: "(1) judgments about a relationship's likely permanence; (2) expectations for avoiding involvement in other relationships; and (3) anticipation of losses if a relationship ends" (Lund, 1985, p. 5). Larson and Goltz (1989, p. 397) state that The theories and existing measures of commitment ... do not adequately connect with the meaning of commitment in religious covenant. ... [M]arriage is entered with the intention of lifelong permanence, sexual and mental fidelity to one's partner, and a personal dedication to marriage-affirming and enriching behaviors. These elements of covenant are much broader and deeper than the meanings in the social sciences used for commitment in the study of marriage. The type of values described by Larson and Goltz reflect those that the new definition of moral commitment should encompass. Thomas and Henry (1985) have a similar position on the necessity of acknowledging the importance of religion in the lives of believers. They suggest that meaning is intrinsically involved with one's perception of emotionally bonding relationships with others in the social world; and ... such relationships with others are inextricably connected with their view of the basic relationship with Deity. (Thomas & 18 Henry, 1985, p. 377) Thomas and Henry (1985) caution family scientists to examine their presuppositions regarding their subjects. A 'blind spot' can develop which prevents researchers from uncovering all the pertinent information in a study area. They suggest that, when a researcher approaches the study with an alternative set of presuppositions, "a very different view of family roles, family functioning, and meaning structures within those families emerges" (Thomas & Henry, 1985, p. 377). It is the intention of this study to correct the imbalance that has occurred in the predominantly secular literature by giving equal treatment to the beliefs of its subjects. This thesis attempts to approach the study of religiosity in terms of the marital bond from a perspective that focuses on the beliefs held by these subjects. This approach allows the researcher to pose more appropriate and relevant questions to the participants, expanding the understanding of the current frameworks. A definition of moral commitment that might be more relevant to Christian subjects is the 'willingness to abandon reciprocity and the comparison of alternatives in favor of a relationship that is not based on a behavioral contract' Leik and Leik (1972, p. 5; cited in Johnson, 1985, p. 14) define commitment as "the extent to which an actor has shifted from 1) interest in a relationship because of the goals it mediates to 2) maintenance of the relationship as the dominant goal." This description, and the other definitions presented here, acknowledge the importance of beliefs to the development of commitment. It is hypothesized that Christian beliefs have a profound impact on marital commitment because of the meaning attributed to marriage in the Christian faith. A person pursues a line of action to its completion because of the importance or value placed on that completion. Robinson (1994, p. 211) found that commitment to the marriage "was facilitated 19 through the value that was placed on the marriage bond." The couple enters into the marriage expecting it to be a permanent arrangement, and this expectation in itself is a powerful deterrent to dissolution (Levinger, Senn and Jorgensen, 1970: Hill et al., 1979). Levinger said that in marriages where partners feel a sense of obligation to the marital bond, "both partners are firmly committed to respect the marriage contract, and divorce is not considered a possibility" (1965, p. 24). It is proposed that these factors contribute to a commitment that is qualitatively different from that experienced by non-Christian populations. In conclusion, moral commitment is defined here as an internal set of beliefs that encompasses the value of marital stability and a sense of person-specific obligation (factors outlined in Johnson's original model). It is an internally-based factor that encourages continuation of the marital relationship, rather than an external constraint which discourages the termination of the marriage. Structural commitment. The following is an examination of the proposed effects of moral commitment on structural commitment. Two components of structural commitment, the monitoring of alternatives and investments, are used as examples here tp demonstrate the proposed impact that moral commitment will have on the external aspects of the commitment framework. These two factors are well established in the literature, but the way in which they might be affected by an individual's belief system has not yet been addressed. The concept of alternatives comes originally from social exchange theory, and is incorporated into commitment theory under the structural commitment construct. The theory outlines two types of comparisons made in relationships: a comparison of what you have with 20 what you think you should have, and a comparison of what you have with what you think you can get in another relationship (Comparison Level and Comparison Level of Alternatives, respectively; Johnson, 1985, p. 10). Even if one is dissatisfied with the current relationship and can foresee a better arrangement elsewhere, a simple exchange cannot be made. One must take into account the lost investments in the current relationship, social reaction, and the cost of termination. Even when all these factors are considered in the cost/reward ratio present in a separation or divorce, prediction and explanation of behavior is not perfect because people become used to exchanging with particular others. Cook and Emerson (1978, p. 728) state that "To the extent that commitments form, the exploration of alternatives is curtailed." It is proposed that this curtailment of the monitoring of alternatives is more profound in the highly religious because, from the beginning of the relationship, there is an expectation of permanence. In addition, there is great value placed on maintaining a single, monogamous relationship throughout one's life and this "commodity" or investment is by definition irreplaceable; you cannot trade one permanent, lifelong marital relationship for another. The unique beliefs that Christians bring to the marital union are hypothesized here to impact how they attend to alternatives. Sabatelli and Shehan (1993) cite a number of authors who believe that "the emergence of commitment... is accompanied by a reduction of attention to alternative relationships" (p, 4505). Leik and Leik (cited in Johnson, 1985, p. 14; 1976b: p. 9), also found that certain types of commitment preclude the awareness of alternatives: "a relationship may reach interpersonal commitment, which means that the members are no longer attending to alternatives." Johnson goes on to discuss which factors in his commitment framework are responsible for this shift away from the monitoring of 21 alternatives. He posits that it is not structural commitment (in terms of the actual availability of alternatives) that is the issue, but rather the attention that the individual pays to what is available. This attention is mediated by personal commitment (Johnson, 1985, p. 15). It is proposed in this thesis that the combination of the Christian view of permanent marriages, and the higher commitment levels associated with such a belief system, will result in a decrease in the monitoring of alternatives. Levinger (1965) noted that "In many marriages, the barriers have trivial importance. The spouses' close attachment precludes that either one would seriously consider breaking the relationship" (p. 20). Johnson (1985, p. 37) suggested that "The symbolic acts in which one engages to display personal commitment serve to restrict the network of other relationships which are available as alternatives." It is hypothesized here that religious couples are less concerned with the alternatives that remain as a result of this restricted network. Rosenblatt (1977, p. 79) found that "a person who is highly committed less often compares his or her spouse with potential alternative spouses." Johnson suggested that the "proposition that commitments lead to the curtailment of exploration of alternatives may apply only when there are high levels of personal commitment" (1985, p. 12). This thesis proposes that it is moral commitment, rather than personal commitment (similar to marital satisfaction) that is the significant factor in predicting the degree to which a subject monitors alternatives. This suggestion is supported by Johnson and Rusbult (1989). These authors found that "the tendency to reject and devalue alternatives was greater under conditions of high commitment" and that this tendency was "more strongly linked to commitment than to satisfaction" (Johnson & Rusbult, 1989, p. 967). 22 The second factor of structural commitment that has been addressed in the literature, but not in connection with the subjects' beliefs, is investment. Investment in a union, (monetary, time, identity, marketable skills, etc.) based on the ideal of lifelong permanence, is proposed to differ from investment in partnerships that are based on behavioral contracts and an acceptance of divorce. The impact of investments on the stability of the relationship is a reciprocal one. The more committed one is, and the more one views the marriage as permanent, the more likely one will be to invest in the relationship. The greater the investment made in the marriage, however, the greater the cost of dissolution, and therefore the more likely one is to maintain the relationship. Investment fosters a more stable marriage, and more stable marriages foster commitment. Another aspect of this reciprocal relationship is based on the attributional perspective. Johnson suggested that within this framework the argument would be "that one's perceptions of investments (or any other form of structural commitment) are inconsistent with perceptions of low personal commitment" (1985, p. 24). In his review of the commitment literature, Johnson found that investment was positively related to commitment, and to behaviors that "were oriented to maintaining the relationship" (1985, p. 23). At the root of the investment discussion is the concept of commitment. It affects, and is affected by, investment. Larson and Goltz stated that "commitment is not the consequence of experiencing a good marriage ... commitment is the individual and relational source of making a good [or even weak] marriage better" (1989, p.397). Moral commitment, in conclusion, is hypothesized to impact both structural and personal commitment as they relate to marital stability. What makes moral commitment such a salient factor is the beliefs that it is based on. These beliefs need not be religiously 23 oriented, as long as they include the aspects of person-specific obligation and the value of marital stability that are found in fundamentalist Christian doctrine. It is suggested that it is not the community or social support that is present in the church that is responsible for the effects of religiosity on stability, as has been proposed in the past. It is possible to have a belief system that is non-religious that reflects high moral commitment, as it is possible to be embedded in a church community but not have a belief system that is consistent with high moral commitment. The beliefs associated with Christian fundamentalism are clear examples of the types of ideas that translate into high moral commitment (such as believing marriage should be permanent and that one has a responsibility to honor the promises made in the wedding vows). It is proposed that external aspects of religiosity, such as church attendance and social pressure, are not as salient in their effects on marital stability as are the internal aspects reflected in the participants' belief systems. Personal commitment. Religiosity has consistently been found to increase marital stability (e.g., Edwards, Johnson & Booth, 1987), but it has usually been suggested that this stability is rooted in external pressures to stay in the relationship. Implicit in the suggestion that the external aspects (i.e., constraints) of religiosity affect stability is the idea that stability is present despite low marital satisfaction. Recently, an increasing number of studies have shown that religiosity is associated with satisfaction and/or adjustment as well (Heaton, 1984; Filsinger and Wilson, 1984; Larson and Goltz, 1989; Robinson, 1994). However, this aspect of religiously-based marriage is often overlooked. It had been suggested that social desirability or conventionality confounded the measures of satisfaction in religious populations, but this 24 argument has also been refuted in several studies (e.g., Filsinger and Wilson, 1984). Heaton (1984), in a study of religious homogamy, found that the greater church attendance associated with religiously homogamous marriages was responsible for a higher level of marital satisfaction. He suggested that "more attention ... be given to the role of husbands' and wives' level of involvement with organized religion" in the study of marital satisfaction (Heaton, 1984, p. 733). Johnson (1985) found that commitment was associated with marital stability, and it is suggested here that commitment is related to satisfaction in addition to increased stability. More specifically, it is proposed that beliefs will affect moral commitment, which will affect satisfaction (one aspect of personal commitment), and both will in turn affect marital stability. In the study of religiosity as it relates to marital stability, there is often an overemphasis on the social constraints, or structural factors. Albrecht and Kunz (1980) found that social pressure was related to stability, and that the opinions of others were barriers to overcome on the road to divorce, but in his review of this literature Johnson found their evidence less than compelling due to flaws in their research. This thesis goes beyond the social constraint/social support literature to offer alternative explanations for both the satisfaction and stability present in committed marriages. Summary of Assumptions in Revised Commitment Model This thesis is concerned with the relationship between religiosity, moral commitment, and marital stability. It is also concerned with modifying Johnson's existing commitment theory so that it includes the belief systems of the test subjects in its examination of stability. In order to do this, the following changes and assumptions have been included in the model: 25 1. The centrality of the external/constraint variable, structural commitment, has been reduced. It is proposed that the importance to the individuals of the external measures is contingent on their level of moral commitment. 2. The factor "value of consistency" that Johnson included in his moral commitment construct has been deleted in order that the belief issues that this thesis proposes are core can be more readily exposed and measured. Johnson (1991) admits that "there has as yet been no serious work completed on the measurement" (p. 121) of the value of consistency. Since this factor does not address the types of beliefs that surround the marital relationship, but rather general societal norms regarding a wide range of behaviors, it has been discarded here for simplicity. 3. The dynamic nature of moral commitment is questioned. Fishbein (1967) views beliefs as relatively stable for individuals. It is proposed that in subjects with high moral commitment, the value of stability and person-specific obligation are stable variables (i.e., moral commitment is not a 'feeling' which may decline, it is a belief that is internalized by the subject). This alternative approach to the concept of moral commitment is reflected in the wording of the questionnaire. 4. Johnson's conception of structural commitment variables, such as availability of acceptable alternatives, as entirely external constraints, is questioned. Although Johnson emphasizes the subjective nature of all his variables, this thesis proposes that the 'external' constructs are mediated by an internal set of beliefs. To reflect this thinking, Johnson's variables have been reworded. For example, the availability of acceptable alternatives is presented as the attention to alternatives, which encompasses both availability and the subjects' perception of the importance of the alternatives. 26 5. Finally, Johnson's use of three stages of marital dissolution (motivation, development of plans, and action to maintain or dissolve the marital relationship) has been replaced with a simple marital instability measure that is more useful in cross-sectional surveys. Hypotheses Hypothesis 1: Christian fundamentalism is positively related to moral commitment. Hypothesis 2: Moral commitment is positively related to marital stability. Hypothesis 3: Personal commitment is positively related to marital stability. Hypothesis 4: Structural commitment is positively related to marital stability. Hypothesis 5: Moral commitment is positively related to personal commitment. Hypothesis 6: Moral commitment is positively related to structural commitment. Hypothesis 7: Moral commitment moderates the effect of personal commitment on marital stability. Hypothesis 8: Moral commitment moderates the effect of structural commitment on marital stability. The following model illustrates the hypothesized relations: 27 Figure 1. Revised Theory of Commitment PERSONAL COMMITMENT - attraction to the relationship - attraction to the spouse - relational identity - religious identity CHRISTIAN FUNDAMENTALISM STRUCTURAL COMMITMENT irretrievable investments social reaction attention to alternative attractions MORAL COMMITMENT - person-specific obligation - value of stability MARITAL STABILITY difficulty of termination The following figures are graphic representations of the moderating effect that moral commitment is proposed to have on personal (Figure 2) and structural (Figure 3) commitment. 28 Figure 2. The Effect of Moral and Personal Commitment on Marital Stability 0 low moral commitment high moral commitment MORAL COMMITMENT HIGH PERSONAL COMMITMENT LOW PERSONAL COMMITMENT 29 Figure 3. The Effect of Moral and Structural Commitment on Marital Stability MARITAL STABILITY 5 0 low moral commitment high moral commitment MORAL COMMITMENT HIGH STRUCTURAL COMMITMENT LOW STRUCTURAL COMMITMENT 30 Method Sample and Data Collection Four Protestant churches were selected to participate in this study. Two Brethren (a traditionally fundamentalist denomination) and two United Churches (traditionally liberal) were sampled. Each of these churches had full-time congregations of approximately 150 people. The two denominations were matched by geographic location to control for economic and ethnic differences. The Brethren churches sampled were Heather Bible Chapel located at 68th Avenue in Vancouver, and the Westminster Gospel Chapel located on 6th Street in Burnaby. The United Churches were the Marpole location at 67th Avenue in Vancouver, and the East Burnaby location on Graham Street. The United and Brethren Churches in Vancouver (Marpole and Heather) were located within 5 blocks of each other, as were the two churches in Burnaby (East Burnaby and Westminster). Contact was initially made with the churches through the ministers at the United churches, and through an elder at each of the two Brethren churches. Permission was given after these initial contact people consulted with members of their church committees. Permission was acquired in all cases over the phone. Questionnaires were given to staff in each of the churches to be made available to the congregations. Fifty questionnaires were distributed to each assembly. Notices were placed on bulletin boards at each church to recruit participants (see Appendix C), and a cover letter was included with each questionnaire (see Appendix D). The cover letter gave instructions on how to complete the questionnaire, and what to do with the finished survey. Each 31 questionnaire was accompanied by a blank envelope, and the respondents were instructed to seal the completed questionnaire in the envelope before returning it to the church staff member. Short announcements were made at all four churches advising the congregation of the study, and telling them how to participate. Participation was voluntary and anonymous, with no remuneration given. Participation was limited to currently married individuals. Of the two hundred questionnaires distributed, sixty three usable surveys were returned (a response rate of approximately 30%). Three questionnaires were discarded because they were incomplete. The questionnaire was 10 pages long, and contained 108 items from various pre-existing and new measures. Measures Dependent Variable: Marital Instability This study is a cross-sectional examination, therefore, the focus for this measure is the degree of instability in currently intact marriages, rather than on subsequent separation or divorce in participants. Edwards, Johnson, and Booth (1987) developed the Marital Instability Index in order to predict marital dissolution, and this scale was selected to measure stability for this study. Edwards, Johnson, and Booth developed the Marital Instability Index because the available measures failed to "detect the severity of the signs of instability ... and yield a prognosis of the future course for a particular marital relationship" (1987, p. 168). The scale they developed measures two main components, cognitions and behavior, in the marital relationship. The authors examined almost 40 items related to these two areas, eliminating 32 those that were not universally applicable, or loaded poorly on factor analysis tests (p. 168). They performed reliability and validity tests on the resultant measure. Reliability was high, with an alpha of .93 (p. 168) and validity was .80 (Spearman correlation). The predictive value of the measure was calculated by comparing the divorce statistics from national probability samples collected in 1980 and 1983. Only 3% of those showing low marital instability in 1980 had divorced in 1983 (p. 169). Of those that scored high on instability, 27% had divorced in the three years between tests (p. 169). Booth, Edwards, and Johnson succeeded in developing a scale which measures the current degree of instability, as well as the future propensity to divorce. They discovered that there is a typical order of events in the dissolution process: moving from unfocused perceptions of marital troubles to specified conceptions that divorce is desirable, to engaging in behaviors dealing with the perceived problems and, finally, taking actions designed to terminate the relationship, (p. 170) A strength of the marital instability scale is that it measures instability even in those marriages that may not end in divorced. The authors who developed this scale defined marital instability as "the propensity to dissolve, even if dissolution was not the final outcome" (Edwards, Johnson, & Booth, 1987, p. 168). The response scale for this measure varies from 'never' to 'very often' on a four-point scale (see Questionnaire, Appendix B). Items 68 to 80 of the questionnaire are from this Marital Instability Index. These questions represent the first section of the measure presented by Booth, Edwards and Johnson (1987). They include another section which they entitle "Attractions and Barriers to Divorce" (p. 170). These items are excluded from this study 33 because other scales that more closely adhere to the model conceptualized by Johnson are available to test these variables. The Marital Instability Index was scored in a slightly different manner for this study. This measure was originally scored based on how many times the respondent chose 'occasionally' or 'never' from the response options, as opposed to 'very often' or 'often.' In effect, the authors scored only two responses, even though the participants were given four options. In this study, numbers were assigned to each of the four response options (e.g. 1 = never, 2 = occasionally, etc.), and the scores from each question were added together. Questions 7 & 14 of the Marital Instability Index were altered slightly so that they were appropriately worded for the married sample (#73 and 80 this questionnaire). The words 'husband's/wife's' and 'husband/wife' were replaced with the word 'spouse's' and 'spouse' respectively. The second minor change that was made was the replacement of the word 'attorney' in the Mil (questions 20 & 21) with the word 'lawyer' in this questionnaire (questions 76 & 77) to be consistent with Canadian as opposed to American terminology. Independent Variable: Personal Commitment The construct of personal commitment, as conceptualized by Johnson, contains the following factors; relational identity, attraction to the relationship, and attraction to the spouse. For two of the three factors, pre-existing scales were available in the literature. The third factor (relational identity) was measured by a question specifically developed for this study. 34 Relational and religious identity. This item (question #7, see Appendix B) was constructed specifically for this study because there were no existing scales in the literature that addressed this concept. This item was pretested before final distribution of the questionnaire. Eight volunteer participants filled out a survey for the pretest. All respondents answered this question with responses that fit the conceptual basis for this variable. All eight participants mentioned their spousal relationship, and all the parents in this pretest sample (six of the eight had children) mentioned their parental roles. The two fundamentalist Christians in the sample mentioned religion (i.e., Christian, Sunday school teacher). All eight respondents also included their occupation in their answers. All the respondents mentioned either familial relationships other than parent or spouse (e.g., aunt, sibling), and/or non-familial relationships (e.g., friend). The following question was used to measure this variable: In the spaces below, please list, in order of importance to you, the roles you occupy that you feel are most strongly linked to your identity (e.g., doctor, parent, friend, student, etc.). I am a ... 1. 2. 3. etc. The concept of relational identity is poorly developed by Johnson. Johnson notes that relational identity's "full incorporation into the commitment framework remains to be accomplished" (1991, p. 121). For the purposes of this study, relational identity is defined 35 as "the extent to which one's involvement in a relationship is incorporated into one's self-concept" (Johnson, 1991, p. 120). For this study, a concept that was originally intended to measure spousal relations was extended to include other intimate relationships (e.g., friend, sibling) and religious identity. Religious identity is defined as being present when one's religious roles (cognitive or behavioral) are an integral part of one's identity. Relational and religious identity were analyzed separately, since the religious component was not part of Johnson's model. Nine pertinent response possibilities were coded for this study. If the participant answered the question but did not give one of the relevant responses, the question was coded as 'not applicable.' If the question was left blank, it was coded as 'missing.' If the participant mentioned church-related roles (e.g. elder, church volunteer, Sunday school teacher) they were coded as having religious identity. If they mentioned an interpersonal role (e.g. spouse, mother), they were coded as having relational identity. In addition, prestige scores were computed from this question using the Pineo, Porter, and McRoberts (1977) scale to measure occupation. Attraction to the relationship. Questions 16 through 25 of the survey measure this aspect of personal commitment. This component of personal commitment was measured using a marital satisfaction index. Questions 21 through 30 of. The Dyadic Adjustment Scale (Spanier & Thompson, 1982, p. 736) were used in this questionnaire. The excluded items address attraction to the spouse and marital stability, which, according to the model being tested in this study, are conceptually different from attraction to the relationship, and are tested by other scales in the survey. 36 Spanier and Thompson evaluated Spanier's original Dyadic Adjustment Scale. They factor-analyzed the scale, and tested it on a sample. The original scale is a 32 item survey used to assess the quality of marital and other dyads. The original scale had a Cronbach's coefficient of .96. Spanier and Thompson retested the reliability of the measure, and also examined it to determine whether Spanier's original four factors were applicable to the new, more homogeneous sample. The Dyadic Adjustment Scale was tested by Spanier and Thompson (1982) using face-to-face interviews. Individuals who were separated between 1975 and 1977 in Centre County, Pennsylvania qualified for their study. A pool of 918 people was identified, with 210 finally completing the questionnaire. The sample included 50 separated and 155 divorced individuals. A Cronbach's alpha of .91 was achieved for this scale for Spanier and Thompson's sample. The majority of the questions used in this survey (see Appendix B) came from the items described by Spanier and Thompson as measuring satisfaction. The remaining questions are from the cohesion or affectional expression measures. The fourth measure identified in the scale was consensus. The authors report that three of the four factors were highly interrelated. This is expected in a global measure like dyadic adjustment, according to Spanier and Thompson (1982, p. 737). Affectional expression was less closely related to the other items. These items were not unrelated to the other three factors, however. The items that were chosen to be included in this survey were related in that they assessed aspects of the relationship, namely patterns of interaction, rather than characteristics of the spouse (more closely related to the concept of attraction to the spouse). Spanier and Thompson concluded that the Dyadic Adjustment Scale is both reliable and valid (p. 737). 37 Additional reliability tests were performed on this scale for this study because the original instrument was altered (see questionnaire, Appendix B). The wording of number 25 in Spanier and Thompson's scale was changed from 'Stimulating exchange of ideas' to 'have a stimulating exchange of ideas' because the instructions ask the respondent to indicate how often they and their spouse engage in the activities listed. Attraction to the spouse. Lund defines love as "positive feelings about a particular person" (Lund, 1985, p. 3). She conceptualizes this variable as having three components: attachment, intimacy, and caring (p. 5). Lund developed this love scale to assess the roles of love and commitment in predicting the continuation of relationships. Lund explores how investment and commitment are interrelated, and how they affect the impact that love has on the longevity of relationships. Lund proposes that the univariate attraction model used to explain the beginning of the relationship is too restricted to explain why some of these relationships continue and others do not. Lund describes commitment, which she proposes will enhance the predictive power of the model, as "an attitude about continuing a relationship that is strengthened by a person's own acts of investing time, effort and resources in that relationship" (1985, p. 4). Lund proposes that commitment leads to investment in the relationship, which in turn leads to a greater expectation that the relationship will continue; the relationship between commitment and investment is seen as circular and reciprocal. To test these propositions about love, commitment, and investments, Lund first ran a study to test the scales, and then a second longitudinal study to test her hypotheses about these variables. Lund used a shortened version of a previously developed scale to measure 38 love. Lund tested her scales on a sample of 30 men and 30 women from college classes. The participants answered open-ended questions about investment in relationships, commitment, relationships iii general, and obligations to continue relationships (p. 8). The items that resulted from these open-ended questions were given to 50 men and 61 women from UCLA. The first part of the longitudinal study was run in February of 1980. College students comprised the sample. The participants were given a 14-page questionnaire. The follow-up survey was administered at the end of June of the same year. A second 14-page questionnaire was mailed out that contained the same scales as the first, but with additional questions on changes in their relationships. The sample consisted of 129 individuals who were graduating from college. Fifty men and 79 women were tested. Items number 8 to 15 of this questionnaire (see Appendix B) are from Lund's Love Scale (1985; adapted from Rubin, 1970). A Cronbach's alpha of .88 was reported by the author for this nine-item scale during pretesting, and .91 in her longitudinal study. Lund found that the measure used for love was conceptually different from her commitment scale. One of Lund's 9 items was omitted from the questionnaire used in this study (question 8) because it does not clearly measure attraction to either the relationship or the spouse. This section of the questionnaire is directed at measuring attraction to the spouse; attraction to the relationship is measured by a different scale. The wording of the remaining questions was changed. Whenever the words 'partner' or 'partner's' appeared in Lund's scale, they were replaced by the words 'spouse' or 'spouse's' respectively, to be consistent with the marital status of the participants. Three of the eight questions used were worded in the negative to prevent a positive response bias. The questions which were reworded in the negative are (in 39 Lund's scale) numbers 1, 6, and 9. The corresponding questions in this survey are 8, 13, and 15 (see Appendix B). Independent Variable: Structural Commitment The construct of structural commitment proposed by Johnson (1991) has four factors: attention to alternative attractions, investment, the perception of the difficulty of termination, and the influence of social reaction. These four components are included in the questionnaire. Investment. Lund describes investment as an accumulation of "time, effort, and resources" (Lund, 1985, p. 6). Numbers 26 through 41 of the questionnaire (see Appendix B) measure this aspect of structural commitment (all taken from Lund, 1985). Lund's Investment Scale was used to determine the degree of investment (monetary, time, emotional, etc.) of the participants in their marriages. Some of the items in Lund's scale were targeting non-marital relationships (e.g., "Making formal agreements about your relationship such as deciding to go steady, get engaged or get married") and were dropped from the measure because they were not relevant to currently married couples in this sample (numbers 2, 7, 8, 9, 10, 11, 14, 16, 17, and 22 of Lund's scale were omitted). The wording of the remaining questions was altered slightly to be relevant for the married sample. Wherever the words 'partner' or 'partner's' appeared, they were replaced with the words 'spouse' or 'spouse's' respectively. In addition, the tense of the sentences was altered (e.g., the word 'spending' was replaced with the word 'spend'). A Cronbach's alpha of .93 is reported by Lund for the Investment Scale during pretesting, and .96 during the 40 longitudinal study. Validity was assessed by examining the scale's association with other measures, such as commitment, and was higher in groups with higher degrees of exclusivity in their relationships. Since not all questions from Lund's scale are included in this study, the validity and reliability of the investment variable were reassessed using item analysis. Attention to alternative attractions. Green's (1983) Inventory of Alternative Attractions was chosen to measure attention to alternatives. It is a six-item measure developed to assess a subject's perception of their attention to relationship alternatives. As outlined by Green and Sporakowski (1983), the alternatives included in this measure are "personal freedom, sexual relations, spending money, enjoyment from friends, respect from other people, and relationships with parents" (p. 81). Green and Sporakowski developed this scale to test the proposition that quality was not the only variable related to marital stability. According to Lewis and Spanier (1979), both external pressures and the attraction of alternatives also impacted stability. Green and Sporakowski tested these propositions on 166 divorced individuals from Richmond, Virginia. The sample consisted of people who had divorced in 1979. A five-point Likert-style format was used for the Inventory of Alternative Attractions, with response options ranging from 'never' to 'always.' The measure does not appear in the 1983 article, but was made available by Sporakowski for use in this study. The scale was developed by Green, who obtained a Cronbach's alpha of .88 for this measure. Concurrent criterion validity was obtained by comparing the scores of divorced (16.9) with married (10.4) participants. Green reports that validity and reliability did not differ by gender. 41 Numbers 59 through 64 of the questionnaire (see Appendix B) are from the Inventory of Alternative Attractions. All the items from Green's inventory were used, and no modifications were made to the wording of the questions. The introduction or instruction section of the scale was modified, however. The second sentence of the introduction was altered. In Green's version it reads; "We are interested in the attractions you had to people and things outside of your marriage during the last six (6) months you and your husband/wife lived together." In this study the second sentence reads; "We are interested in the attractions you have to people and things outside of your marriage." The wording was changed to be consistent with the sample used in this study (i.e., currently married individuals). Influence of social reaction. The influence of social reaction component of structural commitment was measured by Green's (1983) Inventory of External Pressures. This measure was made available for use in this study by Sporakowski. Green and Sporakowski tested an extensive list of external pressures in their study of marital stability (1983). These were: obligations to their marital vows, religious beliefs and obligations to their dependent children, as well as the financial costs of divorce and the pressure from five social networks: their neighbors, their workmates, the people in their church, their relatives, and their friends, (p. 82) Green and Sporakowski used the entire 21-item measure for their study. This study used a reduced, 17-item Likert-style measure (ranging from 'never opposed' to 'always opposed' or 'a great deal' to 'none'). Four questions were omitted because they dealt with obligations, 42 and these variables are measured separately in the personal commitment section of the survey (person-specific obligation). Cronbach's alpha for the entire original measure was reported at .79. Concurrent criterion validity was obtained by Green and Sporakowski from comparisons of divorced (19.3) and married (23.1) subjects. Questions 42 through 58 of this questionnaire are from Green's inventory (see Appendix B). Items 42 through 53 in the questionnaire used for this study represent the first part of Green's inventory. Items 54 through 58 are from the second part of Green's inventory. No changes or omissions were made to the questions, but the instruction section of each part was modified for this study to reflect the differences in the two study's samples (divorced versus currently married). Difficulty of termination. A two-item instrument was developed by the author to test this component of structural commitment (questions 65 & 66 of the survey). The instrument is as follows: How much trouble would ending your relationship be to you personally? Do you worry about the possible financial implications of getting a divorce? The first item listed above is taken from Lund's (1985) Commitment Scale, and is listed as number 65 in the questionnaire (34 in Lund's scale). The response scale for this item ranges from 'none' to 'a great deal' on a four-point scale. The other eight items in Lund's commitment scale were not used because they were confounded with the other variables in the study (e.g., marital stability, attraction to alternatives, and person-specific obligation). The 43 second question has response options ranging from 'never' to 'always' on a 4-point scale. This measure was pretested before final distribution of the questionnaire. Moderating Variable: Moral Commitment There are two components within this variable: person-specific obligation and the value of stability in the marital relationship. This variable is hypothesized to affect or moderate the relationship between the two independent variables (structural and personal commitment) and the dependent variable (marital stability). Value of stability. The scale used to measure this component of the moral commitment variable is Keri Kinnaird's Attitudes Toward Divorce Scale (1986). The author made the measure available for use in this study. This scale was used by Kinnaird and Gerrard (1986) in their study of premarital attitudes and behavior in young women. They studied women from divorced, reconstituted, and intact families to determine whether their dating and sexual experiences, attitudes towards marriage, or attitudes towards divorce differed. Kinnaird and Gerrard sampled.875 females at a university. The researchers selected 30 women from each of their three study categories (intact, divorced, and reconstituted). Kinnaird and Gerrard designed the Attitudes Toward Divorce Scale specifically for their study. This instrument uses a 5-point Likert-style, 12-item format. The questions were designed to gauge the participants' perception of the acceptability of divorce. The response scale varies from 'strongly agree' to 'strongly disagree.' The scale was pretested by the authors on 50 subjects. Cronbach's alpha for the scale was .77, and test-retest reliability was 44 .86. Numbers 86 through 96 of the instrument used for this study (see Appendix B) are from Kinnaird's scale. All but one of Kinnaird's questions was used. Item number 7 in her scale was omitted because it was confounded with the person-specific obligation measure used in this study. Kinnaird's question number 7 reads: "People should feel no great obligation to remain married if they are not satisfied." Kinnaird's question 5 was changed: the word 'American' was replaced with 'Canadian' to be more applicable to the sample (question 90, this questionnaire). Kinnaird's questions are worded in both positive and negative directions to avoid bias. Person-specific obligation. The second component of moral commitment, person-specific obligation, was examined by questions developed by the author specifically for this study (questions 81-85 of the instrument) because there were no pre-existing measures available in the literature. Because it is a new item it was pretested for face validity and test reliability. An example from this measure is as follows (item 81): I do not believe you have an obligation to stick with a bad marriage because of the vows you made to your partner. The response scale ranges from 'strongly agree' to 'strongly disagree' on a four-point scale. 45 Independent Variable: Christian Fundamentalism Altemeyer and Hunsberger (1992, p. 118) describe religious fundamentalism as the belief that there is one set of religious teachings that clearly contains the fundamental, intrinsic, essential, inerrant truth about humanity and deity; that this essential truth is fundamentally opposed by forces of evil which must be vigorously fought; that this truth must be followed today according to the fundamental, unchangeable practices of the past; and that those who believe and follow these fundamental teachings have a special relationship with the deity. In the religious studies literature, extrinsic and intrinsic religiousness are common distinctions (e.g., Donahue, 1985; Hunsberger, 1995). Donahue defines intrinsic religiousness as "a meaning-endowing framework in terms of which all of life is understood," while extrinsic religiousness is defined as "the religion of comfort and social convention, a self-serving, instrumental approach shaped to suit oneself (1985, p. 400). In extrinsic religiosity, religion is part of life, while in intrinsic religiosity, religion is the meaning of life (Donahue, 1985, p. 414). Donahue concludes that "intrinsic religiousness serves as an excellent measure of religious commitment, as distinct from religious belief, church membership," and other religious variables (1985, p. 415). Fundamentalism is measured by a scale designed for this study to reflect the definitions of intrinsic religiousness outlined above. The measure includes questions on prayer time, specific beliefs presented in the Bible, religious homogamy, and the participants' perception of the importance of their religion in their daily lives. The items measuring 46 religious fundamentalism were developed by the author for this study (item numbers 97 through 108). They were tested for face validity and reliability. Control Variables Questions on the following control variables are drawn from various sources: gender, ethnicity, age, length of marriage, previous divorces, and number and age of children. Items 1 through 6 of the questionnaire are from White (1983). White developed a 33-item scale for the "Family Career and Family Structure Research Project." The first 12 questions are general SES-type questions, of which 6 were included, with modifications, in this study. Questions 1,2,3,5,7, and 12 are from White's measure (see Appendix B). An example of this instrument is as follows (item 12): In which of the following groups would you place yourself? Francophone Chinese Japanese Native Indian East Indian Greek Eastern European , British Italian German American Other The category 'American' was replaced by 'Canadian' for this study. Pretesting of the questionnaire determined that subjects preferred this categorization. Question number 5 was altered for this study. White's original question reads: 47 For how many years have you had your present marital status? (1) 0-3 years (4) 11-20 years (2) 4-5 years (5) 21 or more (3) 6-10 years In this study, this question (item 4) was simplified to read: How many years have you been married to your present spouse? The item measuring the respondents' age was also simplified, leaving a single blank for a number, rather than offering six response categories. Respondents were also asked to list the ages of their children, so that one question measured number of children, average age of those children, and the presence of preschool children. Results Description of Sample Seventeen individuals and 22 couples returned the survey. Because couple responses are inherently interdependent, data analysis was performed on males and females separately, rather than on all individuals. It had been hoped that only couples would volunteer to participate in the study. In order to handle the complication of a mixed response group, the author considered doing separate analyses on the couples and individuals. Unfortunately, the small sample size precluded this. It was decided that separating the sample by gender was 48 preferable for retaining respondents and eliminating dependent couple responses. Sixty three people returned complete questionnaires. The distribution of the respondents between the churches was fairly even. Twenty questionnaires were returned from each of the two Vancouver churches (one Brethren and one United), 8 from the United church at the Burnaby location, and the remaining 14 from the Brethren church in Burnaby. Of the sample of 63, 31 were male and 32 female. The mean age of the males in the sample was 60; for females it was 54 (see Appendix F). Of these respondents, 50% were over 50 years old, and 25% were over 70 years old. The respondents had been married between 1 and 58 years. Fifty percent had been married 30 years or more. The men had been married, on average, 32 years; the females averaged 28 years (see Appendix F). Approximately 14% of the respondents did not have children. The age of the sample is reflected in the age of their children — more than 50% of the respondents had children older than 30 (see Appendix F). Ninety-two percent of the sample did not have children under the age of six. Almost 80% of the sample described themselves as 'Canadian,' with the 'British' category describing a further 13%. Two of the 31 men, and three of the 32 women, had previously been divorced. Univariate Distributions Dependent Variable The dependent variable in this analysis, marital stability, was seriously skewed (see Appendix G). A histogram of the frequency revealed that there was insufficient variation in the responses to this measure. The possibility of there being multiple factors within the 49 variable was considered because of the large number of questions in this measure (12). Factor analysis (VARIMAX orthogonal) was performed on the marital stability scale, and 2 dominant factors were discovered. These 2 factors explained 70.1% of the total variance in this measure. A correlation was performed on the two variables resulting from the factor analysis. It was found that, in a simple listing of the variables without weighting by factor weights, the two variables were correlated at .53. When the frequencies were run on these two new factors, the results showed that these factors also contained insufficient variation. These two factors, conceptually, represent two distinct stages in the process of marital dissolution. The first factor measures the thought process that accompanies increasing marital instability (e.g., has the participant or their spouse ever thought their marriage was in trouble). The second factor measures a more advanced stage of dissolution, that which involves action. The questions in this measure tap whether or not the respondent has talked with friends, family, etc. about their dissatisfaction with their marriage. Because the two new factors discovered through factor analysis were conceptually distinct, but still lacked sufficient variation to be useful in analysis, it was decided that one dummy variable would be created. If the respondent had ever thought about divorce or ever talked with others about divorce, they scored a zero on the new dummy variable. If they had not thought or spoken about divorce, they scored a one. Of the 31 men in the sample, 10 had thought or spoken of divorce. Fourteen of the 32 women had also contemplated divorce. Logistic rather than multiple regression was used for the analyses involving this dummy variable. 50 Independent Variables Several manipulations were performed during the analyses to accommodate the limited sample size, and diverse set of measures. One of the first challenges encountered was that, because of the small sample size, the number of missing items was large enough to create problems in the bivariate analyses. In order to combat this situation, those scales which contained more than 4 cases of missing data were examined to determine whether any items could be eliminated without affecting their reliability scores. The eight scales used to measure the independent variables were examined by frequency. The items that showed missing data were removed, and new scales created. These new scales were then subjected to reliability tests. The resulting alpha values were all above .68 (most were over .8), so the new scales were used for the remainder of the analyses. All summative scales with different response categories were computed by calculating z-scores for all the variables required for hypotheses testing. The various measures being used had different response scales. Because the scales were all combined into larger scales, z-scores were compiled for all the independent variables. In one case (the Christian fundamentalism scale), z-scores were calculated on each item of the measure, because of different response scales within this item. Structural commitment. This variable is made up of three sub-scales, and a one-item measure. Reliability tests were performed on the variable, and it was determined that the measures designated to make up the structural commitment scale were not related (alpha = .0155). In the remainder of the analyses these items are tested individually. 51 The three sub-scales used for this variable had normal distributions (see Appendix G), once the problem of missing values had been addressed (eliminating those items with more than 4 missing cases). However, the single item measure, difficulty of termination, was highly skewed (see Appendix G). The item has a four-item response scale, and the majority of the responses fell on only one of these choices. The measure was recoded so that all the low responses were coded as 0, and all the remaining responses were coded as 1. One indicates a high degree of difficulty in terminating the relationship, zero signifies low difficulty. Because this measure was recoded in this manner, all regression analyses performed on it were done by logistic regression. Personal commitment. Reliability tests were performed on this four-item scale. Two of the four items were unrelated. For the remaining analyses religious and relational identity were assessed separately. Attraction to the relationship and attraction to the spouse were related, and constitute the personal commitment scale in the remaining analyses (alpha = .6925). The relational and religious identity scales were taken from the same group of questions (item 7, see Appendix B). Participants were instructed to list five roles that they felt were important to their identity. If the respondents mentioned their spouse, they were coded as 1 on the relational identity item. If they responded with a mention of Christianity or church involvement, they were coded one for religious identity. If they did not mention either of these roles, they were coded as zero. 52 Moral commitment. The two scales that comprise the moral commitment variable, the value of stability and person-specific obligation, were significantly related (alpha = .8929). These two variables did not require any further manipulation, other than to calculate z-scores once the questions with missing values had been removed. Christian fundamentalism. The scale measuring Christian fundamentalism was highly skewed (see Appendix G). To combat this problem, it was recoded into 4 results, low through high, with approximately 25% of the respondents in each category. The measure was also culled of questions with high missing values, and was still highly reliable. In addition, a z-score was calculated on the resulting measure to compensate for the differing response scales found throughout the questionnaire. The two responses that resulted from the z-score calculation were both negative, so a constant of 7.0 was added to both categories to transform it into positive values. Control variables. The following control variables were added to the analyses once the significant relationships in the hypotheses had been determined: age, gender, ethnicity, presence of preschool children, length of marriage, occupation, church region, church group, and religious homogamy. The presence of preschool children was coded 1 for present and 0 for absent. Occupation was taken from the identity question (item 7), and coded into the five general categories in the scale used by Pineo, Porter, and McRoberts (p. 98, 1977). This 53 scale, although not recent, has not changed substantially in terms of the macro-categories. If the participant did not mention an occupation as part of their role or identity, they were assigned the norm for the entire sample. The relationship between marital stability and occupation was assessed using the means for those who did not mention an occupation, and this was compared to the relationship using missing values for these respondents: neither was significant, so the remaining analyses were performed using the means. Region and church group were assigned from the same question. Initially, the four churches received arbitrary numerals. To code for region, the churches in Vancouver were given the same code, and those in New Westminster were given an alternative code. For church group, the fundamentalist churches were coded together, and the United assigned the alternative code. Each of the control variables was tested individually against the dependent variable in the model, marital stability, using logistic regression (the dependent variable is coded with 1 and 0). The only two control variables that were significantly related to marital stability were age (R = .3616, p = .0054) and length of marriage (R = .3646, p = .0051). These two items were only significantly related for females. Because these two variables were highly correlated (correlation coefficient for females: .927), only length of marriage was included in the analyses. Hypothesis Testing Bivariate Analysis Hypothesis 1: Christian fundamentalism is positively related to moral commitment. The moral commitment scale, comprising person-specific obligation and the value of 54 stability, was significantly related to Christian fundamentalism for females (B = .570, R = .322, p = .012) but not for males (B = .104, R = .035, p = .152). This relationship for females was still significant after the control variable, length of marriage, was added to the analysis (B = .714, R = .344, p = .008). This relationship was tested using logistic regression because the dependent variable, moral commitment, is coded with 1 and 0. Individually, the components of moral commitment were also related to Christian fundamentalism. Person-specific obligation was significantly related to Christian fundamentalism for females (b = .341, p = .056), but not for males (p = .278). Multiple regression was used for this test because the dependent variable, person-specific obligation, uses an interval response scale. For females, Christian fundamentalism explained 11% of the variation in person-specific obligation. Adding length of marriage as a control variable to the regression equation removed the significance of the relationship for females (b = .332, p = .069). The value of stability was significantly related to Christian fundamentalism for both females (b = .609, p = .0002) and males (b = .388, p = .031). Multiple regression was used to test this relationship because the dependent variable, the value of stability, is an interval scale variable. Christian fundamentalism explained 15% of the variation in the value of stability for males, and 37% among females. These relationships were still significant when length of marriage was controlled for (females: b = .627, p = .0002; males: b = .390, p = .032). Length of marriage added little to the explained variance between these two variables. 55 Table 1. Summary of Results from Hypothesis 1. Independent variable: Christian fundamentalism Dependent variables: moral commitment, person-specific obligation, value of stability Gender/ Males Females Males Females Dependent without without with controls with controls Variable controls controls moral not significant LR significant not significant LR significant commitment B=.570 B=.714 R=.322 R=344 p=.012 p=.008 person-specific not significant MR not significant not significant obligation b=.341 p=.056 11% of variance explained value of MR MR MR MR stability b=388 b=.609 significant significant p=.031 p=.0002 15% of 37% of variance variance explained explained MR = multiple regression / LR = logistic regression Hypothesis 2: Moral commitment is positively related to marital stability. None of the relationships covered by this hypothesis was significant. The individual components of moral commitment, person-specific obligation, and the value of stability, were not significantly related to marital stability, and neither was the composite scale. All the analyses used to test this hypothesis were done using logistic regression because marital 56 stability, the dependent variable, is coded with 1 and 0. Hypothesis 3: Personal commitment is positively related to marital stability. -A This hypothesis was supported by the study. All analyses performed on this dependent variable, marital stability, are logistic regression because it is coded with 1 and 0. Two of the four items originally intended for use in the personal commitment scale were found to be significantly related to marital stability. Relational and religious identity were not related to marital stability. The personal commitment scale, comprising attraction to the relationship and attraction to the spouse, was significantly related to marital stability for both males [B = 2.550, R (n= 31) = .370, p = .007] and females [B = 3.017, R (n=32) = .336, p = .008]. The relationships were still significant when the control variable, length of marriage, was added to the analysis (females: B = 3.433, R = .306, p = .013; males: B = 2.590, R = .332, p = .012). When the two components of personal commitment were tested individually against stability, males and females both showed significant relationships. For males, attraction to the relationship was significantly related to stability [B = .329, R (n=31) = .367, p = .007]. For females, the relationship was also significant (B = .429, R = .338, p = .008). The results were still significant for both genders when length of marriage was controlled for (females: B = .604, R = .271, p = . 022; males: B = .341, R = .381, p = .006). For females, attraction to the spouse was significantly related to stability [B = .524, R (n = 32) = .384, p = .004], as it was for males (B = .367, R = .311, p = .016). For females (B = .618, R = .284, p = .019) and for males (B = .360, R = .270, p = .028) the relationships were still significant with length of marriage added as a control. 57 Table 2. Summary of Results from Hypothesis 3*. Dependent variable: marital stability Independent variables: relational identity, religious identity, personal commitment, attraction to relationship, attraction to spouse Gender/ Males Females Males Females Independent without without with controls with controls Variables controls controls relational not significant not significant not significant not significant identity religious not significant not significant not significant not significant identity personal B=2.550 B=3.017 B=2.590 B=3.433 commitment R=370 R=336 R=332 R=.306 p=.007 s p=.008 p=. 012 p=013 attraction to B=.367 B=.524 B=360 B=618 the spouse R-.311 R=384 R=.270 R=.284 p=.016 p=.004 p=.028 p=.019 attraction to B=.329 B=.429 B=.341 B=.604 the R=367 R=.338 R=.381 R=.271 relationship p=007 p=.008 p=008 p=.022 all analyses performed are logistic regression Hypothesis 4: Structural commitment is positively related to marital stability. All regression analyses performed on the marital stability variable were logistic regression because of the binomial (1/0) nature of this scale. Investment was significantly related to marital stability for males only [B = 3.113, R (n=26) = .366, p = .007]. For females, there was no significant relationship between stability and investments [B = 11.233, 58 R (n=30) = .000, p = .805]. For males, the relationship between marital stability and investments was still significant once length of marriage was controlled for (B = 3.048, R = .352, p = . 009). Alternative attractions were significantly related to marital stability for females (B = -1.522, R = -.190, p = .059) in a negative direction. For males, this relationship was not significant (B = -.511, R = .000, p = .515). The relationship between alternatives and stability was no longer significant for females once length of marriage was controlled for (B = -.6631, R = .000, p = .490). Social reaction was significantly related to marital stability for females (B = 1.069, R = .230, p = .038), but not after adding length of marriage as a control variable (B = .775, R = .000, p = .398). This relationship was not significant for males (B = .501, R = .000, p=.521). The same gender difference was present in the relationship between the difficulty of termination and marital stability. For males this relationship was not significant (B = .799, R = .000, p = .587), while for females it was (B = 2.420, R = .236, p = .038), even with the control variable (B - 2.777, R = .214, p = .048). 59 Table 3. Summary of Results from Hypothesis 4*. Dependent variable: marital stability Independent variables: investments, alternative attractions, social reaction, difficulty of termination Gender/ Independent Variables Males without controls Females without controls Males with controls Females with controls investments B=3.113 R=.366 p=.007 not significant B=3.048 R-352 p=.009 not significant alternative attractions not significant B=-1.522 R=-.190 p=.059 not significant not significant social reaction not significant B=1.069 R=.230 p=.038 not significant not significant difficulty of termination not significant B=2.420 R=.236 p=.038 not significant B=2.777 R=.214 p=.048 all analyses were performed using multiple regression Hypothesis 5: Moral commitment and personal commitment are positively related. Males and females differed in how these variables were related. Four factors comprise this dependent variable: attraction to the spouse, attraction to the relationship, religious identity, and relational identity. 60 Religious identity. The first test was done using multiple regression on religious identity, an interval scale dependent variable. For males, religious identity was not related to the moral commitment scale, or to the two individual measures within it. The most significant relationship for females was between religious identity and the composite moral commitment scale (b = .383, p = .027). Moral commitment explained 15% of the variation in religious identity among females. The second test was done on the same dependent variable (religious identity), but against one component of the moral commitment scale, person-specific obligation. Multiple regression was used for this analysis because religious identity is an interval-type scale. Person-specific obligation was not strongly related to religious identity for females (b = .063, p = .085), and was unrelated for males (p = .594). Person-specific obligation explained approximately 10% of the variation in religious identity for females, although the significance was low for this relationship. The value of stability was significantly related to religious identity for females, however (b = .044, p = .040). Approximately 13% of the variance in religious identity was explained by the value of stability for females. For females, the relationships between religious identity and moral commitment (b = .382, p = .031) and the value of stability (b = .044, p = .045) were still significant once length of marriage was added to the analyses. 61 Relational identity. There were no significant relationships between relational identity and any of the moral commitment variables for either males or females. Relational identity is an interval-type dependent variable, so all the analyses for this variable were done using multiple regression. Personal commitment scale. The composite personal commitment scale was not related, for males or females, to either of the two individual components of moral commitment. The composite scale of moral commitment was close to being significantly related to the personal commitment scale for females (B = 1.435, R = .181, p = .066), but the relationship for males was not significant (B = .486, R = .000, p = .532). The significance of the relationship between personal and moral commitment for females increased when the control variable, length of marriage, was added (B = 1.784, R = .223, p = .043). These relationships were tested using logistic regression because the personal commitment scale (dependent variable) is coded with 1 and 0. Attraction to the relationship. Attraction to the relationship was not related to the moral commitment scale, or its two individual factors (value of stability and person-specific obligation). This dependent variable is scored using an interval response scale, and was analyzed using multiple regression. 62 Attraction to the spouse. For females, attraction to the spouse (an interval dependent variable) was related to person-specific obligation (b = .414, p = .018) using multiple regression analyses. Person-specific obligation explained 17% of the variation in attraction to the spouse. The relationship was still significant when length of marriage was included (b = .740, p = .023). With the control variable, 32% of the variation was explained. Males did not show a significant effect between person-specific obligation and attraction to the spouse (b = .050, p = .876). 63 Table 4. Summary of Results from Hypothesis 5. Dependent variables: attraction to the spouse, attraction to the relationship, religious identity, relational identity, personal commitment Independent variables: moral commitment, person-specific obligation, value of stability Indep. Variables Moral Commit ment Moral Commit ment Person-specific Obligafn Person-specific Obligafn Value of Stability Value of Stability Gender/ Dep. Variables Males Females Males Females Males Females religious identity not significant MR b=383 p=027 15% explained not significant not significant not significant MR b=044 p=.040 13% explained relational identity not significant not significant not significant not significant not significant not significant personal commit ment not significant not significant not significant not significant not significant not significant attraction to the relation ship not significant not significant not significant not significant not significant not significant attraction to the spouse not significant not significant not significant MR b=.414 p=.018 17% explained not significant not significant MR = Multip e Regression 64 Hypothesis 6: Moral commitment is positively related to structural commitment. As outlined in the univariate analysis section, the four components of structural commitment were found to be unrelated, and were therefore analyzed separately in the bivariate analyses. The four components of structural commitment are investment, attention to alternative attractions, social reaction, and difficulty of termination. Investment. When tested against the moral commitment scale and its two components, person-specific obligation and the value of stability, investment showed no significant relationships. These analyses were performed using logistic regression because investment is coded with 1 and 0. Social reaction. Social reaction was not related to the moral commitment composite scale, but was related significantly to both person-specific obligation and the value of stability. These relationships were significant for females only, and were tested using logistic regression; the dependent variable, social reaction, is coded as 1 and 0. The relationship was strongest between social reaction and person-specific obligation (B = .600, R = .267, p = .023). When length of marriage was added as a control variable, the relationship was still significant (B = .578, R = .213, p = .045). The value of stability showed a significant effect with social reaction only after the control variable was added (B = .305, R = .219, p = .042). Before the control, the relationship was not significant (B = .181, R = .154, p = .081). 65 Attention to alternative attractions. Attention to alternative attractions was not significantly related to moral commitment or the value of stability. For females, attraction to alternatives was significantly related (B = -.720, R = -.233, p = .037) to person-specific obligation in a negative direction. This relationship was analyzed using logistic regression because the dependent variable, attention to alternatives, is coded as 1 and 0. For males, the relationship was not significant (B = .106, R = -.000, p = .625). Adding the control variable, length of marriage, removed the significance of this relationship for females (B = -.764, R = -.160, p = .078). Difficulty of termination. Difficulty of termination was not related to the composite moral commitment scale, but for females both person-specific obligation and the value of stability were significantly related to difficulty of termination. For the value of stability, the relationship was slightly stronger (B = .260, R = .267, p = .038) than for person-specific obligation (B = .390, R = .261, p = .040). Both of these relationships were still significant when the control variable, length of marriage, was added (value of stability: B = .276, R = .271, p = .036; obligation: B = .377, R = .260, p = .041). These analyses were performed using logistic regression because the dependent variable is a dummy variable.1 'Tests following Tabachnick and Fidell (1983) did not confirm suppressor effects on the analyses that resulted in stronger relationships between the independent and dependent variables once the control variable had been added. The author concludes that the differences in these values between non-control and control analyses is the result of the two distinct regression lines (e.g., analyses with the control variable explain considerably more of the variance). 66 Table 5. Summary of Results from Hypothesis 6. Dependent variables: investments, alternative attractions, social reaction, difficulty of termination Independent variables: moral commitment, person-specific obligation, value of stability Indep. Variables Moral Commit ment Moral Commit ment Person-specific Oblig'n Person-specific Oblig'n Value of Stability Value of Stability Gender/ Dep. Variables Males Females Males Females Males Females invest. not significant not significant not significant not significant not significant not significant social reaction not significant not significant not significant LR B=.600 R=.267 p=.023 with control: B=.578 R=.213 p=.045 not significant LR B=.181 R=.154 p=.081 with control: B=.305 R=.219 p=.042 alternat. attract'ns not significant not significant not significant LR B—.720 R=-.233 p=.037 not significant not significant difficulty of terminat. not significant not significant not significant LR B=.390 R=.261 p=.040 with control: B=.377 R=.260 p=.041 not significant LR B=.260 R=.267 p=.038 with control: B=.276 R=.271 p=.036 LR = Logistic Regression 67 Multivariate Analysis Hypothesis 7: Moral commitment moderates the effect of personal commitment on marital stability. The following equation describes the variables used to test this hypothesis: marital stability = moral commitment + personal commitment + (moral commitment x personal commitment) There were no significant relationships in the test of this hypothesis. Logistic regression was used to test this hypothesis because marital stability, the dependent variable, is coded with 1 and 0. Hypothesis 8: Moral commitment moderates the effect of structural commitment on marital stability. The following equation describes the variables used to test this hypothesis: marital stability = moral commitment + structural commitment factors 1 through 4 + (moral commitment x structural commitment factors 1 through 4) Each individual factor within the structural commitment construct was tested separately (a total of four tests were performed to test this hypothesis). A specific example of one of these equations is as follows: 68 marital stability = moral commitment + investments + (moral commitment x investments) To test this hypothesis, each of the four sub-categories of structural commitment had to be tested with moral commitment against marital stability. For females, difficulty of termination was significantly related to marital stability in this interaction equation. No other significant relationships occurred. Summary of Significant Relationships Table 6 summarizes the significant relationships that resulted from testing the eight hypotheses. Only one of the hypotheses showed no significant relationships (hypothesis 2). The results of the analyses are discussed below. 69 Table 6. Summary of Significant Relationships Variables Christian Fundamentalism Marital Stability , Moral Commitment Moral Commitment Hypothesis 1: mc + cf (f) pso + cf (f) vs + cf (f & m) Hypothesis 2: no significant relationships not tested Personal Commitment not tested Hypothesis 3: pc + ms (f & m) atrel + ms (f & m) atsp + ms (f & m) Hypothesis 5: rigid + mc (f) rigid + vs (f) atsp + pso (f) Structural Commitment not tested Hypothesis 4: inv + ms (m) alt + ms (-ve, f) sr + ms (f) dt + ms (f) Hypothesis 6: sr + pso (f) sr + vs (f) alt + pso (-ve, f) dt + vs (f) pc = personal commitment / atrel = attraction to the relationship / atsp = attraction to the spouse / rigid = religious identity / rltid = relational identity / inv = investments / alt = attraction to alternatives / sr = social reaction / dt = difficulty of termination / ms = marital stability / mc = moral commitment / pso = persona-specific obligation / vs = value of stability / f = females / m = males / -ve = negative relationship 70 Discussion The three purposes of this study are to a) balance the emphasis on external variables in the marital stability literature with a focus on beliefs, b) revise Johnson's model to reflect this emphasis on beliefs, and c) determine whether these beliefs, and the construct of moral commitment used to measure them, are viable and salient components of the commitment model. The hypotheses developed to test the new model and the importance of moral commitment, are outlined below, along with a summary of the significant results. Following this discussion of the findings, the revised commitment model is discussed. The relationship between beliefs, in this case religious beliefs, and marital stability is not a simple one. The type of church and its belief system play an important role in determining the beliefs held by an individual. Furthermore, gender appears to influence the strength of these beliefs, and the impact they have on the other constructs in the commitment model. The relationships contained within the new commitment model devised for this study are described in Figure 4. 71 Figure 4. Commitment Model PERSONAL COMMITMENT attraction to the relationship attraction to the spouse relational identity religious identity CHRISTIAN MORAL MARITAL FUNDAMENTALISM STRUCTURAL COMMITMENT irretrievable investments social reaction attention to alternative attractions COMMITMENT - person-specific obligation STABILITY value of stability difficulty of termination 72 The causal ordering implicit in Figure 4 suggests that the individual's belief system (i.e. fundamentalist or liberal) effects the moral commitment variable. Moral commitment influences both structural and personal commitment. The three commitment constructs (moral, personal, and structural) impact marital stability. In addition, moral commitment moderates the effect of personal and structural commitment on marital stability. These specific relationships were tested as series of hypotheses (below). The following is an examination and an interpretation of findings for each hypothesis. Hypothesis Testing Hypothesis 1. Christian fundamentalism was hypothesized to be positively related to moral commitment. This hypothesis was partially supported by this study. Christian fundamentalism was positively related to the moral commitment scale and its two components: the value of stability and person-specific obligation. An unexpected finding was that only the value of stability was significantly related to Christian fundamentalism for both males and females. The other two relationships, involving moral commitment and person-specific obligation, were only significant for females. This hypothesis tests the first relationship in Figure 4; the effect of an individual's value system (fundamentalist or liberal Christianity) on moral commitment. The results indicate that this relationship is different for males and females. Fundamentalist men and women place equal value on stability within their marriages. 73 However, religious beliefs relate to feelings of obligation for women towards their spouses, but not for husbands towards their wives. In addition, the combination of the value of stability and obligation (the moral commitment scale) does not relate to men's belief systems but it does for women. A number of factors may influence these gender differences. Women may place more value on the spiritual aspects of their lives, internalizing religious belief systems to a greater extent than do men. The role that women traditionally hold as the socializers of children may mean that they place more importance on moral issues. Hypothesis 2. Hypothesis 2 assumed that moral commitment would be positively related to marital stability. This hypothesis was not supported by this study. Neither the moral commitment scale, nor its two components (person-specific obligation and the value of stability) were related to marital stability. Indeed, moral commitment for either gender had no apparent affect on marital stability. This implies that if moral commitment has any affect on marital stability, it will be as mediated by either personal commitment or structural commitment. As outlined in Figure 4, a direct relationship was predicted between moral commitment and marital stability, as well as moderating effects between moral commitment and structural and personal commitment. The direct relationship was not supported by the findings. The moderating effect of moral commitment is tested in hypotheses 7 and 8. This finding may be due in part to the lack of variation in the marital stability measure. The original scale for this variable was highly skewed, necessitating the construction of a dummy variable. This religious sample contained minimal amounts of instability. In a larger, more diverse religious sample, this hypothesis may find more support. 74 Hypothesis 3. This hypothesis predicted that personal commitment would be positively related to marital stability. There was partial support for this hypothesis. The personal commitment scale and its two components, attraction to the spouse and attraction to the relationship, were indeed positively related to marital stability. These relationships were significant for both males and females. The two other factors in this construct, relational and religious identity, were not related to stability. These results support the suggestion that satisfaction is related to marital stability. The degree to which one incorporates either the spousal relationship or one's religious roles (cognitive or behavioral) into one's self-concept does not appear to affect stability. Hypothesis 4. Hypothesis 4 tested whether structural commitment was positively related to marital stability. This hypothesis was partially supported. Each of the four components of structural commitment were significantly related to marital stability, but none of these relationships were significant for both males and females. For males, only irretrievable investments were significantly related to marital stability. The costs, in terms of lost time, energy, and resources do not appear to affect women's marital stability. Although this study does not distinguish in its analysis between the types of investments, future studies may determine if certain resources have more impact on men's decisions to divorce (e.g., do perceived financial costs weigh more heavily than time or energy?). 75 For females, both social reaction and the difficulty of termination were positively related to stability. According to this study, women appear to be more influenced than men by both the financial consequences of divorce and the opinions of others. Attention to alternative attractions was negatively related to marital stability for females. In other words, the more they attended to alternatives, the less stable their relationships were. This result is not surprising, as the commitment literature predicts this relationship. What is interesting is that this relationship was not supported for the males in this sample. Apparently, a man's perception of comparatively better situations outside his marriage does not affect the stability of his relationship. Hypothesis 5. This hypothesis predicted that personal commitment would be positively related to moral commitment. This hypothesis was partially supported by this study. The analyses testing this hypothesis were extensive because the independent and dependent variables in the equation are both composed of several factors. Personal commitment, the dependent variable, is composed of attraction to the relationship, attraction to the spouse, religious identity, and relational identity. The independent variable, moral commitment, is composed of the value of stability and person-specific obligation. Each of the components of the dependent variable were tested against each of the components of the independent variable. The composite scales for these two variables were also tested against one another. These tests produced three significant relationships, all for females. Religious identity was positively related to the moral commitment scale, and to the individual factor of the value of stability. Women place more importance on stability in their 76 marital relationships when they incorporate their religious roles into their self-concepts. Men do not appear to need to feel that Christianity is an important aspect of their identity in order for them to achieve high marital stability. Attraction to the spouse was positively related to person-specific obligation for women. Satisfaction with their spouse is reflected in greater feelings of obligation towards that spouse for wives. The degree of satisfaction that men feel with their spouses does not affect their level of obligation. Hypothesis 6. Structural commitment was hypothesized to be positively related to moral commitment. There were several significant relationships between structural and moral commitment for females, but none for males. As in hypothesis 5, the variables involved in testing this hypothesis each have several factors. The four components of structural commitment (social reaction, difficulty of termination, investments, and attention to alternative attractions) were each tested against the three components of moral commitment (the composite scale, the value of stability, and person-specific obligation). According to this study, women's levels of moral commitment influence how much attention they pay to the costs involved in leaving their relationships, the reactions of others towards their divorce, and whether they are conscious of the alternative situations that might result from a break-up. Social reaction was positively related to both person-specific obligation and the value of stability for females. This indicates that the degree of obligation that a woman feels towards her spouse, and how much she values stability in the marital relationship, are related 77 to her perception of others' opinions about divorce. Difficulty of termination was positively related to both the value of stability and person-specific obligation. Termination was viewed as more difficult by women who valued stability and felt obligated to their spouses. Traditionally, women are more negatively affected by divorce. They tend to suffer financially, carry the burden of dependent children, and often have sacrificed educational and occupational opportunities for their families. The finding that there is no relationship between moral commitment and the difficulty of termination for men may be explained by these gender-specific consequences. Attention to alternative attractions was negatively related to person-specific obligation. The greater a woman's feelings of obligation towards her spouse, the less she attended to the alternative relationships and resources outside her marriage. How valuable a woman found stability in her marriage was not related to whether or not she attended to alternatives. Hypothesis 7. This hypothesis tested the moderating effect of moral commitment on the relationship between personal commitment and marital stability. There was no support for this hypothesis. Moral commitment does not have either a direct effect on marital stability, or a moderating effect through personal commitment. It indirectly affects stability, because there are significant relationships between personal commitment and moral commitment. The effect is perhaps a mediating (enhancing) effect rather than a moderating effect. 78 Hypothesis 8. This hypothesis examined the moderating effect of moral commitment on the relationship between the four components of structural commitment and marital stability. It was not supported by this study. As in hypothesis 7, neither the direct nor the moderating effect of moral commitment on marital stability was observed in this sample. There were significant relationships between moral commitment and each of the four structural commitment factors, however. This suggests that moral commitment may mediate the effect of these components on marital stability. Commitment Model The commitment model, as outlined by Johnson (1991), was modified for this study to test the salience of the moral commitment component of the theory. By placing the moral commitment variable in the center of the model, value systems were given greater importance within the theory. The results of the analyses performed for this study point to the need to re-examine the original model. Aspects of both the structural and personal commitment constructs were called into question by this study. Moral commitment. This component of Johnson's model was poorly developed in the literature. Rusbult criticized the lack of empirical evidence supporting the inclusion of this construct in the commitment model. Moral commitment was found, in this study of a religious population, to be an important variable in the study of marital stability. It was not directly related to 79 stability, and did not moderate the effects of personal and structural commitment on stability, but several significant relationships were found between moral, personal, and structural commitment. The majority of the significant relationships occurred only for females, however. Structural commitment. Reliability tests showed that the four components of structural commitment, as conceptualized by Johnson (1991), were not significantly related to each other. The lack of correlation between these variables highlights a need to rethink the original construct. In addition, although several significant relationships were found between structural commitment and marital stability, none of them was significant for both males and females. In Johnson's original model structural commitment was of central importance and was proposed to have a greater effect on stability than either personal or moral commitment. In this study, personal commitment and its two components, attraction to the relationship and attraction to the spouse, were significantly related to marital stability for males and females. These findings indicate that structural commitment may not be the most important variable in predicting stability, at least within a religious sample. Personal commitment. Three components made up Johnson's original construct of personal commitment. These factors were attraction to the relationship, attraction to the spouse, and relational identity. For this study a fourth variable was added: religious identity. Relational and 80 religious identity were not related to the other two factors in the personal commitment construct. Relational identity did not significantly relate to any of the variables in the commitment model. Its relevance to commitment theory is doubtful based on the results of this study. Religious identity, the fourth, new, component of personal commitment which was added to determine the relationship between an internalization of one's religious role and commitment, was significantly related to two of the components in the original model: moral commitment and the value of stability. Future studies may examine this variable as a component of moral commitment rather than personal commitment. The values inherent in a religious self-concept may be more appropriately associated with the moral commitment construct. This variable did not relate to any of the other variables in the model, however, and future studies may determine that its addition to the moral commitment construct does not add anything to the predictive value of the theory. An alternative commitment model, based on the findings of this study, is presented in Figure 5. It outlines the following modifications: 1. The relocation of the religious identity variable from personal to moral commitment. 2. The placement of all commitment components (moral, personal, and structural) as relatively equal in importance. 3. The separation of the structural commitment variables. 81 Figure 5. Potential Revision for the Model of Commitment BELIEF SYSTEM MORAL COMMITMENT -value of stability - person-specific obligation - religious identity STRUCTURAL PERSONAL COMMITMENT COMMITMENT - irretrievable investments attraction to the relationship - attention to alternative attractions attraction to the spouse social reaction difficulty of termination MARITAL STABILITY 82 Implications for Research Theoretical implications. The results of this research show that religious beliefs are significantly related to moral commitment. In previous studies differences between religious and non-religious populations have been attributed to external barriers and constraints against divorce in church communities. This study has shown that there is considerable variation in commitment variables in religious populations, despite similarities in doctrine (church group was not a significant control variable). The fact that beliefs (as measured by moral commitment) significantly affected both structural and personal commitment suggests that this factor needs to be included in the model. This construct indeed "contributes to desire to maintain a relationship ... [and] ... contributes to our understanding of commitment" as it was challenged to do by Rusbult (1991, p. 165). Although the results of this study suggest that moral commitment, and belief systems in general, have a place in the stability literature, the findings did not support a central role for this variable (see Figure 5). Moral commitment does, however, deserve an equal place in the commitment model along with satisfaction and constraint factors. A final implication for future research based on this study is the need to differentiate between males and females when examining belief systems, specifically religious beliefs. Gender differences were prominent in the findings of this study, suggesting that men and women differ in the way that they integrate their beliefs into their lives and relationships. 83 Future research. Several question arose from this study that may be addressed in future research. The first of these involves determining if moral commitment might relate more strongly to marital stability in a more diverse religious sample. The limited relationship between these two variables in this study might be the result of a lack of variation in the marital stability index. Another area to explore in future research is gender differences in commitment in religious populations. The determinants of these differences could be studied and a deeper examination made of the exact relationships (e.g., what types of investment are significantly related to stability for men). The model developed as a result of this study (Figure 5) could be examined in relation to the repositioning of the religious identity variable. In addition, future research may test whether moral commitment does have a mediating, rather than moderating, effect on personal and structural commitment as they relate to marital stability. Limitations The small sample size limits the applicability of this study. The purpose of the study was, however, not an exhaustive examination of an established model, but rather an exploration of new concepts and relationships. The total number of respondents (63) was sufficient for the purposes of the study. However, because there were both couple and individual respondents, controlling for correlations between spouses reduced the sample size. The original sample was split by gender, resulting in two samples of 31 males and 32 females. 84 The absence of a control variable specifically for occupation is a limitation of this study. The author expected, based on pretesting, that the relational and religious identity question (# 7, see Appendix B) would encompass this measure. Unfortunately, many respondents did not mention their occupation in response to this question. From the data that is available from this question, occupation was not significantly related to marital stability. The importance of this omission may be negligible, based on this lack of significance. The age structure of the sample is another limitation of this study. A more diverse religious population which included, for example, campus organizations, would improve the applicability of the results. Cohort effects and other such problems will only be revealed if the sample varies sufficiently in age. The confounding effects of age on length of marriage, presence of children, etc. can only be unraveled if enough people of similar age vary in these categories. A further limitation of the study is the lack of sufficient variation in the dependent variable; marital stability. The small sample size and select population meant that there were only a few participants with high marital instability. This lack of variation affects the number and strength of the other relationships in the model. Finally, the generalizability of the findings is limited because the respondents are not from a probability sample. The religious sample is not intended to be representative of the general population, but neither is it suggested that it describes religious groups in general. Indeed, it was a sample of select religious groups based on assumptions of maximizing differences in beliefs about moral commitment and marital stability (i.e., a liberal and fundamentalist comparison). 85 Summary The three purposes of this study were: a) to balance the emphasis on external barriers in the stability literature with an internal, belief-based model b) to revise Johnson's model of commitment to reflect this focus on internal beliefs c) to determine whether the measure of these beliefs, moral commitment, was a useful component of commitment theory. Despite the lack of a direct effect for moral commitment on marital stability in this study, the results might still indicate that moral commitment is indeed a valuable construct in the study of commitment and stability. Moral commitment neither exerted a direct effect on marital stability nor moderated the effects of personal and structural commitment on commitment. However, moral commitment affected both structural and personal commitment. This study was done on a select religious group who hold specific beliefs about marital stability. Despite the fact that the entire sample was religious, distinctions could still be made in the types of beliefs, and the effects these beliefs had. If sufficient variation is present in a relatively homogamous sample to suggest that belief systems are important to stability, a study comparing the beliefs of religious and non-religious groups may show that this is an extremely salient variable. The results of this study also revealed that gender differences are an important aspect in the study of stability. In addition, tests performed on the structural commitment variable revealed that its four components (investments, social reaction, termination, and alternatives) are not related. Future researchers should consider incorporating these variables separately 86 in their analyses. In conclusion, although the central position of moral commitment within the model must be questioned based on the results of this study, it is clear that moral commitment, and the belief systems reflected by this measure, are definitely valuable components in the prediction of marital stability. 87 References Adams, J. F. & Sprenkle, D. H. (1990). Self-perception and personal commitment: A challenge to current theory of marital dissolution and stability and implications for marital therapy. The American Journal of Family Therapy. 18(2). 131-140. Albrecht, S. L. & Kunz, P. A. (1980). The decision to divorce: A social exchange process. Journal of Divorce. 3(4). 319-337. Altemeyer, B., & Hunsberger, B. (1992). Authoritarianism, religious fundamentalism, quest, and prejudice. The International Journal for the Psychology of Religion. 2. 113-133._ Blumstein, P. & Schwartz, P. (1983). American Couples. New York: William Morrow and Company, Inc. Booth, A., Johnson, D. & Edwards, J. N. (1983). Measuring marital instability. Journal of Marriage and the Family. 45. 387-394. Chan, L. Y. W. & Heaton, T. B. (1989). Demographic determinants of delayed divorce. Journal of Divorce. 13(1). 97-112. Coleman, J. & Coleman, S. (1992). Friends for life: Building a loving and lasting relationship with your mate. Vancouver: Word Publishing. Cook J. & Emerson, R. E. (1978). Power, equity, and commitment in exchange networks. American Sociological Review. 43(5), 721-739. Donahue, M. J. (1985). Intrinsic and extrinsic religiousness: Review and meta analysis. Journal of Personality and Social Psychology. 48(2). 400-419. Edwards, J. N, Johnson, D. R. & Booth, A. (1987). Coming apart: A prognostic instrument of marital breakup. Family Relation. 36. 168-170. Filsinger, E. E. & Wilson, M. R. (1984). Religiosity, socioeconomic rewards, and family development: Predictors of marital adjustment. Journal of Marriage and the Family. 46(3), 663-670. Fincham, F. D. & Bradbury, T. N. (1987). The assessment of marital quality: A reevaluation. Journal of Marriage and the Family. 49(4). 797-809. 88 Fishbein, M. (1967). The AB Scales: An operational definition of belief and attitude. In M. Fishbein (Ed.), Readings in attitude theory and measurement (pp. 183-189). New York: John Wiley and Sons. Glenn, N. D. (1991). The recent trend in marital success in the United States. Journal of Marriage and the Family, 53, 285-295. ' Green, R. G. & Sporakowski, M. J. (1984). The dynamics of divorce: Marital quality, alternative attractions and external pressures. Journal of Divorce. 7(2), 77-88. Heaton, T. B. (1984). Religious homogamy and marital satisfaction reconsidered. Journal of Marriage and the Family. 46. 728-733. Hill, C. T., Rubin Z., & Peplau, L. A. (1979). Breakups before marriage. In G. Levinger & O. C. Moles (Eds.). Divorce and separation: Context, causes and consequences. New York: Basic Books. Hunsberger, B. (1995). Religion and prejudice: The role of religious fundamentalism, quest, and right-wing authoritarianism. Journal of Social Issues. 51(2), 113-129. Jenkins, K. W. (1991). Religion and Families. In Family Research: A sixty-year review, 1930-1990, Volume 1. Stephen J. Bahr (Ed.). Toronto: Levington Books. Johnson, D. J. & Rusbult, C. E. (1989). Resisting temptation: Devaluation of alternative partners as a means of maintaining commitment in personal relationships. Journal of Personality and Social Psychology, 57(6). 967-980. Johnson, M. P. (1985). Commitment, cohesion, investment, alternatives, constraint. Why do people stay together when they really don't want to? Paper presented at the Theory Construction and Research Methodology Workshop, National Council on Family Relations Meeting, Dallas, TX. Johnson, M. P. (1991). Commitment to Personal Relationships. Advances in Personal Relationships, 3, 117-143. Johnson, M. P. (1991). Reply to Levinger and Rusbult. Advances in Personal Relationships. 3, 171-176. Kinnaird, K. L., & Gerrard, M. (1986). Premarital sexual behavior and attitudes toward marriage and divorce among women as a function of their mother's marital status. Journal of Marriage and the Family. 48. 757-765. 89 Larson, L. E., & Goltz, J. W. (1989). Religious participation and marital commitment. Review of Religious Research. 30(4). 387-400. Leik, R. & Leik, S. (1972). Interpersonal commitment as a balancing mechanism in social exchange. Paper presented at the annual meeting of the Pacific Sociological Association, Scottsdale, AR. In Johnson, M. P. (1985). Commitment, cohesion, investment, alternatives, constraint. Why do people stay together when they really don't want to? Paper presented at the Theory Construction and Research Methodology Workshop, National Council on Family Relations Meeting, Dallas, TX. Leik, R. & Leik, S. (1976b). Methodological paradigms for studying developing relationships. Paper presented at the Conference on social exchange in developing relationships, The Pennsylvania State University. In Johnson, M. P. (1985). Commitment, cohesion, investment, alternatives, constraint. Why do people stay together when they really don't want to? Paper presented at the Theory Construction and Research Methodology Workshop, National Council on Family Relations Meeting, Dallas, TX. Levinger, G. (1965). Marital cohesiveness and dissolution: An integrative review. Journal of Marriage and the Family. 27. 19-28. Levinger, G. (1991). Commitment VS. cohesiveness: Two complementary perspectives. Advances in Personal Relationships. 3, 145-150. Levinger, G, Senn, D. J., & Jorgensen, B. W. (1970). Progress toward permanence in courtship: A test of the Kerckhoff-Davis hypothesis. Sociometry, 33. 427-443. Lewis, R. A. & Spanier, G. B. (1979). Theorizing about the quality and stability of marriage. In Contemporary Theories about the Family. Volume 1. Eds. Wesley R. Burr, Reuben Hill, F. Ivan Nye & Ira L. Reiss. New York: The Free Press. Lund, M. (1985). The development of investment and commitment scales for predicting continuity of personal relationships. Journal of Social and Personal Relationships. 2, 3-23. Maneker, J. S. & Rankin, R. P. (1993). Religious homogamy and marital duration among those who file for divorce in California, 1966-1971. Journal of Divorce and Remarriage. 19(1/2). 233-247. McRae, W. J. (1980). Preparing for your marriage. Michigan: The Zondervan Corporation. 90 Murstein, B. I. & MacDonald, M. G. (1983). The relationship of 'exchange orientation' and 'commitment' scales to marriage adjustment. International journal of Psychology. 18. 297-311. Pineo, P., Porter, J., & McRoberts, H. (1977). The 1971 census and the socioeconomic classification of occupations. Canadian Review of Sociology and Anthropology. 14(1). Robinson, L. C. (1994). Religious orientation in enduring marriage: An exploratory study. Review of Religious Research. 35(3). 207-218. Rosenblatt, P. C. (1977). Needed research on commitment in marriage. In G. Levinger & H. Rausch (Eds.). Close relationships: Perspectives on the meaning of intimacy. Amherst: University of Massachusetts Press. Rusbult, C. E. (1991). Commentary on Johnson's 'Commitment to personal relationships': What's interesting, and what's new? Advances in Personal Relationships. 3. 151-169. Sabatelli, R. M. & Shehan, C. (1993). Exchange and resource theories. In P. G. Boss, W. J. Doherty, R. LaRossa, W. R. Schumm, & S. K. Steinmetz (Eds.), Sourcebook of family theories and methods: A contextual approach (pp. 385-411). New York: Plenum Press. Schumm, W. R. & Bugaighis, M. A. (1985). Marital quality and marital stability: Resolving a controversy. Journal of Divorce. 9(1). 73-77. Spanier, G. B. & Thompson, L. (1982). A confirmatory analysis of the Dyadic Adjustment Scale. Journal of Marriage and the Family. 44. 731-738. Swensen, C. H. & Trahaug, G. (1985). Commitment and the long-term marriage relationship. Journal of Marriage and the Family. 47. 939-945. Tabachnick & Fidell (1983). Using Multivariate Statistics. San Francisco: Harper & Row. Thomas, D. L., & Henry, G. C. (1985). The religion and family connection: Increasing dialogue in the social sciences. Journal of Marriage and the Family. 47. 369-379. Touliatos, J., Perlmutter, B. F., & Straus, M. A. (1990). Handbook of Family Measurement Techniques. Newbury Park: Sage Publications. 91 Turner, R. H. (1970). Family Interaction. New York: Wiley. In Swensen, C. H. & Trahaug, G. (1985). Commitment and the long-term marriage relationship. Journal of Marriage and the Family. 47. 939-945. Udry, J. R. (1981). Marital alternatives and marital disruption. Journal of Marriage and the Family. 43. 889-897. Udry, J. R. (1983). The marital happiness/disruption relationship by level of marital alternatives. Journal of Marriage and the Family. 45. 221-222. 92 Appendix A Definitions Availability of Acceptable Alternatives: alternatives that extend beyond the replacement of the current marriage to economic consequences, social ties, etc. (Johnson, 1991, p. 124). Difficulty of Termination: includes the costs of ending the relationship, in terms of monetary and other resources (Johnson, 1991, p. 124). Investments: an accumulation of time, effort, and resources (Lund, 1985), defined by Johnson (1991, p. 123) as irretrievable "to the extent that they were made in anticipation of a future return that has not been realized, and the expenditure was not returnable." Love: positive feelings about a particular person — composed of attachment, intimacy and caring (Lund, 1985) Marital Instability: "the propensity to dissolve [the marital relationship], even if dissolution was not the final outcome" (Edwards, Booth, Johnson, 1987) Moral Commitment: an internal set of beliefs that encompass the value of marital stability and a sense of person-specific obligation (factors outlined in Johnson's original model). It is an internally-based factor that encourages continuation of the marital relationship, rather than an external constraint which discourages the termination of the marriage. Person-specific Obligation: feelings of obligation felt by individuals stemming from contracts between them and other individuals. This contract is not necessarily legally binding, or even articulated (Johnson, 1991, p. 122). Relational Identity: "the extent to which one's involvement in a relationship is incorporated into one's self-concept" (Johnson, 1991, p. 120). Religious Identity: present when one's religious roles (cognitive or behavioral) are an integral part of one's identity. Social Reaction: involves others' feelings about dissolution, and the pragmatic consequences of a breakup on others (e.g., children or friends) should the relationship dissolve. 93 Appendix B Questionnaire This questionnaire is composed of several sections. There are instructions accompanying each section. Keep in mind that your first response is best. Do not take a long time with the questions or try to go back and change answers. Your assistance is very much appreciated. Completion of this questionnaire constitutes your consent for the researcher to use the data in statistical form only. Your personal responses will remain strictly confidential. You may decide not to complete this questionnaire at any time. Please give only one answer per question. For certain questions you are asked to write an answer. For these questions, please print. Mark an 'X' next to the single best answer for each question. 1. Sex. male female 2. Age. 3. List the ages of your children (if you do not have children, please skip this question): (1) (2) (3) (4) (5) (6) (7) (8) (9) 4. How many years have you been married to your present spouse? 94 5. Have you ever been divorced or separated? yes no 6. In which of the following groups would you place yourself? Francophone Chinese Japanese Native Indian East Indian Greek Eastern European British Italian German - Canadian Other 7. In the spaces below, please list, in order of importance to you, the roles you occupy that you feel are most strongly linked to your identity (e.g., doctor, parent, friend, student, etc.). I am a ... 1. 2. 3, 4. 5. 95 Please circle the answer that best describes your feelings about the following statements: 8. I do not feel that I can confide in my spouse about virtually anything. 1 2 3 4 5 strongly strongly agree disagree 9. I would do almost anything for my spouse. 1 2 3 4 5 strongly strongly agree disagree 10. If I could never be with my spouse, I would feel miserable. 1 2 3 4 5 strongly strongly agree disagree 11. If I were lonely, my first thought would be to seek my spouse out. 1 2 3 4 5 strongly strongly agree disagree 12. One of my primary concerns is my spouse's welfare. 1 2 3 4 5 strongly strongly agree disagree 96 13. I feel that there are things I would be unable to forgive my spouse for. 1 2 3 4 5 strongly strongly agree disagree 14. I feel responsible for my spouse's well-being. 1 2 3 4 5 strongly strongly agree disagree 15. It would be easy for me to get along without my spouse. 1 2 3 4 5 strongly strongly agree disagree Circle the answer that best describes how often you and your spouse engage in these activities: 16. quarrel 1 2 3 4 5 very frequently occasionally never 17. get on each other's nerves 1 2 3 very frequently 97 4 5 occasionally never 18. kiss 1 2 very frequently 3 4 occasionally 5 never 19. engage in outside interests together 1 2 3 4 5 very frequently occasionally never 20. have a stimulating exchange of ideas 1 2 3 4 5 very frequently occasionally never 21. laugh together 1 2 3 4 5 very frequently occasionally never 22. calmly discuss something 1 2 3 4 5 very frequently occasionally never 23. work together on a project 1 2 3 4 5 very frequently occasionally never 98 Please circle the answer that best reflects how often the following issues are a problem in your relationship: 24. being too tired for sex 1 2 3 4 5 very frequently occasionally never 25. not showing love 1 2 3 4 5 very frequently occasionally never Please circle the word that best describes how often in your relationship you engage in the following activities: 26. Spend your free time with your spouse rather than doing other things or seeing other people. never occasionally frequently always 27. Spend continuous time alone together such as on dates, weekend outings, or vacations, never occasionally frequently always 28. Buy gifts for your spouse or pay for entertainment, never occasionally frequently always 29. Share important personal feelings, problems, and beliefs with your spouse, never occasionally frequently always 99 30. never Reveal your sexual experiences/preferences to your spouse, occasionally frequently always 31. Contribute financially to your spouse or your relationship in general, never occasionally frequently always 32. Try to develop interests and activities in common with your spouse, never occasionally frequently always 33. Tell your spouse your true feelings about the relationship such as whether you love him or her. never occasionally frequently always 34. Integrate your spouse into your family. never occasionally frequently always 35. Consciously make time to see your spouse. never occasionally frequently always 36. Do favours for or help your spouse. never occasionally frequently always 37. Change things about yourself to please your spouse, such as your attitudes, habits or appearance. never occasionally frequently always 100 38. Change your career plans or other interests to maintain your relationship, never occasionally frequently always 39. Put effort into making the relationship work where there are problems, never occasionally frequently always 40. Try to encourage and support your spouse. never occasionally frequently always 41. Invest emotionally in your spouse. never occasionally frequently always We are interested in the opinions about divorce in general of the people who know you. If the person listed below is always opposed to divorce no matter what, circle the number for always opposed. If they are somewhere in between, circle the number that best describes them. If some of the items do not apply to you (if, for example, you do not have an oldest brother or religious leader, or a parent or grandparent is deceased) please circle not applicable (N/A). 42. mother 1 2 3 4 5 never sometimes always N/A opposed opposed opposed 101 43. father 1 2 never opposed 3 4 sometimes opposed 5 always opposed N/A 44. grandfather 1 2 never opposed sometimes opposed 5 always opposed N/A 45. grandmother 1 2 never opposed sometimes opposed 5 always opposed N/A 46. oldest brother 1 2 never opposed sometimes opposed 5 always opposed N/A 47. oldest sister 1 2 never opposed sometimes opposed 5 always opposed N/A 102 48. next oldest brother 1 2 never opposed sometimes opposed 5 always opposed N/A 49. next oldest sister 1 2 never opposed sometimes opposed 5 always opposed N/A 50. closest male friend 1 2 never opposed sometimes opposed 5 always opposed N/A 51. closest female friend 1 2 3 never opposed sometimes opposed 5 always opposed N/A 52. minister, rabbi, or priest 1 2 3 never sometimes opposed opposed 5 always opposed N/A 103 53. counselor or therapist 1 2 3 4 5 never sometimes always N/A opposed opposed opposed The groups people belong to also have different opinions about divorce. For each of the groups listed below, please circle the number provided that best describes that group's opinion about divorce. Although the members of groups differ in their opinions, we are interested in the groups as a whole - how most of the members of a group feel about divorce. If you are not a member of a particular group (for example, if you are not employed) please circle not applicable (N/A). 54. neighbours 1 2 never opposed sometimes opposed 5 always opposed N/A 55. people where you work 1 2 3 never opposed sometimes opposed 5 always opposed N/A 56. people in your church 1 2 3 never opposed sometimes opposed 5 always opposed N/A 104 57. your relatives (as a whole) 1 2 3 4 5 never sometimes always N/A opposed opposed opposed 58. your friends (as a whole) 1 2 3 4 5 never sometimes always N/A opposed opposed opposed Married people differ in the attractions they have to options outside of the marital relationship. We are interested in the attractions you have to people and things outside of your marriage. Please circle the answer that best describes how often you are attracted to each of the following: 59. The personal freedom you might have if you were single or married to someone else, never seldom sometimes frequently always 60. Having sexual relations with someone other than your spouse. never seldom sometimes frequently always 61. The spending money you might have if you were single or married to someone else. never seldom sometimes frequently always / 62. The enjoyment from friends you might have if you were single or married to someone else. never seldom sometimes frequently always 105 63. The respect from other people you might have if you were single or married to someone else. never seldom sometimes frequently always 64. The relationship you might have with your parents if you were single or married to someone else. never seldom sometimes frequently always Please circle the answer that best describes your feelings about the following statements: 65. How much trouble would ending your relationship be to you personally? none a little some a great deal 66. Do you worry about the possible financial implications of getting a divorce? never occasionally frequently always 67. Sometimes married people think they would enjoy living apart from their spouse. How often do you feel this way? very often often occasionally never 68. Even people who get along quite well with their spouse sometimes wonder whether their marriage is working out. Have you thought your marriage might be in trouble in the last three years? very often often occasionally never 69. As far as you know, has your spouse ever thought your marriage was in trouble? very often often occasionally never 106 70. Have you talked with family members, friends, clergy, counselors, or social workers about problems in your marriage within the last three years? very often often occasionally never 71. As far as you know, has your spouse talked with relatives, friends, or a counselor about problems either of you were having with your marriage? very often often occasionally never 72. Has the thought of getting a divorce or separation crossed your mind in the last three years? very often often occasionally never 73. As far as you know, has the thought of getting a divorce or separation crossed your spouse's mind in the last three years? very often often occasionally never 74. Have you or your spouse seriously suggested the idea of divorce in the last three years? very often often occasionally never 75. Have you talked about dividing up the property? very often often occasionally never 76. Have you talked about consulting a lawyer? very often often occasionally never 107 77. Have you or your spouse consulted a lawyer about a divorce or separation? yes no 78. Because of the problems people may be having with their marriage, they sometimes leave home either for a short time or as a trial separation. Has this happened in your marriage within the last three years? very often often occasionally never 79. Have you talked with your spouse about filing for divorce or separation? very often often occasionally never 80. Have you or your spouse filed for either a divorce or separation petition? yes no Please circle the response that most closely reflects your level of agreement with each statement. 81. I do not believe you have an obligation to stick with a bad marriage because of the vows you made to your spouse. strongly agree disagree strongly agree somewhat somewhat disagree 82. I think that when you get married, you are promising your spouse that you'll never leave them. strongly agree disagree strongly agree somewhat somewhat disagree 108 83. I believe people ought to stay together even if one of the spouses has been unfaithful, strongly agree disagree strongly agree somewhat somewhat disagree 84. I believe you have an obligation to your spouse to put the marriage before your own personal wants. strongly agree disagree strongly agree somewhat somewhat disagree 85. I believe that when you get married, you're promising your spouse that you will stay in the marriage until one of you dies. strongly agree disagree strongly agree somewhat somewhat disagree 86. When people marry, they should be willing to stay together no matter what happens, strongly agree disagree strongly agree somewhat somewhat disagree 87. If people are not happy in their marriage, they owe it to themselves to get a divorce and try to improve their lives. strongly agree disagree strongly agree somewhat somewhat disagree 88. The marriage vow "till death do us part" represents a sacred commitment to another person and should not be taken lightly. strongly agree disagree strongly agree somewhat somewhat disagree 109 89. The negative effects of divorce on children have been greatly exaggerated, strongly agree disagree strongly agree somewhat somewhat disagree 90. In the long run, Canadian society will be seriously harmed by the high divorce rate, strongly agree disagree strongly agree somewhat somewhat disagree 91. Many people who get divorced are too weak to make personal sacrifices for the good of their families. strongly agree disagree strongly agree somewhat somewhat disagree 92. Even if people are unhappy with their marriage, they should stay together and try to improve it. strongly agree disagree strongly agree somewhat somewhat disagree 93. These days, the marriage vow "till death do us part" is a formality. It doesn't really mean that people should stay in an unsatisfactory marriage. strongly agree disagree strongly agree somewhat somewhat disagree 94. Most children of divorced parents experience negative effects of the divorce for the rest of their lives. strongly agree disagree strongly agree somewhat somewhat disagree 110 95. The fact that most individuals no longer feel that they have to stay in unhappy marital relationships will benefit society as a whole. strongly agree disagree strongly agree somewhat somewhat disagree 96. Most people who get divorces do so as a last resort - only after trying other solutions to the problems in their marriage. strongly agree disagree strongly agree somewhat somewhat disagree Please circle the answer that best describes your feelings about the following statements: 97. Do you presently feel part of a religious group? definitely yes somewhat not really definitely not 98. How great an impact do you feel that your faith has on your marriage? large impact moderate impact little impact no impact N/A 99. How often do you read the Bible? daily weekly monthly rarely never 100. How often do you pray? daily weekly monthly rarely never 101. How often do you and your spouse have shared devotional times? daily weekly monthly rarely never 111 102. I believe that everything in the Bible is true, strongly agree disagree agree somewhat somewhat strongly disagree 103. I believe that marriage is instituted by God. strongly agree disagree strongly agree somewhat somewhat disagree 104. I believe that sexual intimacy is acceptable to God only within the institution of marriage. strongly agree disagree strongly agree somewhat somewhat disagree 105. I believe that Jesus is the Son of God. strongly agree disagree strongly agree somewhat somewhat disagree 106. I believe in Creation, not Evolution. strongly agree disagree strongly agree somewhat somewhat disagree 107. I use the scripture as a guide when I have to make difficult decisions, strongly agree disagree strongly agree somewhat somewhat disagree 112 108. Are you and your spouse: both Protestants both Catholics both Jewish both agnostic both atheists both another religion (e.g., Hindu) we have different religious beliefs Thank-you for completing this questionnaire. Please seal your completed questionnaire inside the blank envelope provided, and return it to your church. If both you and your spouse filled out a copy, please enter any five digit number in the space below. The same number should appear on each spouse's copy. Couple # 113 Appendix E Table 7. Exclusion of Questions Based on Missing Values. Variable Included Items Excluded Items Christian Fundamentalism 97 99 100 101 103 105 106 107 98 102 104 Value of Stability 86 88 89 90 92 93 94 87 91 95 96 Obligation 82 84 85 81 83 Attention to Alternatives 60 61 62 59 63 64 Attraction to the Spouse 9 10 12 13 14 15 8 11 Attraction to the Relationship 18 20 21 22 23 25 16 17 19 24 Investment 26 27 28 29 32 35 36 40 41 30 31 33 34 37 38 39 Social Reaction 42 43 44 45 46 47 48 49 50 51 52 54 55 56 57 58 53 116 Appendix F Tables 8 & 9. Descriptive Statistics for Males Variable/ Statistic Age Average Age of Children Mean 59.667 41.452 Median 66.000 39.000 Mode 66.000 88.000 Standard Deviation 17.629 24.350 Kurtosis -.967 -.048 SE of Kurtosis .833 .821 Skewness -.321 .829 SE of Skewness .427 .421 Missing Cases 1 0 Variable/ Statistic Number of Years Married Occupation Mean 32.129 58.077 Median 31.000 62.000 Mode 3.000 48.000 Standard Deviation 17.646 12.831 Kurtosis -1.241 -1.029 SE of Kurtosis .821 .887 Skewness -.300 -.192 SE of Skewness .421 .456 Missing Cases 0 5 Tables 10 & 11. Descriptive Statistics for Females Variable/ Statistic Age Average Age of Children Mean 53.656 36.375 Median 49.500 33.500 Mode 46.000 88.000 Standard Deviation 18.519 26.842 Kurtosis -1.245 -.145 SE of Kurtosis .809 .809 Skewness .067 .839 SE of Skewness .414 .414 Missing Cases 0 0 Variable/ Statistic Number of Years Married Occupation Mean 28.156 56.526 Median 25.000 48.000 Mode 3.000 48.000 Standard Deviation 18.180 11.217 Kurtosis -1.434 -1.399 SE of Kurtosis .809 1.014 Skewness .018 -.206 SE of Skewness .414 .524 Missing Cases 0 13 118 Appendix G Tables 12, 13 & 14. Descriptive Statistics for Dependent and Independent Variables Variable/ Statistic Marital Stability Christian Fundamentalism Personal Commitment Mean -.644 -5.752 .476 Median -.263 -4.133 .000 Mode -.263 -5.133 .000 Standard Deviation .490 5.664 .503 Kurtosis -1.806 3.199 -2.057 SE of Kurtosis .595 .595 .595 Skewness -.502 -1.751 .098 SE of Skewness .302 .302 .302 Missing Cases 0' 0 0 Variable/ Statistic Moral Commitment Difficulty of Termination Investment Mean .524 .852 .460 Median 1.000 1.000 .000 Mode 1.000 1.000 .000 Standard Deviation .503 .358 .502 Kurtosis -2.057 2.226 -2.039 SE of Kurtosis .595 .604 .595 Skewness -.098 -2.038 .163 SE of Skewness .302 .306 .302 Missing Cases 0 2 0 119 Variable/ Statistic Social Reaction Attraction to Alternatives Mean .492 .557 Median .000 1.000 Mode .000 1.000 Standard Deviation .504 .501 Kurtosis -2.066 -2.011 SE of Kurtosis .595 .604 Skewness .033 -.237 SE of Skewness .302 .306 Missing Cases 0 2 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0099194/manifest

Comment

Related Items