UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The relationship of empathy and self-esteem to the level of paternal involvement in child care Schein, Howard Walter 1994

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

Item Metadata

Download

Media
831-ubc_1994-0612.pdf [ 5.66MB ]
Metadata
JSON: 831-1.0054031.json
JSON-LD: 831-1.0054031-ld.json
RDF/XML (Pretty): 831-1.0054031-rdf.xml
RDF/JSON: 831-1.0054031-rdf.json
Turtle: 831-1.0054031-turtle.txt
N-Triples: 831-1.0054031-rdf-ntriples.txt
Original Record: 831-1.0054031-source.json
Full Text
831-1.0054031-fulltext.txt
Citation
831-1.0054031.ris

Full Text

THE RELATIONSHIP OF EMPATHY AND SELF-ESTEEM TO THE LEVEL OF PATERNAL INVOLVEMENT IN CHILD CARE by HOWARD WALTER SCHEIN B.A. Simon Fraser University, 1985 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in THE FACULTY OF GRADUATE STUDIES {DEPARTMENT OF COUNSELLING PSYCHOLOGY) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA October 1994 © Howard Walter Schein, 1994 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. (Signature) Department of cOUNSELLING pSYCHOLOGYThe University of British Columbia Vancouver, Canada DATE: OCT. 13, 1994 DE-6 (2/88) Abstract There is a large body of work on the study of the relationship between parenting and child personality, but very little on the relationship between parenting and parental personality characteristics, and more specifically fathers (Baruch & Barnett, 1986b; Demo et al., 1987). The present research focuses on the relationship between paternal involvement in child care and the father's level of empathy and self esteem. A sample of 52 subjects was drawn largely from the academic community including 36 fathers and 16 non-fathers. This study is correlational in nature, but the comparison sample of non-fathers was used to determine any possible differences on the dependent variables among father's level of involvement, and between non-fathers and father's levels of paternal involvement. A questionnaire administered to the fathers requested demographic information, also contained the Empathic Concern (EC) and Perspective-taking (PT) scales of the Interpersonal Reactivity Index (Davis, 1980), the Rosenberg (1979) Self-esteem Scale (SE), and the Paternal Involvement in Child Care Index (PI) by (Radin & Harold-Goldsmith, 1985). The non-fathers received the same package, but were not requested to answer the PI scale. Sub-hypotheses for a relationship between recollection of paternal involvement and present paternal involvement, and the relationship between desire to become a parent and EC, PT, and SE, are also explored. The results indicated that there was not a significant positive relationship between paternal involvement and any of PT, ii EC, or SE. Analyses of the within father, and father versus non-father groups indicates some support for higher self-esteem in the high paternal versus low paternal involvement groups. An inverse relationship is suggested between empathic concern and paternal involvement. Near significant results support these findings in the correlational analyses with paternal involvement. Findings for sub-hypotheses include support for a positive relationship between desire to father and perspective-taking for non-fathers. Results suggest enough support for further research in this area, including longitudinal research to determine potential direction of effect for any relationship between father's personality development specifically, or parents generally. iii Table of Contents Abstract ii Table of Contents iv List of Tables vii Acknowledgement xi CHAPTER ONE 1 Introduction 1 CHAPTER TWO 10 Research Literature Review 10 Empathy Literature 23 Empathy Measures 26 Self-Esteem Literature 32 Self-Esteem Measures 35 Paternal Involvement in Child Care Measures 3 8 State Versus Trait 42 Hypotheses 43 Main Hypotheses 43 Hypothesis One: 43 Hypothesis Two: 44 Hypothesis Three: 44 Hypothesis Four: 44 Hypothesis Five: 44 Hypothesis Six: 45 Hypothesis Seven: 45 Hypothesis Eight: 45 Hypothesis Nine: 46 Sub Hypotheses 46 Hypothesis Ten: 46 Hypothesis Eleven: 46 Hypothesis Twelve: 47 Hypothesis Thirteen: 47 Hypothesis Fourteen: 47 Hypothesis Fifteen: 47 Hypothesis Sixteen 48 CHAPTER THREE 49 Method 49 Design 49 Sample Selection 49 Measures 51 Paternal Involvement in Child Care 51 Empathy 51 Self-Esteem 52 iv Demographic Information 53 Procedures 54 Statistical Analysis 56 Hypotheses One Through Six: Comparing Father's Level of Paternal Involvement Between Fathers and Against Non-Fathers ....56 Level One Analysis 56 Level Two Analysis 56 Hypotheses Seven Through Nine: Analyzing the Relationship Between Paternal Involvement and the Dependent Variables for Fathers Only 57 Hypotheses Ten Through Twelve: The Relationship Between The Desire to Parent and the Dependent Variables 57 Hypotheses Thirteen Through Sixteen: The Relationship Between Recollection of Father Involvement and the Dependent Variables 58 Analysis of Relationships of Demographic Variables to the Research Variables 59 CHAPTER FOUR 62 Results 62 Sample Characteristics 62 Relationships of Demographic Data to Research Variables 66 Comparative Samples 70 Research Hypotheses Results 76 Hypotheses One Through Six 77 Level one analysis 77 Level two analysis 82 Hypotheses Seven, Eight, and Nine 86 Hypotheses Ten, Eleven, and Twelve 88 Hypotheses Thirteen Through Sixteen 88 Level one analysis for recollection of involvement 90 Level two analysis for recollection of involvement 92 CHAPTER FIVE 93 Summary and Conclusions 93 Restatement of Problem, Purpose and Findings ....93 Summary of Main Hypotheses Results 95 Summary of Sub-hypotheses Results 97 Conclusion 98 Limitations 105 Implications for Counselling and Future Research 107 v References 115 Appendix A 122 Appendix B 13 0 Appendix C 143 Appendix D 150 vi List of Tables Table 1 Mean Age for all Subjects, Non-father, and Father Groups 63 Table 2 Means for Length of Relationship for all Subjects, Non-father, and father groups 63 Table 3 Correlations of Age and Relationship Length and Research Variables for Whole Sample 67 Table 4 Crosstabulation of parental involvement class by Partner's birthplace 69 Table 5 Summary Table for other research using variables PI, PT, EC, and SE 73 Table 6 Results of t-tests Between Sub-samples and variables, PI, PT, EC, SE, and RINV 76 Table 7 Means and Standard Deviations of PT, EC, and SE for the Three Father Groups 78 Table 8 Results of t-tests Between Non-fathers and High Paternal Involvement Fathers 79 Table 9 Results of t-test analyses with Low PI and High PI fathers for PT, EC, and SE 80 Table 10 Medium and high parental involvement means t-tests for Perspective-taking, Empathic Concern, and Self-esteem 81 Table 11 Medium and low parental involvement means t-tests on Perspective-taking, Empathic Concern, and Self-esteem 81 Table 12 F-ratios for the dependent variables and the "two" father groups 82 vii Table 13 Results of t-test analyses for non-fathers and low PI fathers for the two father groups 83 Table 14 Results of t-test analyses for high PI and non-fathers, for the two father groups 84 Table 15 Results of t-tests on low PI and high PI fathers for the two father group comparison 85 Table 16 Correlation coefficients for PI with PT, EC and SE 86 Table 17 Correlations for sub-scales of PICCI and PT, EC, and SE 87 Table 18 Kendall correlation Table for PDF with PT, EC, and SE 88 Table 19 Pearson correlation coefficients for RINV with PT, EC, and SE for the whole sample 89 Table 20 Pearson correlation coefficients for RINV with PT, EC, SE, and PI for fathers 89 Table 21 Pearson Correlation coefficient for RINV with PT, EC, and SE for non-fathers 90 Table 22 Three father group t-test results compared with non-fathers on variable RINV 91 Table 23 Two father group t-test results for non-fathers versus high and low PI fathers on variable RINV 92 Table C.l Tabulation of Ethnicity and Religion for Subjects and Partners 144 Table C.2 Hours of Work per Week and Percentage Done at Home for Both Subjects and Their Partners 145 viii Table C.3 Hours of Study and Percentage Done at Home for Both Subjects and their Partners 146 Table C.4 Table of Frequencies for Personal Inlcome 147 Table C.5 Table of Frequencies for Household Income 147 Table C.6 Frequency Table for Main Activity of Subjects and Their Partners 148 Table C.7 Frequency Table for Age and Sex of Children 148 Table C.8 Frequency Table for Years of Schooling for Fathers, Non-fathers, and Their Spouses 149 Table D.l Spearman Coefficients for Independent and Dependent Variables with Demographic Variables (Fathers Only) 151 Table D.2 Results of One-Way Analysis of Variance Comparing the Average Number of Children in Each of the Paternal Involvement Groups . . . 152 Table D.3 Kruskal-Wallis One-way Anova for Years of Schooling and Non-Father, and High, Medium, and Low Involvement Groups 153 Table D.4 Kruskal-Wallis One-way Anova for Religiosity by Non-father, High, Medium, and Low Involvement Groups 153 Table D.5 Kruskal-Wallis One-way Anova for Study Time per Week by Non-father, High, Medium, and Low Involvement Groups 154 Table D.6 Kruskal-Wallis One-way Anova for Percentage of Study Time at Done at Home by Non-father, High, Medium, and Low Involvement Groups 154 Table D.7 Kruskal-Wallis One-way Anova for Hours of Work per Week by Non-father, High, Medium, and Low Involvement Groups 155 ix Table D.8 Kruskal-Wallis One-way Anova for Percentage of Work done at Home by Non-father, High, Medium, and Low Involvement Groups .... 155 Table D.9 Kruskal-Wallis One-way Anova for Personal Income by Non-father, High, Medium, and Low Involvement Groups 156 Table D.10 Kruskal-Wallis One-way Anova for Household Income by Non-father, High, Medium, and Low Involvement Groups 156 Table D.ll Kruskal-Wallis One-way Anova for Partner's Years of Schooling by Non-father, High, Medium, and Low Involvement Groups 157 Table D.12 Kruskal-Wallis One-way Anova for Partner's Religiosity by Non-father, High, Medium, and Low Involvement Groups 157 Table D.13 Kruskal-Wallis One-way Anova for Partner's Study Time per Week by Non-father, High, Medium, and Low Involvement Groups 158 Table D.14 Kruskal-Wallis One-way Anova for Partner's Percentage of Study Time Done at Home by Non-father, High, Medium, and Low Involvement Groups 158 Table D.15 Kruskal-Wallis One-way Anova for Partner's Work Hours per Week by Non-father, High, Medium, and Low Involvement Groups .... 159 Table D.16 Kruskal-Wallis One-way Anova for Partner's Work Hours per Week by Non-father, High, Medium, and Low Involvement Groups .... 159 x Acknowledgments During the long process of writing this thesis a number of people have been significant in supporting me. Marv Westwood, who gave me the room to stamp my signature on this work. Beth Haverkamp, who provided me with meticulous and conscientious feedback to produce my best work. To James White for his creative input especially at the beginning of this work, and always exhorting me to get the most out of my work. To Nettie, my wife, who both nursed our first child, and sometimes me through some of the most difficult moments of the thesis writing process. To my son Nathan, who was concieved and spent the first year of his life watching his Daddy struggle through. Finally my Dad, for giving me the tools to who gave me the basis to be a father. X* The Relationship Between Empathy and Self-esteem to the Level of Paternal Involvement in Child Care CHAPTER ONE Introduction The industrial revolution brought about changes in the family that are still with us today. Men went to work away from the home leaving mothers to raise children at home. Families have decreased in size since agrarian times because large families are not necessary for labor. Women spend less time raising children and more recently are seeking employment outside the home. Society, however, has not fully caught up to these changes because although women work outside the home much more now, they are still the primary care takers of children (Lamb & Bronson, 1980). Although this sketchy scenario may not do justice to hundreds of years of social history, it captures some of the broad developments in family structure and lays the groundwork for this study. At the same time and possibly partly as a result of this movement, the structure of the family is changing. These new family structures have received attention from researchers. For example, we are seeing dual-earner families (Volling & Belsky, 1991), role reversed families (Russell, 1978), shared care giving families (Russell, 1989), and men winning custody of children in divorce proceedings (Robinson & Barret, 1986). These developments may or may not have an impact on the male role, but there is some evidence that the man's role in the family is changing (Baruch & 1 Barnett, 1986a), and it is certainly a topic of much exploration. This research concerns itself with the role of the father in the family, his level of involvement, and how it may be related to his personality. The degree to which men are parenting, or whether their participation is increasing is debatable (Cowan & Bronstein, 1988) and the evidence is contradictory (Lewis & O'Brien, 1987). Findings in a Swedish study of couples on parental leave indicated that one third of the fathers took a major role in infant care (Lamb et al., 1982). A similar study by Radin (1986) found that only 25% of men who were originally primary care givers with their new born were so after a 4 year follow-up. Many of these studies report paternal involvement in minutes or hours and show that fathers' participation has not greatly increased over the years (Lewis & O'Brien, 1987). Although this research seems to contradict a significant or discernible shift in men's involvement in parenting, there is some suggestion that it is changing in the last decade (Daniels & Weingarten, 1988), and enough evidence to suggest that there are some highly involved fathers. To set the context for the changing role of the father it would be useful to examine his "traditional" role during the past century. The traditional role describes a style of parenting that has been the norm and has been unchallenged and unexamined until the 1960's, but still prevails today (Feldman, 1990). This role is characterized by the "husband-breadwinner/wife-homemaker nuclear family" (Fein, 1978, p.124). More precisely, this is the traditional twentieth century family structure, but it implies a 2 dichotomy where the father and mother perform duties that are largely complimentary, but exclusive. This dichotomy suggests that possessing certain personality characteristics is advantageous or deleterious to each side of this split. Characteristics that describe the traditional father are protector, provider, assertive, aggressive, competitive, and rational (Coleman & Coleman, 1981). Coleman and Coleman (1981), further suggest that these fathers deny the importance of emotions, intuitions and nurturing. They generally have low involvement in the day to day care of their children. The profile created here is a common one cited throughout the fathering literature (Feldman, 1990; Russell, 1978). The mother side in the mother/father dichotomy is the antithesis of the traditional father (the homemaker, child minder, and being more nurturant). The attributes suggested by the traditional mother role does not mean that they are exclusively the female domain. Recent research supports the contention that men are becoming more involved in the nurturing aspects of parenting (Cowan & Bronstein, 1988; Fein, 1978; Radin, 1986; Russell, 1978). One analogy is that of the traditional father as a "sky father" whereas "earth father" is his antithesis (Coleman & Coleman, 1981). The earth father's activities are nurturing and focused on intimate parenting behaviors that sustain relationships within the family. These activities also describe the non-traditional father (Fein, 1978), or highly involved father (Russell, 1978). He is characterized as more emotionally responsive, warm, and affectionate (Bronstein, 1988). It is 3 debatable whether this emergent perspective on men has spawned an upsurge in the non-traditional father (Cowan & Bronstein, 1988). Problem Area The above description of new family structures and father involvement in child care demonstrates a new focus on the father's role in the family. To date much research on fathering explores the father's contribution to child development (Radin, 1986). Very little research examines the father's experience of being a highly or traditionally involved in child care (De Lucci & Davis, 1991a). The present research focuses on relationships between level of paternal involvement and characteristics of the father. There is an apparent lack of developed theories on the experience of fathering on male development. One theory asserts that fatherhood is an opportunity for qualitative adult development, specifically facilitating the process of differentiation and individuation (Cowan, 1988). Cowan (1988) further argues that a man's self-concept can change in fatherhood, specifically in the areas of identity, locus of control, and self-esteem. For example, he found that new fathers realized that different qualities were necessary in being a father than those needed to compete in the work world, and they experienced a yearning to be caring and empathic. Although many men still find it unsafe to express feelings of vulnerability, Cowan (1988) further found that some men went against the stereotype and reported that having a child brought them more in touch with their feelings and that they felt more comfortable with self disclosure. 4 Although Cowan's (1988) work is descriptive in nature, it supports the idea that the experience of fathering has an impact on men. George Herbert Mead states that the human self arises through the ability to take on the attitude of the group he or she belongs to. The individual is continually reacting back to society causing an adjustment on the part of the community to which the individual adjusts (Strauss, 1956). This strongly suggests a reciprocal process in the development of human identity. On a larger scale he argues that when groups in a community are separate from other groups, that this creates hostility or enmity towards each other (Strauss, 1956). The idea of an interactive response between persons appears also to fit with Cowan's (19 88) findings that some men report changes in themselves from the experience of fathering. Mead suggests that some type or amount of contact between people is an essential ingredient in human development. Donaldson (1980) uses this idea in research on changing the attitudes of non-disabled persons towards disabled persons. He further asserts that non-handicapped persons exposed to handicapped persons may develop a more empathic response and a positive attitude change through this exposure. Although this is a different population than is addressed in the present research, and there is contradictory evidence to the above assertion (Bell, 1962), perhaps a similar process is also at work for fathers who are more highly involved in raising their children (i.e., more contact) than those less involved. Cowan (1988) also cites this possibility in his work referred to above and Russell (1982), who 5 suggests that parental values may shift in highly participant fathers, who may place more value on expressiveness and interpersonal relations. However, Russell believes that fathers who are highly committed and sensitive to children are more likely to have these characteristics before they became highly involved fathers. This brings into doubt whether certain predispositions are in place before being involved in fathering, or whether contact with one's children promotes changes. Another theory with relevance to this area of study is Erickson's stages of psychosocial development. Of particular interest to the present research is Erickson's sixth stage of development (loosely referred to as generativity versus stagnation, Monte, 1980). In this stage mature adults begin to feel a concern for the next generation reflected in the care of their own offspring. For this stage "care" is defined by Erickson as "the widening concern for what has been generated by love, necessity, or accident: it overcomes the ambivalence adhering to irreversible obligation" (Erickson as quoted in Monte, 1980, p. 257). John Kotre (1984) refers to a type of generativity called parental generativity occurring from young adulthood to old age. Nurturing and basic care of children are parental expressions of generativity. Kotre (1984) notes that more men are becoming involved in the care of their children and that "communal impulses" are being released as a result. It is unclear what communal impulses are, but it is clear that Kotre (1984) believes that involvement in child rearing is beneficial in the resolution of the generativity stage. 6 Research done to examine paternal involvement in child care largely looks at antecedents that influence the level of involvement. For example, Radin and Harold-Goldsmith (1989) found that fathers in their study were more involved in child care if they were jobless, the mother worked, and they held a flexible view of the male role. Baruch and Barnett (1986b) found that, if the mother had a flexible attitude toward the male role, the father was likely to be more highly involved in child care. Other work has shown that the age (De Lucci & Davis, 1991b), and sex of the child (Radin, 1985), were predictors of levels of paternal involvement. Russell (1978) on the other hand found no significant effects for number and sex of children, fathers' age, socioeconomic status, or the mothers' employment status (although an interaction may exist), in his work on father involvement and sex-role. Although there are methodological concerns with some of the above works, they do point to variables related to paternal involvement in child care. Demo and associates (1987) assert that most research examining parenting and self-esteem has been unidirectional, usually examining for parental effects on child self-esteem. Instead, they assume that there is some measure of reciprocity of effect between children and their parents. In fact, in their literature review they claim they could find no research examining the possibility that children effect parental self perceptions (Demo et al., 1987). Some researchers assume that fathers who are highly involved in parenting already possess the qualities associated with this level of involvement (Russell, 1982), and 7 that these characteristics antedate levels of involvement. Lamb (1982) warns us against making this assumption and cautions us against reaching conclusions about causality in what is largely correlational research. A previous warning was issued by Richard Bell (19 68) who found that research to this time emphasized parents as a major determinant of human development. In a review of the literature Bell found support for the assumption that there is a genetic basis that will activate parental responses and reinforce parental behavior previously evoked (Bell, 1968). The purpose of the following research will be to measure father's involvement in parenting their children. Also measured will be the personality variables of empathy and self-esteem, in particular, their relationship to the level of paternal involvement. A group of non-fathers will act as a comparison group to increase descriptive and explanatory opportunities. Their empathy and self-esteem scores will be compared to the father's scores by level of paternal involvement. It is expected that as the level of paternal involvement increases, so will the characteristics of empathy and self-esteem. Much of the following literature review will highlight the characteristics of empathy and self-esteem (as well as other personality variables) as related to paternal involvement. Also emphasized will be literature that supports the possibility that children or contact with children through parenting may influence adult development. The present research cannot determine directionality, but a contrast group of non-fathers is used to 8 consider the possibility that children can effect parents as well as the often researched parent to child effect. CHAPTER TWO Research Literature Review Heath (1978) explored the idea that becoming a parent facilitates men's resolution of Erickson's generativity stage, in which a person develops the dimensions of stability, integration, and autonomy. In his study, Heath used professional men who were part of another longitudinal study and measured fathers versus non-fathers. The men were administered a battery of personality tests when they entered college and then re-tested some years later as adults. Statistical analyses revealed that the fathers showed significantly higher levels of empathic concern (as measured by the Valuator Test for Values, which is a test about values more than about empathy), were more considerate, sensitive and sympathetic, and rated themselves as more understanding and intimate in relationships to persons they felt close to (Heath, 1978). Note that the above research utilized a small sample of men for the number of variables measured, and there was dropout over the years of the original study. Caution must be used in interpreting the significance of these results. Further, rather than measuring level of paternal involvement, it was fathers versus non-fathers only, who were being compared. In spite of these limitations, this work can lend some credence to the notion that parenthood does have an impact on men. Also, along with Cowan's (1988) work, it is one of the few studies that examines the impact parenting may have upon the father. 10 In a largely anecdotal work, Greenberg and Morris (1974) studied the experience of being a first time father. The authors observed that there were specific aspects of the developing bond in the first few days after the child's birth. In a process they term "engrossment" the fathers report increased feelings of esteem, adequacy, and worth. It should be noted that this study is descriptive and limited to the first three days after birth. It also uses inference to make leaps in logic about the process of engrossment, not supported by theory or data. In a study that furthered the work of Greenberg and Morris (1974), fathers were examined in two parental contact conditions. Keller and associates (1985) randomly assigned new fathers into two groups. Those who were provided intense contact with their newborns at the hospital (the authors did not provide details of what extended contact was), and a group of fathers who were also present at delivery, but had only "traditional" contact with their newborns. At the prenatal stage fathers were given a demographic questionnaire and the Tennessee Self-Concept Scale, and then re-administered the surveys at six weeks post partum, as well as a survey that requested other information including child characteristics, attitudes concerning the birth, and the father's relations with his spouse. Results showed those fathers in the extended contact condition provided a greater proportion of time to the care of children than fathers in the traditional condition. Scores at the postpartum stage between the two father groups were also significant at the .05 level on the self-concept scale. The extended contact fathers showed a significant increase at the six 11 week observation over the traditional contact fathers on the Total Positive Self scale. The authors speculated that experimentally increasing the amount of contact fathers had with their infants had the effect of increasing self-esteem (Keller et al., 1985). Although the authors took some steps to follow a rigorous quasi-experimental design (e.g., double blind technique, using mother's scores of paternal participation as validation), there were some questionable aspects of the study. For example, there was little description of the operationalization of the extended versus traditional contact groups. It is also notable that there were differences in self-concept scores between the two groups at time one, and it is unclear as whether this difference is significant or not. Finally, it is notable that in their literature review of research on fathering, they could not find in similar research, self-esteem as measured by a psycho-metrically valid, standardized measure. Another study assessed the attributes associated with male care givers of elderly people. Kaye and Applegate (1990) found that these men described themselves as possessing affective traits typically associated with women, as measured by a sex role inventory and a battery of other personality tests. The findings of this study are limited, particularly in its sampling procedures (a biased sample drawn from volunteers in one institution), but supports an association between involvement in care giving and nurturing characteristics. Other anecdotal research, as an adjunct to work examining determinants of paternal involvement in child care, found that men highly involved in parenting exhibit 12 less sex differentiated behavior and felt more involved and competent as parents (Baruch & Barnett, 1986a). Van de Water and Mc Adams (1989) conducted research examining the relationship between nurturing and generativity as defined by Erickson. They hypothesized that care was reflected through the trait of nurturing and that generative adults (those who scored high on tests that reflect generativity according to the authors) will score high on nurturing. They found that there was a significant correlation between high scores on measures of generativity and the nurturance scale from Jackson's Personality Research Form (Van de Water and Mc Adams, 1989). Salient to the idea of generativity according to these authors was also the idea of adults taking on a role of nurturing, supporting, and guiding the next generation. The works and ideas presented to this point have touched upon sex-role orientation and the qualities or adjectives associated with "masculinity" and "femininity". Work by Bern (1981) suggests that stable traits are associated with masculinity, femininity, or a high mix of both, called "androgyny". She found that more men score as masculine whereas women score as feminine on her scales. The descriptors that compose femininity that are salient to this work are adjectives such as compassionate, warm, sympathetic, sensitive, and understanding (Bern, 1981) and are similar to the types of descriptors cited in the research associated with parenting. One study did measure the relationship between level of paternal involvement and sex-role classifications (Russell, 1978). 13 Russell found that androgynous scoring men participated in more nurturing behavior towards their children than masculine scoring men, and participated more in the day to day care of their children. Although there are sampling biases in this study, it is methodologically sound and lends support to the nurturant type of characteristics of interest to the present research. For example, empathy has often been cited as a motivation for prosocial behavior that is "associated with a tendency to help" (Hoffman, 1977, p. 199). Radin and Sagi (1982) examined correlates of child rearing fathers in the US and Israel. Citing Aldous' (1974) work, they saw participant fathers as "role makers" where individuals who create new roles in families are expected to have high esteem, internal locus of control, and interpersonal sensitivity. They compared low and high involved fathers in these countries and found that highly involved fathers in Israel had children who had greater empathy and internal locus of control. These constructs were inferred from a structured interview. This is typical of research that examines the impact of fathering on children and the antecedents to level of paternal involvement, but could also suggest possible differential effects of levels of paternal involvement. It is suggested that parents who have an externally oriented locus of control are engaged less in direct family involvement than those with an internally oriented locus of control (Swick, 1986). In a later review of literature of parental efficacy and related variables, Swick (1988) concluded that parents having an 14 internal locus of control had positive relations with their children. While this work addresses parenting in general, and quality of relationship (through a construct called "positive relations") versus quantity of involvement, it points to a possible relationship between quantity of parenting and locus of control. Another work by Mondell and Tyler (1981) measured warmth and planning activities with children as related to competence. These authors equated self efficacy to locus of control and found that competence and locus of control were significantly correlated. This association was later tested by Sherer and Maddux (1982), finding that those with an internal locus of control (as measured by Rotter's I-E scale) were more likely to have high self-efficacy regarding their ability to parent. They found that parents with an internal locus of control expressed warmer affect and were more helpful in problem solving. Again this research is only suggestive of a link between quantity of involvement (e.g. planning activities and problem solving with children) and locus of control. Russell (1982) interviewed fathers in two types of family structures: the traditional family and the shared care giving family. The traditional family had fathers who went to work every day and had limited involvement with their children, while the shared care giving families had fathers and mothers who both worked and apparently shared in the care of the children. Although in this latter family structure the mothers spent the most time on parenting tasks, Russell found that the fathers in the shared care giving families had become more sensitive, 15 understood their children better, had more physical contact with their children, and reported greater satisfaction with their role as a parent. Later work by Russell (1989) confirmed that shared care giving fathers reported greater satisfaction with their parental involvement and that they scored higher on competence and self-esteem as parents than did traditional family fathers. These constructs were measured by self-report instruments created by the author, so validity has not been established. The shared care giving and traditional family dichotomy in both studies (Russell, 1982 & 1989) was based also on maternal work patterns and subsumes a larger set of variables than merely paternal involvement. Any associations to paternal involvement may be generous at best and misleading at worst. David Demo and associates (1987) hypothesized that higher levels of reported communication and support between parents and children, and parental control over decisions concerning the adolescent, results in a higher level of self-esteem being reported by the parents and adolescents. Self-esteem was measured by the Rosenberg Self-Esteem Scale (Rosenberg, 1965) while the other variables were researcher developed scales. Data was analyzed using correlation and regression coefficients. It was found that perceived support of parents by adolescents, and perceived level of communication between fathers and adolescents was significantly correlated with paternal self-esteem. Because not all variables significantly correlated with parental self-esteem, but a larger number did so for adolescents, the authors 16 posited that perhaps adults derive self-esteem from more sources than children, who would derive much from their parents (Demo et al., 1987). A recent study measured perceived closeness of parents to their children and parenting satisfaction (Paulson et al., 1991). Here the variables of closeness, warmth, self-esteem and expressiveness were measured in father-son and father-daughter dyads. These variables were measured for the child and the parent. It was found that closeness in the parent-child relationship predicted parental satisfaction with parenting ability (a positive relationship) although parents' reports of closeness to their children did not predict any variance in self-esteem, expressiveness, or instrumentality. Although this finding is not supportive of the present study's hypotheses, it is worth noting that closeness was defined largely in a qualitative sense (e.g., intimacy and self-disclosure) as opposed to a quantitative level of involvement. Related work examining the effects of paternal involvement on children found that the children of fathers who were highly involved in their care, after a 26 year follow-up, scored higher on empathic concern than those whose fathers were less involved in their care (Koestner et al., 1990). Paternal involvement was measured by spousal reports and so may not necessarily be reliable. Also, this study examined the effect of paternal involvement on child development. It is unfortunate that fathers in the original sample were not measured for empathic concern as well. It does however, lend support for the notion that empathy 17 may be linked to paternal involvement and raises the question of whether the inverse of this relationship may be true or whether there is some sort of interaction effect between fathers and children and the development of empathy. Jay Belsky (1984) has developed a model of three sources of parental functioning culled from the parenting literature including parental personality characteristics, the child's individual characteristics, and contextual sources within which the parent-child relationship is embedded. The model also assumes various reciprocal relationships between developmental histories, marital relations, social networks, and jobs, with individual personality characteristics and psychological well-being of parents. He further assumes that parenting is multiply determined, that all factors are not equally influential in determining parental functioning. Personality characteristics influence parenting indirectly by influencing the social context in which parenting exists. Belsky (1984) cautions that most of the studies are based on correlational data and causality cannot be determined. Research by Crouter and associates (1987) found that fathers in dual-earner families were more involved with their children than fathers in single-earner families (contextual factors), but often at the expense of marital relations (dual-earner fathers reported more marital conflict than single-earner fathers). This shows that work and marital relations are related to the level of involvement in child care of at least some fathers. While this study was methodologically sound, sample sizes were small (20 18 dual-earner and 20 single-earner families) limiting the generalizability of the findings. Further research by Crouter and Crowley (1990) also found that dual-earner and single-earner fathers had different levels of involvement with their children. Single-earner fathers spent significantly more time with their sons than daughters, while dual-earner fathers spent equal amounts of time with sons and daughters (Crouter & Crowley, 1990). Radin and Harold-Goldsmith (1989) found that if the father was jobless, held a flexible view of the male role, and the mother was working, he was more likely to be highly involved in child care than if these factors did not exist. These studies are correlational in nature, but still support the idea that contextual factors are related to parental involvement in child care as depicted in Belsky's (1984) model. Later research by Volling and Belsky (1991) also supported Belsky's (1984) above cited work. Volling and Belsky conducted longitudinal correlational research on 54 dual-earner and 65 single earner families exploring for determinants of paternal involvement in child care. Based on Belsky's (1984) work, the authors define determinants as being either characteristics of the father, characteristics of the child, or contextual factors. Paternal involvement in child care was measured as two separate dimensions. One was "father-infant interaction," observed in the home and included father's responding behaviors such as stimulating the child, physical care, and/or positive affection. A second dimension was termed "father responsibility for child care," and was composed of a series of child care 19 related tasks. The degree to which the father was responsible for these tasks was rated jointly by the mother and father (Volling and Belsky, 1991). Volling and Belsky (1991) employed regression weights to examine the relationship of father-infant interaction and father responsibility on each of the determinant variables. This was done at both 3 and 9 months of age of the child for the total sample. Findings can be summarized as follows. Measuring all fathers for father-infant interaction, fathers who were older, had higher self-esteem, and had higher interpersonal affect, engaged in more of the activities defined as father-infant interaction with their children, but this held true only at the 3 month age measurement of the child. For the variable of Father Responsibility, fathers were categorized as single-earner and dual-earner (based on Crouter and Crowley's 1990, and Crouter and associates 1987 work). At both the 3 and 9 month age-of-child intervals, fathers in single-earner families who scored higher on interpersonal affect (but not self-esteem) had higher paternal responsibility scores. However, dual-earner fathers participated more in child care than single-earners over all, but quality of the marital relationship for dual-earner fathers predicted involvement. Of note, it seemed that personal characteristics of the single-earner father predicted paternal responsibility in child care, while contextual factors such as pre-marital division of labor (e.g., fathers who were more involved in traditionally female household tasks participated more in child care), and marital relations (more negative marital 20 relations meant more father involvement in child care), predicted paternal responsibility in child care for dual-earner fathers. There were no significant effects for the demographic or child characteristic variables measured. These results led Belsky and Volling (1991) to speculate that perhaps single-earner fathers became more involved as a matter of choice and therefor those who evinced more concern for others' feelings (interpersonal affect) became more involved in the care of their children as a matter of choice. Dual-earner fathers on the other hand became more involved in the care of their children out of necessity and therefor it is contextual factors among this sub-group that predicts differences in father responsibility. Some of the conclusions made by the authors in terms of cause and effect based on the longitudinal design of the study should be taken with caution. Parents are involved in the care of their children for many years, while this study examined only a 6 month interval at a very early stage in the child's life. Radin's (1985) four year follow-up study of paternal involvement leads us in this caution, because after four years she found that the highly involved fathers were significantly less involved after a four year follow-up. Taken together, the literature cited above seems to indicate a number of personality characteristics that are relevant to fathers and their involvement in child care. For example, Radin and Sagi (1982), and Russell (1989) found that more involved fathers exhibited higher self esteem while Keller and associates (1985), found this true for self-concept. Self-esteem is one 21 variable that is worthy of reinvestigation in the present study. However, most research is suggestive of an empathic, nurturing, or affective process at work (Heath, 1978, Russell, 1978 & 1982; Greenberg and Morris, 1974; Kaye and Applegate, 1990; and Cowan, 1988) . Volling and Belsky (1991) found that the variable called Interpersonal Affect (as taken from the Jackson Personality Form devised by Jackson in 1980) was related to paternal involvement. Interpersonal Affect is defined as the degree to which an individual is oriented towards the feelings and concerns of others (Jackson, 1976). A high scoring subject on Jackson's scale is defined as a person who "tends to identify closely with other people and their problems, values close emotional ties with others", and is concerned about others (Jackson, 1976, p.10). On the other hand the definition of Empathic Concern from the Interpersonal Reactivity Index (Davis, 1980) is the degree to which the respondent "experiences feelings of warmth, compassion and concern for the observed individual (Davis, 1980, p. 10). Although it is beyond the scope of this research to assess the concurrent validity between these measures, intuitively at least, there seems to be a similarity between the variables. Empathy also appears to be a characteristic or process that could be beneficial for parenting, and in this study will be examined in relation to paternal involvement with children. The above studies of variables correlated to levels of paternal involvement in child care suggest that there are differences between men who are highly involved and those who are 22 not. Although variables such as empathy and self-esteem are cited among the different studies, their measurement is largely through interview or non-standardized tests (Greenberg & Morris, 1974; Russell, 1982; Radin & Sagi, 1982; Cowan, 1988). This points to the need for more rigorous measurement of these variables. Following is an examination of the literature related to empathy and self-esteem. Empathy Literature Webster's Dictionary (1984, p. 229) defines empathy as the "identification with, and understanding of the thoughts or feelings of another." Gladstein (1983) and Chlopan, McCain and Carbonell (1985), suggest a dual model of empathy. The first, similar to the dictionary definition, also referred to as the cognitive definition, is the ability to understand another's thinking or feeling and accurately perceive their world. A second model, referred to as affective empathy, is conceptualized as an emotional contagion, a person's emotional response to observing another's actual or anticipated condition. Empathy appears to be an ability as well as a process. Mead refers to role-taking (widely recognized as empathy, as cited in Hogan's 19 69 work) as a social and individual process necessary for a developing self and adjusting to society. Further it is seen as an emotional experience in which we participate in the experience of another (Katz, 1963). Katz further argues that empathy is an innate and primitive form of communication, but that civilization puts emphasis on other sense modalities such as sight and hearing to communicate. The ability to empathize can be 23 diminished and as well enhanced and relearned. If this is indeed possible, then could the intimacy of an involved relationship between parent and child be such an enhancement? Judith Jordan (1984) views empathy as an interactive process between two people and as developmental for individuals. It is an experienced connection and similarity to other people. She further argues that empathy is often confused with a contagion-like phenomenon where the individual experiencing another's state loses his or her own sense of self. However, for one to empathize, he or she must have a well-differentiated self with a sensitivity to the similarities and differences of another. Empathy requires a balance between affective and cognitive processes and boundary flexibility in order to experience another's state while also maintaining a sense of self. She also describes empathy as an interactive process where, for example, a mother feeding a child is providing mirroring cues for her baby, but is also experiencing some identification with the eating child (Jordan, 1984). This little vignette, while hypothetical, suggests an interactive process where the mother, who is making subtle mirroring cues for the baby, is responding herself to a feeling about the baby's needs. Although she is demonstrating an empathic ability by her focus on the child, she is also being influenced by the interactive process and perhaps the baby's cues. This again may suggest where the connection and focus between parent and child may influence the parent as much as the child and could be an empathy enhancement as referred to by Katz (19 63). 24 Goldstein and Michaels (1985) argue that empathy is multi-dimensional. They outline four components to empathy: cognitive, affective, perceptual, and communicative. Affective empathy is the emotional response one has to the emotions of another person. Whether that emotional response needs to be identical or accurate is less certain for Goldstein and Michaels. The cognitive component of empathy is an ability of the perceiving person to understand the thoughts, feelings, perspectives, or behaviors of another person. The perceptual and communicative components refer more to the process of first, attending to a target person's cues regarding his or her feelings, and second, verbalizing the understanding of, or experiencing vicariously, the target person's feelings (Goldstein & Michaels, 1985). These later two aspects of empathy appear to refer to a process rather than an ability. Goldstein and Michaels (1985) speculate that these components (including another called "socialization of empathy") are interactive and probably difficult to operationalize, but are presented as directions for further research. With respect to parental empathy, Goldstein and Michaels (1985) in a review of existing literature found some support for at least a reciprocal nature of child-parent empathy development. However, in their review they conceded that empathy was not always clearly operationalized. Moreover, most studies researched empathy related variables such as love, nurturance, and morality. Perception, reverberation, comprehension, and communication of another's emotion are not directly addressed by these variables, as is empathy in Goldstein and Michael's view (Goldstein and 25 Michaels, 1985). While Goldstein and Michaels (1985) hypothesize that close involvement with one's own child over a long period of time may enhance a parent's empathic ability, they also concede that the body of research on parental empathy is very small. The above work by Goldstein and Michaels (1985) supports the possibility that fathers who are more involved in the care of their children may then benefit from the increased connection. However, it must also be stated that it could be that men who are initially more empathic are inclined to be more involved with their children. The relationship between paternal involvement and empathy is one that still needs to be measured in a standardized way before exploring directionality. Empathy Measures As indicated earlier two basic models of empathy arise out of the literature; affective and cognitive (Gladstein, 1983). Affective empathy is the ability to respond to another's emotions with the same emotion. This happens when the observer has acquired the same emotional state of another. It describes a contagion effect. This definition means that a father has the ability to feel his child's emotions, and that the more highly involved in the care of the child, the more he has this ability. The second model is cognitive empathy. This is the ability to understand another's thinking or feeling, and generally being able to perceive that other's world (Gladstein, 1983). This is consistent with the Johnson and associates (1983) view of empathy as the ability to apprehend another's condition and a sensitivity to the person's affective experiences, as well as Hogan's idea 26 that empathy is the act of construing another's mental state (Chlopan et al., 1985). Chlopan and his colleagues (1985) reviewed the Hogan Empathy Scale (representing a measure of cognitive empathy) and the Questionnaire Measure of Emotional Empathy (representing a measure of affective empathy) and found that both scales had adequate validity. They suggested that both instruments might be measuring two different aspects of empathy. The major difference is that in the cognitive model the individual is aware of perceptions of others, while this is not necessarily so for the affective model of empathy where one can respond to another's feelings without being aware of their own reactions (Gladstein, 1985). It appears that the affective model of empathy may be a harder construct to measure, because it could be out of the individual's awareness. Robert Hogan (1969) developed a scale to measure a cognitive construct of empathy. His test is designed to be unidimensional, and to measure the cognitive model of empathy (Chlopan et al., 1983) and the characteristics of a highly empathic person (Hogan, 1969). These characteristics include imaginative play and pretending; awareness of the impression one makes on others; insight into one's own motives; and being socially perceptive (Johnson et al., 1983). Hogan (1969) defined empathy as the ability to construe the mental state of another. However, in an examination of the psychometric properties of the Hogan Empathy Scale, Cross and Sharpley (1982) found a reliability score of .605, which they claimed was not sufficient for a published test. 27 They also found that the items did not posses sufficient standards of homogeneity for researchers. Many scales measuring empathy have been developed with relative degrees of success (Chlopan et al., 1983). Some have been developed for specific purposes, such as measuring nurses' empathy levels (Layton & Wykle, 1990), and counsellor empathy (Gladstein, 1985). Others attempt to measure different dimensions of empathy and create different sub-scales for this purpose such as the Mehrabian Epstein Scale (Chlopan et al., 1983), and the Interpersonal Reactivity Index (Davis, 1983). Indeed, there does appear to be some differences in the way that empathy is defined and measured. Rather than choosing a cognitive or an affective measure of empathy this research will measure both. The Interpersonal Reactivity Index (Davis, 1980) contains a sub-scale for each. Mark Davis (1980) has conceived empathy as a multidimensional construct. He notes in his work the distinctions between a test like Mehrabian-Epstein's (emotional reactivity) versus the work of Hogan's which has a cognitive emphasis, as being representative of two traditions. One tradition sees empathy as a type of emotional contagion, where one experiences the feelings of another, while the other tradition reflects a cognitive understanding of another's perspective. Davis sees these as legitimate distinctions, but argues that while both the Mehrabian-Epstein, and Hogan scales focus on a different tradition, they each also contain some items that are more representative of the other tradition. 28 Davis (1980) felt that a measure that tapped into both notions of empathy would yield more precise results of this construct. In developing his instrument he used a pool of 50 items from various tests combined with those he created, to tap into responses one might have to others' experiences. In a factor analysis he found the existence of four major factors. These included, empathic concern (the ability to experience feelings for others), perspective-taking (ability to adopt or understand point of view of others), fantasy (ability to identify with characters in books, novels, or plays), and personal distress (experiencing discomfort or anxiety when witnessing troubling experiences of others). Davis further refined the item pool to fit into the four emerging categories keeping only those that loaded most heavily for both sexes, and those that did not load heavily into more than one scale. The result was a four scale, 28 item test (seven items per scale) called the Interpersonal Reactivity Index (Davis 1980). In examining the inter-correlations of the four scales, Davis (1980) found that the perspective-taking and empathic concern scales were moderately positively correlated (.33 for males and .30 for females) demonstrating a relationship between the cognitive and emotional aspects of empathy, but also indicating that they are not measuring the same construct (Davis, 1980). The empathy and fantasy scales are also moderately correlated, but the perspective-taking and fantasy scales are only marginally correlated for males and females (.10 and .12 respectively). The empathy and the personal distress scale are not significantly correlated for females (.04) and only marginally correlated for 29 males (.11). The personal distress scale and the perspective-taking scales are low to moderately negatively correlated for females (-.16) and males (-.29). The four scales are titled Fantasy Scale, Perspective-Taking Scale, Empathic Concern Scale, and the Personal Distress Scale. The two scales most salient for this work are the Empathic Concern Scale (EC) and the Perspective-Taking Scale (PT). The former refers to "the degree to which the respondent experiences feelings of warmth, compassion, and concern for the observed individual" (Davis, 1980, p. 9-10). The latter refers to "an ability or proclivity to shift perspectives -- to step outside the self --when dealing with other people" (Davis, 1980, p. 9). These definitions together appear to capture the essence of empathy that a parent, or in this case father, may feel toward a child in his care. Since his original work, Davis has tested the multidimensional aspect of empathy, as well as it's convergent and divergent validity with the Hogan Empathy Scale and the Mehrabian and Epstein Questionnaire Measure of Emotional Empathy QMEE (Davis, 1983). As would be expected, the Perspective-Taking scale (From the IRI) and the Hogan Empathy Scale were significantly positively correlated with coefficients of .42 for males and . 37 for females. Significant (although lesser) correlations exist between the Empathic Concern scales and the Hogan Empathy Scale (HES) for both males and females (.11 and .25 respectively). A positive relationship of .63 for males and .56 for females also exists as expected between the Mehrabian and Epstein QMEE and the 30 empathic concern scale. Lesser positive correlations also exist between the QMEE and the perspective-taking scale for males and females (.22 and .17 respectively). The expected positive correlation between the PT scale and the HES, and the QMEE and the EC scale on the Davis test lend support to definitions of the cognitive and affective components measured by the respective scales on the Davis test. The low to moderate positive correlations between the emotional and cognitive scales versus the cognitive or emotional emphasis of each test could support Davis' (1980) contention that the HES and QMEE contain items that measure aspects of each other. Subsequent examinations of the IRI have upheld Davis' (1980) work. In a study utilizing a medical school student population, Coman and Evans (1988) found that student scores were generally similar to the original normative sample findings. However, it should be noted that this is only an informal judgment, as tests of significance could not be computed because standard deviations were not supplied in Davis' (1980) original work. Carey and associates (1988) conducted a study designed to replicate the factor structure of the IRI. In this study they used a random sample of dietitian and dietetic interns. Through a factor analysis they found the same four discernible dimensions of empathy found by Davis (1980). Although the sample utilized in this study was not necessarily representative of the general population because it was a different type of sample than those it has been tested on thus far, it supported the validity of the factor structure of the IRI across different populations. The 31 authors concluded that this instrument is a valuable tool in counselling practice and research. What the IRI seems to afford the present work is an opportunity to acknowledge the importance of both the emotional contagion and cognitive perspective-taking aspects that seem to enter into the definition of empathy. However, no attempt will be made to enter into justifying one as more "true" of empathy than the other. Davis (1983) asserts that empathy can best be understood as a set of constructs that are related, in that they all concern responsiveness to others. If it is acknowledged that both aspects of empathy are important, it can also be accepted that both be measured within the present study as possibly related to parenting. It is reasonable to propose that a father who takes care of a child will have a skill to understand the perspective of that child, while at the same time being able to feel the pain or joy of his own child. The more involved the father is with the child, the more he may be able to take the perspective of, or feel the feelings of the child. For these reasons the IRI seems to be a very appropriate instrument to capture these processes and/or abilities. Testing for the effects of both the emotional and cognitive aspects of a larger construct (if it is a valid construct on its own) called empathy also relieves us of the debate and possible pitfalls of not fully or validly capturing a construct. Self-Esteem Literature Self-esteem (or lack of it) is an idea one hears throughout contemporary and popular psychology as a cause for many of the 32 problems that effect humankind (Robin, 1989). Because it is such a commonly used term it is very likely that it has taken on a life of its own and can mean different things to many people. It will then be important to define it for the purposes of this study and establish its salience to paternal involvement in child care. A closely related term to self-esteem and one often referred to in relation to it is the notion of self-concept. Martin and Coley (1984) define self-concept as self knowledge that one possesses regarding one's strengths and weaknesses. Rosenberg (1979) postulates self-concept as the totality of thoughts and feelings referring to oneself. Self-esteem on the other hand is the desire to think well of oneself. It is a motivator of behavior and, along with the concept of self-consistency, it enhances and maintains self-concept (Rosenberg, 1979). The distinction between self-concept and self-esteem may be subtle, but it is an important one. Do fathers who are more or less involved with their children have more or less knowledge of themselves or more precisely do they have a higher or lower evaluation of themselves? Self-esteem has been described as a kind of attitude (e.g., approval or disapproval of the self) or an aspect of self-attitude. Campbell (1984) breaks self-esteem into two components. The first is what he terms a "good opinion of the self." The second is the cause of that good opinion and consists of the perceived possession of desirable objects or characteristics. The possession of the desirable qualities may be actual or potential, direct or vicarious, conscious or unaware. By "good," Campbell 33 does not imply a moral standard, but a subjective desirability for the self. This definition is similar to Rosenberg's, although Rosenberg (1979) conceives of self-esteem as a motivating force, while Campbell discusses "causes" of self-esteem. The distinction is muted when Campbell (1984) argues that the appetite for self-esteem is insatiable and people have the urge to achieve the objects or characteristics that lead to self-esteem. This is similar to Rosenberg's (1979) idea that self-esteem is a motivator for behavior. Campbell (1984) writes that sources of self-esteem are both other-dependent and self-dependent. The other-dependent sources of esteem such as acceptance, love, friendship, respect, etc., are given to us by others. He argues that esteem from others (social esteem) is sought because it implies qualities that are worthy of that esteem. David Demo (1985) views self-esteem as a "fluctuating self-attitude" with a baseline that may fluctuate given certain circumstances or different times. Further, measures of self-esteem seem to tap either experienced self-esteem or presented self-esteem. Presented self-esteem is that which is ascribed by others to a given person, while experienced self-esteem is that which is self ascribed. Both Demo's (1985) and Campbell's (1984) works are supportive of the notion that self-esteem is affected by relations with others. Despite the work by Demo and associates (1987) cited earlier, the contention that children can effect the self-esteem of parents is not a common one in research literature. Coopersmith's (1967) work on the antecedents of self-esteem quite naturally makes 34 connections between childhood experiences and circumstances, and level of self-esteem. For example, one interesting association he found was that fathers of high self-esteem pre-adolescent boys, take a more active and supportive role in rearing their boys. He concludes that children are able to influence their parent's behavior, but that the influence is usually in the other direction, particularly with younger children. As well, Mead suggests that it is more difficult for an adult to take the attitude of a child because of different levels of self-consciousness, but that it is necessary to do so to educate the child (Strauss, 1956). Again, this seems suggestive of a possible child-parent influence (the child influencing the parent's behavior, or in Mead's case that the parent should allow him or herself to be more influenced by the child), but there is little direct evidence to suggest this link does actually exist. Self-Esteem Measures Rosenberg (1979) postulates that self-esteem is one aspect of self-concept and perhaps the most important variable in personality research. Whereas self-concept is the totality of one's thoughts about oneself, self-esteem is an evaluation of whether one considers oneself a person of worth, having self respect, or thinking of the self as unworthy, and lacking self respect. For purposes of measuring this characteristic, Rosenberg developed the Rosenberg Self-Esteem Scale (Rosenberg, 1979) . Coopersmith, who has done extensive work on self-esteem, also offers a definition of self-esteem. "In short, self esteem is a personal judgment of worthiness that is expressed in the attitudes 35 the individual holds toward himself or herself. It is a subjective experience which the individual conveys to others by verbal reports and other overt expressive behavior" (Coopersmith, 1967; p. 5). Coopersmith further argues that this estimate of general self-esteem is enduring, and a subjective judgmental evaluation. His belief that, at least in part, self-esteem is conveyed by verbal reports lends itself well to being measured by a paper and pencil test. However, the Coopersmith Self-Esteem Inventory (SEI) is fraught with poor psychometric properties, including modest construct validity (Johnson et al., 1983), range restriction, and an unstable factor structure (Blascovich & Tomaka, 1991). The self esteem instrument that appears most appropriate for this research is the Rosenberg Self-Esteem Scale (SES) developed by Rosenberg (1979). Demo and associates (1987) used this instrument in their research examining relationships between relationship-style variables and self-esteem in a sample of adolescents and their parents. In this particular research the authors obtained a .86 test-retest reliability coefficient. Further, Demo (1985) researched this scale in an earlier work assessing the validity of a number of self-esteem measures. Although the SES demonstrated only moderate convergent validity with the other self-esteem measures (.55 with the Coopersmith self-esteem inventory and .32 with peer rating scales), in a factor analysis it was found that the SES showed a significant relationship with both experienced and presented self-esteem 36 thereby supporting the SES as a valid measure of self-esteem (Demo, 1985). In a review of the scale, Blascovich and Tomaka (1991) found that it possessed adequate reliability, convergent and discriminant validity, and that it had utility as a unidimensional measure of self-esteem. In terms of internal consistency, Dobson et al. obtained a Cronbach alpha of .77, while Flemming and Courtney (1984) obtained a Cronbach alpha of .88. In terms of test-retest reliability, Flemming and Courtney obtained a correlation of .82 after one week, while in their work Silber and Tippet (1965) obtained a correlation of .85 after 2 weeks. The SES demonstrates convergent validity showing a .72 correlation with the Lerner Self Esteem Scale ( Savin-Williams & Jaquish, 1988). The SES demonstrated discriminant validity in that negative correlations were found with anxiety (-.64) and with depression (-.54). Blascovich and Tomaka (1991) do warn that the SES is susceptible to socially desirable responding and that college students' scores tend to be negatively skewed; "low" self-esteem scores are found even within groups that have relatively high self-esteem. However, absolute self-esteem is not being measured in this research, but self-esteem relative to the different comparison groups. It is conceded that to describe one group as having lower self-esteem than another, when in fact there is not much real difference between the groups, calls in to question the significance of the findings. This point will be kept in mind when examining the results of this research. 37 Paternal Involvement in Child Care Measures Graham Russell's (1978) study of the father's sex role orientation measured aspects of parenting, including feeding, dressing, diaper changing, bathing, dressing for school, playing, and reading stories. Parents were interviewed with respect to the time in a week he or she spent performing these activities. Generally, it was found that fathers spent far less time performing any of these activities than mothers. Although Russell (1978) felt that the measurement of the amount of time spent on each activity was important, there are drawbacks to measuring only these activities. There is no doubt that they are necessary child care functions, but they are very specific and encompass only a few tasks necessary in caring for a child. Other research has attempted to measure the quality as well as the quantity of paternal involvement (Grossman et al., 1988). The quantity is measured by time spent in play and care as well as overall time spent with children on weekdays and weekends, but this is vague and does not necessarily capture the full responsibilities of parenting. Quality is measured by the child character constructs of "autonomy" and "affiliation". Although quality is a very important aspect of parenting, there can be many definitions of parenting quality and it is beyond the scope of this research to examine these qualities. A far more satisfying scale of paternal participation is one developed by Baruch and Barnett (1986b). They measured five different dimensions. The first is called "interaction," which captures quantity and intensity of the time spent with children. 38 Was the father just present, or was he interacting on a more involved level? They also measured absolute time the fathers spent in child care chores, solo versus joint participation with the mother, performance of household chores, and ultimate responsibility for tasks as opposed to simply performing them. These scales seem to encompass a fuller range of parenting tasks, although concentrating on physical tasks. This scale also involves extensive interviews, the keeping of time diaries (Baruch & Barnett, 1986b), and reports of parenting by the child (Levant et al., 1987). Also emphasized are perceived feminine or masculine aspects of some the tasks measured. This is not relevant for the present research. Finally, the rigorous recording demanded of the participants could be a deterrent to volunteering for research. Radin and Harold-Goldsmith (1989) developed a scale designed specifically to measure paternal involvement in child care. They developed the items on the questionnaire based on the family role literature (Radin and Harold-Goldsmith, 1985). The resulting test contains five sub-scales: (1) a statement of general father involvement; (2) a scale of child care responsibility (e.g., basic care); (3) responsibility for socialization; (4) overall influence in child rearing decisions; and (5) the father's availability to the child(ren). Taken together, the total scores for all the scales yield a statement of paternal participation in child care. The test is designed for both the father and the mother to complete and the composite scores represent paternal involvement. However, in their work Radin and Harold-Goldsmith (1985) found a 39 .77 correlation between the scores of both parents, indicating some general agreement on the perceived amount of paternal participation in child care between the parents. Some research on the psychometric properties of this test is examined below. In terms of concurrent validity, Radin and Harold-Goldsmith (1983) point to the high correlation coefficient (.76 in an earlier study) between parents in their initial sample of parents of pre-school children, and a .74 correlation between the same parents in the follow-up for the same study four years later. All correlations were significant to the .001 level. Radin (1985, paper) argues that it is unlikely that parents discussed their answers with each other because the PICCI was imbedded in a much larger test and it would not have been obvious how paternal involvement was computed from among all the questions. For construct validity, Radin (1985) points to works by Radin and Epstein (1975) and Radin and Glasser (1972) where a significant positive correlation was found between PICCI scores and the Cognitive Home Environment Scale which assesses the amount of cognitive stimulation in a home by the responding parent. It is evident that further research with other parental involvement measures would be useful in establishing the construct validity of the PICCI. Test-retest reliability had not been established when the test was administered the first time. Radin (1985) discussed the correlation between the two times involved in her research, which consisted of initial data collection and a 4 year follow-up. The four year correlation between time one and time two was .52 for 40 the entire scale, but when examining scores by the sex of the child, a correlation of .80 was obtained for female children, and for families with sons the correlation was not significant. However, it should be noted that there were theoretical reasons for this pertaining to the research question. Limited test-retest reliability was established in 1988 when only three of the PICCI scales were administered one week apart to 19 fathers of preschool children. The test-retest correlation of the Statement of Involvement scale was .78; the Decision Making scale .99; and the Availability scale .78. All scores were significant to the .001 level. These scores seem to indicate adequate reliability; however, further research on the whole scales with a larger sample would lend greater confidence to these findings and allow greater confidence in assuming adequate test-retest reliability. Internal consistency was measured using data from the above study and it was found that for fathers at time one there was a .67 Cronbach alpha value among all the scales on the test, while at time two the correlation was .68. The research above does not build an overwhelming case for the psychometric properties of the PICCI, however this test was developed relatively recently, as were each of the parenting tests cited above. None have rigorously researched psychometric properties, but the PICCI is one of the more promising measures available. The PICCI is specific in its measurement of "paternal" involvement, and is a short and simply completed paper and pencil test, but it does rely on memory and perception. However, because it is easy to complete and does not require the keeping of diaries 41 and elaborate record keeping, it should make it more palatable for subjects and perhaps increase compliance. State Versus Trait Research that examines change in personality characteristics assumes that they can change. The trait and state debate to explain differences in behavior and personality has existed since the early part of the nineteenth century (Krahe, 1992, Eysenck & Eysenck, 1985). Krahe (1992) regards this debate as two opposing views of behavior and personality. The state or situationalist position is that behavior is dependent on specific situations (Krahe, 1992), while trait theorists argue that traits are "important semi-permanent personality dispositions" (Eysenck & Eysenck, 1985, p.33), relatively stable across time and situation (Krahe, 1992). Considering empathy and self-esteem within this framework implies that at one extreme these characteristics would be relatively enduring and unchangeable, even across situation and time, while at the other extreme it is arguable that these characteristics are very situation specific. Applying either of these positions rigidly has implications for the present research. For example, a rigid trait stance means that any relationships between parenting and personality traits results from a certain level of a trait or combination of traits, predisposing one to a certain level of parenting. On the other hand a rigid situational stance implies that the specific situation --e.g., parenting-- determines the level and/or type of behavior. However, even staunch adherents of the trait position 42 claim that there has always been an assumed interaction between trait and situation (Eysenck and Eysenck, 1985). It is not the purpose of this research to enter into this debate, but to acknowledge that it exists and has some implications for the findings of this or any personality research. A view that seems reasonable to adopt for this research is that forwarded by Barbara Krahe (1992), who envisioned a continuous interaction between personal characteristics (traits) and situation as a paradigm for understanding individual behavior. This position then allows for a possible interaction between parenting situation and personality traits as applied in this research. Hypotheses This purpose of this research is to measure the relationships between variables. The working hypotheses can be stated as follows (the scales referred to will be fully discussed in the constructs section): Main Hypotheses Hypothesis One: Ho: Fathers categorized as having "high" involvement in parenting (PI) will not score significantly higher on perspective-taking (PT) than non-fathers. Hi: Fathers categorized as having "high" involvement in parenting will score significantly higher on perspective-taking than non-fathers. 43 Hypothesis Two: Ho: Fathers categorized as having "high" involvement in parenting will not score significantly higher on empathic concern (EC) than non-fathers. H2: Fathers categorized as having "high" involvement in parenting will score significantly higher on empathic concern than non-fathers. Hypothesis Three: Ho: Fathers categorized as having "high" involvement in parenting will not score significantly higher on self-esteem (SE) than non-fathers. H3: Fathers categorized as having "high" involvement in parenting will score significantly higher on self-esteem than non-fathers. Hypothesis Four: Ho: Fathers categorized as having "high" involvement in parenting will not score significantly higher on perspective-taking than fathers defined as having "low" involvement in child care. H4: Fathers categorized as having "high" involvement in parenting will score significantly higher on perspective-taking than fathers defined as having "low" involvement in child care . Hypothesis Five: Ho: Fathers categorized as having "high" involvement in parenting will not score significantly higher on empathic 44 concern than fathers defined as having "low" involvement in child care. H5: Fathers categorized as having "high" involvement in parenting will score significantly higher on empathic concern than fathers defined as having "low" involvement in child care. Hypothesis Six: Ho: Fathers categorized as having "high" involvement in parenting will not score significantly higher on self-esteem than fathers defined as having "low" involvement in child care. H6: Fathers categorized as having "high" involvement in parenting will score significantly higher on self-esteem than fathers defined as having "low" involvement in child care. Hypothesis Seven: Ho: There will not be a significant positive relationship between the level of paternal involvement and empathic concern. H7: There will be a significant positive relationship between fathers' level of paternal involvement in child care and empathic concern. Hypothesis Eight: Ho: There will not be a significant positive relationship between the level of paternal involvement and perspective-taking. 45 H8: There will be a significant positive relationship between fathers' level of paternal involvement in child care and perspective-taking. Hypothesis Nine: Ho: There will be no significant positive relationship between the level of paternal involvement and self-esteem . Hq: There will be a significant positive relationship between fathers' level of paternal involvement in child care and self-esteem. Sub Hypotheses Hypothesis Ten: Ho: Non-fathers will not show a significant positive relationship between their perspective-taking scores with their expressed level of desire to father children. Hio: Non-fathers will show a significant positive relationship between their perspective-taking scores with their expressed level of desire to father children. Hypothesis Eleven: Ho: Non-fathers will not show a significant positive relationship between their empathic concern scores with their expressed level of desire to father children. H n : Non-fathers will show a significant positive relationship between their empathic concern scores with their expressed level of desire to father children. 46 Hypothesis Twelve: Ho: Non-fathers will not show a significant positive relationship between their self-esteem scores with their expressed level of desire to father children. H12: Non-fathers will show a significant positive relationship between their self-esteem scores with their expressed level of desire to father children. Hypothesis Thirteen: Ho: There will not be a significant positive relationship between men who score higher on how involved their own fathers were in their care, and perspective-taking. H13: There will be a significant positive relationship between men who score higher on how involved their own fathers were in their own care and perspective-taking. Hypothesis Fourteen: Ho: There will not be a significant positive relationship between men who score higher on how involved their own fathers were in their care, and empathic concern . H14: There will be a significant positive relationship between men who score higher on how involved their own fathers were in their own care and empathic concern. Hypothesis Fifteen: Ho: There will not be a significant positive relationship between men who score higher on how involved their own fathers were in their care, and self-esteem. 47 H15: There will be a significant positive relationship between men who score higher on how involved their own fathers were in their own care and self-esteem. Hypothesis Sixteen Ho: There will not be a significant positive relationship between fathers who score higher on how involved their own fathers were in their care, and paternal involvement. Hi6: There will be a significant positive relationship between fathers who score higher on how involved their own fathers were in their own care and paternal involvement. 48 CHAPTER THREE Method Design This proposed study is a correlational field study. This section includes the details regarding how this research will be conducted, including the definition of variables (constructs), how they will be measured, the methodology to be followed, analysis of data, and limitations of this research. Sample Selection There are three residential housing units at UBC which are reserved for student couple and family housing. Within these complexes are approximately 53 0 residences. The goal was to access the male residents of the housing, whether father or non-father. The only restrictions were that they be in a common-law relationship with their partner with non-adult children living at home with them. A notice was sent to each household introducing the principal researchers, outlining the nature of the research, contact information, and requesting that both fathers and non-fathers participate (see Appendix A). To further entice participants, a draw for a prize was offered (equivalent to $75.00). On May 26/93, notices were delivered throughout the day to every household of student family housing. On the notice a cellular phone number was posted to be used to contact the researcher on that day only. Other numbers were listed for contact after May 26. Between 4:00 PM and 8:00 PM that day (also 49 posted on the notice) the researcher was in the residence community center providing those interested in participating with copies of the questionnaires. Interested participants could drop by and pick up a questionnaire or phone the cellular number and request one be delivered to their home. By the night of the delivery, 22 men or spouses of men responded to the notice and either picked up, or had a copy of the questionnaire delivered. Between this date and June 14 (the date of the prize draw), a further 15 men called the researcher to express their interest in participating. All received a copy of the questionnaire. Of this sample of 37, only 4 did not return their questionnaire, or informed the researcher that they would not complete the questionnaire. Of these, 2 were single (not in a common-law relationship) students visiting from the United States who happened to be in the Commons Block on the day the researcher was there. Another one indicated time constraints (his spouse picked up a questionnaire out of interest), and two never responded to inquiries into their progress. After June 14 further participants were solicited through academic and personal contacts. The requirements were the same, but this sample was less voluntary in that personal requests were made on the researcher's behalf to participate. This method was used between June 14 and early September 1993 and yielded a further 22 responses. The total sample gathered was 56, with only 4 being inappropriate because, while the respondent may have been a father, he was not in a common-law relationship and/or all his children were adults and had left home. The total usable sample 50 for research purposes was 52 respondents; 3 6 fathers and 16 non-fathers . Measures An attempt has been made in this research to develop a rationale for a relationship between the level of paternal involvement in the care of children, and the variables of empathy and self-esteem. For measurement purposes these variables need definition and then operationalization. Following is a description of the instrument to measure the dependent and independent variables. Paternal Involvement in Child Care The Paternal Involvement in Child Care Index (PICCI) is a paper and pencil questionnaire consisting of five sections with a total of 23 questions (Radin & Harold-Goldsmith, 1985). Each section represents one of the five sub-scales for the PICCI and asks the respondent to check an appropriate answer or to estimate a proportion of time that various tasks are done with the child. Each section has instructions for the participant on how to complete it. Completion time for the questionnaire is approximately five minutes. Empathy The measure of empathy chosen for this study is Davis' (1980) Interpersonal Reactivity Index(IRI) consisting of four separate measures that tapped into four distinctive yet related constructs of responsiveness to others. Of interest to the present research are two of the scales; Perspective-Taking (PT) and Empathic Concern(EC). These two scales are related by definition and 51 statistically and permit to measurement of two aspects of empathy; an emotional contagion aspect (by EC), and a cognitive understanding of another's situation (by PT). The entire scale consists of 28 questions with seven questions devoted to each scale. At the end of each question on the questionnaire, is a blank space where the respondent writes a letter ranging from "A" to "E", corresponding to a range of "Does not describe me well" to "Describes me very well". These letters correspond to a five point numerical likert-type scale ranging from one through five. This scale takes approximately five minutes to complete. Mark Davis gave this researcher permission to use the Interpersonal Reactivity Index. Two scales that were scored for this research were the aforementioned EC and PT scales. Davis (1980) reported internal consistency alpha coefficients of .75 for males on the PT scale and .72 on the EC scale. Test-retest reliability coefficients ranged from .61 to .79 over a 60 to 75 day period. Self-Esteem The Rosenberg Self-Esteem Scale (SES) is the instrument selected to measure self-esteem for this study (Rosenberg, 1965) . The SES is a short test (10 items), scored on a 4-point Guttman-type scale. Other authors have adapted this test using a 5 or 7 point Likert-type scale resulting in a broader range of scores (Blascovich & Tomaka, 1991). The present research will use a 5-point scale to measure the degree to which each statement applies to the test taker. Each of the ten questions consists of a self-52 statement where the subject must circle a response ranging from "Strongly Agree" to "Strongly Disagree". These responses are converted into corresponding numerical scores ranging from one to five and then added together for total Self-Esteem score. Permission to use this scale was granted by Princeton University Press. The scale takes the respondent approximately five minutes to complete. The SES demonstrates adequate internal consistency with Dobson et al. obtained a Cronbach alpha of .77, while Flemming and Courtney (1984) obtained a Cronbach alpha of .88. In terms of test-retest reliability, Flemming and Courtney obtained a correlation of .82 after one week, while Silber and Tippet (1965) obtained a correlation of .85 after 2 weeks. Demographic Information Each questionnaire also contains an information section to obtain pertinent demographic information that is intended to yield scores allowing them to be statistically analyzed with the research variables. This section was developed by the researcher based on Canadian census questions (Statistics Canada,). The questionnaire consists of 3 6 question; 19 pertaining to the subject, 11 that pertain to the spouse (answered by the subject), 6 more that are answered only by fathers, and two that are answered by non-fathers. Most of the questions are answered by the subject checking the category that applies to himself, and then for analysis were converted into numerical data yielding interval scores. The two questions pertaining to age and length of relationship status required the participants to indicate numerical data. 53 Finally, there were a short series of questions that yielded data comprising two research variables for the sub-hypotheses. One question for all subjects asked them to indicate on a five point Likert-type scale their recollection of the degree to which their father was involved in parenting them as a child. The score ranged from one ("not at all") to five ("very much"). Second, there were two questions that asked non-fathers only whether they desired to become a father in the future, and to estimate the probability that this will happen. The first question was checked by the respondent as "yes" or "no" and then scored as "1" or "2" by the researcher. The second question asked the respondent to estimate the probability based on 20% intervals. These responses were scored ranging from "1" to "5" from lowest to highest. Neither of these questions were subjected to psychometric testing therefor reliability and validity cannot be established. These variables are included for descriptive purposes only. Procedures Those agreeing to participate received a package containing a letter of introduction, including a brief description of the study, contact information, researcher information, the questionnaire package, and an "end page" thanking the respondent and offering a copy of the results of the study upon completion. There was also a "face page" with completion instructions and a removable "ticket" for the prize draw. The tickets which requested name and phone number were removed immediately and deposited into a draw box. Once this was completed only a code 54 number remained on the questionnaire. The initial sample of UBC residents also had their name and phone number recorded into a separate log book so that they could be contacted if information was left unanswered. When this process was completed the pages of the book containing personal information was shredded. Once sampling was completed, the PICCI, IRI, and SES, sections were scored. This data was then transcribed into an statistical system file for the SPSSX (version 4.0) statistical package. There were two recurring problems on the questionnaires that needed to be reconciled in the recording of data. For some of the demographic questions, participants indicated "na" or "not applicable" although these were not offered choices. These responses were ignored and no value was indicated in the data file. Second, the demographic question that requested information regarding the participant's and spouse's main activity for the previous year often had two categories checked. Generally, these were the categories "taking care of the children" and "keeping house". When this occurred a decision was made to include the first response only. This was an arbitrary decision made to make statistical analysis more meaningful in light of the relatively small sample size. Of the father sub-sample one respondent did not provide information about his children, (e.g. sex, age, and number of children), however his scores on the PT, SC, PI, and SE were included in the analysis. 55 Statistical Analysis As this research largely examines relationships between variables, statistical analyses that measured strength of relationships, direction of relationships, and significance of relationships, were employed. Below are a discussion of statistical analyses employed both for answering the hypotheses and for descriptive purposes. The discussion is partitioned by hypotheses and theme. Hypotheses One Through Six: Comparing Father's Level of Paternal Involvement Between Fathers and Against Non-Fathers Level One Analysis In order to statistically compare level of parental involvement and it's relationship to the personality variables, among both fathers and with non-fathers, it was decided to divide the fathers into groups based on their level of involvement. The father sample size allowed for 3 equal groups of 12 based on their score ranking. These were called the "high," "medium," and "low" paternal involvement groups. These means were analyzed through one-way analysis of variance and then a least-significant difference (LSD) multiple comparison procedure was performed to determine any significant differences between paternal involvement and the father and non-father groups. Individual t-tests were also performed to determine the specific hypotheses probability statistic. Level Two Analysis In addition, an examination of the frequency distribution of self-esteem by paternal involvement scores indicated a narrow 56 range of paternal involvement scores for the medium involvement group. Based on this, a second stage of analysis was conducted where the sample of fathers was also divided into two groups of "high" and "low" parental involvement categories (18 in each category) based on their score ranks. To test for potential differences in PT, EC, and SE means between "high" PI and "low" PI, and between "high" PI and non-fathers a one-way analysis of variance with a follow-up LSD multiple comparison procedure to check for significant differences. T-tests were also performed on each of the pairs of father groups. Hypotheses Seven Through Nine: Analyzing the Relationship Between Paternal Involvement and the Dependent Variables for Fathers Only The measurement of the relationship of PT, EC, and SE with Paternal Involvement allowed for a simple yet elegant analysis between variables that yielded continuous data. A Pearson product-moment correlation was performed to measure the direction and strength of relationship between Paternal Involvement and each of PT, EC, and SE. Because the hypotheses specified a direction of relationship, a one-tailed test of signifcance was used to determine the level of significance of the relationship between each pair of variables. Hypotheses Ten Through Twelve: The Relationship Between The Desire to Parent and the Dependent Variables To test the hypotheses relating to desire to parent (dichotomous data), and the probability that one estimates being a father in the future (interval data), Kendall's tau correlation statistic for ranked data was used to measure it's relationship to 57 PT, EC, and SE. The demographic questionnaire included two questions relating to this topic. The first question asked whether the non-father wanted to be a father, yielding a dichotomous "yes" or "no" answer. This variable was not used in statistical analyses. The second question pertained to the probability that one would become a father in 20% intervals. This variable was considered the probability that one desired to become a father in the future (PDF) and was analyzed for relationships with PT, EC, and SE, using Kendall's tau. Hypotheses Thirteen Through Sixteen: The Relationship Between Recollection of Father Involvement and the Dependent Variables Recollection of one's own father being involved in parenting (RINV) yielded continuous data and the Pearson product-moment correlation statistic was used to measure its relationship with the main variables. The analysis took place on several different levels. To answer hypotheses thirteen through sixteen, all men were grouped together to determine the relationship between RINV and each of PT, EC, and SE. A one-tailed analysis was used to determine level of significance of relationship because of the hypothesized relationship between the variables being specified. For the Relationship between RINV and PI, only the father group was compared also using the Pearson statistic and a one-tailed level of significance. The prior division of fathers into three and two groups based on level of paternal involvement also allowed for level one and level two analyses to be performed as described for hypotheses one through six. This was done primarily for descriptive purposes and 58 further utilizing the data available to that end. The level of recollection of father involvement could also be compared between the father and non-father groups for both the "three" father groups (level one analysis), and the "two" father groups (level two analysis). To accomplish this a one-way analysis of variance was performed also utilizing a follow-up LSD procedure and t-tests where the F ratio demonstrated a level of significance of p. < .05. Analysis of Relationships of Demographic Variables to the Research Variables Relationships between demographic information and research variables were analyzed both within the father group, and between fathers and non-fathers. The purpose is to identify potential relationships between demographic and research variables that confounds any differences between the father groups on the dependent research variables (PT, EC, SE, and RINV) as well as paternal involvement. Some of the data yielded in the questionnaire was categorical and the responses were so varied that it was difficult to generate meaningful statistical results. These variables included the religion and ethnicity of both the subject and partner. Also, questions pertaining to whether participants received child care on a regular basis and if so for what activity, did not yield sufficient data to employ statistical analyses (the majority responded "not applicable" or "na" to these questions). The question asking whether the partner was born in North America or not, A chi square statistical analysis was used to determine the relationship between paternal involvement and the 59 partner's birthplace. The categorization of paternal involvement into three levels made this analysis possible. The variables age and relationship length both yield continuous data, as do the dependent and independent variables. This indicated the use of the Pearson product-moment correlational analysis to examine the relationship between these two demographic variables and the research variables. This is particularly salient to hypotheses seven through nine which also use this statistic to test the hypotheses. For fathers only, Spearman's rho was used to analyze the relationship of interval data to the dependent variables and to paternal involvement. While correlational statistics are appropriate for continuous or interval data and measuring the relationship of variables, very few of the demographic variables yielded continuous data. Also, the fathers were divided into non-fathers and paternal involvement categories, necessitating the use of other statistics. When analyzing fathers versus non-fathers, the same groups (i.e. high involvement, medium involvement, and low involvement) that were used for hypothesis testing were used for level one and two analyses. The various types of data collected required that appropriate statistics were used to analyze the relationships between the subject groups and the demographic data. Continuous data were analyzed for the different father groups using a one-way analysis of variance, and then followed up with t-tests if indicated by a significant F ratio. These variables are age, and relationship length. Interval data were analyzed with the father groups using the Kruskal Wallis one-way analysis of variance. 60 This analysis allows for the measurement of interval data and does not require a normal distribution of data. These variables are subject and partners' years of schooling, study time per week, percentage of study time performed at home, work time per week, and percentage of work performed at home. Also included were the subject's personal income and household income. 61 CHAPTER FOUR Results Sample Characteristics This section will present total sample and sub-sample demographic information. Sub-samples include fathers versus non-fathers, and characteristics for 3 categories of fathers according to level of paternal involvement in child care; high involvement, medium involvement, and low involvement groups. Following will be a description of selected sample characteristics. A full tabulation of demographic characteristics can be found in Appendix C. The men ranged in age from 23 to 51 with the average age being almost 3 5 years old. 45 were married, 6 reported living common-law, while one reported being divorced. The years they reported having this status ranged from 1 to 21 years with the average being approximately 7.6 years (see Tables 1 and 2). Most of the men were born in North America (37), although 15 said they were born outside North America. The group of non North Americans largely came out of the father sample, from the student resident housing. The ethnic origin most often reported was of British descent, while 5 reported being of Chinese origin, 4 Eastern European, and 3 each of Francophone, and German. 62 Table 1 Mean Acre for all Subject Group Whole group Non-fathers Fathers 3, Non-fa n 52 16 36 ther, and m 34.79 32.13 35.97 Father sd 5.56 3.65 5.89 Groups minimum 23 23 24 maximum 51 36 51 Missing cases = 0 The most often reported religious background was "no religion" (22), while 6 reported being Roman catholic, and 5 Anglican. Other lesser reported religions were, United Church (1), Baptist (2), and Jewish (2). It is worthy noting that 32 of the 52 subjects reported that they did not feel part of any religion. Twenty five subjects reported that in the past year apart from weddings, funerals, or baptisms they did not attend any religious service, while 11 reported doing so at least once per week. The others did so somewhere between these two extremes. Table 2 Means for Length of Relationship for all Subiects, Non-father, and Father Groups Group Whole group Non-fathers Fathers n 52 16 36 ni 7.59 3.4 9.33 sd 5.88 2.77 5.97 minimum 1.0 1.0 1.0 maximum 21 11 21 Note. Missing cases = 0 Half of the sample reported that being a student was their main activity in the previous 12 months, while 23 reported that working at a job was their main activity. Only 2 reported that taking care of their children was their main activity, while one was looking for work. 63 The rest of the demographic data was recorded in intervals and so lends itself to less precise description. For example, personal income was reported in $10,000 increments ranging from less than $10,000 to more than $40,000. The largest number (13) reported earning over $40,000 in the last year, although the numbers were fairly evenly distributed over the intervals. Household income was also reported in $10,000 increments with a category added for those earning over $50,000. This category had the most responses (17), with the $40,000-$50,000 and $20,000-$30,000 categories being next most reported with 9 respondents each. Other interval data described the number of hours one studied per week in the past year, and of that, what percentage was done at home. The most reported category was more than 40 hours per week (12), although 17 responded that this was not applicable to them. Of those in study, 16 reported that 41-60% of their study hours were done at home, 9 reported 0-20%, and 5 each in the 0-20% and 81-100% categories. In terms of work time, 17 reported that this category did not apply to them, while 9 reported working only 0-5 hours per week in the past year. Of those working, 42 reported that 0-20% of their work was done at home, while 5 did 41-60% of their reported work hours at home. With respect to the children, of the 3 6 first or only children the average age was 5.9, with 19 being females. Only one of these children were not living with the family. 16 of the fathers had a second child with the average age being 6 years of age, 9 being male. All were living with the father. There were 64 seven respondents who had a third child with the mean age being 4.4 years and 5 being female. None had four children or more. Two questions concerning receiving child care were included, but only 19 reported that they received any child care at all, and of those 8 indicated receiving less that 15 hours of day care per week, while 7 received more than 35 hours per week. Participants were also asked for demographic information of their spouses. With respect to religion, the largest response was that the men did not know what their spouse's religion was (16). Of those indicating their partners religion, Roman Catholic and Anglican were indicated by 8 respondents. Of those who considered their spouse religious, 10 reported that their partners attended religious services at least once per week in the last year, although 23 responded that their partners did not attend a service in the past year. In terms of ethnicity, the largest number of responses indicated that their partners were of British ancestry, while the others were spread amongst 18 other categories, some researcher created, others indicated by the respondents. Thirty eight of the spouses were born in North America, 13 outside of North America. The largest group of spouses (22) had completed 15-16 years of schooling, while a further 10 had completed 19-2 0 years of schooling. The largest number of men (23) indicated that their spouse's main activity in the past year was working at a job, 16 were students, while 10 were taking care of their children. Of those who were in studies, 20 indicated that they studied only between 65 0-5 hours per week, while 10 studied 6-10 hours per week. Of study time, 22 (42%) did between 21-40% of their studying at home. Of those who worked in the past year, 21 worked between 0-15 hours per week, while 19 worked 36-40 hours per week. Of that work time, 38 (73%) indicated that only 0-20% percent of that work was done at home. Relationships of Demographic Data to Research Variables The data for all research variables as well as age and marriage length yielded continuous data. This allowed for a Pearson product-moment correlation to be performed on the research variables with age and length of relationship. These analyses are of particular importance to hypotheses seven through nine where correlational statistics are used to determine the relationships between PI and each of PT, EC, SE, and RINV. Analyses were done for the entire sample correlating age and relationship length with PT, EC, SE, and RINV. A separate analysis was also done for fathers only, including the variable PI. There was a negative significant relationship between recollection of paternal involvement (RINV) and age for the whole group. The negative correlation between relationship length and RINV approached significance. This result seems logical in that it might be guessed that those who are involved in longer relationships may also be older. There were no significant relationships between the variables age and length of relationship and any of PI, PT, EC, or SE for the fathers only group. The results are displayed in Table 3. 66 Table 3 Correlations of Age and Relationship Length and Research Variables for Whole Sample Variable PT EC SE RINV Age .0802 -.0667 .1107 -.2793* Relationship .1560 -.0601 .1169 -.2539** Length n = 52 **p_ < .05 * p. < .10 Correlations of Age and Relationship Length and Research Variables for Fathers Only Variable PI PT EC SE RINV Age -.0635 .1130 .1133 .0873 -.1668 Relationship -.0387 .1026 .0335 .0610 -.1104 Length n = 36 Level of significance based on two-tailed analyses A Spearman correlation statistic was used to determine whether demographic variables yielding interval data demonstrated significant relationships to the research variables. This analysis was performed for the whole sample for PT, EC, SE, and RINV with 12 demographic variables. Out of the 48 pairs of variables only the relationship between RINV and percentage of work done at home by the subjects resulted in a significant relationship at p_ < .05 (correlation coefficient = .31, p = .039) . This number of significant relationships would be expected by chance at the set level of significance. This indicates that 67 largely none of these demographic variables have significant relationships that may have some explanatory power in the relationships between the dependent and independent variables. The full results are displayed in appendix D. The relationships between demographic information and research variables of fathers versus non-fathers (largely pertaining to hypotheses one through six) was accomplished using the categories developed for hypothesis testing. The salient groups are: non-fathers, low (Low PI), medium (Medium PI), and high (High PI) paternal involvement fathers based on the rank scores from the PICCI. This creates the groups for level one and level two analyses. The Kruskal-Wallis one-way analysis of variance revealed no significant (p < .05) differences among the demographic data for the subject groups themselves or as reported for their partners. One result did approach significance (p « .08), where the spouses in the non-father group appeared to work more than the fathers in the low parental involvement category. The demographic variables in this analysis are those yielding continuous and interval data including partner and subject's years of schooling, study time per week, percentage of study time performed at home, work time per week, and percentage of work performed at home. Also included were the subject's personal income and household income. The lack of relationship between the dependent variables and the demographic data cited above, supports a contention that the demographic variables do not influence the relationships between 68 the dependent and independent variables. The full results are displayed in appendix D. There were some differences among the father groups with respect to marital status and whether the spouse was born in North America or not. A chi square cross tabulation analysis for the collapsed ethnicity variable revealed that there was a significant difference between the group of non-fathers and the father groups (see Table 4). Out of 16 non-fathers, 15 of the partners were born in North America while for the low paternal involvement group, had 7 out of 12 partners who were born outside of North America (p < .05). For the medium and low parental involvement groups each had 9 partners born in North America, while the medium involvement group had 2 born outside North America, and the low involvement group had three. Table 4 Crosstabulation of Parental Involvement Class by Partner's Birthplace Born in NA? | Row | Yes| No| Total Parental + + + Involvement Non- | 15| 1| 16 Father | j j 31.4% + + + High PI | 5| 7| 12 I | | 23.5% Med PI | 9| 2| 11 I | | 21.6% Low PI | 9| 3| 12 | j j 23.5% Column 38 13 51 Total 74.5% 25.5% 100.0% Pearson Chi-Square =10.2447 df = 3 69 Significance = .01660 All but one respondent replied as either being married or common-law. Regarding marital status among the four groups, the chi square analysis approached significance (p ~ .08) with the non-father group having 4 common-law relationships among its members while the high paternal involvement group had 2 common-law relationships out of 12, and the other two groups were all married only. Finally, a one-way analysis of variance was performed for each of the father involvement categories with the number of children in each category. The means for each of the three groups from low to high involvement were 1.5, 1.67, and 1.58 respectively. The one-way analysis of variance produced an F ratio of .1218 which was not significant. Therefore it appears that for these fathers the number of children in the family was not related to paternal involvement. The full data is presented in Appendix D. The above descriptive data has given a profile of the subjects who participated in the study, and of some of the differences among them with respect to the different groups of analysis. The implications of these differences will be discussed in the concluding section. Comparative Samples Because of the largely voluntary nature of the subjects participating in this study, it is difficult to know how much they may differ from the entire population of men in relationships, both father and non-father. In fact this sample may differ from 70 the population from which it was largely selected (UBC family housing). The differences that exist between volunteer samples and the general population have been well documented (Rosenthal & Rosnow, 1975). Although it would be very difficult to ascertain how this sample is distinguished from the population attempting to be accessed, the literature from which the instruments were developed provide some other sample characteristics to which the subjects for this research may be compared. Before presenting and comparing sample means and norms for the variables it is important to note that it is difficult to make such comparisons because of different target populations sampled for the various studies. Not all sample characteristics are provided in the various studies and it is generally impossible to determine whether differences in observed norms for example, are significant or not. Below is a presentation table of studies and the means found using the same instruments as the present research for the research variables. One must be cautious when interpreting any differences between score means or alpha coefficients, but it appears that in general, scores on PT for the present study are somewhat higher than other research findings. Similarly the EC scores for the present study appear to be lower than other research. It is not known whether these differences are significant, but it should be noted that one of the studies used an adolescent population (Davis & Franzoi, 1991), while Davis' (1980) normative sample utilized students from undergraduate courses. It is important to remember that not all of the subjects in the present research (including 71 those who lived in student housing) are students. Davis and Franzoi (1991) found that the increase over time for perspective-taking was significant over the three years the test was administered. It is not known, but possible that this effect continues with age. 72 Table 5 Summary Table for Other Research Using Variables PI, PT, EC, and SE Study PT EC SE PI Comments PRESENT STUDY 2 0 . 2 1 1 7 . 0 4 4 0 . 6 5 3 5 . 9 7 average age of whole sample = 34.78 Davis' (1980) normative scores Coman et al.(1988) Coman et al. (1988) Davis & Franzoi(1991) Davis (1980] Davis (1980) Demo et al. (1980] Radin (1981) Radin (1981) Radin and Harold-Goldsmith (1983) 16 .78 19.04 17 .48 19.46 17 .05 18.05 15 .05 18.43 15.96 18.55 15.96 18 .55 35 .91 4 0 . 5 8 3 8 . 2 6 3 4 . 3 3 scores for males in an undergraduate psychology course 4th year male medical students 6th year male medical students adolescent male scores for year one and three of a longitudinal study Correlation between PT and EC for normative male sample =.37. For the present study alpha =.33 longitudinal study scores for parents, father mean age = 43.2 data gathered for time one in 1977 primarily middle class data gathered for time two in 1981 men in working class families 73 The intercorrelation alpha between PT and EC for this study seems similar to the findings for the normative sample. The intercorrelation alpha for the Davis (1980) study was .37, while for the present research alpha = .33. Again these seem fairly similar although whether the similarity or difference is significant is unknown. In terms of self-esteem, Demo and associates (1987) used a four point scale for each of the 10 question on the Rosenberg Self-esteem Scale (Rosenberg, 1979). The present research used a five point likert-type scale allowing for higher raw scores. This may explain the higher raw mean scores of the present research over the Demo and associates (1987) scores. If the same point scales were used for both studies the mean scores may be more similar. Finally, for parental involvement the scores for the present research appear to be lower than those for three of the four quoted studies in the table. The only study with a somewhat lower overall score sampled working class fathers (Radin & Harold-Goldsmith, 1883). It could be argued that the large number of males sampled from student housing (especially the fathers) could be classed as lower socio-economic status based on their income and need for student housing. However it is difficult to do a true comparison between the two samples without more information form the Radin and Harold-Goldsmith (1983) study. It is difficult to draw any conclusions from comparing the means of different studies. While some data seems to be similar to the means of the present study (Radin & Harold-Goldsmith, 1983; 74 Coman et al., 1988), many are arguably different. Whether these differences are due to the previously discussed different data collection and/or sample demographics, may or may not explain these differences However, the generally lower PI scores for the present research could detract from any argument that the present research sample consists of fathers who chose to be involved in this study because they are inherently more involved in parenting than those who chose not to (or at least as compared to the other research samples). A key level of analysis was to determine the extent of the differences among the two main subject samples that were collected for this research. There was the sample collected from UBC family residence (n = 28), and those collected through personal contacts (n = 24) . A major threat to the findings of this research is the possibility that the two samples differ significantly from each other, particularly concerning the research variables. To accomplish a measurement of this possibility the UBC and non-UBC groups were compared with each research variable (PI, PT, EC, SE, and RINV) using £-tests. The findings indicated that the sub-samples did not significantly differ on the independent variable (PI) or dependent variables (PT, EC, SE, and RINV). One caveat to this assertion is that there appeared to be a near significant relationship between sub-sample groups and RINV pointing to a lower group mean for the non-UBC group. The results are displayed in the Table 6 below. 75 Table 6 Results of t-tests Between Sub-samples and variables, PI, PT, EC, SE, and RINV UBC Variable n Mean sd 24 38 8.8 Parental Involvement 28 20.2 4.7 Perspective-taking 28 17.1 3.6 Empathic Concern 28 40.2 6.5 Self-esteem 28 2.8 1.0 Recollection of Involvement Non-UBC n Mean sd 12 40 5.0 24 20.3 4.7 24 16.9 3.8 24 41.2 6.2 24 3.3 1.0 df t p. value 33.4 -.83 .41 48.9 -.05 .96 48 .22 .83 49.4 -.54 .59 48.2 -1.9 .06 Note, t values, and significant probabilities are separate variance estimates These findings indicate very little significant difference between the sub-samples which supports their use together as a whole sample. This also further contributes to the validity of the findings. Research Hypotheses Results The analyses performed on the data for the purpose of hypotheses testing will be presented. There were several different levels of analysis of the data. These will be made explicit below. The first level of analysis involved comparing non-father scores on PT EC and SE with fathers by the three different levels of paternal involvement (high, medium, and low). The sample size of non-fathers was 16. From the group of 36 fathers, three father groups were created based on their level of paternal involvement 76 (PI scores). For this they were simply ranked by PICCI score, divided into the lower, middle, and top scores and assigned the categories of low (low PI), medium (medium PI), and high (high PI) paternal involvement (12 in each group). This level of grouping is referred throughout as the "three father group" and is considered the "level one" analysis. The "level two" analysis involves the creation of two categories of father involvement (referred to as "two father group"). In this level of analysis the fathers were divided into two equal groups of 18 subjects based on their PICCI scores. Hypotheses One Through Six Level one analysis. The first three hypotheses deal with comparing the scores of non-fathers and the three father groups on the variables of PT, EC and SE. Hypotheses four through six compare the PT, EC and SE scores of the three father groups amongst themselves. Because measuring these relationships uses the same statistical analyses, they will be dealt with as a group. Based on the literature review findings, each of the six hypotheses assert that high involvement fathers will exhibit significantly higher PT, EC, and SE scores than non-fathers and those categorized as low involved fathers. These specific assumptions indicate the use of t-tests to analyze the mean differences of each pair of groups. The means for each subject group by PT, EC, and SE are displayed in Table 7. 77 Table 7 Means and Standard Deviations of PT, EC, and SE for the Three Father Groups FATHER CLASS Non-fathers Low PI fathers Medium PI fathers High PI fathers Mean 19.88 18.75 22.67 19.67 PT SD 4.67 4.23 5.14 4.19 Mean 18.44 16.17 17.50 15.58 EC SD 3.63 4.28 3.68 2.58 Mean 40.31 38.50 40.42 43.50 SE SD 5.33 7.49 6.82 5.42 Non-father n = 16 Father groups n =12 A one-way analysis of variance resulted in non-significant F ratios among fathers and non-fathers on all dependent variables. The results of a multiple range test (LSD procedure) on the variable EC however, indicated a significant difference between non-fathers and high PI fathers. The results are displayed in Table 8. 78 Table 8 Results of t-tests Between Non-fathers and High Paternal Involvement Fathers Variable Perspective-taking Empathic Concern Self-Esteem Non-n 16 16 16 -Fathers Mean 19.9 18.4 40.3 n 12 12 12 Hicrh PI Fathers Mean 18.8 16.2 38.5 t value 0.12 2.43 -1.55 P. .90 .022 .13 df 25.1 25.9 23.6 F ratio for PT = 1.63, £ = .194 F ratio for EC = 1.76, p_ = .162 F ratio for SE = 1.33, p_ = .277 Note: results are for separate variance estimates As is clear from Table 7 above, highly involved fathers do not have higher overall means on PT, EC, or SE. In fact, it appears that non-fathers have higher empathic concern than fathers highly involved in parenting. This result is supported by the results of analysis on hypothesis ten. The next set of analyses involved a series of jt-tests within the father groups (hypotheses four through six). This analysis examines the relationships of PT, EC, and SE between high paternal involvement fathers (High PI), and low involvement fathers (Low PI). The results are displayed in Table 9. 79 Table 9 Results of t-test Analyses With Low PI and High PI Fathers for PT, EC, and SE Low PI Fathers High PI Fathers Variable Perspective-taking Empathic Concern Self-Esteem n 12 12 12 Mean 18.75 16.17 38.50 n 12 12 12 Mean 19.67 15.58 43.50 t value -.53 .40 -1.87 E .60 .69 .076 df 22. 18. 20. F ratio for PT = 2.42, E = -10 F ratio for EC = 0.90, p. = .41 F ratio for SE = 1.73, E = .19 Note: Separate variance estimate, two-tailed probability. As is evident from the table, the high PI fathers do not have significantly greater mean scores on the personality variables than low PI fathers. It is worth noting that on the variable self-esteem, the mean differences (the high PI fathers have a higher mean) approached significance (p < .10). For descriptive purposes all possible relationships were explored through one-way analysis of variance, followed with t.-tests comparing mean differences on pairs of variables between the high PI, and both the low PI and medium PI father groups. The results are displayed in Tables 10 and 11. As is evident from these tables, there are no significant differences in the means of any of the groups. However, the 80 differences in means on the variable PT between the low PI and medium PI fathers approach significance (p = .055). The medium paternally involved fathers appear to score higher on PT than low paternally involved fathers. Table 10 Medium and High Parental Involvement Means t-tests for Perspective-taking, Empathic Concern, and Self-esteem Medium PI High PI Fathers Fathers Variable Perspective-taking Empathic Concern Self-Esteem n 12 12 12 Mean 22.67 17.50 40.42 n 12 12 12 Mean 19.67 15.58 43.50 t value 1.57 1.48 -1.23 E .13 .16 .23 dj 21. 19. 20. Note: Separate variance estimate, two-tailed probability Table 11 Medium and Low Parental Involvement Means t-tests on Perspective-taking, Empathic Concern, and Self-esteem Low PI Medium PI Fathers Fathers Variable Perspective-taking Empathic Concern Self-esteem n 12 12 12 Mean 18.75 16.17 38.50 n 12 12 12 Mean 22.67 17.50 40.42 t value -2.03 -.82 -.66 E .055 .42 .52 df 21. 21. 21. Note: Separate variance estimate, two-tailed probability 81 Level two analysis. This analysis involves comparing non-fathers with fathers for the high and low paternal involvement division. The predicted relationship was that the high involvement group would demonstrate higher overall means for PT, EC, and SE. To accomplish this analysis t-tests were performed. Also, a one-way analysis of variance was conducted for each dependent variable (PT, EC, and SE) and the father groups to examine all possible relationships for descriptive purposes. The results are displayed below in Table 12 and indicate no significant differences between each of the variables and the father groups. The probability level for the self-esteem F ratio approached significance, and a follow-up multiple comparison test (LSD) indicated that there was a significant difference between low and high paternally involved fathers. Table 12 F-ratios for the Dependent Variables and the "Two" Father Groups Variable F-ratio Probability Perspective-Taking Empathic Concern Self-esteem n=18 for each father group. 0 . 4 4 1 .98 2 . 4 5 0 . 6 4 0 . 1 5 0 . 0 9 6 82 The results of individual t-tests are displayed in Tables 13 and 14. Table 13 Results of t-test Analyses for Non-fathers and Low PI Fathers for the Two Father Groups Variable Perspective-taking Empathic Concern Self-esteem Non-n 16 16 16 -fathers Mean 19.86 18.44 40.31 n 18 18 18 Low PI fathers Mean 21.06 16.83 38.56 t value 0.79 1.19 -0.70 E .486 .241 .436 df 32. 32. 30. Note. Separate variance estimate, two-tailed significance The results of the comparison of non-fathers and low PI fathers for the two father groups, indicate there are no significant mean differences on the dependent variable scores. As the results of the next series of t-tests demonstrate (shown in Table 14), there is a significant difference in the mean scores of EC, where non-fathers demonstrate a significantly higher EC score than the high PI fathers ( p < .05). This suggests that in the present sample, non-fathers attribute to themselves a higher level of empathic concern than those fathers highly involved in the care of their children. 83 Table 14 Results of t-test Analyses for High PI and Non-fathers, for the Two Father Groups Variable Perspective-taking Empathic Concern Self-esteem Non-n 16 16 16 -fathers Mean 19.86 18.44 40.31 n 18 18 18 Hiah PI fathers Mean 19.67 16.00 43.06 t value 0.13 2.15 -1.53 E .89 .04 .14 df 30. 28. 31. Note. Separate variance estimate, two-tailed significance Finally in a comparison of high PI and low PI fathers of the two father groups there is a significant difference between the level of self-esteem between the two groups ( p < .05). The high PI fathers demonstrate a higher mean score on self-esteem than the non-fathers. The results are displayed in Table 15 below. 84 Table 15 Results of t-tests on Low PI and High PI Fathers for the Two Father Group Comparison Variable Perspective-taking Empathic Concern Self-esteem n 18 18 18 Low PI Fathers Mean 21.06 16.83 38.56 n 18 18 18 Hicrh PI Fathers Mean 19.67 16.00 43.06 t value .88 .69 -2.09 E .39 .49 .045 df 33. 30. 29. Note. Separate variance estimate, two-tailed significance The results of the preceding t-tests for "three" father (level one) and "two" father (level two) comparisons suggest some interesting results. Generally, where there are significant mean differences, the scores for the empathy scales of the IRI are opposite of the expected direction (i.e. the low father involvement and non-father groups tend to have higher scores than the high involvement groups). The exception to this is the medium PI group having a higher mean PT score than the low involvement group. For the variable self-esteem, the results supported the expected position, where for the two father comparison the high PI group had a significantly higher mean score than the low PI group (p < .05). This did not hold up for the three father comparison, although the difference approached significance (p = < .10). The implications of these findings will be discussed in the concluding section. 85 Hypotheses Seven, Eight, and Nine The next set of data analyses tested hypotheses seven through nine regarding the assumptions about the relationships between fathers paternal involvement (PI), and the personality variables Perspective-taking (PT), Empathic Concern (EC), and Self-esteem. The hypotheses predicted that as the level of PI increased so would the scores for PT, EC, and SE. The statistic chosen to test these hypotheses was the Pearson product-moment correlation for pairs of variables. In this analysis paternal involvement (PI) was correlated with Perspective-taking (PT), Empathic Concern (EC), and Self-esteem (SE). The results shown in Table 16 were not significant, indicating that there is no direct linear relationship between the level of paternal involvement and the dependent variables. However, the level of significance for both EC and SE approach significance (p_ < .10). The result for empathic concern suggests a possible inverse relationship between it and paternal involvement. The approaching significant result for self-esteem suggests a positive linear relationship between self-esteem and paternal involvement. Table 16 Correlation Coeff Variable Parental Involvement ici ents for PT .0649 p - .71 PI wi th PT, EC and EC -.2265 p = .092 SE SE .2564 p = .066 n=3 6 Note. Results reflect one-tailed significance Although the results were not significant, the PICCI has five sub-scales affording greater discrimination of analysis with 86 respect to type of paternal involvement in child care. There are five sub-scales on the PICCI; an overall statement of involvement (SI), overall child care responsibility (CR), responsibility for socialization (RS), influence in child rearing decisions (ID), and availability to the children (AV). A Pearson product-moment correlation coefficient was used to analyze these relationships. Table 17 Correlations for Sub-scales Variables Statement of involvement Child care responsibility Responsibility for socialization Influence in child rearing decisions Availability to the children P P P P P of PICCI PT .21 = .21 .17 = .34 .09 = .62 -.15 = .37 -.04 = .80 and PT, EC EC -.07 p = -.67 -.15 p = .37 -.33** p = .049 -.28* p = .09 -.24 p = .15 and SE SE .26 p =.12 .37** p = .026 .10 p = . 55 .08 p = .64 .22 p =.19 n = 36 **p_ < .05 *p_ < .10 Note. Results reflect two-tailed significance There were some significant (p < .05) results in this analysis which are displayed in Table 17. There was a positive significant correlation between the level of overall responsibility (CR) one has in raising the children and self-esteem. There was a significant negative correlation between responsibility for socialization (RS) and empathic concern. One 87 other relationship approached significance (p < .10). There was a negative correlation between influence in child rearing decisions and empathic concern. Hypotheses Ten, Eleven, and Twelve The next set of analyses involved comparing one's expressed level of desire to become a father in the future with the main variables PT, EC, and SE (hypotheses ten through twelve). The expressed level of desire to become a father (PDF) was recorded as interval data (in intervals of 20% ranging from 0-20% through to 80-100%). Kendall's correlation coefficient was used to accomplish this. None of the results showed significance, however Perspective-taking did approach significance at p = .051 with a positive relationship with PDF. The results are displayed in Table 18. Table 18 Kendall Correlation Table for PDF with PT, EC, and SE Variable Level of desire to father n=16 PT EC SE .3416 .0673 .1097 E = .051 E = -376 E =.300 Hypotheses Thirteen Through Sixteen Included in the questionnaire was a question pertaining to the recollection all subjects had of their own father's involvement in their parenting (RINV). Did the level of that recalled involvement have any relationship now to father's own involvement in parenting, or to their level of empathy as measured by the variables PT and EC? This question is the basis for hypotheses 13 through 16. Because the data yielded by this 88 question was continuous, a Pearson product-moment correlation coefficient was used to determine these relationships. There were no significant findings in this analysis which are displayed in Table 19. Table 19 Pearson Correlation Coefficients for RINV with PT, EC. and SE for the Whole Sample Variable PT EC SE Recollection of .0200 .0153 .0208 father p_ = .444 p_ = .457 p = .442 i nvo1vement n = 52 Note. One-tailed significance To determine the relationship between recollection of one's father's level of involvement in his parenting for the group of fathers the same statistic was computed for the variables PT, EC, SE, and PI. The results are displayed in Table 20 and indicate that there were no significant relationships among these variables. Table 20 Pearson Correlation Coefficients for RINV with PT, EC. SE, and PI for Fathers Variable PT EC SE PI Recollection of .0439 -.1144 .1324 -.0328 father p_ = .40 p_ = .253 p = .44 p = .425 involvement n=3 6 Note. One-tailed significance A similar analysis was also done for non-fathers, except for the variable PI. Again a Pearson product-moment correlation coefficient was used to examine the relationship between the 89 variable RINV, with PT, EC, and SE. The results of this analysis also showed no significant relationship between subject's recollection of his own father's involvement in parenting, and the dependent variables, perspective-taking, empathic concern, and self-esteem. The results are displayed in the Table 21. Table 21 Pearson Correlation Coefficient for RINV with PT, EC, and SE for Non-fathers Variable PT EC SE Recollection of .0315 -.0178 -.2573 father p = .456 p = .475 p = .17 involvement n = 16 Note. One-tailed significance Level one analysis for recollection of involvement. A further level of analysis for one's recollection of his own father's involvement in child care, was to examine how RINV related to the various father groups. One-way analyses were done for both the two and three father groups on RINV. The analysis for the three father group produced a significant F ratio (F = 2.86, p = .047). The LSD procedure indicated that the medium PI and the high PI groups had significantly lower mean RINV scores than the non-fathers. The follow up t-tests verified these results, finding that the difference between the high PI and non-fathers was significant to the p_ < .05 level and that the difference between the medium PI and non-fathers was significant to the p_ < .01 level. Generally speaking, it appears that non-fathers remember their own fathers being more highly involved in their parenting than those presently more highly involved as 90 fathers themselves. The results of the t-test results are presented in Table 22. Table 22 Three Father Group t-test Results Compared with Non-fathers on Variable RINV Variable Recollection of one's father's involvement Variable Recollection of one's own father's involvement Variable Recollection of one's own father's involvement Non-n 16 Non-n 16 Non-El 16 -fathers Mean 3.56 -fathers Mean 3.56 -fathers Mean 3.56 Medium PI fathers n Mean 12 2.58 Hicrh PI fathers n Mean 12 2.58 Low PI fathers n Mean 16 3.0 t value 2.95 t value 2.18 t value 1.46 E .007 E .041 E .16 df = 25.72 Note, t values are separate variance estimates 91 Level two analysis for recollection of involvement. A one-way analysis was also performed for the two father group. This analysis produced a significant F ratio (F = 4.49, p_ = .016). A further multiple range test (an LSD procedure) indicated a significant mean difference between high PI fathers and non-fathers, where non-fathers had a higher mean score on how highly they thought their fathers were involved in their care as children. The follow-up t-test value was significant (p = .005). The results are displayed in Table 23. Table 23 Two Father Group t-test Results for Non-fathers Versus High and Low PI Fathers on Variable RINV Variable Recollection of one's father's involvement Variable Recollection of one's father's involvement Non-n 16 Non-n 16 -fathers Mean 3.56 •fathers Mean 3.56 n 18 Low n 18 Hicrh PI fathers Mean 2.55 PI fathers Mean 2.94 t value 3.01 t value 1.83 E .005 E .076 df =31.67 Note, t values are separate variance estimates This section has outlined and presented statistical analyses done for the data pertaining specifically to testing the hypotheses, but also including some descriptive analyses to add to 92 a discussion about the results of hypotheses testing. This discussion will take place in the next chapter. 93 CHAPTER FIVE Summary and Conclusions The purposes of this chapter are to review the problem and purpose of this research, to draw conclusions based on the analyses within limitations, and then to present a discussion with emphasis on implications for counselling and future research. Restatement of Problem, Purpose and Findings As stated in the literature review, much of the research on fathering focuses on factors that influence paternal participation in child care (Radin & Harold-Goldsmith, 1989; Baruch and Barnett, 1986b). A further problem cited by Demo and associates (1987) is that most research that examines parenting, focuses on how the parent(s) or parenting influences child development. The present research attempts to present alternate relationships. This is accomplished by focusing not on what influences paternal involvement, but by focusing on involvement as an independent variable and examining possible relationships between paternal involvement and personality variables. Yet, because of design limitations, no directionality can be inferred; the intention is to raise alternate possibilities. The design of the current study is correlational; however a contrast group was employed to support or detract from any plausible indication of causality. The position that paternal involvement be an "independent" variable and the personality variables deemed "dependent" was not to infer causality, but to create reference points for examining the variables, particularly 94 as a plausible alternative to the assumed directionality of the bulk of the research. The exploration of the relationship between paternal involvement and paternal personality variables was dealt with through the testing of hypotheses four through nine. Hypotheses one through three deal with the comparison of fathers to non-fathers on the dependent variables. Sub-hypotheses were created for the purpose of complementing the findings. For example, sub-hypotheses ten through twelve deal with non-fathers' expressed level of desire to become fathers as related to the empathy and self-esteem. Findings here could lend support to positions that men with certain personality attributes do or do not intend to become fathers. Sub-hypotheses thirteen through sixteen, linking recollection of one's own father's involvement in the present subject's care, were included to further test the possibility that one's present level of involvement in child care might somehow be related to recollection of his father's involvement, or to present empathy, self-esteem, or paternal involvement levels. In particular, research showing that present empathy levels may be determined by past parental involvement in one's own care (Koestner, Weinberger & Franz, 1990) was the impetus for these hypotheses. While no connection between desire to parent and eventual level of involvement can be made based on the finding of these hypotheses, they could support or detract from the directionality implications of the Koestner and associates (1990) findings implying that if one's father was highly involved in parenting (in combination with 95 other factors), it may lead to higher levels of empathy in the male offspring. Generally speaking, the findings of the sub-hypotheses will create more questions than they will answer, and have been included to be exploratory rather than explanatory. Summary of Main Hypotheses Results While the above discussion focused on the results of hypotheses testing, it also included other analyses of a descriptive nature for the research variables. This section will summarize the results of the hypotheses tests and the assertions that can be made based on those results. Hypotheses were accepted or rejected at p < .05 level of significance. For the hypotheses that used correlational tests and predicted a direction of relationship, one-tailed tests of significance were accepted. For jt-tests, two-tailed results were accepted. Results were considered approaching significance if p < .10. The first three hypotheses dealt with comparing the personality variable mean scores of perspective-taking, empathic concern, and self-esteem of the high paternal involvement group with the non-fathers. Level one analysis divided the fathers into three groups based on their paternal involvement and compared non-fathers to highly involved fathers. Level two analysis divided the fathers into three groups based on the same criteria as level one analysis and compared the same sub-samples. The assertion was that high PI fathers would exhibit higher mean scores on the personality variables than non-fathers. The resultant t-test values were not significant at the p < .05 level so hypotheses 96 numbers one through three were rejected and the null hypotheses accepted. The next series of analyses dealt with hypotheses four through six. The mean scores of low paternally involved fathers were compared with the highly involved fathers on the same personality variables. Again level one and level two analyses were performed on the stipulated father groups. It was expected that the high PI fathers would exhibit significantly higher mean scores. None of the t values were significant at the p < .05 level, so the hypotheses were rejected. However, the self-esteem scores of high PI fathers were greater than those of the low PI fathers and did approach significance at p_ = .076. For the three father categorization hypotheses one, two, and three were rejected and the null hypotheses accepted. The level two analysis demonstrated no significant differences for the variables PT and EC, therefor hypotheses four and five were rejected and the and the null hypotheses accepted. For hypothesis 6, high PI fathers had a significantly (p_ = .045) higher mean score for self-esteem than did the low PI fathers (this difference had approached significance for the "three" father groups). On the basis of this analysis, the prediction that high involvement fathers would score higher on self-esteem than low involvement fathers (hypothesis six) is supported. The null hypothesis is rejected. Hypotheses seven, eight, and nine dealt with the proposal that there would be a significant positive relationship between father's PI scores and the variables PT, EC, and SE. The 97 correlation analysis produced no significant results at the p < .05 level so the hypotheses were rejected and the null hypotheses accepted for hypotheses seven through nine. However, it is notable that a positive relationship between paternal involvement and self-esteem did approach significance. Also of note is that the negative correlation between empathic concern and paternal involvement approached significance. Summary of Sub-hypotheses Results Hypotheses ten, eleven, and twelve dealt with non-father's level of desire to become a father in the future as related to their scores on perspective-taking, empathic concern, and self-esteem. The assumption here is that the more one desires to father, the higher his scores on PT, EC, and SE will be. None of the results of the correlational analyses proved to be significant at the p_ < .05 level, so these hypotheses were rejected. It is worth noting that the relationship between desire to father and PT approached significance (p_ = .051). For hypotheses ten through twelve the null hypotheses were accepted. The final level of analysis dealt with the relationship between one's recollection of his own father's involvement in his care and the variables perspective-taking, empathic concern, and self-esteem. These hypotheses assert that there will be a direct linear relationship between PT, EC, and SE and how involved one's own father was in his care. The correlational analysis for these variables on the whole sample did not produce any significant correlation coefficients at p < .05. Hypotheses thirteen, fourteen, fifteen, and sixteen were rejected, and the null 98 hypotheses accepted. Although the hypotheses did not indicate the use of the father group comparisons for RINV, level one and two analyses were done for RINV to allow for comparisons between father involvement and non-fathers. Significant results indicated that highly involved fathers remembered their fathers as being less involved in their own care than non-fathers. This applied for level one and two analyses; further, for level one analysis, the medium paternally involved fathers also remembered their own fathers as less involved in their care than the non-fathers did. Conclusion This study compared the relationship between paternal involvement and the personality characteristics of empathy and self-esteem. On the basis of the results, no linear relationships exist between these variables for this sample. However, it is of value to note that when the subscales of the paternal involvement scale were analyzed, there appeared a significant negative correlation between responsibility for socialization and empathic concern and a finding approaching significance between influence in childrearing decisions and empathic concern. Conversely, there was a significant positive correlation between child care responsibility and self-esteem. The division of the sample into subgroups allowed this research to meet two objectives. One is that it enables non-fathers to be compared to fathers by level of involvement. Second, it examined the possibility that the relationships between these groups and the personality characteristics might be non-linear. For the two father group comparison that included 99 comparisons with non-fathers, the high paternal involvement group demonstrated a significantly lower mean score on empathic concern than non-fathers This finding is consistent with the negative correlation between responsibility for socialization of children (a sub-scale of paternal involvement) and empathic concern. The analysis did not result in any significant mean differences between the father groups and the perspective-taking (for both two and three father groups). The near significant higher mean for medium paternally involved subjects over the low involvement group appears to contradict the findings for empathic concern. This appears curious because although both empathic concern and perspective-taking reflect different aspects of empathy, they are significantly positively correlated subscales of the IRI (Davis, 1980) . Although not a statistically significant result, this finding does suggest the possibility that any relationship between paternal involvement and empathy may be non-linear. If one examines the raw mean scores of both empathy subscales for the three father group comparison, they appear to indicate the possibility of a curvilinear relationship where the medium involvement group has higher raw means for both empathic concern and perspective-taking than both low and high involved fathers. It is reiterated that this was not proven through statistical analysis, but only supported by the one near significant t-test result. Analysis between parental involvement groups and self-esteem indicated that the high paternal involvement group had a 100 significantly higher mean score on self-esteem than the low paternal involvement group for the two father comparison groups. The findings for the three father groups did not support this finding, but the mean difference did approach significance with the highly involved fathers demonstrating higher average self-esteem than the low involvement group. Further correlational analysis also indicated a near significant finding for this relationship. This finding is indirectly supported by other research that suggests that higher levels of participation are associated with feeling more competent as a parent (Baruch & Barnett, 1986) . This result appears to indicate that those men who are more involved in parenting have higher self-esteem, but that as an even higher involved group is isolated, this relationship becomes less true. This suggests that when involvement is more extreme it does not follow that self-esteem increases in relative amounts. This further suggests that the relationship may not be a linear one. When examining empathic concern, it is evident that higher involvement in at least some aspects of parenting are negatively correlated with empathic concern. A near significant correlational relationship was also found to support this finding. Regarding perspective-taking, there is no relationship between it and the level of parental involvement, or when comparing these levels to non-fathers. The above findings for empathic concern and perspective-taking seem to be somewhat contradictory. However, Davis (19 83) predicted and found that there would be little or no relationship 101 between self-esteem and empathic concern (the ability to experience sympathy and concern for others). On the other hand, Davis also expected and found that those with higher perspective-taking ability would have higher self-esteem due to the generally better social functioning of those who can understand the perspective of others. In the present research, the finding that empathic concern is negatively correlated to some aspects of paternal involvement while self-esteem seems to be higher in the higher paternal involvement group is not an unexpected result based on Davis' findings. The results of this research support Davis' (1983) findings. Concerning the sub-hypotheses, some interesting results were revealed. Although only approaching significance (p_ = .051), perspective-taking appears positively correlated to non-father's level of desire to become a father in the future. This did not apply to empathic concern or self-esteem. This indicates a relationship between a potential father's ability to take the perspective of another and his level of desire to become a father in the future. Russell (1982) argues that fathers who are active participants in parenting are more committed to being so before they became highly involved. It is then also possible that non-fathers who desire to become fathers, are higher on perspective-taking before becoming fathers. When analyzed in the context of the present fathers, it is does not follow that highly involved fathers are more able to take the perspective of others. This result does support the findings of Koestner and associates 102 (1992), where men whose father's were more involved in their care was one of the factors that lead to higher empathy levels as adults. Because this sample of fathers did not exhibit higher empathy levels (as measured by perspective-taking) the findings of Russell are not necessarily supported by these findings although it is possible that those fathers with higher perspective-taking or empathic concern exhibited these levels before becoming fathers. In keeping with other results in this research, the relationship between either perspective-taking and empathic concern with the independent variable is not related to the other's relationship to the same variable. This again in spite of the significant intercorrelation between these two variables both in Davis' (1980) findings and this research. It is also consistent with Davis' 1983 research where, although intercorrelated, perspective-taking and empathic concern are not always similarly related to other personality variables (i.e., self-esteem). In this research there was no significant relationship between self-esteem and the level of desire to become a father in the future. One might intuitively hypothesize that a relationship may exist between the level of desire to become a father and one's actual future involvement. The findings of the present study that there is no relationship between desire and self-esteem, but there is between self-esteem and paternal involvement raises some interesting possibilities. One possibility is that desire to become a parent and later involvement are not related. Another 103 possibility is that those with higher self-esteem do not sort themselves into a higher level of parenting. Self-esteem could possibly develop as a result of ones higher involvement. This idea is somewhat supported by the findings of Demo and associates (1987) who found that fathers experiencing higher levels of communication with their adolescent, also demonstrated higher self-esteem. Again this relationship is correlational and not causal although Demo urges us to consider the possibility that parental self-esteem could at least be partly derived from relationships with one's children. Cowan (1988) also asserts that a man's self-concept (including self-esteem) can change during fatherhood particularly with the new demands and qualities demanded by the role of being a father. Generally speaking, one's recollection of how involved one's own father was in one's care as a child, was not correlated to perspective-taking, empathic concern, self-esteem, or the present level of paternal involvement for fathers. However, although not explicitly included in the hypotheses, an analysis of the father comparison groups indicated that those men in the high and medium paternal involvement groups demonstrated lower means on how involved they remembered their father's were in their care, than non-fathers. There were no significant differences between the father groups for mean scores on paternal involvement, empathic concern, and perspective-taking with recollection of father involvement. These results are curious yet compelling. How is it that those fathers more highly involved in parenting remember their own 104 fathers as less involved than non-fathers? Yet, this does not hold for low involvement fathers. One possibility is that those fathers more highly involved now are determined to be more involved parents than their own fathers were to them. This possibility is supported by Lamb and associates (1987) based on the results of interview reports of men who were dissatisfied with the lack of involvement of their own fathers expected to strive to be more involved parents than their fathers were. It is also possible that until a man becomes a parent himself, he has less concept of how involved a parent could or should be until he has his own experience as a frame of reference. In other words, non-fathers only "think" that their own fathers were more highly involved than those who are middle to highly involved parents now who know the experience of parenting. Once non-fathers themselves become parents, they may change their view on how involved their fathers were. One must also keep in mind that the medium involved fathers were significantly older than non-fathers. Although no significant difference exists between low involvement fathers and non-fathers, age may be a confounding factor for these results. Sub-analyses for relationships between the research variables revealed some significant (albeit few), differences among the father groups. One finding is that the non-father group (largely non-UBC) had significantly fewer spouses born outside of North America than the fathers taken as a whole or as sub-groups. This finding certainly could have implications for cultural differences between the fathers and non-fathers, including values about which 105 parent is more involved in child care. It should also be pointed out that these differences were not substantiated within the father groups where personality variables were measured against relative levels of paternal involvement. Finally, and perhaps as would be expected, non-fathers were involved in their spousal relationships for a significantly shorter time than the fathers. It is unclear what effect this would have on the results, particularly considering the results that the analyses of the hypotheses revealed no significant differences on dependent variable scores between non-fathers and any of the father sub-groups, or in correlational analyses. Limitations The results of this research must be interpreted with some caution. The most notable limitation is the sample and sampling methodology. Approximately half the sample is derived from student housing from the UBC campus. This in itself suggests a the possibility of a homogenous population. The demographic information suggests that it is a fairly diverse group concerning ethnicity. In fact, this particular sample may be more ethnically diverse than the population of the Greater Vancouver, perhaps all of Canada. Because the sample was drawn from family housing, the average age tended to be older (almost 35 years of age) than one might expect to find in the overall university population. Sampling methodology creates a more formidable limitation most notably in that it is largely a volunteer sample. This is true of those sampled from student residence, although somewhat less true of the rest of the sample in that those solicited 106 through personal contacts may have agreed to participate as a "favor" where they might not have if there was no personal contact. This may have made the non-UBC part of the sample in a sense less voluntary than the UBC sample. The danger of course was that these two samples may be different particularly concerning the research variables. This did not prove to be the case based on the statistical analysis. Regardless of these tests for confounding relationships between the research and demographic variables, the generalizability of the results found for this sample must be exercised with caution. Related to the above limitation, was the creation of artificial divisions of the level of paternal involvement. These divisions may in no way represent "true" differences among fathers of their level of involvement. It must be stated that the intention for doing this was to compare relative levels of parenting involvement among this group of subjects only. In no way is there an attempt to imply that these typical divisions in the general population. It was also necessary to form these groups for statistical analysis, specifically to be able to compare the fathers to the non-fathers on the dependent variables. Dividing the groups into paternal involvement also created the problem of comparing small sample sizes. This further limits the ability to generalize the results in spite of statistical controls. Where possible, the various groups were analyzed for differences on other demographic information to identify possible confounding variables. Mostly, these analyses revealed that the various groups did not differ on most demographic variables, 107 thereby supporting the results. However, because of the small sub-group sample sizes, caution must be exercised when interpreting significant differences on dependent variables, in spite of analyses for potential confounding differences among the sub-groups. Conversely, the significant findings based on the small sample size may also indicate greater confidence because it was found in a relatively small sample size. Even with significant mean differences found between the father groups, one must question whether these differences practically significant. For example, does someone who scores 43 on a self-esteem scale really have more self-esteem than one who scores 38? The variables of the recollection of one's own father's involvement in parenting and the level of desire to become a father in the future are exploratory only, and have not been instrumentally validated. They are meant to be taken at face value only, with all the inherent limitations that retrospective and projective data posses. Furthermore, each variable is based on one or two questions only, further threatening validity and potential reliability. They were included merely to add further context and discussion generation to the results of the main hypotheses analyses. Finally, the results of this research can only demonstrate whether a relationship exists between variables or not. Although a case has been made to consider effects of parenting on the father, the design of this research does not allow us to draw conclusions about causality. A comparison group was used to support or oppose potential causality, not to prove it. 108 Longitudinal research would be indicated to best measure any changes in men who become fathers. Implications for Counselling and Future Research We know that for this sample of men, those grouped together as more highly involved in parenting exhibit higher self-esteem than those who are less involved in parenting. What we do not know is what direction this effect is in. In terms of counselling, this distinction may not be important. If higher involvement in parenting did result in higher self-esteem it is unlikely that a therapist would prescribe that a father who is not highly involved in parenting to become more involved to feel better about himself. However, the result of this research combined with other research that points to higher involved fathers feeling more competent and self-confidence (Russell, 1989), and Cowan and Cowan's (1988) finding that those satisfied with their role in the care of their children had higher self-esteem before the birth of their children, could be used as information to develop hypotheses for therapy. For example, very little involvement with one's children, may be an indicator that a father is not feeling good about himself then. What caused the low involvement or possible low feeling of self worth is less important than dealing with specific children's issues or the feelings of low self worth. In a systemic approach to the study of fathering, Pauline Boss (1986) argues that physically and or psychologically absent fathering could be imbedded in a family system where, for example, the under functioning of one member in the system could be 109 compensated for by the over functioning of one or more members of that system. This in a sense would strengthen the notion that the under functioning member (a less involved father in this case) is not needed, which could result in continued under involvement. Boss further states that when this absence of involvement occurs, there is less communication in the system. Communication between fathers and adolescents has been shown to be positively related to father self-esteem (Demo et al., 1987). The results of this research also seemed to point to empathic concern being negatively correlated to the paternal involvement subscales, responsibility for socialization of the children (significant), and influence in childrearing decisions (approaching significance). These findings are not necessarily surprising as the tasks involved in these roles (they are largely defined figuratively) do not by definition require that one feel another's feelings, although the significant negative correlation is somewhat more puzzling. In some senses the idea of over involvement in some aspects of childrearing seems consistent with the idea of an over-reaction or resolve to undo what one might believe was an under involvement in his care as a child by his father. It is further supported by the findings of this study where the medium and higher paternally involved fathers on average remember their fathers as less involved in their parenting than non-fathers. Gordon (1990) believes that unresolved emotions about one's own parental relationships could be a reason for estrangement from one's children. Perhaps then one could be over involved in at 110 least some aspects of parenting, but not necessarily because one is in touch with the feelings of his children, but as a reaction to an unresolved past. These scenarios could be viewed as possible avenues of exploration in therapy, not as definite relationships. Generally speaking, the results of this research with regard to limitations, must be viewed with caution when exploring implications for counselling. The findings that higher involved fathers appear to report higher self-esteem is generally supported by other research. Its usefulness in counselling however, should be limited to developing hypotheses about a client, and one should be cautioned against assuming that a less involved father has low self-esteem. These findings can point to areas of exploration if the subject matter is related to fathering and self-esteem or empathic concern. Indications of direction or causality cannot be supported by this research. If we enter into a narrow debate about which causes the other we may be missing the possibility that the effect could be an interactive one. Demo and associates (1987) raise this point in their research, while Bell (1968) was more vigorous in challenging the accepted assumption of directionality in much of the psychological and sociological literature. The present researcher asks the reader to consider an interactive stance when looking at the result of this research, but also concedes a sense of dissatisfaction that many may experience when limited to correlational research. While one may muse at the possibility of an interactive effect between variables, it would be far more 111 satisfying to be able to pinpoint possible direction of effect through other research designs. Directions for further research resulting from this study and general topic area are vast. As has been pointed out by various authors, the possibility that men are developmentally and profoundly effected by the experience of fathering is under researched (Demo et al., 1987; Lamb et al., 1987). Correlational research such as this (even with a comparison group) is not able to assess the impact (or not) of the experience of fathering on men. This potential impact has been theorized (Lamb, et al., 1987), but unfortunately this research has not been able to test these particular causal theories. Various studies have researched potential immediate change shortly after the birth (usually months) of a child (Cowan & Cowan, 1988; Volling and Belsky, 1991; Keller et al., 1985; Frodi et al, 1983), or change in involvement over a longer period of time, but not necessarily measuring father personality characteristics (Radin, 1985; Crouter & Perry-Jenkins, 1987). This researcher was unable to find longitudinal research that measures the effects on fathers over a longer period of time. A longitudinal design over a longer period of years measuring potential changes in the characteristics of fathers is indicated. Such a group could be compared to a contrast group over the same time. In the review of the literature there appears to be an implied debate emerging. On one side are those that argue that the experience of fathering somehow can effect and change a man, while on the other end are those who argue that it is men with 112 certain characteristics that become more or less involved with their children (Belsky 1984; Volling & Belsky, 1991; Crouter et al., 1987). This research may look at some personal characteristics, but is often concerned with social factors such as dual versus single earner families, employment, maternal attitudes on fathering, and others. While this research is valuable, it does not subsume the type of question this study hopes to answer. Do men who are more involved in parenting differ on key personality characteristics? This study has had only limited success at answering this question compounded by the limitations of this study. However, some association between level of paternal involvement and self-esteem and even empathic concern has surfaced. As a result of the small sample sizes, correlation coefficients that appeared significant were not statistically so. Because significance is tied to sample size, the significant and near significant results may also be interpreted with even more confidence because of he relatively small sample size for analysis. This suggests that replication of this type of research with a larger sample size could yield statistically significant results. The finding that linear relationships do not always exist (i.e., self-esteem) while significant relationships do when the men are divided into relative father involvement categories, indicates the value of expanding the research horizons when examining personality characteristics. While curvilinearity was not supported by the results, the possibility that non-linear 113 relationships do exist could also be undertaken in future research. Even for the relationships that were found to be significant or near significant, the actual correlation is fairly low. This suggests that there are other variables that are related for example, to self-esteem. Future research could expand on the traditional versus non-traditional father roles as suggested by Lamb and associates (1982) and Aldous' work on role taking to examine the relationship of these variables taken together with self-esteem. How would being a more involved father and stepping out of traditional sources of validation effect self-esteem, or paternal involvement? Demo and associates (1987) point out that adult self-esteem would be derived from more sources than their child if this effect did in fact exist. Whether fathers are actually more paternally involved now than at anytime in the recent past is inconclusive based on contradictory evidence (Cowan & Bronstein, 1988; Lewis & O'Brien, 1987). However, a recent Statistics Canada release indicates that the number of dual-earner families increased steadily until 1988 when it leveled off to about 61% of all families (Statistics Canada, 1992). This suggests that traditional father only breadwinner household is actually in the minority. It appears that the division of labor in the family may be changing from the stereotypical roles of the last century and possible that fathers are becoming more involved in parenting perhaps through necessity. This is echoed pertinently by Levant and associates (1987, p.156) in the following statement: 114 A very large challenge faces the current generation of fathers. It is now expected that fathers will learn a new role, one that departs radically from the roles of their fathers, involves a new definition of masculinity, and requires housework skills such as cooking and cleaning and child care skills such as sensitivity to children, nurturance, expressivity, and the ability to manage children's problems. 115 References Aldous, J. (1974). The making of family roles and family change. The Family Coordinator, 23, 231-235. Ambert, A. M. (1992) The Effects of Children on Parents. Barush, K., & Barnett, R. (1986a). Consequences of fathers' participation in family work: Parents' role strain and well-being. Journal of Personality and Social Psvcholoav, 54, 983-992. Barush, K., & Barnett, R. (1986b). Fathers' participation in family work and children's sex-role attitudes. Child Development, 57, 1210-1223. Bell, A. H. (1962). Attitudes of selected rehabilitation workers and other hospital employees toward the physically disabled. Psychological Reports, 10, 183-186. Bell, R. Q. (1968). A reinterpretation of the direction of effects in studies of socialization. Psychological Review. 75(2). 81-95. Bern, S. (1981). Sex-role Inventory: Professional manual. Palo Alto: Consulting Psychologists Press. Blascovich, J., & Tomaka, J. (1991). Measures of self-esteem. In J. Robinson; P. Shaver, & L. Wrightsman (Eds.), Measures of Personality and Social Psychological Attitudes: (Vol. 1).(pp. 115-137). San Diego: Academic Press. Borg, W. R., & Gall, M. D. (1989). Educational Research: An Introduction (5th ed.). New York: Longman. Boss, P. (1986). Psychological absence in the intact family: A systems approach to a study of fathering. Marriage and Family Review. 10, 11-39. Bronstein, P. (1988). Marital and parenting roles in transition: An overview. In P. Bronstein & P. Cowan (Eds.), Fatherhood today: Men's changing role in the family (pp. 3-10). New York: John Wiley & Sons. Campbell, R. N. (1984). The New Science: Self-Esteem Psychology. Lanham, MD: University Press of America. Carey, J., Fox, E., & Spraggins, E. (1988). Replication of structure findings regarding the Interpersonal Reactivity Index. 116 Measurement and Evaluation in counseling and development. 21, 102-105. Cellini, J. V., & Kantorowski, L. A. (1982). Internal-external locus of control: New normative data. Psychological Reports. 51, 231-235. Chlopan, B. E., McCain, M. L., Carbonell, J. L., & Hagen, R. L. (1985). Empathy: Review of available measures. Journal of Personality and Social Psychology, 48, 635-653. Coopersmith, S. (1967). The Antecedents of Self-Esteem. San Francisco: W. H. Freeman. Cowan, C , & Bronstein, P. (1988). Fathers' roles in the family: Implications for research, intervention and change. In P. Bronstein & P. Cowan (Eds.), Fatherhood today: Men's changing role in the family (pp. 341-348). New York: John Wiley & Sons. Cowan, P. (1988). Becoming a father: A time of change, an opportunity for development. In P. Bronstein & P. Cowan (Eds.), Fatherhood today: Men's changing role in the family (pp. 13-35). New York: John Wiley & Sons. Coleman, A., & Coleman L. (1981). Earth father/sky father. Englewood Cliffs, NJ: Prentice-Hall. Coman, G., Evans, B., & Stanley, R.(1988). Scores on the interpersonal reactivity index: A sample of Australian medical students. Psychological Reports. 62, 943-945. Cross, D. G., & Sharpley, C. F. (1982). Measurement of empathy with the Hogan Empathy Scale. Psychological Reports. 50, 62. Daniels, P., & Weingarten, K. (1988). The fatherhood click: The timing of parenthood in men's lives. In P. Bronstein & P. Cowan (Eds.), Fatherhood today: Men's changing role in the family (pp. 36-52). New York: John Wiley & Sons. Davis, M. H. (1983). Measuring individual differences in empathy: Evidence for a multidimensional approach. Journal of Personality and Social Psychology. 44, 113-126. Davis, W. L., & Phares, E. J. (1969). Parental Antecedents of internal-external control of reinforcement. Psychological Reports. 24. 427-436. 117 De Lucci, M. , & Davis, A. (1991a). Do men's adult life concerns affect their fathering orientation?. The Journal of Psychology. 125. 175-188. De Lucci, M., & Davis, A. (1991b). Father-child relationships from the preschool years through mid-adolescence. The Journal of Genetic Psychology, 152. 225-238. Demo, D. H. (1985). The measurement of Self-Esteem: Refining our methods. Journal of Personality and Social Psychology. 48, 1490-1502. Demo, D., Small, S., & Savin-Williams, R. (1987). Family Relations and the self-esteem of adolescents and their parents. Journal of Marriage and the Family. 49. 705-715. Donaldson, J. (1980). Changing attitudes toward handicapped persons: A review and analysis of research. Exceptional Children, 7, 504-514. Eysenck, H. J. & Eysenck, M. W. (1985). Personality and Individual Differences: A Natural Science Approach. New York: Plenum Press. Fein, R. (1978). Research on fathering: Social policy and an emergent perspective. Journal of Social Issues. 34, 122-135. Feldman, L. (1990). Fathers and fathering. In R. Meth & R. Pasick (Eds.), Men in therapy: The challenge of change (pp.88-107). New York: The Guilford Press. Gladstein, G. A. (1983). Understanding empathy: Integrating counseling, developmental, and social psychology perspectives. Journal of Counseling Psychology. 30. 467-482. Goldstein, A. P., & Michaels, G. Y. (1985). Empathy: Development, Training, and Consequences. Hillsdale, NJ: Lawrence Erlbaum Associates. Gordon, B. (1990). Men and their fathers. In R. Meth & R. Pasick (Eds.), Men in Therapy: The challenge of change (pp. 234-246). New York: The Guildford Press. Greenberg, M., & Morris, N. (1974). Engrossment: The newborn's impact upon the father. American Journal of Orthopsychiatry, 44, 520-531. Grossman, F., Pollack, W., & Golding, E. (1988). Fathers and children: Predicting the quality and quantity of fathering. Developmental Psychology. 24. 82-91. 118 Halpin, G., Halpin, G. , & Whiddon, T. (1980). The relationship of perceived parental behaviors to locus of control and self-esteem among American Indian and white children. The Journal of Social Psychology. 111, 189-195. Heath, D. (1978). What meaning and effects does fatherhood have for the maturing of professional men?. Merrill-Palmer Quarterly. 24, 265-278. Hogan, R. (1969). Development of an empathy scale. Journal of Consulting and Clinical Psychology. 33. 307-316. Hoffman, M. L. (1977). Empathy, Its Development and Prosocial Implications. Nebraska Symposium on Motivation. American Psychological Association. Isaac, S., & Michael, W. (1981). Handbook in research and evaluation (2nd ed.). San Diego: Edits Press. Johnson, J., Cheek, J., & Smither, R. (1983). The structure of empathy. Journal of Personality and Social Psychology, 45., 1299-1312. Jordan, J. V. (1984). Empathy and self-boundaries (Work in Progress: No. 16). Wellesley, MA: Stone Center for Developmental Services and Studies. Kaye, L., & Applegate, J. (1990). Men as elder caregivers: A response to changing families. American Journal of Orthopsychiatry. 60., 86-95. Katz, R. (1963). Empathy: Its Nature and Uses. New York: Free Press. Keller, W., Hildebrandt, K., & Richards, M. (1985). Effects of extended father-infant contact during the newborn period. Infant Behavior and Development, 8., 337-350. Koestner, R. , Franz, C , & Weinberger, J. (1990). The family of origins of empathic concern: A 2 6-year longitudinal study. Journal of Personality and Social Psychology. 58, 709-717. Kotre, J. (1984). Outliving the Self: Generativitv and the Interpretation of Lives. Baltimore: John Hopkins University. Krahe, B. (1992). Personality and Social Psychology: Towards a Synthesis. London: Sage Publications. 119 Lai, C. (1990). UBC SPSSX: Statistical Package for Social Sciences. Extended Version, Release 3.0 (Under MTS). Vancouver: University Computing Services. Lamb, M. E. & Bronson, S. K. (1980). Fathers in the context of family influences: Past, present and future. School Psychology Review. 9_, 336-353. Lamb, M., Frodi, A., Hwang, C., & Frodi, M. (1982). Varying degrees of paternal involvement in infant care: Attitudinal and behavioral correlates. In M. Lamb (Ed.), Nontraditional families: Parenting and child development. Hillsdale, NJ: Lawrence Erlbaum Associates. Lamb, M., Pleck, J., & Levine, J. (1987). Effects of paternal involvement on fathers and mothers. In C. Lewis & M. O'Brien (Eds.), Reassessing fatherhood: New observations on fathers and the modern family (pp. 109-125). Newbury Park, CA: Sage Layton, C. (1985). Note on the stability of Rotter's I-E scale. Psychological Reports. 57, 1165-1166. Layton, J., & Wykle, M. (1990). A validity study of four empathy instruments. Research in Nursing and Health. 13. 319-325. Levant, R., Slattery, S., & Loiselle, J. (1987). Fathers' involvement in housework and child care with school aged daughters. Family Relations. 36, 152-157. Martin, J. D., & Coley, L. A. (1984). Intercorrelations of some measures of self-concept. Educational and Psychological Measurement. 44, 517-521. Mondell, S. & Tyler, F. B. (1981). Parental competence and styles of problem solving/play behavior with children. Developmental Psychology. 17. 73-78. Oyserman, D., Radin, N., & Benn, R. (In Press). Dynamics in a Three-Generational Family: Teens, Grandparents, and Babies. Developmental Psychology. Radin, N. (1986). The influence of fathers on their sons and daughters. Social Work in Education, 8., 77-91. Radin N., & Sagi, A. (1982). Childrearing fathers in intact families, II: Israel and the USA. Merrill-Palmer Quarterly. 28, 111-136. 120 Radin, N., & Goldsmith, R. (1985). Caregiving fathers of preschoolers: Four years later. Merrill-Palmer Quarterly, 31, 375-383. Radin, N., & Harold-Goldsmith, R. (1989). The involvement of selected unemployed and employed men with their children. Child Development, 60. 454-459. Radin, N. (1990). Validity and Reliability of the Paternal Child Care Index (PICCI). Unpublished manuscript, University of Michigan, School of Social Work, Ann Arbor. Radin, N., & Williams, E. (in press). Predictors of adolescent academic achievement and expectations in a white middle-class population: An 11-year follow-up. American Journal of Orthopsychiatry. Robinson, B., & Barret, R. (1986). The developing father. New York: The Guilford Press. Robson, P. (1989). Development of a new self-report questionnaire to measure self-esteem. Psychological Medicine, 19, 513-518. Rosenberg, M. (1979). Conceiving the Self. New York: Basic Books. Rosenthal, R., & Rosnow, R. (1975). The Volunteer Subject. New York: Wiley. Rotter, J. B. (1966). Generalized expectancies for internal versus external control of reinforcement. Psychological Monographs, 80(Whole Number). Rotter, J. B. (1990). Internal versus external control of reinforcement: A case history of a variable. American Psychologist, 45, 489-493. Russell, G. (1978). The Father role and its relation to masculinity, femininity, and androgyny. Child Development, 49, 1174-1181. Russell, G. (1982). Highly participant Australian fathers: Some preliminary findings. Merrill-Palmer Quarterly. 28, 137-158. Russell, G. (1989) . Work/family patterns and couple relationships in shared caregiving families. Social Behavior, 4, 265-283 . Sherer, M., & Maddux, J. E. (1982). The self-efficacy scale: Construction and validation. Psychological Reports, 51, 663-671. 121 Statistics Canada. (1992). Characteristics of Dual-earner Families. (Catalogue No. 13-215). Strauss, A. (1956). The Social Psychology of George Herbert Mead. Chicago: Phoenix Press. Van de Water, D., & McAdams, D. (1989). Generativity and Erickson's "belief in the species". Journal of Research Personality, 23, 435-449. Volling, B., & Belsky, J. (1991). Multiple determinants of father involvement during infancy in dual-earner and single-earner families. Journal of Marriage and the Family, 53, 461-474. Webster's II: New riverside dictionary. (1984). Boston, MA: Houghton Mifflin. Wells, E. L., & Marwell, G. (1976). Self-Esteem: Its Conceptualization and Measurement. Beverly Hills: Sage. Wiehe, V. R. (1987). Empathy and locus of control in child abusers. Journal of Social Services Research, 9., 17-30. 122 Appendix A This appendix contains the notices that were sent out to UBC family residence for the initial sample collection. Also included is the complete information package as handed to those agreeing to participate in the study. This includes a letter of introduction, a face page including draw information, instructions for completing questionnaire, and an end page thanking including delivery instructions. Finally, there are also copies of a revised letter of introduction, face page, and end page for subjects solicited after the UBC residence sampling was complete. These documents do not include draw information (as the draw had already been completed) and provides new delivery instructions. 123 ATTENTION MALE RESIDENTS! HI! MY NAME IS HOWARD SCHEIN, AN M.A. CANDIDATE IN COUNSELLING PSYCHOLOGY. I AM CONDUCTING RESEARCH ON FATHERING AND NEED SOME HELP FROM BOTH FATHERS AND NON-FATHERS IN COMMON-LAW OR MARRIAGE RELATIONSHIPS. THE STUDY OF FATHERHOOD, AND MEN'S INVOLVEMENT IN FAMILY IS VITAL TO OUR UNDERSTANDING OF FAMILY STRUCTURE AND DEVELOPMENT. WHAT I AM ASKING FOR IS ABOUT 20-30 MINUTES OF YOUR TIME TO FILL OUT SOME QUESTIONNAIRES. INFORMATION YOU SUPPLY IS COMPLETELY CONFIDENTIAL AND PARTICIPATION IS VOLUNTARY. SHOULD YOU CHOOSE TO PARTICIPATE, YOUR NAME WILL BE ENTERED FOR A DRAW AT THE END OF THE SURVEY FOR A $75.00 PRIZE (DETAILS CONTAINED IN QUESTIONNAIRE, OR CALL ME). ALL YOU DO IS PHONE THE NUMBER BELOW AND I WILL DROP OFF THE QUESTIONNAIRE AND PICK IT UP WHEN COMPLETED. I APPEAL TO YOUR SENSE OF IDENTIFICATION WITH A STRUGGLING STUDENT ATTEMPTING TO COMPLETE MY DEGREE AND ASK FOR YOUR HELP! SHOULD YOU DESIRE TO PARTICIPATE, WISH FURTHER DETAILS, OR HAVE CONCERNS PLEASE CALL ME OR LEAVE A MESSAGE AT 521-8302; OR CALL DR. M.J. WESTWOOD IN THE DEPARTMENT OF COUNSELLING PSYCHOLOGY, (PROJECT SUPERVISOR) AT 822-6457. PLEASE REPLY BY MAY 31. ***T0DRY, MflV 28 ONLY, I WILL BE RT 521-8302 FROM 3PM TO 6PM TO TAKE CALLS, RND WILL DELIUER R QUESTIONNAIRE TO V0UR DOOR THIS EUENING OR RNV OTHER TIME YOU WISH TO AARANGE.*** 124 Dear Participant, Thank-you for taking the time to participate in the Father Survey Project. My name is Howard Schein, M.A. candidate in Counselling Psychology and the student researcher for this project. My home number is 521-8302. The research supervisor is M. J. Westwood, Ph.D, and Associate Professor in the Department of Counselling Psychology (822-6457). The purpose of this project is to survey father's participation in child care and related personality traits of these fathers compared with non-fathers. With the roles of men and women in society and the family changing rapidly this research is cutting edge and vital to assessing potential impacts of these changes on fathers specifically, and men in general. Your participation in this research will make a small demand on your valuable time, but is vital to the research. Enclosed in this package is a questionnaire package including instructions for each section. The entire package should take 20-30 minutes to complete. The next page contains instructions on the return of the questionnaire once completed. We request that you return or have them ready for pick-up by May 31, 1993. Please note that you have the right to refuse participation in this survey without jeopardy to yourself. Completion of this questionnaire will be taken as consent of your participation in this project. Information given by you will be completely confidential and accessible only to Dr. Westwood and myself. Results will be kept anonymous by assigning a number to each questionnaire and results will be recorded by number, not by name. The completed questionnaires will be stored in a secure location in the Department of Counselling Psychology and destroyed at the end of the research. Finally, at some point after the questionnaires have been processed, I would like to invite you to participate in an informal interview (one hour in length) about your experiences as a father . However, your participation in the questionnaire does not obligate you to participate in the interview. If you are interested in the interview you may indicate so on the final page of the questionnaire package. Again I would very much like to thank-you for your participation in this research and invite you to call me or Dr. Westwood at the above numbers should you require any clarification of instructions or information about the study. Thank-you, Howard Schein, M.A. Candidate & Student researcher M. J. Westwood, Ph.D, Associate Professor, & Thesis Supervisor 125 Dear Participant, Thank-you for taking the time to participate in the Father Survey Project. My name is Howard Schein, M.A. candidate in Counselling Psychology and the student researcher for this project. My home number is 521-8302. The research supervisor is M. J. Westwood, Ph.D, and Associate Professor in the Department of Counselling Psychology (822-6457). The purpose of this project is to survey father's participation in child care and related personality traits of these fathers compared with non-fathers. With the roles of men and women in society and the family changing rapidly this research is cutting edge and vital to assessing potential impacts of these changes on fathers specifically, and men in general. Your participation in this research will make a small demand on your valuable time, but is vital to the research. Enclosed in this package is a questionnaire package including instructions for each section. The entire package should take 20-30 minutes to complete. The next page contains instructions on the return of the questionnaire once completed. We request that you return or have it completed within one week of receiving it. Please note that you have the right to refuse participation in this survey without jeopardy to yourself. Completion of this questionnaire will be taken as consent of your participation in this project. Information given by you will be completely confidential and accessible only to Dr. Westwood and myself. Results will be kept anonymous by assigning a number to each questionnaire and results will be recorded by number, not by name. The completed questionnaires will be stored in a secure location in the Department of Counselling Psychology and destroyed at the end of the research. Again I would very much like to thank-you for your participation in this research and invite you to call me or Dr. Westwood at the above numbers should you require any clarification of instructions or information about the study. Thank-you, Howard Schein, M.A. Candidate & Student researcher M. J. Westwood, Ph.D, Associate Professor, & Thesis Supervisor 126 FATHER SURVEY PROJECT START PAGE The following pages contain a number of separate sections, each with a separate set of instructions to follow. Please try to complete every question as instructed, and it should take no more than 30 minutes. We request that you have the questionnaires completed by Monday May 31, when we will be making an initial pick up of the completed questionnaires. Should you be unavailable this day, or unable to complete the questionnaires by this date, then I will be making one more pick up between June 11-13 (you will be notified regarding a specific date and time). You may also drop the completed forms off at the front desk of the Counselling Psychology Dept. (5780 Toronto Rd. at Acadia) care of Marv Westwood. The draw date will be June 14 at 7:00 PM in the commons block lounge. I will be in the commons lounge from 6:00 on June 14. Completed questionnaires and tickets (at bottom of this page) must be received by 6:30 on this date to be eligible for the draw. Howard Schein, Project Coordinator Please indicate the prize option you desire and then enter your name and number below. $75.00 Safeway gift certificate $75.00 certificate for a dinner for two at A Kettle of Fish Cash option, $75.00 Name; Phone#: 127 FATHER SURVEY PROJECT START PAGE The following pages contain a number of separate sections, each with a separate set of instructions to follow. Make sure you check each page for changes in instructions so as not to miss questions pertaining to you. Please try to complete every question as instructed, and it should take no more than 30 minutes. Once you have completed the questionnaire if you have not already arranged for a pick-up you may call me at 521-8302, or drop it off at 5780 Toronto Rd., Vancouver (Department of Counselling Psychology, UBC). The questionnaire begins on the following page. Howard Schein, Project Coordinator 128 THE END! Thank-you for your support and participation in the father survey project! If you have completed the questionnaire please refer to the "START" page for return of the questionnaire, or you may call me or leave a message at 521-8302 for pick-up of the materials. Just two more pieces of information... 1. If you would like a summary of the results of the research please check the space below. Please send me a summary of the results. 2. Once the statistical research is complete, I would like to put a more personal face on this research by interviewing some of you about your fathering experiences (for approximately one hour). If you are interested in participating enter your name and phone number in the space below. Name: Phone #: TO ENTER IN THE DRAW PLEASE RECORD YOUR NAME AND NUMBER ON THE START PAGE! THE DRAW WILL BE MADE ON MONDAY JUNE, 14 AT 7:00 PM IN THE COMMONS BLOCK LOUNGE. THE WINNER WILL BE CONTACTED BY PHONE OR YOU MAY BE PRESENT AT THE DRAWING. ONCE THE DRAW HAS BEEN MADE THE TICKETS WILL BE DESTROYED. 129 THE END! Thank-you for your support and participation in the father survey project! If you have completed the questionnaire please refer to the "START" page for return of the questionnaire, or you may call me or leave a message at 521-8302 for pick-up of the mater ia ls . 1. If you would like a summary of the results of the research please check the space below and leave your name and address in the space provided. (This page will be removed). Please send me a summary of the results. 130 Appendix B This appendix includes a copy of the entire questionnaire package as handed out to potential subjects. Included in the package is the demographic questionnaire, the Parental Involvement in Child Care Index (Radin, 1990), the Interpersonal Reactivity Index (Davis, 1980), and the Rosenberg Self-esteem Scale (Rosenberg, 1965). It should be noted that the PICCI, IRI, and SES were formatted or reformatted to fit this package. 131 FATHER SURVEY PROJECT SECTION (1) To be filled out by all participants This section is composed of four parts (A,B,C,D). A and B are to be filled out by all participants. C is to be filled out by non-fathers only, while D is to be filled out by fathers only. Most questions require only a check mark adjacent to the appropriate answer. However, a few questions do require a short one word answer or numerical data. Please answer in the available space. Part A: To be filled out by all participants 1) Are you a father? yes no 2) Your Age 3) What is your present marital status? married divorced common-law widowed separated single 4) For how many years have you had your present marital status?. 5) How many years of schooling did you complete? 0-12 years 17-18 years 13-14 years 19-20 years 15-16 years 21 or more 6) What, if any, is your religion? no religion Roman Catholic United Church Anglican Presbyterian Lutheran Baptist Eastern Orthodox Jewish Other, please specify 7) Other than special occasions such as weddings, funerals, or baptisms, how often did you attend services or meetings connected with your religion in the last 12 months? at least once a week at least once a month a few times a year at least once a year not at all 132 8) Do you presently feel a part of some religious group? yes no 9) Which of the following groups is closest to your ethnicity? Francophone Greek Chinese _ British .Japanese Italian _Native Indian German _East Indian American .Eastern European other, please specify. 10) Were you born in North America? yes no 11) During the past 12 months, what best describes your MAIN activity? working at a job or business looking for work _a student .keeping house .taking care of your children _retired .other, please specify 12) How many hours per week did you study in the past 12 months (including class time)? not applicable 21-25 hours 0-5 hours 26-30 hours 6-10 hours 31-35 hours 11-15 hours 36-40 hours 16-20 hours more than 40 13) Of the above hours spent in study, approximately what percentage were done at home? 0 - 2 0 % 6 1 - 8 0 % 2 1 - 4 0 % 8 1 - 1 0 0 % 4 1 - 6 0 % 14) How many hours per week did you work in the past year? not applicable 0-5 hours 21-25 hours 6-10 hours 26-30 hours 11-15 hours 31-35 hours 16-20 hours 36-40 hours more than 40 133 15) Of the above work hours, approximately what percentage were done at home? 0-20% 61-80% 21-40% 81-100% 4 1 - 6 0 % 16) What type of work do you do? 17) Which of the following categories is closest to your total PERSONAL income per year? less than $10,000 $10,001-$20,000 $20,001-$30,000 $30,001-$40,000 more than $40,000 18) Which of the following categories is closest to your total HOUSEHOLD income per year? less than $10,000 $ 1 0 , 0 0 1 - $ 2 0 , 0 0 0 $20,001-$30,000 $30,001-$40,000 $40,001-$50,000 more than $50,000 19) To the best of your recollection, how involved was your father in parenting you as a child? (please Circle one) 1 2 3 4 5 not at all not much some much very much Part B: Questions about your spouse/partner. To be filled out by all par t ic ipants 1) How many years of schooling did your spouse complete? 0-12 years 17-18 years 13-14 years 19-20 years 15-16 years 21 or more don't know 2) What, if any, is your spouse's religion? don't know no religion Roman Catholic United Church Anglican Presbyterian Lutheran Baptist Eastern Orthodox Jewish other, please specify 134 3) Other than special occasions such as weddings, funerals, or baptisms, how often did your spouse attend services or meetings connected with her religion in the last 12 months? at least once a week at least once a month a few times a year at least once a year not at all 4) Which of the following groups is Francophone Chinese Japanese Native Indian East Indian Eastern European closest to your s[ Greek British Italian German American don't know other, please specify 5) Was your spouse born in North America? yes no 6) During the past 12 months, what best describes your spouse's MAIN activity? working at a job or business looking for work a student keeping house taking care of your children ret i red other, please specify 7) How many hours per week did your spouse study in the past 12 months (including class t ime )? not applicable 21-25 hours 0-5 hours 26-30 hours 6-10 hours 31-35 hours 11-15 hours 36-40 hours 16-20 hours more than 40 8) Of the above hours spent in study, approximately what percentage were done at home? 0 - 2 0 % 6 1 - 8 0 % 2 1 - 4 0 % 8 1 - 1 0 0 % 41-60% 135 9) How many hours per week did your spouse work in the past year? .not applicable _0-5 hours _6-10 hours _11-15 hours 16-20 hours _21-25 hours _26-30 hours _31-35 hours 36-40 hours 10) Of the above work hours, approximately what percentage were done at home? 0 - 2 0 % 2 1 - 4 0 % 4 1 - 6 0 % . 6 1 - 8 0 % 8 1 - 1 0 0 % 11) What type of work does your spouse do?. Part C: to be filled out by non-fathers only. 1) At this time, do you desire to become a father in the future? yes no 2) Please estimate the probability that you will become a father? . 0 - 2 0 % 2 1 - 4 0 % 4 1 - 6 0 % 6 1 - 8 0 % 8 1 - 1 0 0 % Part D: to be filled out by fathers only. 1) The items (a through i) below ask for information about each of your children for up to seven children (#1 through #7). Please enter information as directed in the question. QUESTIONS a. Please indicate age in years b. Please indicate sex "m"=male, "f"=female c. If the child is your bioloqical child please check d. If the child is adopted please check e. If the child is a step-child please check f. If the child is a foster child please check g. If the child is living with you please check h. If the child is living elsewhere please check i. If the child is special needs and requires special attention please check child #1 child #2 child #3 child #4 child #5 child #6 child #7 136 2) If you have fathered any child(ren) not living with you, how often did you see the child(ren) during the past year? not applicable daily at least once a week at least once a month at least once a year less than once a year 3) In the past year did your child(ren) living with you receive child care on a regular basis so that you could do one of the following? (Exclude care provided by your spouse). not applicable work at a job study do volunteer work provide care to another family member do something else, please specify 4) If the child(ren) living with you receive care on a regular basis, approximately how many hours per week does this entail? less than 5 hours 21-25 hours 5-10 hours 26-30 hours 11-15 hours 31-35 hours 16-20 hours 36-40 hours more than 40 hours 137 SECTION (2) To be filled out by fathers only This questionnaire will ask you a few questions about how you and your spouse arrange for the care of your child(ren). Please take a few moments to consider each question and answer the question as directed by the bold type. 1) How involved are you in caring Check the most appropriate category for your children? a. very involved b. involved c. neutral d. uninvolved e. very uninvolved 2) Not counting the hours your youngster is in a school or center, with a sittei or asleep for the night, what percentage of the remaining time are you the child's prime caregiver? %. Your spouse? %. (By prime caregiver is meant the person who must be available to attend to the child's needs.) Estimate a percentage for each you and your spouse. 3)This question consists of two parts in each of the two columns divided by the bold line. TASKS a. Feeding the child b. Having sole respon-sibility for the child ren c. Disciplining the children d. Setting limits for the children's behavior e. Helping children with personal problems f. Bathing and dressing the children g. Putting the children to bed h. Helping the children to learn a) How frequently are the following parenting tasks done in your family? Check one of the following Freq. Sometimes In f req . b) What percentage of these tasks are done by: For each category estimate amount Self Spouse Other 138 4) Who in your family generally makes decisions about the following, and how frequently? For each question please check one box. DECISIONS 1. When children should be disciplined 2. When children are old enough to try new things father always father more than spouse father & spouse equally Spouse more than fa ther spouse always 5) How available are you to the children? For each question please check one box ITEMS a. away from home and children weeks and months at a time. b. away from home days at a time c. away from home on weekends d. out in the evening at least 4 nights per week e. out in the evening at least 2 nights per week f. misses supper with children at least 2 nights per week g. has breakfast during the week with children and family h. home during the week for lunch i. home afternoons when children come home from school j . home all day during the week with children and family Frequently Sometimes Inf requent ly SECTION (3) To be filled out by all participants The following statements inquire about your thoughts and feelings in a variety of situations. For each item, indicate how well it describes you by choosing the appropriate letter on the scale at the top of the page: A, B, C, D, or E. When you have decided on your answer, fill in the letter in the answer space following the item. READ EACH ITEM CAREFULLY BEFORE RESPONDING. Answer as honestly and as accurately as you can. Thank-you. ANSWER SCALE A B C D E Does not describe Describes me me well very well ITEM 1. I daydream and fantasize, with some regularity, about things that might happen to me. 2. I often have tender, concerned feelings for people less fortunate than me. 3. I sometimes find it difficult to see things from the "other guy's" point of view. 4. Sometimes I don't feel very sorry for other people when they are having problems. 5. I really get involved with the feelings of the characters in a novel. 6. In emergency situations, I feel apprehensive and ill-at-ease. 7. I am usually objective when I watch a movie or play and I don't often get completely caught up in it. 8. I try to look at everybody's side of a disagreement before I make a decision. 9. When I see someone being taken advantage of, I feel kind of protective towards them. 10. I sometimes feel helpless when I am in the middle of a very emotional situation. 1 1 . 1 sometimes try to understand my friends better by imagining how things look from their perspective. 12. Becoming extremely involved in a good book or movie is somewhat rare for me. 1 3. When I see someone get hurt, I tend to remain calm. 1 4. Other people's misfortunes do not usually disturb me a great deal. 15. If I'm sure I'm right about something, I don't waste much time listening to other people's arguments. 1 6. After seeing a play or movie, I have felt as though I were one of the characters. 1 7. Being in a tense emotional situation scares me. 140 ANSWER SCALE A B C D E Does not describe Describes me me well very well 18. When I see someone being treated unfairly, I sometimes don't feel very much pity for them. 19. I am usually pretty effective in dealing with emergencies. 20. I am often quite touched by things that I see happen. 2 1 . I believe that there are two sides to every question and try to look at them both. 22. I would describe myself as a pretty soft-hearted person. 23. When I watch a good movie, I can very easily put myself in the place of a leading character. 24. I tend to lose control during emergencies. 25. When I'm upset at someone, I usually try to "put myself in his shoes" for a while. 26. When I am reading an interesting story or novel, I imagine how I would feel if the events in the story were happening to me. 27. When I see someone who badly needs help in an emergency, I go to pieces. 28. Before criticizing somebody, I try to imagine how I would feel if I were in their place. 141 Section (4): To be filled out by all participants This questionnaire assesses how you feel about yourself. Please circle the number for each question that represents what you feel is true. Move quickly through the questions without dwelling too long on an answer. 1. I feel I am a person of worth, at least on an equal basis with others. 1 2 3 Strongly Agree Neutral Agree feel that I have a number of good qualities. 1 2 3 Strongly Agree Neutral Agree 4 Disagree 4 Disagree 5 Strongly Disagree 5 Strongly Disagree 3. All in all, I am inclined to feel that I am a failure. 1 Strongly Agree 2 Agree 3 Neutral 4 Disagree 5 Strongly Disagree 4. I am able to do things as well as most other people. 1 Strongly Agree 2 Agree 3 Neutral 4 Disagree 5 Strongly Disagree 5. I feel I do not have much to be proud of. 1 Strongly Agree 2 Agree 3 Neutral 4 Disagree 5 Strongly Disagree 6. I take a positive attitude toward myself. 1 Strongly Agree 2 Agree 3 Neutral 4 Disagree 5 Strongly Disagree 7. On the whole, I am satisfied with myself. 1 Strongly Agree 2 Agree 3 Neutral 4 Disagree 5 Strongly Disagree 142 8. I wish I could have more respect for myself. 1 Strongly Agree 2 Agree 3 Neutral 4 Disagree 5 Strongly Disagree 9. I certainly feel useless at times. 1 Strongly Agree 2 Agree 3 Neutral 4 Disagree 5 Strongly Disagree 10. At times I think I am no good at all. 1 Strongly Agree 2 Agree 3 Neutral 4 Disagree 5 Strongly Disagree 143 Appendix C This appendix includes contains a series of tables displaying descriptive results of the research. Included are full tabulation of demographic information that was used for any analysis in this research. 144 Table C.l Tabulation of Ethnicity and Religion for Subjects and Partners Ethnicity Francophone Chinese Japanese Native Indian East Indian Eastern European Greek British Italian German American don't know Canadian Dutch Israeli Mexican Brazilian Iranian African Malay Russian Singaporean Portuguese Swedish Subject 3 5 2 1 1 4 0 18 0 3 1 0 4 3 1 1 1 1 1 1 1 1 1 0 Partner 3 4 3 0 1 5 1 11 2 6 0 2 4 1 1 1 2 1 1 1 0 0 0 2 Religion don't know no religion Roman Catholic United Church Anglican Presbyterian Lutheran Baptist Eastern Orthodox Jewish Christian Natural World Agnostic Protestant Native Traditional Islam Buddhist Baha'i Health Personal Spiritualization Unitarian Mennonite Greek Orthodox Subject na 22 6 1 5 0 2 0 2 2 4 1 1 1 1 1 1 1 1 1 0 0 0 Partner 1 16 8 1 8 0 1 2 1 0 4 1 1 1 1 1 1 1 0 0 1 1 1 145 T a b l e C.2 Hours of Work Subiects and Subiects Fath 7 6 2 0 3 1 4 3 10 0 | Non 2 0 1 1 1 0 2 1 7 0 : cer ' Their Week and Partner Partners Fath 1 7 6 5 2 2 1 1 10 0 | Non 2 0 1 0 0 1 2 9 0 0 Percentage Done s Hrs of Work na 0-5 5-10 11-15 16-20 21-25 26-30 31-35 36-40 over 40 at Subiects Fath 1 0 30 1 3 0 0 Non 0 12 0 2 1 0 Home for Both Partners Fath | 0 25 1 3 4 0 Non 13 1 0 0 1 0 % done at home na 0-20 21-40 41-60 61-80 81-100 146 T a b l e C.3 Hours their of Studv and Partners Subiects Fath 0 10 3 2 2 0 2 2 5 9 | Non 7 0 1 1 0 3 0 1 0 3 Percentage Done Partners Fath 1 15 7 1 2 1 1 2 2 4 1 Non 5 3 3 0 0 0 0 0 2 3 Hrs of Studv na 0-5 6-10 11-15 16-20 21-25 26-30 31-35 36-40 over 40 at Home Subiec Fath 1 0 7 4 13 1 5 for :ts Non 0 4 1 2 4 4 Both Subiects Partners Fath 1 18 2 6 1 3 | Non 0 4 1 2 4 4 and % done at home na 0-20 21-40 41-60 61-80 81-100 147 Table C.4 Table of Frecaiencies for Personal Income Personal Income in Thousands Entire sample 11 10 7 11 13 Fathers 8 7 5 6 10 Non-fathers 3 3 2 5 3 less than $10 $10 - $20 $20 - $30 $30 - $40 $40 - $50 Table C.5 Table of Frecaiencies for Household Income Household Income in Thousands less than $10 $10 - $20 $20 - $30 $30 - $40 $40 - $50 over $50 Entire Sample 3 8 5 10 9 17 Fathers Non-Fathers 2 6 5 6 8 9 1 2 0 4 1 8 148 Table C.6 Frecaiencv Table for Main Main Activity Act Working at job or business looking for work student keeping house caring for children other ivitv of Subj ec 23 1 26 0 2 0 Subiects :ts and Their Partners Partners 23 1 16 1 10 1 Table C.7 Frecaiencv Table for Age and Sex of Children Age and Sex of Child n n n mean age male female total age range 1st Child 2nd Child 3rd Child 16 19 36 16 5.9 1-14 6.0 1-11 4.4 0-8 149 Table C.8 Frequency Table for Years of Schooling for Fathers, Non-fathers, and Their Spouses Fathers Subiects Partners 1 1 9 5 14 6 1 3 17 6 5 3 1 Years School 1-12 13-14 15-16 17-18 19-20 21+ don't of ing know Non-fathers Subiects Partners 1 2 3 6 2 2 1 2 5 2 5 1 150 Appendix D This appendix includes contains a series of tables displaying information important to this research, but not crucial to relaying salient analytic results of the research. The results of statistical analysis not essential for hypotheses or descriptive results, or that which would be too unwieldy for the main text body are included. The following tables include the use of variable names to use space more efficiently. Below is a table indicating the what these variable names refer to. PICLASS AGE(1,2,3) AGE MALEN STU PSTU WORK PWORK PINC HINC SSTU PSSTU SWORK SPWORK Father Involvement Level. Age of first, second, and third child Age of subject Length of Relationship Study Hours per Week Percentage of Study Hours Done at Home Hours of Work per Week Percentage of Work Done at Home Personal Income Household Income Spouse's Study Hours per Week Percentage of Spouse's Study Done at Home Spouse's Work Hours per Week Percentage of Spouse's Work Done at Home 151 Table D.l Spearman Coefficients for Independent and Dependent Variables with Demographic Variables (Fathers Only) YSCH STU PSTU WORK PWORK PINC HINC SYSCH PI .1066 (n 36) sig .268 .1480 (n 35) sig .198 .0691 (n 27) sig .366 -.2402 (n 36) sig .079 .2064 (n 34) sig .121 -.0909 (n 36) sig .299 .0891 (n 36) sig .303 .0772 (n 36) sig .327 PT -.0143 (n 36) sig .467 .0860 (n 35) sig .312 .1568 (n 27) sig .217 -.0680 (n 36) sig .347 .0295 (n 34) sig .434 -.1344 n( 36) sig .217 .0994 n( 36) sig .282 .2803 (n 36) sig .049 EC -.0265 (n 36) sig .439 .0902 (n 35) sig .303 .2076 (n 27) sig .149 .0314 (n 36) sig .428 -.2338 (n 34) sig .092 .1671 (n 36) sig .165 .1197 (n 36) sig .243 .2317 (n 36) sig .087 SE -.1820 (n 36) sig .144 -.2641 (n 35) sig .063 -.1169 (n 27) sig .281 .0436 (n 36) sig .400 .1184 (n 34) sig .253 .0394 (n 36) sig .410 .2442 (n 36) sig .076 -.0760 (n 36) sig .330 RINV .1162 (n 36) sig .250 -.0872 (n 35) sig .309 .1213 (n 27) sig .273 -.0012 (n 36) sig .497 ** .3070 (n 34) sig .039 .1574 (n 36) sig .180 .2390 (n 36) sig .080 -.0760 (n 36) sig .330 SWORK SPWORK SSTU PSSTU PI .1400 (n 35) sig .211 .0349 (n 33) sig .424 -.0531 (n 36) sig .379 .1710 (n 31) sig .179 PT .2200 (n 35) sig .102 -.1422 (n 33) sig .215 .0380 (n 36) sig .413 .0947 (n 31) sig .306 EC .0704 (n 35) sig .344 -.0599 (n 33) sig .370 .0261 (n 36) sig .440 -.1140 (n 31) sig .271 SE .1942 (n 35) sig .132 .1683 (n 33) sig .175 -.0786 (n 36) sig .324 .0435 (n 31) sig .408 RINV .0664 (n 35) sig .352 -.1023 (n 33) sig .286 -.2986 (n 36) sig .038 -.0272 (n 31) sig .442 *E ^ -05 152 T a b l e D.2 R e s u l t s of One-Wav A n a l v s i s of V a r i a n c e Comparing t h e Average Number of C h i l d r e n i n BETWEEN WITHIN TOTAL GROUP LowPI H i P I MedPI TOTAL SOURCE f GROUPS GROUPS COUNT 12 12 12 36 D . F . 2 33 35 MEAN AGE 1 .5000 1 .6667 1 .5833 1 .5833 Each of t h e SUM OF SQUARES .1667 2 2 . 5 8 3 3 2 2 . 7 5 0 0 P a t e r n a l MEAN SQUARES .0833 .6843 STANDARD STANDARD DEVIATION ERROR MINIMUM .7977 .2303 . 8876 .2562 .7930 . 2289 . 8062 .1344 1 .0000 .0000 1 .0000 .0000 Involvement Groups F F RATIO PROB. . 1 2 1 8 . 8857 MAXIMUM 95 PCT CONF 3 . 0 0 0 0 . 9932 TO 3 . 0 0 0 0 1 .1027 TO 3 . 0 0 0 0 1 .0795 TO 3 . 0 0 0 0 1 . 3 1 0 5 TO INT FOR MI 2 . 0 0 6 ! 2 . 2 3 0 1 2 . 0 8 7 : 1 . 8 5 6 : 153 Table D.3 Kruskal-Wallis One-wav Anova for Years of Schooling and Non-Father, and High, Medium, and Low Involvement Groups Mean Rank Cases 22 25 31. 28 .00 .92 .38 .21 16 12 12 12 Non-fathers Low PI Medium PI High PI 52 Total Corrected for ties Cases Chi-Square Significance Chi-Square Significance 52 2.8228 .4198 2.9852 .3939 Table D.4 Kruskal-Wallis One-way Anova for Religiosity bv Non-father. High-Medium, and Low Involvement Groups Mean Rank 26.69 29.25 23.75 23.91 Cases 16 12 12 11 Non-fathers Low PI Medium PI High PI 51 Total Corrected for ties Cases Chi-Square Significance Chi-Square Significance 51 1.1002 .7770 1.2664 .7371 154 T a b l e D.5 K r u s k a l - W a l l i s One-way Anova fo r Study Time p e r Week bv Non-f a t h e r . High, Medium, and Low Involvement Groups Mean Rank 22.47 26.33 26.91 29.54 Cases 16 12 11 12 Non-fathers Low PI Medium PI High PI 5 1 T o t a l C a s e s 51 C h i - S q u a r e 1.6310 S i g n i f i c a n c e .6524 C o r r e c t e d f o r t i e s C h i - S q u a r e S i g n i f i c a n c e 1.7210 .6323 T a b l e D.6 Kruskal-Wallis One-way Anova for Percentage of Study Time at Done at Home bv Non-father. High. Medium, and Low Involvement Groups . Rank 23.77 16.81 17.45 18.94 Cases 11 8 10 9 Non-fathers Low PI Medium PI High PI 38 Total Cases Chi-Square Significance 38 2.4567 .4832 Corrected for ties Chi-Square Significance 2.7067 .4391 155 Table D.7 Kruskal-Wallis One-wav Anova for Hours of Work per Week by Non-father. High, Medium, and Low Involvement Groups Mean Rank 30.40 28.04 23.42 21.04 Cases 15 12 12 12 Non-fathers Low PI Medium PI High PI 51 Total Cases Chi-Square Significance 51 3.2377 .3564 Corrected for ties Chi-Square Significance 3.3959 .3345 Table D.8 K r u s k a l - W a l l i s One-wav Anova f o r P e r c e n t a g e of Work done a t Home bv N o n - f a t h e r . High. Medium, and Low Involvement Groups Mean Rank 26.60 23.54 23.73 25.68 Cases 15 12 11 11 Non-fathers Low PI Medium PI High PI 49 Tota l Cases 49 Chi-Square .4254 S ign i f i cance .9349 Corrected for t i e s Chi-Square S ign i f i cance 1.1518 .7646 156 Table D.9 Kruskal-Wallis One-way Anova for Personal Income by Non-father, High, Medium, and Low Involvement Groups Mean Rank 26.59 26.46 29.67 23.25 Cases 16 12 12 12 Non-fathers Low PI Medium PI High PI 52 Total Cases Chi-Square Significance 52 1.0765 .7827 Corrected for ties Chi-Square Significance 1.1258 .7709 Table D.10 K r u s k a l - W a l l i s One-way Anova f o r H o u s e h o l d Income b y N o n - f a t h e r , H i g h , Medium, a n d Low I n v o l v e m e n t G r o u p s Mean Rank 29.00 22.54 27.21 26.42 Cases 16 12 12 12 Non-fathers Low PI Medium PI High PI 52 T o t a l Cases 52 C h i - S q u a r e 1.2807 S i g n i f i c a n c e .7337 C o r r e c t e d f o r t i e s C h i - S q u a r e S i g n i f i c a n c e 1.3485 .7176 157 Table D.ll Kruskal-Wallis One-way Anova for Partner's Years of Schooling by Non-father. High. Medium, and Low Involvement Groups Mean Re 27, 21 29, 27. ink .06 .50 .54 .71 Ca ses 16 12 12 12 Non-fathers Low PI Medium PI High PI 52 Total Cases Chi-Square Significance 52 1.8880 .5960 Corrected for ties Chi-Square Significance 2.0691 .5582 Table D.12 Kruskal-Wallis One-wav Anova for Partner's Religiosity by Non-father, High. Medium, and Low Involvement Groups Mean Rank 30.38 26.88 20.63 24.55 Cases 16 12 12 11 Non-fathers Low PI Medium PI High PI 51 Total Cases Chi-Square Significance 51 3.1013 .3763 Corrected for ties Chi-Square Significance 3.4700 .3247 158 Table D.13 Kruskal-Wallis One-wav Anova for Partner's Study Time per Week by Non-father. High, Medium, and Low Involvement Groups . Rank 29.47 25.25 26.38 23.92 Cases 16 12 12 12 Non-fathers Low PI Medium PI High PI 52 Total Corrected for ties Cases Chi-Square Significance Chi-Square Significance 52 1.0452 .7903 1.1204 .7722 Table D.14 Kruskal-Wallis One-wav Anova for Partner's Percentage of Study Time Done at Home bv Non-father. High, Medium, and Low Involvement Groups Mean Rank 23.77 16.81 17.45 18.94 Cases 11 8 10 9 Non-fathers Low PI Medium PI High PI 38 Total Corrected for ties Cases Chi-Square Significance Chi-Square Significance 38 2.4567 .4832 2.7067 .4391 159 T a b l e D.15 K r u s k a l - W a l l i s One-way Anova f o r P a r t n e r ' s Work Hours p e r Week by N o n - f a t h e r . High, Medium, and Low Involvement Groups . Rank 32.53 18.13 25.23 24.33 Ca ses 15 12 11 12 Non-fathers Low PI Medium PI High PI 50 Tota l Cases 50 Chi-Square 6.6440 S ign i f i cance .0842 Corrected for t i e s Chi-Square S ign i f i cance 7.0998 .0688 T a b l e D .16 Kruskal-Wallis One-wav Anova for Partner's Work Hours per Week by Non-father. High, Medium, and Low Involvement Groups L Rank 22.60 24.80 23.59 27.46 Ca ses 15 10 11 12 Non-fathers Low PI Medium PI High PI 48 Total Cases Chi-Square Significance 48 .8631 .8343 Corrected for ties Chi-Square Significance 1.7172 .6331 160 

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:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0054031/manifest

Comment

Related Items