UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The associations between childhood factors and adult attachment : a study of gay men Landolt, Monica A. 1999

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

Item Metadata

Download

Media
831-ubc_1999-463702.pdf [ 5.77MB ]
Metadata
JSON: 831-1.0089285.json
JSON-LD: 831-1.0089285-ld.json
RDF/XML (Pretty): 831-1.0089285-rdf.xml
RDF/JSON: 831-1.0089285-rdf.json
Turtle: 831-1.0089285-turtle.txt
N-Triples: 831-1.0089285-rdf-ntriples.txt
Original Record: 831-1.0089285-source.json
Full Text
831-1.0089285-fulltext.txt
Citation
831-1.0089285.ris

Full Text

THE ASSOCIATIONS BETWEEN CHILDHOOD FACTORS AND ADULT ATTACHMENT: A STUDY OF GAY MEN by Monica A. Landolt B.A. (Honours), Queen's University at Kingston, 1993 M.A., The University of British Columbia, 1995 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Department of Psychology) We accept this thesis as'conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA September 1999 © Monica A. Landolt, 1999 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. D e p a r t m e n t o f D The University of British Columbia Vancouver, Canada Date -6 (2/88) ABSTRACT There are a number of childhood factors reported to be associated with gay male sexual orientation including childhood gender non-conformity and socialization factors such as paternal and peer rejection. Attachment theory suggests that poor quality childhood relationships may be related to anxiety and avoidance attachment dimensions in adulthood. The purpose of this study was to explore the association between these childhood factors and anxiety and avoidance in gay men's close relationships. A community sample of 192 self-identified gay men completed questionnaires and a 2 hour attachment interview. No major findings related to the avoidance dimension were significant. In terms of attachment anxiety, results were partially consistent with attachment theory: paternal and peer, but not maternal, rejection independently predicted anxiety. Quality of peer relationships largely mediated the association between parental rejection and anxiety. In addition, quality of peer relationships mediated the association between gender non-conformity and anxiety. Good quality relationships in one domain did not compensate for poor quality relationships in another domain. The importance of fathers and peers to gay men's current relationship functioning is discussed. ii TABLE OF CONTENTS Abstract ii Table of Contents iii List of Tables iv List of Figures v List of Appendices vi Acknowledgements vii Introduction 1 Method 26 Results 37 Discussion 51 References 75 Tables 89 Figures 93 Appendices 100 iii LIST OF TABLES Table 1 Demographic Characteristics of the Follow-up Participants 89 Table 2 Mean Scores for Self-report and Interview Measures 90 Table 3 Inter-correlations among Key Variables 91 Table 4 Correlations between Health-related and Key Variables 92 iv LIST OF FIGURES Figure 1 Two-dimensional Model of Attachment 93 Figure 2 Path Model of Parental and Peer Rejection to Anxiety 94 Figure 3 Path Model of Femininity to Rejection to Anxiety 95 Figure 4 Path Model of Masculinity to Rejection and Warmth to Anxiety 96 Figure 5 Alternate Path Model: Relationship Quality to Femininity to Anxiety 97 Figure 6 Alternate Path Model: Relationship Quality to Masculinity to Anxiety 98 Figure 7 Interaction of Paternal and Peer Rejection on Anxiety 99 v LIST OF APPENDICES Appendix 1 Recollections of Early Child Rearing Scale (EMBU) 100 Appendix 2 Modified Boyhood Gender Non-conformity Scale (BG(N)CS) 102 Appendix 3 Epstein Peer Scale (E-PEER) 103 Appendix 4 Inventory of Peer Attachment (IP A) 104 Appendix 5 Internalized Homophobia Scale (IHS) 106 Appendix 6 History of Attachment Interview (HAI) 107 Appendix 7 History of Attachment Interview—Coding Form 119 Appendix 8 Suggestions for the Reduction of Volunteer Bias 123 vi ACKNOWLEDGEMENTS I would like to thank the individuals who graciously participated in this project. This study would not have happened without their time, dedication, and honesty. I must also acknowledge Doug Oram, my colleague in the West End Relationship Project. Doug brought unfailing support, drive, effort, energy, and insights to both the project and me. I am very lucky to have worked with someone who is as kind as Doug is. I would like to thank Mariana Brussoni for her efforts in data collection and coding but also because she has been a wonderful friend and supporter of me. I must thank my supervisory committee: Don Dutton, Dan Perlman and Kim Bartholomew. There are few graduate students who have been as fortunate as I am to have such support. My committee allowed me freedom to explore issues and ideas that I found interesting and relevant. They were more than generous with their time and knowledge. I have learned so much from these three individuals and I am truly grateful. I would like to thank my friends who have been there for me, and with me, over the years including: Christina Bews, Ardis Krueger, Francine Mahoney, Michael Pollex, Debra Sauve, Christine Stager, and Cynthia van Ginkel. I would also like to thank my family for their unfailing faith and love, especially, Jack and Gwen Landolt, Lydia and John Ridd, Carol and Peter Butzelaar, and Margaret Philip. Finally, I would like to thank my inspiration during the final push, Joseph Peschisolido. vii Childhood Factors and Attachment 1 THE ASSOCIATIONS BETWEEN CHILDHOOD FACTORS AND ADULT ATTACHMENT: A STUDY OF GAY MEN Interpersonal relationships are an important area of study under the Social Psychology rubric. This broad research domain includes diverse topics and many possible relationship configurations. For example, this area has covered research on parent/child relationships, sibling relationships, peer relationships, work place and professional relationships, and romantic relationships to name a few. One of the most wide spread theoretical perspectives that underscore the study of interpersonal relationships has been the attachment perspective (Simpson, Rholes, & Phillips, 1996). The State of Same-Sex Relationship Research The social sciences, of which social psychology and its interpersonal relationships sub-category is included, has been negligent in the realization that gay relationships are a worthy area of study. As noted above, whereas so many diverse lines of inquiry have been generated in the field of interpersonal relationships, the populations of study have primarily been heterosexual to the exclusion of same-sex populations. This is despite the fact that evidence of gay coupling has long been established, first appearing in early classical writings (Boswell, 1980). Indeed, same-sex marriages were legally recognized until 342 A.D. (Solomon, 1986). The latter point indicates that in antiquity, gay relationships must have been acknowledged as being both viable and relatively prevalent, to justify legislation. Dynes (1987) thoroughly reviewed the extant gay and lesbian literature. He found that of 5,000 citations, only 36 looked at gay and lesbian coupling. Research including that from the earlier part of this decade does not bode any better. Allen et al., (1995) showed that between 1980 and 1993, less than 1% of articles in relevant psychological journals focused on issues related to lesbian and gay men. Childhood Factors and Attachment 2 Moreover, the methodology used in research focusing on lesbian and gay issues has lagged behind that found in the general interpersonal relationships field (Deenen, Gijs, & van Naerssen, 1995). In the 1950s, research with gay men relied on clinical samples (i.e., samples of men in treatment to change their sexual orientation), or samples of men imprisoned because of their sexual behaviour. Sell and Petrulio (1996) have claimed that methods of selecting representative samples of gay men have not yet significantly improved. Common recruitment settings include gay and lesbian organizations, clinical settings (both AIDS and non-AIDs related), advertisements (e.g., newspapers, magazines, or newsletters), and educational settings. These authors concluded that such recruitment methods are not necessarily comparable and that it is impossible, without caveats, to generalize using samples obtained from these settings. In the final analysis, Sell and Petrulio (1996) recommended use of probability or random sampling techniques as a means to obtain more representative samples. Deenen et al., (1995) have reported that in over 35 years of research on gay relationships there is no evidence of improvement in methodology. Clement (1990, cf. Deenen et al., 1995) has claimed that researchers restrict their attention to easily obtained populations and to questions of "how much" and "how often". Deenen et al., (1995) have stated that if variables relevant to gay men were correlated with other parametres, this may contribute to the understanding of gay relationships and to theory building, which are both lacking. In their recommendations, they suggested that gay relationship studies might be improved by broadening theory and research topics and by connecting with the general field of interpersonal relationships research. This thesis can clearly be categorized in the interpersonal relationship domain of inquiry. It looks at how recollected quality of relationships with parents and peers might predict adult attachment outcomes for gay men. The empirical exploration of gay male relationships to the exclusion of other populations is both necessary and legitimate. The goal of such an endeavor is Childhood Factors and Attachment 3 to rectify the current gap in the literature. To date, the research on gay male relationships is mostly comparative in nature. That is, most interpersonal relationship research that has included gay men in its empirical focus has contrasted gay relationships with heterosexual relationships. One can argue that this contrast effort is inherently biased. It suggests that the study of gay relationships requires a normative control group (i.e., heterosexuals), but this is in no way the case. The empirical focus outlined below only includes gay men. Thesis Introduction In this thesis, I explore the associations between childhood factors and anxiety and avoidance in gay men's close adult relationships. I begin with an overview of the principles of attachment theory and introduce two dimensions underlying individual differences in adult attachment: attachment anxiety and avoidance of intimacy. This is followed by a review of the literature on sexual orientation differences in recollected quality of father and son and mother and son relationships. I then discuss gender non-conformity, defined as the relative presence of feminine traits and/or relative absence of masculine traits during boyhood, which many gay men report as having experienced in childhood. I suggest that gender non-conformity appearing as early as age 2 may be associated with paternal, but not necessarily maternal, rejection. The review of the literature is extended to include recollected quality of peer relationships, as these relationships can serve an attachment function. The association between gender non-conformity and quality of peer relationships is also explored. Stigma is introduced as a possible mechanism underlying the associations between gender non-conformity and paternal and peer rejection. Bringing together the parental and peer domains, I discuss how good childhood relationships in one domain may be able to compensate for poor relationships in another domain. In the case of gay men, one might expect that good relationships with mothers could compensate Childhood Factors and Attachment 4 for rejecting relationships with fathers or peers. I conclude with a summary of the research premise and outline five primary research questions and one secondary question that are examined empirically in the thesis. Principles of Attachment Theory Attachment theory proposes that parenting quality has a powerful influence on child development and subsequent adult attachments (Bowlby, 1988; Collins & Read, 1990; Hazan & Shaver, 1987). This point has been empirically validated in a number of studies. Notably, Ainsworth, Blehar, Waters, and Wall (1978) found that quality of parent-child interactions (parental sensitivity and responsiveness) was associated with differences in the quality of attachment shown by young children toward their caregivers. They identified three distinct attachment patterns in 9-to-18 month old children: secure, anxious-resistant, and avoidant. Hazan and Shaver (1987) have subsequently shown that these three patterns occur with approximately equivalent frequency in adult romantic relationships. Internal working models, which develop as a result of daily parent and child interactions, are considered to be the mechanisms underlying attachment patterns and their continuity across the life-span (Bowlby, 1979). As such, internal working models are understood to be mental constructs containing beliefs and expectations about the self as an individual worthy of care and attention (the Self-Model), as well as beliefs and expectations about the availability and trustworthiness of significant others (the Other-Model) (Bowlby, 1973). Building on the work of Ainsworth et al., (1978) and Hazan and Shaver (1987), Bartholomew (Bartholomew, 1990; Bartholomew & Horowitz, 1991) dichotomized the Self-Model and Other-Model to better capture Bowlby's conceptualization of attachment models (see Figure 1). Her work yielded four-attachment patterns including secure (positive Self-Model, positive Other-Model), dismissing Childhood Factors and Attachment 5 (positive Self-Model, negative Other-Model), preoccupied (negative Self-Model, positive Other-Model) and fearful (negative Self-Model, negative Other-Model). A complimentary way of conceptualizing the Self- and Other-Models is in terms of the dimensions of anxiety and avoidance (Bartholomew, 1990; Griffin & Bartholomew, 1994). The anxiety dimension refers to an emotional response expressed in the context of an attachment relationship. That is, it refers to the degree of sensitivity to threats of relationship security and potential loss and rejection experienced by individuals. Individuals in relationships can have an internalized sense of relationship confidence and well-placed faith in their partners (low anxiety), or they can spend excessive amounts of time monitoring their relationships, being hyper-vigilant to rejection cues (high anxiety). The avoidance (versus closeness) dimension describes the behavioural strategy that individuals engage in to regulate anxiety generated in the context of an attachment relationship. Individuals who feel threatened can either seek out and make contact with their attachment figures to reduce felt anxiety (closeness or high approach), or they can withdraw and avoid closeness in relationships as a means of reducing anxiety (high avoidance). Previous research has shown that the two dimensions captured individual differences in both childhood and adult attachment (Brennan, 1998; Griffin & Bartholomew, 1994) and that they were the most parsimonious way to explain individual differences in adult attachment (Fraley & Waller, 1998). Given this, the anxiety and avoidance attachment dimensions were used through out the remainder of the thesis (unless otherwise specified). Internal working models are theorized to be cognitive/emotional/behavioural systems through which information regarding new relationships is filtered (Collins & Read, 1994). Continuity of attachment is achieved because new relationships are assimilated, and ambiguous situations are interpreted, in accordance with previously learned scripts (Feeney & Noller, 1996; Childhood Factors and Attachment 6 Sroufe & Fleeson, 1986). The end result is that similarities in attachment patterns are found in and across parent/child relationships, friendships, and romantic relationships. There is a fair amount of evidence attesting to the consistency of attachment patterns across various relationships. In a longitudinal study, Shulman, Eliker, and Sroufe (1994) found that attachment to parents at age 12 and 18 months was significantly associated with subsequent peer competence in pre-adolescence. Hunter and Youniss (1982) found that adolescents judged their peers to be similar to their parents in terms of important relationship features. Armsden and Greenberg (1987) and Bartholomew and Horowitz (1991) found that family attachment ratings were moderately correlated with peer attachment ratings in college samples. Collins and Read (1990) found a positive association between recollections of parent's caregiving style and attachment in dating couples as did Hazan and Shaver (1987) in their more diverse community sample. Peer Influence and Attachment Evidence suggests that, in addition to parental relationships, peer relationships have an independent influence on current attachment strategies. Epstein (1983), for example, showed that while maternal acceptance was important for lovability, an important component of self-esteem, peer acceptance was more important for all other components of self-esteem. As well, Bartholomew and Horowitz (1991) found that peer attachment uniquely contributed to the prediction of interpersonal problems over and above that of parental attachment (i.e., in the case of dominance). Weiss (1982) and Ainsworth (1989) suggested that peers eventually replace parents as primary attachment figures. Hazan and Zeifman (1994) found that by late adolescence, peers (including romantic partners) become full-blown attachment figures. Trinke and Bartholomew (1997) reported that for a sample of young adults, attachment relationships were ranked in the Childhood Factors and Attachment 7 following order of importance: partners (if the participant had a partner), mother, fathers, siblings, and peers. Fraley and Davis (1997) also reported that young adults were predominantly attached to parents, but began to use peers as attachment figures. These authors suggested that internalized working models of self and other facilitate the transfer of attachment from parents to peers. This would be congruent with Bowlby's (1979) thinking as he asserted that internal working models (developed as a result of continuous interactions with parents) predispose individuals to replicate previously established patterns in new relationships. Further evidence comes from Elicker, Englund, and Sroufe (1992) who demonstrated that quality of caregiver and infant relationship was an important predictor of peer relationships and peer competence. However, they also showed that insecure pre-adolescents were just as likely as secure pre-adolescents to form friendships (although relationship quality was compromised in the former group). Cooper and Cooper (1992) asserted that while adolescents may seek out friendships to compensate for difficulties in the family, they may not have the skills necessary to sustain beneficial peer relationships. It must be noted that parents influence the development and quality of peer relationships. To some degree, this is because parents manage the social lives of their children both directly, by facilitating access to peers, and indirectly, by imparting social skills, values and expectations that are then used in social interactions with peers (Hartup, 1979). However, in moving from childhood to adolescence, the importance of peers as agents of socialization grows. This is partially a function of the amount of time spent interacting with peers, which is often inversely related to time spent with parents (Hartup, 1983). Whereas both parents and peers are important sources of socialization, over the course of development, peer norms may come to validate or Childhood Factors and Attachment 8 filter parent norms (Siman, 1977). Therefore, it may be the case that peer relationships mediate the associations between parental relationships and subsequent adult attachment. Attachment and Gay Men To date, there has been little work published on attachment relationships among gay men. A notable exception is Mohr's (1999) theoretical paper on attachment and gay men. His unique contribution was that that he used attachment theory to help understand same-sex relationships from an evolutionary perspective. There are also a limited number of published empirical studies. Some of these studies focus on "dyadic attachment" which is actually something quite separate from any construct conceptualized or grounded in attachment theory (Deenen, Gijs, & van Naerssen, 1995). Dyadic attachment is used as one measure of relationship quality (along with autonomy and equality) and specifically refers to the value placed on a given relationship (e.g., Kurdek, 1995; Peplau & Cochran, 1982). An example of research grounded in attachment theory is that by Landolt and Dutton (1997) who confirmed that insecure attachment was predictive of perpetrating domestic violence in a community sample of 52 gay male couples. Their results replicated findings first observed with heterosexual men (e.g., Dutton, 1994; Dutton, Saunders, Starzomski & Bartholomew, 1994). Couples in the Landolt and Dutton (1997) study were recruited by advertisements in two local gay and lesbian papers. Only couples that had been together for at least 6 months were included. The authors of this study indicated that selection criteria and volunteer bias might have reduced the generalizability of their findings. Kurdek (1997) conducted a study looking at the mediational role of attachment in the association between neuroticism and relationship commitment. His study was based on data collected from 33 gay, 40 lesbian, and 70 heterosexual couples. Findings related to attachment in Childhood Factors and Attachment 9 this study were two-fold: 1) depression was linked to having a negative self-model (the anxiety attachment dimension) and a negative other-model (the avoidance attachment dimension), and 2) individuals in couples that were composed of members who both had positive self-models were more committed to their relationships. Even though Kurdek (1997) had data from gay, lesbian, and heterosexual couples, sexual orientation did not seem to be a variable of interest in this study. The only explicit mention of sexual orientation, indicating that gay partners had more positive other-models compared to heterosexual wives, was found in a footnote. Holtzen, Kenny, and Mahalik (1995) looked at the associations between being securely attached to mothers and fathers, disclosure of sexual orientation, and negative cognitions. Their sample included 72 gay and 41 lesbian young adults and adults recruited through the support group, Parents and Friends of Lesbians and Gays (PFLAG), or through a university gay and lesbian student support group. Results indicated that gay men and lesbians who were more securely attached to their parents were more likely to disclose their sexual orientation to their parents and had fewer dysfunctional cognitions (which co-occur with depressive symptoms). Thus, this study suggested that current relationship quality with parents was associated with mental health outcomes for gays and lesbians. Perhaps most relevant to this thesis are the following two published reports, one theoretical and one empirical. The theoretical report by Colgan (1987) closely parallels the thinking underlying this thesis. He suggested that gay men experience difficulties in intimate relationships because of developmental influences of family and peers. Specifically, negative responses from family and peers related to perceived gender non-conformity may lead to gay men's over-attachment and over-separation in relationships. Over-attachment is expressed in a number of ways: 1) by a continued search to find the right partner met by constant disappointment of not finding one who is loving enough, or loving Childhood Factors and Attachment 10 in the right way, 2) by an inability to end unsatisfying relationships, and 3) by low sexual assertiveness (Colgan, 1988). Features of over-attachment share important characteristics with the anxiety dimension (Bartholomew, 1990; Bartholomew & Horowitz, 1991). Therefore, over-attachment and anxiety can be seen as parallel constructs. Over-separation is expressed in a number of ways: 1) emphasis on physical aspects of a relationship rather than emotional aspects, 2) poor interpersonal skills, and 3) use of drugs and alcohol in conjunction with sexual behaviour. Features of over-separation share important characteristics with the avoidance dimension (Bartholomew, 1990; Bartholomew & Horowitz, 1991). Therefore, over-separation and avoidance can be seen as parallel constructs. Interestingly, then, based on clinical interests and experiences, Colgan (1987) theorized that gender non-conformity leading to rejection was predictive of anxiety and avoidance in adult attachment relationships. Colgan (1987) did not empirically test his ideas. However, his speculations form one of the major hypotheses of this thesis, although both lines of inquiry were derived independently. Equally relevant to this thesis is a study by Ridge and Feeney (1998) who looked at the associations between recollection of early parenting quality and adult attachment styles in a sample of 77 gay, 100 lesbian, 39 heterosexual male and 111 heterosexual female college students. These authors found that relative frequencies of attachment styles were similar irrespective of sexual orientation. They also found that there was no link between early parenting and adult attachment styles. There are a number of serious limitations to the Ridge and Feeney (1998) study. For example, the authors used a self-report measure of attachment that may have poor validity among gay men. In an unpublished Masters thesis, Callander (1999) tested the validity of self-report measures of attachment with gay men. She reported that problems arose with the validity of the Childhood Factors and Attachment 11 attachment patterns, related to the preoccupied and dismissing insecure patterns, when self-report measures were involved. Moreover, Ridge and Feeney (1998) used a categorical measure of attachment that required participants to indicate their most prototypical attachment style. This method of measurement is problematic because it presupposes individuals have only one attachment style. In actual fact, two individuals who report the same primary attachment style may be very different depending on their secondary or tertiary strategies (Griffin & Bartholomew, 1994). This difference and any associated sexual orientation difference would never be captured using the categorical method. Furthermore, analyses based on discrete measures of group membership have lower power than those based on continuous measures (Griffin & Bartholomew, 1994). Thus, whereas there may indeed be a link between attachment and early relationships with parents, the Ridge and Feeney (1998) study might not have the statistical power to find the effect. Another limitation of the Ridge and Feeney (1998) study is that they used a 16-item adjective checklist to measure the recollected quality of parent-child relationships. This measure was probably too brief to capture the range of experiences associated with recollected relationship quality. Moreover, it had a dichotomous choice response structure that forced participants to either only agree or disagree as to whether a given adjective described their relationships. Thus, the limited range of responses probably did not reflect the more variable nature of parent-child relationships. A particular limitation of all the aforementioned studies on attachment and gay men is that they were based on relatively small (N < 100), convenience samples. Irrespective of whether participants were recruited through advertisements, support organizations, or university clubs, non-representative samples and volunteer-bias may have reduced the generalizability of findings (Rosenthal & Rosnow, 1975). Indeed, as previously noted, these methodological inadequacies Childhood Factors and Attachment 12 are characteristic of most of the research conducted with gay men (Deenen et al., 1995; Sell & Petrulio, 1996). Notwithstanding the limited number and/or quality of published studies on attachment relationships among gay men, it is likely that consistent with findings in heterosexual samples, the quality of parenting received in childhood would have an impact on gay men's current attachment strategies. However, research on gay men's recollected relationships with fathers and mothers suggests that the relative importance of fathers and mothers may differ. A review of this literature is outlined in the next two sections. There is reason to believe that the attachment relationship hierarchy could be different for gay men. Kurdek (1988) reported that for gay men in cohabiting relationships, the most frequent social support providers, in order, were friends, partners, family, and co-workers (see also Kurdek & Schmitt, 1987). He concluded that this might be due to family members' difficulty dealing with homosexuality among their relations (Blumstein & Schwartz, 1983; McWhirter & Mattison, 1984). Thus, if peers become especially important attachment figures for gay men, it is possible that in a gay sample, peer relationships may mediate the association between the quality of childhood relationships with parents, and anxiety and avoidance attachment dimensions in adulthood. Sexual Orientation Differences in Recollection of Father and Son Relationships Many studies have been conducted comparing gay and heterosexual men's recollections of their relationships with their fathers during childhood. Interestingly, only one study, an unpublished dissertation by Greenblatt (1966, cf. Hooker, 1969), showed that both gay and heterosexual participants rated their fathers as equally good, generous, pleasant, dominant, and unprotective. However, while this study indicated that both gay and heterosexual participants Childhood Factors and Attachment 13 described their fathers as generally fine individuals, it did not measure directly the recollected quality of the father and son relationship. More common in the literature are studies that showed systematic differences in gay and heterosexual men's recollections of their relationships with their fathers. For example, both Phelan (1996) and Milic and Crowne (1986) found that gay men were more likely than heterosexual men to report that their fathers were rejecting and less loving. Apperson and McAdoo (1968) found that gay participants reported that their fathers were more critical, impatient, and rejecting, than their heterosexual counterparts. Thompson, Schwartz, McCandless, and Edwards (1973) confirmed that gay men were more likely than heterosexual men to report recollections of their fathers as being hostile and detached. Evans (1969) found that gay participants reported that they spent less time with their father, felt less accepted or respected by their father, were more aware of hating their father, and were more afraid that their fathers would physically harm them. Pillard (1990) conducted an interesting investigation of father/son relationships and sexual orientation. Unlike previous researchers, Pillard not only gathered data from gay and heterosexual participants, but he also collected data from participants' brothers. He found that heterosexual brothers were more likely to rate themselves as distant from their fathers if they came from a family with a gay brother. Paternal alcoholism was also found to be more common in the gay sample as compared to the heterosexual sample, which may help explain the higher overall ratings of paternal distance reported in families with gay sons. Pillard's general findings might suggest that it was not factors related to sexual orientation per se that were associated with negative father and son relationships. However, when gay brothers' ratings were compared to their heterosexual brothers' ratings, in 81% of cases the gay brother rated himself as more distant from the father. Childhood Factors and Attachment 14 Finally, a number of related studies using clinical samples of heterosexual and gay men have been conducted. They have confirmed that gay participants compared to heterosexual participants were more likely to report that their fathers were more rejecting and distant (e.g., Bieber & Bieber, 1979; Buhrich & McConaghy, 1978; Freund, Langevin, Steiner, & Zajac, 1974; Isay, 1989; Sipova & Brzek, 1983). Sexual Orientation Differences in Recollection of Mother and Son Relationships Whereas sexual orientation differences in reports of father and son relationships are quite consistent across studies, comparable studies focusing on mother and son relationships yield more equivocal results. For example, as part of the same study reviewed above, Greenblatt (1966, cited in Hooker, 1969) found that his gay and heterosexual samples reported that their mothers were equally good, generous, pleasant, dominant, and unprotective. Milic and Crowne (1986), however, found that gay participants were more likely than heterosexual participants to report that their mothers were rejecting. Apperson and McAdoo (1968) found that gay participants were less likely than heterosexual participants to report that their mothers were involved or restrictive in their upbringing. This finding can be contrasted with Evans' (1969) finding that gay men, as compared to heterosexuals, recalled that their mothers were less encouraging of their masculinity and were more enmeshed and role-reversed in their relationship with their sons. Thompson, Schwartz, McCandless, and Edwards (1973) also found that gay participants were more likely than heterosexual participants to report having "close binding", intimate mothers (i.e., mothers who were more enmeshed with their sons). However, the gay participants in this latter study did not report feeling any more identified with their mothers, which suggests that they did not necessarily accept the relationship dynamic imposed on them by their mothers. Childhood Factors and Attachment 15 Finally, research using clinical samples also shows widely discrepant results. Bieber and Bieber (1979) found that gay men recalled their mothers as being overly close, dominant and restrictive. Similarly, Freund, Langevin, Steiner, and Zajac (1974) found that gay men reported feeling closer to their mothers, and that their mothers were more involved during their childhood. Isay (1989) reported that he saw no differences between heterosexuals' and gay men's mother and son relationships. It should be noted, though, that both Bieber and Bieber (1979) and Isay (1989) relied on clinical observations to substantiate their claims. Finally, Sipova and Brzek (1983) found that non-effeminate gay men recalled their mothers as being less loving and less dominant when compared to heterosexuals; however, no differences were found between effeminate gay men and heterosexuals in their recollections of maternal love and dominance. In summary, previous research seems to indicate that gay men are more likely than heterosexual men to recollect having poor relationships with their fathers. Findings regarding their recollected quality of mother and son relationships are more equivocal. Gender Non-conformity Researchers have suggested that sexual orientation per se does not become manifest in gay men until the age of 19 or 20 (Harry & Devall, 1978). While most adult gay men state that their same-sex attraction started somewhere between the ages of 8 to 14 (Isay, 1990), it is not until their heterosexual peer group begins to engage in opposite sex dating that gay men fully recognize their sexual orientation. Because homosexuality only becomes apparent after childhood, it is unlikely that there could be any causal relationship between homosexuality and negative parent and child relations. However, gender identity and gender appropriate behaviour emerge between the ages of 2 and 4 (Fagot, 1985; Huston, 1983). It could be gender non-conformity, defined as the relative presence of feminine traits and/or the relative absence of masculine traits during boyhood, a correlate of t Childhood Factors and Attachment 16 homosexuality, that is linked with negative parent and child relations. Indeed, many gay men reported that as early as age 4 they felt "different from their peers" and that they saw themselves as more sensitive, less aggressive and less competitive (Isay, 1990). Retrospective studies comparing gay and heterosexual men have consistently found that gay men reported experiencing greater gender non-conformity as children. For example, Saghir and Robins (1973) found that 67% of gay men reported experiencing gender non-conformity in childhood compared to 3% of heterosexuals. Whitam (1977) found that 96% of gay men versus 26% of heterosexuals reported possessing at least one gender non-conforming trait. A meta-analysis of 41 studies by Bailey and Zucker (1990) demonstrated that gay participants recalled substantially more gender non-conforming behaviour in childhood than their heterosexual counterparts.1 Moreover, these authors concluded that the observed effect sizes (mean effect size = 1.31, SD = .43) were among the largest ever reported in the realm of sex-dimorphic behaviour. A number of researchers have criticized retrospective studies of gender non-conformity. They have argued that because adult gay men know that their sexual orientation is atypical, memory bias could lead them to exaggerate possible differences in childhood to be consonant with their present experience (Ross, 1980; Peplau, Garnets, Spalding, Conley, & Veniegas, 1996). However, a few studies are available which help refute this argument. Harry (1983), for example, replicated the basic finding that there are sexual orientation differences in recollections of childhood gender non-conformity. In addition, he looked at rates of gender non-conformity in adult gay men and found that while 67% of his adult gay sample had defeminized (i.e., became indistinguishable from heterosexual males), the majority of these men still recalled experiencing gender non-conformity in childhood. Thus, if present experiences are assumed to bias childhood memories, one would expect that the defeminized men would be more likely to recall memories Childhood Factors and Attachment 17 of childhood gender conformity as this would be consonant with their current defeminized experience. Moreover, an interesting study by Bailey, Miller, and Willerman (1993) assessed sexual orientation differences in recollections of childhood gender non-conformity ratings by sampling heterosexual and gay participants, as well as their mothers. These authors found that, as with past studies, there were large, significant differences in self-ratings between gay men and heterosexuals; as well, mothers of gay men remembered their sons as being less masculine. Importantly, Bailey et al., (1993) could rule out the possibility that mothers' recollection of their sons' gender non-conformity was associated with knowledge of their sons sexual orientation, as the correlation between awareness of sons' sexual orientation and recollections of sons' gender non-conformity was not significant (r = .12). Thus, this latter study helps negate the reconstructive bias argument because it replicated the findings of the retrospective studies using corroborative third party data. Finally, prospective studies that follow gender non-conforming boys through adolescence to adulthood when sexual orientation is more established have confirmed findings from retrospective studies. Green (1985, 1986), for example, followed "feminine" boys and a demographically matched comparison group of boys aged 2>lA to 11 years (average age was VA years) for over 15 years. Of the boys in the "feminine" sample who could be reached for follow up, 75% reported having more than incidental gay fantasies, as compared to none of the boys in the comparison group. Similarly, of the feminine boys from whom information on sexual orientation could be collected, 80% reported they were bisexual or gay, whereas all boys in the comparison group reported that they were heterosexual (Roberts, Green, Williams, & Goodman, ' The meta-analysis included studies of both male and female participants. Childhood Factors and Attachment 18 1987). Zucker (1990) summarizing the results of six studies, in which a total of 55 effeminate boys were followed, revealed that 63% of them had a subsequent gay sexual orientation. Gender Non-conformity, Sexual Orientation, and Familial Relationships The research cited above indicates that gender non-conformity in childhood is far more likely to characterize gay men than heterosexuals. This research does not, however, show whether gender non-conformity is a factor that may account for the previously observed relationship between sexual orientation and recollections of poor parent and child relations. Fortunately, a number of studies are available which directly address this issue. Seigelman (1974), for example, measured the relationship between masculinity, femininity, and quality of parent and child relationships in a gay sample. He found that high masculine gay men had more positive family backgrounds than did low masculine gay men, and that low feminine gay men reported having close, less casual, more loving fathers than did high feminine gay men. Level of femininity or masculinity was not associated with recollected quality of mother and son relationships for this gay sample. Langlois and Downs (1980) found that mothers rewarded and fathers punished their children's cross-gendered behaviour. Freund and Blanchard (1983) found that gay men, as compared to heterosexuals, were more likely to recollect significantly poorer father and son relationships. Moreover, the quality of father and son relationships was inversely related to childhood gender non-conformity but not sexual preference. Whitam and Mathy (1986) looked at factors associated with the gay male sexual orientation in four different cultures: Brazil, Guatemala, the Philippines, and the United States. They reported that while the specific content of the gender non-conforming behaviour might ' differ depending upon the given culture, equivalent behaviour was acted out to the same degree of frequency across cultures. They also asserted that although there was cross-cultural consistency in gender non-conformity, there was variation in parental reaction to gender non-Childhood Factors and Attachment 19 conforming boys. American parents were more likely to view gender non-conformity in children as unacceptable whereas Latin American and Filipino parents were more accommodating. Indeed, Filipino families were found to be the most tolerant, viewing homosexuality and its precursor, gender non-conformity, as natural and inevitable in certain individuals. Thus, societies vary greatly with respect to their attitudes to gender non-conformity. Whiman and Mathy (1986) speculated that the differences were due to the society's general sexual culture and openness towards sexuality. Harry (1989) conducted a study on sexual orientation differences in the recollection of physical abuse during adolescence. He found that gay men were more likely than heterosexuals to report being abused, and that the abuse was related to a history of childhood femininity and poor relationships with fathers. Harry (1989) concluded that a history of childhood femininity might actually provoke paternal abuse. Schilder and Kort (1996) asserted that gender non-conforming boys were more likely to be targets of emotional, physical, or sexual abuse and this may make them more vulnerable to suicide and high risk sexual behaviour leading to HIV infection. Cameron and Cameron (1995) found that gay men were far more likely than heterosexual men to report having had sexual relations with family members during childhood (29% and 7%, respectively). Doll et al., (1992) found that in a sample of 1001 bisexual and gay men, 37 percent experienced "abusive sexual contact during childhood and adolescence". They concluded that the risk of childhood sexual abuse was highest for male youth that exhibit stereotypical feminine characteristics (i.e., gender non-conformity). Gender Non-conformity and Peer Relationships There are reasons to believe that gay men may be more likely than heterosexual men to experience poor quality peer relationships in childhood, perhaps as a result of their exposure to Childhood Factors and Attachment 20 negative parent/child relationships. First, although rarely explored in gay samples, working models may predispose individuals to interact in new relationships using previously established patterns. Second, as with parent and child relations, gender non-conformity during childhood may have a negative influence on quality of peer relations. Langlois and Downs (1980) found that, at a very early age (ages 3 and 5 years), peers punished boys who exhibited cross-gendered behaviours (gender non-conformity). Carter and McCloskey (1984) reported that elementary school children were particularly rejecting and derogatory to boys who violated gender norms. Moreover, these children indicated that they would prefer not to associate with other children who violated gender-role norms. Saghir and Robins (1973) reported that many gay respondents in their study recalled being teased mercilessly during childhood because of their gender non-conformity, and that it was peer (and parental) pressure which instigated their defeminization. Bieber and Bieber (1979) found, based on clinical observations, that gay men recalled experiencing difficulties with their childhood same-sex friendships because they could not cope with aggressive, combative peer dynamics. Unfortunately, gay men are also likely to report that peer rejection in childhood continues well into adolescence. For example, Remafedi (1987) found that 30% of his bisexual and gay adolescent sample reported being physically abused by peers (half of which occurred on school property), and 50% reported being verbally abused by peers. Similarly, Rotheram-Borus, Rosario, and Koopman (1991) found that over half of their bisexual and gay adolescent sample had been ridiculed by peers because of their sexual orientation, and that gender non-conforming boys received the greatest abuse. Stigma It is possible that, to some degree, stigma is the mechanism underlying parental and peer reactions to gender non-conformity. Stigma is a difficult construct to define because it is abstract Childhood Factors and Attachment 21 and ambiguous (Ainlay, Coleman, & Becker, 1986). The classic definition of stigma comes from Goffman (1963) who wrote that, "stigma ... refer(s) to an attribute that is deeply discrediting, but it should be seen that a language of relationships, not attributes, is really needed" (p. 3). Goffman (1963) also wrote that individuals with discrediting qualities are seen as less than human because of these qualities. To a certain extent, stigma is difficult to define because it is a multi-faceted construct that occurs at three levels: 1) the societal level, 2) the dyadic level, and 3) the individual level. At the societal level, for stigma to occur there must be social consensus that a certain quality, attribute, or mark is discrediting and violates social norms (Stafford & Scott, 1986). At the dyadic level, persons coming into contact with possessors of the discrediting quality must be cognizant of the norm violation. Further, these persons' impressions of the possessor must be eclipsed by the mark itself (Jones et al., 1984). Finally, at the individual level, the marked person must perceive the derogation, attribute it to the mark, and internalize the derogation in their self-concept (Jones et al., 1984). This internalizing of stigma, or self-stigma, is a relative construct. Take, for example, the case of a man and a woman who are both overweight. For the sake of argument, one could claim that there is a societal proscription against corpulence, irrespective of gender. However, even if the negative societal view regarding excess body weight is gender-free, the woman may feel more stigmatized than the man since judgements of physical appearance are often more central to the female self-concept (Jones et al., 1984). It is possible that stigma may mediate the association between gender non-conformity and rejection to the extent to which there is a societal proscription against gender non-conformity. Following this, one would have to assume that parents and peers were aware of the proscription and rejected gender non-conforming boys based on this attribute. In fact, it would need to be shown that evaluations of gender non-conforming boys by parents and peers were dominated by Childhood Factors and Attachment 22 the stigmatic attribute. This evaluation process is also known as identity engulfment (Jones et al., 1984). Finally, one would have to show that gender non-conforming boys perceived that parental and peer rejection was a result of their gender non-conformity and internalized the rejection. Measured in adult gay men, internalized stigmatization due to gender non-conformity would take on the form of internalized homophobia (Meyer & Dean, 1998). Evidence for the possible role of stigma in the relationship between gender non-conformity and rejection partially comes from the aforementioned Whitman and Mathy (1986) cross-cultural study which showed that whereas gender non-conformity was cross-culturally consistent, attitudes and reactions to it were socially determined. That is, various societies' attitudes towards sexuality in general and homosexuality (and its precursor, gender non-conformity), in particular, mediated the associations between gender non-conformity and parental rejection. Gay male sexual orientation has long been known as a potential stigmatizing variable. Goffman (1963) wrote that in the United States (and by extension, Canada) the "ideal", or non-stigmatized person is "young, married, white, urban, northern, heterosexual Protestant father of college education, fully employed, of good complexion, weight, and height, and a recent record in sports" (p. 128). Anyone, then, who fails to match these characteristics is stigmatized to some degree. Jones et al., (1984) have outlined six dimensions to stigma. Four of the six dimensions seem particularly relevant to gay men and attest to their stigmatization. Origin and perceived responsibility is a first factor. In this regard, gay male sexual orientation is particularly stigmatizing to the degree to which people perceive it as a choice, since that which is unavoidable is considered less egregious. Childhood Factors and Attachment 23 Permanence is a second factor contributing to the degree of stigmatization. A permanent discrediting quality is more likely to contribute to stigmatization. It ensures that a greater number of potential stigmatizers have access and exposure to the possessor. Indeed, research has shown that the majority of gay males, even those who were particularly gender non-conforming in childhood, defeminize by adulthood (e.g., Whitam, 1977; Saghir & Robins, 1973), primarily as a reaction to persistent social and familial pressure (Harry, 1982). For example, Harry (1982) conducted a study of 1,556 urban gay men and found that 46% defeminized, 25% were never effeminate, 26% of the men were persistently effeminate, and 3% newly effeminate (see also Bell Bell, Weinberg, & Hammersmith, 1981, and Kelly & Warshafsky, 1987 for similar findings). Concealability, or the extent to which the discrediting quality is hidden, is a third factor. Concealable attributes are less noticeable, affording less opportunity for exposure and, therefore, stigmatization. Unlike members of racial or ethnic minorities, lesbian, gay men and bisexuals possess a potentially discrediting mark that is not usually visible until they chose to reveal it (Diplacido, 1998). Research shows that the majority of gay men describe themselves as being masculine. Moreover, even within the gay male community there is a premium placed on masculinity and the ability to "pass" or to be straight acting (Bailey, Kim, Hills, & Linsenmeier, 1997). Thus, one can imagine that gay men who intentionally avoid the gay lifestyle and successfully "pass" in heterosexual society would be less stigmatized by others. Passing, however, does not reduce "closeted" gay men's level of self-stigmatization. In fact, individuals who are explicitly "out" to others about their same-sex attraction have lower internalized homophobia scores compared to those who are not explicitly "out" (Meyer & Dean, 1998). Peril or danger is a fourth dimension of stigma. Gay male sexual orientation is perceived to be dangerous to the extent that it is unfamiliar. Studies show that negative attitudes to gay men decrease as a function of exposure to gay individuals (Herek & Capitanio, 1996). There are Childhood Factors and Attachment 24 two additional dimensions that seem less relevant to stigmatization of gay men. These include disruptiveness or hampering of communication and aesthetic quality, the extent to which the mark makes the possessor physically repellent (Jones et al., 1984). Compensation It should now be quite apparent that gay men (possibly because of their childhood gender non-conformity) are more likely than heterosexual men to have experienced poor relationships with both fathers and peers, although research on the quality of mother and son relationships is less consistent. Since the quality of childhood attachment relationships is expected to influence current use of attachment strategies (in terms of anxiety and avoidance), the corollary is that the likelihood of being insecurely attached in adulthood is increased for gay men. However, the presence in childhood or adulthood of a supportive, loving, person who is not the primary attachment figure, has been found to moderate the effects of receiving poor parenting as a child (Rutter & Quinton, 1984). A few studies are available that explore the nature and function of compensatory relationships, although they are based on samples of children. Rutter (1979) found that children who had a positive relationship with at least one parent had fewer behavioural problems than children who did not have a positive relationship with either parent, van Aken and Asendorpf (1997) looked at support by parents, classmates, friends and siblings. They found that low support from one parent could only be compensated by a supportive relationship with the other parent. Support from parents could not compensate for low supportive relationships with peers (friends, classmates and siblings). In contrast, Patterson, Cohn, and Kao (1989) found that children rejected by peers who had warm relationships with their mothers had significantly better adjustment outcomes than did rejected children who did not have warm relationships with mothers. Stocker (1994) found that while there was no difference in adjustment measures for Childhood Factors and Attachment 25 children who either had warm friendships and/or warm relationships with mothers, these children were better adjusted than children who had low levels of friendship and maternal warmth. Considering the literature review on compensation and the earlier literature review on gay men's recollected relationship quality with fathers, mothers and peers, it would be likely that if compensation were to occur, it would be good quality relationships with mothers that would moderate the impact of poor quality relationships with fathers or peers. However, it may be that exposure to any significant positive relationship (mother and/or father and/or peer) may help ameliorate the effects that other negative childhood relationships may have on current attachment strategies. Summary and Rationale There are a number of childhood socialization factors that seem to be associated with male homosexuality or, perhaps more specifically, boyhood gender non-conformity. These factors include paternal and peer rejection, which are known correlates of insecure attachment. However, findings from studies on gay men's recollected quality of mother/and son relationships are more ambiguous. Moreover, although findings indicate that fathers and peers tend to be rejecting of gender non-conforming behaviours in boys, this association between rejection and gender non-conformity does not necessarily hold true for mothers' treatment of sons. The purpose of the present study was to explore the relationship between childhood gender non-conformity, quality of relationships with fathers, mothers and peers, and gay men's current levels of anxiety and avoidance in close relationships. More specifically, the following questions were explored: Primary Research Questions 1 ) Are the associations between parental and peer relationships and anxiety and avoidance dimensions consistent with attachment theory? That is, do paternal, maternal, and peer Childhood Factors and Attachment 26 relationships each independently predict anxiety and avoidance? In addition, do peer relationships (at least partially) mediate the relationship between parental rejection and anxiety and avoidance? 2) Do gay men report differences between recollected quality of relationships with fathers and with mothers? 3) Are the associations between (a) gender non-conformity and paternal rejection and (b) gender non-conformity and peer rejection stronger than (c) the association between gender non-conformity and maternal rejection? 4) Does relationship quality (with mothers and/or fathers and/or peers) mediate the associations between gender non-conformity and anxiety and avoidance? 5) Does a positive relationship in one domain (i.e., either with mothers or father or peers) compensate for a poor relationship in another domain (either with mothers, fathers or peers)? Secondary Research Question 1) This study was not designed as a stigma analysis. But since this formulation seems relevant and a measure of self-stigmatization, internalized homophobia, has been included, the following question can be asked: Is the relationship between gender non-conformity and rejection (paternal, maternal, or peer) mediated by self-stigma? Method The present study was one component in a larger project not reported here. The overall project was an investigation of gay and bisexual men's relationship experiences funded by a Wayne F. Placek Award given to Dr. Kim Bartholomew. This award is administered by the American Psychological Foundation, a non-profit society associated with the American Psychological Association. The purpose of the award is to further understanding of social issues relevant to gay men and lesbians. Childhood Factors and Attachment 27 Participants Three hundred self-identified gay and bisexual men in the West End of Vancouver, British Columbia, Canada completed a telephone survey. From the initial survey, 266 respondents (87%) indicated their willingness to be re-contacted and told about the follow-up study. Of these 266 who gave permission to be re-contacted, 192 participants completed both follow-up components (questionnaire and interview), two participants completed the questionnaire but not the interview, and one participant completed the interview but not the questionnaire component. Thus, 65% (195/300) of those who were originally surveyed participated in the follow-up study. Analyses and results from this study were based on data from the 192 participants who fully completed both follow-up components of the project (unless otherwise specified). Table 1 lists the demographics of the follow-up participants. Of the 192, 180 indicated they were gay and 12 indicated they were bisexual in the telephone survey. In the questionnaire, participants were asked to indicate their sexual orientation on a 7-point continuum ranging from 1-exclusively gay to 7-exclusively heterosexual. Eighty-one percent of the sample self-reported being exclusively gay, 15% self-reported being predominantly gay, 3% self-reported being somewhat more gay than heterosexual, 1% self-reported being bisexual, and .5% (one individual) self-reported being somewhat more heterosexual than gay. Participants ranged in age from 20 to 71 years, with a mean age of 39 years (SD = 9.37). The majority of participants reported identifying with the following ethnic groups: 45.8% British ethnicity, 27.1%o other European ethnicity, 5.2% French Canadian ethnicity, and 12.5% other Canadian (excluding First Nations) ethnicity (please see Table 1 for a detailed ethnic description of the sample). It should be noted that many men who identified themselves as having a British background may be second or third generation Canadian. Childhood Factors and Attachment 28 In terms of relationship status, 33% of participants in the sample were in a serious, committed relationship. As expected, an overwhelming majority of individuals in relationships had male partners. Only two participants reported having opposite-sex relationships: one individual reported being married/living with a woman and one individual reported being in a dating relationship with a woman. The sample was well educated: 13.6% had some or all of a post-graduate education, 39.6% had some or all of a university education, 29.7% had some or all of a community college education, 15.1% had some or all of a high school education, and 1.5% had some or all or a elementary school education. Most of the sample were working full time (77.1%), 5.7% were working part time, 1% were students, 1% were working at home without pay, 4.2% were on leave from their jobs, 4.7% were retired, and 6.3% were unemployed. In terms of gross annual income, 24% earned $50,000 or more, 16.1% earned between $40,000 and $49,000, 24% earned between $30,000 and $39,999, 18.2% earned between $20,000 and $29,999, and 16.7% earned less than $20,000. Comparisons on all demographic variables and three additional health-related variables were conducted between those participants who completed the follow-up component of the study (N = 192) and those individuals who gave permission to be re-contacted but did not complete the follow-up (N = 74). The health-related variables included: 1) HIV status, 2) participation in unprotected anal sex with a male partner in the past year and, 3) degree of physical abuse ever experienced in a same-sex relationship. The health-related variables were included to assess whether those who completed the follow-up component were generally more vulnerable or "at risk" compared to those who did not complete the follow-up component. No demographic differences between these two groups were found. However, in terms of the health-related comparisons, there was one significant between group difference. Respondents Childhood Factors and Attachment 29 who completed the follow-up component were more likely to report being HIV positive (27%) compared to those respondents who gave permission to be re-contacted but did not complete the follow-up (11%), x2 (1) = 4.12, p < .05. Demographic and health-related comparisons between the follow-up sample (N = 192) and those individuals who did not give permission to be re-contacted (N = 34) were also conducted. In terms of demographics, one significant difference emerged. Fewer men in the follow-up sample identified themselves as bisexual (6%) compared to the latter group of men (29%), % 2 (1) = 17.64, p. < .0001. In terms of the health-related variables, respondents who completed the follow-up component were more likely to report being HIV positive (21%) compared to those respondents who did not give permission to be re-contacted (3%), % 2 (1) = 5.66, p_ < .05. There was also a significant between group difference in degree of physical abuse ever experienced in a same-sex relationship. Those respondents who completed the follow-up component were more likely to have experienced a greater degree of physical abuse (M = 1.86, SD = 3.09) compared to those respondents who did not give permission to be re-contacted (M = .81, SD = 1.59), t (60) = 2.74,p <.01. Procedures Survey. Initial recruitment involved a telephone survey of a community sample of 300 gay and bisexual men in the West End of Vancouver. The survey was conducted by Canadian Facts, a professional survey company. The survey utilized a book-plus design. The sample was drawn from the latest telephone listing in the identified West End exchanges and a digit was added to each listing to maximize the chance of reaching newly listed or unlisted respondents. This random sampling methodology attempted to maximize the diversity of a community sample of gay men as all households in the West End had an equal probability of being included in the Childhood Factors and Attachment 30 sample. The overall response rate, calculated as the number of respondents who completed interviews divided by the number of known eligible respondents, was 49% for this survey. Households were first screened for men 18 years or older. If more than one adult man resided in the household, a potential respondent was randomly selected using the most recent birthday method (Oldendick, Bishop, Sorenson & Tucker, 1988). Surveyors ensured that the respondents were able to complete the survey in private. All men were asked about their sexual orientation. If participants stated that they were heterosexual, they completed a short version of the survey that assessed demographics. Bisexual or gay men were asked to complete the full survey which took approximately 15-20 minutes and included a series of demographic, abuse, and health-related questions. At the end of the full survey, respondents were asked if they could be re-contacted and told about the follow-up component of the study which would explore their childhood and current relationship experiences in further detail. Questionnaires and in-person interviews. Attempts to re-contact respondents were made within 2 weeks of the original telephone survey date, although some respondents could not be reached within that period. In the re-contact telephone conversation, respondents were informed about, and were invited to participate in, the two components to the follow-up study. The follow-up included a questionnaire about recollections of family and peer relationships, and a one-to-one, in-person attachment interview that was conducted at the West End Relationship Project (WERP) office located in Vancouver's West End. If respondents indicated a willingness to participate in the follow-up, their addresses were recorded so that questionnaires could be mailed to their homes and interview dates were scheduled. If they preferred, they could also pick up sealed questionnaires at the WERP office. Questionnaires were mailed to participants within 3 days of the re-contact telephone conversation. Interview dates were usually scheduled 10 days later to allow time for mail Childhood Factors and Attachment 31 delivery and completion of questionnaires. Participants were called the day before their interview and reminded of the interview time and location. Questionnaires took approximately 1 hour to complete. Participants brought their completed questionnaire package with them to their in-person interview. Interviews were conducted either by a gay male psychotherapist or by one of two female doctoral students knowledgeable about gay male experiences (one being the author). There were no interviewer effects on attachment ratings. Each interview lasted approximately 2 hours. At the start of the interview session, individuals were paid $20.00 for their time and asked to sign a consent form. Following the interview session, participants were debriefed and thanked for their involvement in the study. Referrals to various counseling and community agencies were offered at that time to ensure that participants could seek help if they wanted to discuss further issues that arose during the interview. Participants were also told that members of the WERP team would facilitate connection with any counseling or community agencies on their behalf. Measures Recollections of Early Childrearing. (Egna Minnen Betraffande Uppfostran, EMBU; Perris, Jacobsson, Lindstrom, von Knorring, & Perris, 1980). This scale measures respondents' memories of their upbringing. The EMBU was originally developed in Sweden (Gerslma, Emmelkamp, & Arrindell, 1990) and has been translated and widely used with English speaking samples (Ross, Campbell, & Clayter, 1982). This 80-item scale describes particular cognitive, emotional, or behavioural scenarios that are rated on a 4-point Likert scale ranging from "never occurred" to "always occurred." For purposes of this study, only sub-scales assessing recollections of paternal and maternal warmth, and paternal and maternal rejection were assessed. An example of an item from the warmth sub-scale is "When I was sad, I could seek comfort from my father (or mother)". The rejection sub-scales included items pertaining to both physical and Childhood Factors and Attachment 32 psychological abuse. An example from the rejection sub-scale is "My father (or mother) would punish me hard, even for little things". The correlation between paternal warmth and paternal rejection was r = -.68,_p_ < .001. The correlation between maternal warmth and maternal rejection was r = -.64,_p < .001. For this sample, both Cronbach alphas for the paternal and maternal warmth sub-scales were .96. Cronbach alphas for the paternal and maternal rejection sub-scales were .94 and .96 respectively. Because rejection rather than warmth is more central to understanding anxiety and avoidance, analyses reported below were based on the rejection sub-scales for both fathers and mothers (unless otherwise specified). Notably, for most analyses, results did not differ when the warmth sub-scales for mothers and fathers were used (exceptions are specified). Boyhood Gender (Non) Conformity Scale. (BG(N)CS; Hockenberry & Billingham, 1987). The boyhood gender non-conformity scale originally included 20 statements designed to assess boyhood gender non-conforming (and conforming) behaviour and identity. Each statement is rated on a 7-point Likert scale ranging from "never or almost never true" to "always or almost always true". One item from the original scale, "As a child, I felt like a girl", was dropped because it seemed offensive. One item, " As a child, I liked to engage in rough-and tumble play", was added because it was theoretically relevant. Three additional items were dropped because they contained peer content that overlapped with the quality of peer relationships measures. Although this scale has been used as a single measure of gender non-conformity, a principal component analysis was used to test the scale's unidimensional structure. Results indicated a two-factor structure.2 The two factors were clearly interpretable as femininity and masculinity. An example of an item with a high loading on the femininity factor is "As a child, I Childhood Factors and Attachment 33 preferred girls' games and toys". An example of an item with a high loading on the masculinity factor is "As a child, I imagined I was the male character in stories I read or watched on TV". Three items loading highly on both factors were dropped as they could not be used to distinguish factors. An example of such an item is "As a child, I preferred to play with girls". The results of the principal component analysis were used to guide the composition of two sub-scales labeled femininity and masculinity. Thus, relatively high femininity and/or relatively low masculinity were both used as measures of gender non-conformity in this study. The Cronbach alphas for the femininity and masculinity sub-scales were .78 and .70 respectively. Mother-Father-Peer Scale. (E-PEER; Epstein, 1983) is a 70-item scale that assesses acceptance and rejection by mothers, fathers, and peers. Items are ranked on a 5-point Likert scale of agreement ranging from "strongly disagree" to "strongly agree". Only the 10 items pertaining to peer acceptance and rejection were used in this study. An example of a peer acceptance item is "When I was a child, other children liked to play with me". The Cronbach alpha for this scale was .88 in this sample. An example of a peer rejection item is "when I was a child, other children picked on me and teased me". The Cronbach alpha for this scale was .90 in this sample. The correlation between the acceptance and rejection sub-scales was, r = -.75, p < .001. Inventory of Peer Attachment. (IPA; Armsden & Greenberg, 1987). This is a 25-item scale used to measure recollections of the quality of childhood/adolescent friendships and attachments. Each item is rated on a 5-point Likert scale ranging from "almost never or never true" to "almost always or always true". Three sub-scales were first generated from the inventory. The sub-scales include trust (e.g., "As a child/adolescent, I trusted my friends"), communication (e.g., "As a child/adolescent, I liked to get my friends' point of view on things I A Scree plot broke between the second and third component so two components were retained. The Eigenvalue for Childhood Factors and Attachment 34 was concerned about"), and alienation (e.g., "As a child/adolescent, I felt alone or apart when I was with my friends"). The Cronbach alphas for the trust, communication, and alienation sub-scales were .94, .88, and .82, respectively, for this sample. The three sub-scales strongly correlated with each other: r = .81, p_ < .001 between trust and communication, r = -.81, p_ < .001 between trust and alienation, and r = -.66, p < .001 between communication and alienation. Items for the inventory were therefore combined to create a total score. The Cronbach alpha for the total score was .95 in this sample. Peer Relationships Composite Measure. The IPA total scale correlated significantly with both the E-PEER acceptance and rejection sub-scales, r = .72, p < .001 and r = -.62, p < .001. As such, all three were combined using principal components analysis to create a composite index of peer rejection. All analyses reported below for peer relationships use the composite measure unless otherwise specified. Internalized Homophobia Scale. (IHS; Wagner, Serafini, Rabkin, Remien, & Williams, 1994). Internalized homophobia was used as the measure of self-stigma. This is a 20-item scale designed to assess the extent to which an individual is disapproving of homosexuality in general, and uncomfortable with their own homosexuality in particular. Each item is rated on a 5-point Likert scale ranging from "strongly disagree" to "strongly agree". An example of an item from this scale is "Whenever I think a lot about being gay, I feel critical about myself. The Cronbach alpha for this scale was .92 in this sample. History of Attachment Interview (HAD. This 2 hour semi-structured interview asks participants to describe their recollection of their relationship with their parents as children, as well as their current relationship with their parents. Participants are also asked to describe their friendships, romantic relationships, and feelings about the importance of close relationships. If the first component was 5.2 and the Eigenvalue for the second component was 2.10. Childhood Factors and Attachment 35 participants have not been involved in romantic relationships, they are asked the reasons. They are also asked about loneliness, shyness, degree of trust of others, impressions of other people's evaluations of themselves, and hope for any changes in their social lives. In addition to the above questions, gay men are asked about early recollections of their sexual orientation (at what age they were first aware of their same sex attraction, the extent to which they felt different in childhood and adolescence), their coming out experiences and their possible AIDS-related-loss experiences. The interview is designed to assess both participants' characteristic experiences and feelings in close relationships, as well as the internal coherence and consistency of their relationship accounts. Trained raters coded the interviews using the attachment framework proposed by Bartholomew (Bartholomew, 1990; Bartholomew & Horowitz, 1991). Each participant's degree of correspondence to each of the four prototypic attachment patterns (secure, fearful, preoccupied, dismissing) was rated on a 9-point scale (ranging from 1 (no correspondence) to 9 (excellent fit)). The secure prototype is characterized by valuing intimate relationships, the capacity to maintain close friendships without sacrificing personal autonomy, and coherence and thoughtfulness in discussing relationships and related issues. The dismissing prototype is characterized by a downplaying of the importance of emotional relationships, restricted emotionality, an emphasis on independence and self-reliance, and a lack of clarity or credibility in discussing relationships. The preoccupied prototype is characterized by anxiety, an over-involvement in close relationships, a dependence on other peoples' acceptance for a sense of personal well-being, a tendency to idealize others, and incoherence and exaggerated emotionality during discussion of relationships. The fearful prototype is characterized by anxiety resulting in the avoidance of relationships for fear of rejection, a sense of personal insecurity, and a distrust of others. : was Childhood Factors and Attachment 36 Continuous ratings of the four attachment patterns were generated. Each participant also classified according to their best fitting attachment category by taking the highest of the four continuous ratings as an indicator of the participant's predominant pattern. A second trained coder rated 53 of the interviews so that rating reliability could be assessed. Inter-rater reliability for the secure, fearful, preoccupied and dismissing attachment prototypes were as follows: .69, .69, .69, .60. The two attachment dimensions, anxiety and avoidance, were derived from the four attachment prototypes using the procedure outlined by Griffin and Bartholomew (1994). Anxiety was derived by summing preoccupied and fearful scores and subtracting secure and dismissing scores. Avoidance was derived by summing dismissing and fearful scores and subtracting secure and preoccupied scores. Inter-rater reliability for anxiety and avoidance was .74 and .70. All analyses reported below pertaining to attachment use the anxiety and avoidance measures unless otherwise specified. Because the dimensions were derived using difference scores, some might argue that the reliability of these scores should be lower than the reliability scores for each of the four prototypes (e.g., Cronbach & Furby, 1970). This argument is primarily based on the special case of difference scores in pretest and posttest scenarios which assumes data collected from parallel tests resulting in limited individual variation between tests. Notably, it has now been shown that even difference scores generated from pre- and posttest are in fact highly reliable (Rogosa & Willett, 1983; Zimmerman & Williams, 1982). Notwithstanding this new evidence, derivation of the attachment dimensions from the four attachment prototypes is not equivalent to taking pre-and post-test difference scores since the prototypes are not parallel constructs. In fact, the attachment dimensions show greater inter-rater reliability than that observed with the prototypes. Childhood Factors and Attachment 37 Results Descriptive Analyses of Key Measures Data were checked for univariate and multivariate normality. Only the paternal and maternal rejection scores were positively skewed, z = 5.80, p_ < .001 and z = 7.52, p < .001, respectively. Logarithmic transformations of these two variables were performed and were used in all correlational analyses (Tabachnick & Fidell, 1989). The means and standard deviations of all measures are reported in Table 2. EMBU. The mean score for paternal warmth was 2.18 and the mean score for maternal warmth was 2.80. These scores indicated that, on average, fathers were seen as occasionally warm, and on average, mothers were seen as often warm. The mean score for paternal rejection was 1.76 and the mean score for maternal rejection was 1.57. These scores indicated that, on average, participants remembered their fathers and mothers as occasionally rejecting. All 192 participants completed the EMBU for their mothers but only 181 participants completed the EMBU for their fathers. This is because 11 participants either did not know their fathers at all, or did not feel that they had enough experiences to answer questions addressed in the EMBU about their relationships with their fathers. Using a methodology similar to that used in the present study, Henderson (1998) administered the EMBU in a sample of primarily heterosexual men living in the Lower Vancouver Mainland (9.7% of the sample self-identified as gay, N = 60). She reported the following mean scores for paternal warmth, maternal warmth, paternal rejection, and maternal rejection: 2.23 (SD = .86), 2.71 (SD = .82), 1.45 (SD = .46), 1.39 (SD = .37). It can be noted that compared to the Henderson sample, there does not appear to be any difference in recollections of parental (mother and father) warmth but the gay sample was more likely to remember both their fathers and their mothers as more rejecting. Childhood Factors and Attachment 38 BG(N)CS. The mean score for the femininity sub-scale was 3.39. This score indicated that the men in the sample, on average, recalled that they sometimes engaged in feminine behaviours during childhood. The mean score for the masculinity sub-scale was 3.22. Again, this score indicated that the men in the sample, on average, recalled that they sometimes engaged in masculine behaviours during childhood. Scores on the two sub-scales were correlated with two measures coded from the interview: age at which participants were first aware of their same sex-attraction,3 and the extent to which participants felt that they were different in childhood (rated on a 9-point scale). There was no significant correlation between recollection of feminine behaviour and age of sexual orientation awareness, but recollection of feminine behaviour was positively correlated with feeling different in childhood (coded from the interview), r = .29, p < .001. That is, those participants who were more likely to recall engaging in feminine behaviour were also more likely to feel different in childhood. Recollection of childhood masculinity was positively correlated with age of sexual orientation awareness, r = .17, p < .05. That is, those participants who recalled engaging in more masculine behaviours in childhood were less likely to become aware of their same-sex attraction at an early age. Conversely, recollection of childhood masculinity was negatively correlated with feeling different in childhood, r = -.16, p < .05. Specifically, those participants who recalled engaging in more masculine behaviour were less likely to feel different in childhood. Taken together, these associations add to the construct validity of the BG(N)CS as well as help elucidate the nature of gender non-conformity experienced by this sample. 3 Twenty-eight percent of the sample were aware of their same sex attraction before the age of six, 51% first became aware of their same sex attraction between the ages of 7 and 12, and 17% first became aware of their same sex attraction between the ages of 13 and 18. Childhood Factors and Attachment 39 Measures of peer relationships. The mean score for the Epstein peer acceptance measure was 3.10 and the mean score for Epstein peer rejection measure was 2.78. These scores indicated that participants, on average, sometimes felt accepted and sometimes felt rejected by their peers during childhood and adolescence, although acceptance scores were significantly higher than rejection scores, t (191) = 2.56, p < .01. Epstein (1983) validated this measure in a sample of 355 men. The mean score for peer acceptance and rejection in his sample were 3.81 and 2.36, respectively. Thus, it can be seen that, compared to the Epstein (1983) sample, participants in the present sample recalled lower levels of peer acceptance and higher levels of peer rejection. In terms of the IPA, mean scores for peer trust, communication, and alienation were 3.35, 2.83, and 2.68 respectively. These scores indicated that participants sometimes felt they had trusting friendships with good communication during childhood and adolescence, but to a lesser extent, they felt some alienation. As with comparisons between the Epstein (1983) measure of peer acceptance and rejection, comparisons between peer trust and peer alienation scores show that participants were more likely to recall trusting rather than alienating friendships, t (191) = 6.05, p_ = .00. Comparisons between this sample and other samples on the IPA measures were not possible, as data collected from comparable samples are not available. IHS. The mean score for internalized homophobia was 1.8. This score indicated that, on average, participants disagreed with negative attributions regarding gay male sexual orientation. In fact, 70% of the sample had an average IHS score below 2. Wagner et al., (1994) reported average total IHS scores for a number of samples. These include a sample of gay men at a national Dignity conference4, a New York Dignity service, and a community sample of gay men who were raised in Catholic households but were not members of a gay-identified church organization. Average total scores for the three samples were: 33, 37, and 36, and there were no Childhood Factors and Attachment 40 significant between-group differences (Wagner et al., 1994). The average total IHS score in this sample was 36 indicating that levels of internalized homophobia were comparable with those found by Wagner et al., (1994). History of Attachment Interview (HAD. The mean scores for continuous ratings of secure, fearful, preoccupied and dismissing patterns were 3.47, 4.00, 3.43, and 2.70, respectively. To further understand the extent of security/insecurity in the sample, participants were classified according to their best-fitting attachment category by taking the highest of the four continuous ratings as an indicator of the participant's predominant pattern. Using this method, 20% of the sample was categorized as predominantly secure, 38% as predominantly fearful, 22% as predominantly preoccupied, 14% as predominantly dismissing, and 5% had a mix of two patterns that equally predominated. This can be contrasted with research by Mickelson, Kessler, and Shaver (1997) who found in a nationally representative sample that 59% of respondents self-reported having a predominantly secure attachment style. Self-reports of attachment, compared to expert ratings of attachment interviews, are prone to social desirable responding, and, therefore, the Mickelson et al., (1997) findings may have to some degree over-estimated prevalence of security. Interview ratings of attachment have not been used in any nationally representative sample; however, research with college samples using this method showed that 47% to 51% of participants have been categorized as secure (e.g., Bartholomew & Horowitz, 1991; Scharfe & Bartholomew, 1994). Thus, the apparent lack of attachment security in this community sample of gay men is noteworthy. The two attachment dimensions, anxiety and avoidance, were derived from the four attachment prototypes. The mean rating for anxiety and avoidance were 1.28 and -.20, respectively. The former score suggest that participants, on average, had relatively high levels of 4 Dignity is an organization of gay men and lesbians of Catholic background. Dignity members strive to integrate Childhood Factors and Attachment 41 anxiety. Avoidance scores were close to the midpoint of the scale, which indicated, on average, that participants were neither avoidant nor approach oriented. Scharfe et al., (1994) reported anxiety and avoidance scores of -1.10 and -.10, respectively in their college sample. Thus, compared to the Scarfe et al., (1994) study, participants in this study were more anxious. Analyses of Research Questions Zero-order correlations for all key self-report measures and the two attachment dimensions derived from interview codings are presented in Table 3. Key findings from this table will be discussed below in conjunction with the research questions. Research Question 1 This question explored how paternal, maternal and peer relationship quality relate to anxiety and avoidance. Further, it tested the possibility that peer relationships partially mediate the relationship between parental rejection and anxiety and/or avoidance. To test a mediational model, three conditions must be satisfied. First, the predictor must be associated with the criterion variable, i.e., paternal and maternal rejection must each be associated with anxiety/avoidance. Second, the mediator must be associated with the criterion, i.e., peer rejection must be associated with anxiety/avoidance. Third, the predictor and the mediator must be associated, i.e., paternal and maternal rejection must be associated with peer rejection. If these conditions are met, one can test for mediation by regressing the criterion variable on the predictor and mediator. Mediation holds if the effect of the predictor variable on the criterion variable is reduced when the mediator is controlled. Perfect mediation holds if the predictor variable has no effect after the mediator is controlled; however, this is difficult to achieve since measurement error in the mediator tends to underestimate the effect of the mediator and overestimate the effect of the predictor (Barron & Kenny, 1986). their faith and same-sex attraction. Childhood Factors and Attachment 42 Goodness-of-fit measures are often reported as an index in mediational models. A measure of fit involves comparing all observed and implied correlations, rather than just focusing on a few relevant associations. Goodness of fit is not reported when models are fully recursive because, in this case, there is always a perfect fit between model and observed data. Fully recursive models are those that include a direct link between each variable and all variables further down the chain (Klem, 1995). Anxiety. I first looked at the zero-order correlations. Paternal and maternal rejection were significantly correlated with each other (r = .53, p < .001) and they were significantly correlated with anxiety (paternal and anxiety: r = .28, p < .001, maternal and anxiety (r = .17, g < .001), satisfying the first condition necessary for a mediational model. Next, I looked to see that peer rejection was correlated with anxiety, and it was (peer and anxiety r = .39, p < .001), thus satisfying the second condition. Finally, I looked at the zero-order correlations between paternal and peer rejection and maternal and peer rejection and found both to be significant (r = .37, p < .001 and r = .35, p < .001, respectively). I was also interested in the combined contribution of paternal rejection and maternal rejection in predicting peer rejection. A multiple regression showed that the combined predictor variables accounted for 17% of the variance in peer rejection (R change= -17, Flange (2, 178) = 17.87, p<.001). A path analysis depicted in Figure 2 was used to test if peer rejection mediated the relationship between parental rejection and anxiety. The path coefficients for paternal rejection (Beta = .17, t = 2.10, p < .05) and peer rejection (Beta = .32, t = 4.10, p< .001) were found to be significant, but the path coefficient for maternal rejection was not significant. Thus, rejection from both fathers and peers independently predicted contributed to anxiety, but maternal rejection did not have an independent influence. In addition, the association between paternal Childhood Factors and Attachment 43 rejection and anxiety (r = .28) dropped when peer rejection was added to the model (Beta = .17). The significant association between maternal rejection and the anxiety (r = .17) ceased to be significant when paternal and peer rejection was added to the model (Beta = .02). Consequently, it can be seen that peer rejection largely mediated the association between parental (father and mother) rejection and anxiety. Avoidance. As reported above, the inter-correlations among the three forms of rejection were all significant. However, in looking at the zero-order correlations between the three forms of rejection and avoidance, the only significant correlation was between peer rejection and avoidance (peer and avoidance, r = .23, p < .01). That is, neither paternal rejection nor maternal rejection was associated with avoidance. Because the zero-order correlations between paternal rejection and avoidance and maternal rejection and avoidance were non-significant, the first condition for a mediational model was not satisfied. Therefore, there was no need to assess whether peer relationships served as a mediator between parental rejection and avoidance. Research Question 2 Research question 2 explored whether gay men in this sample reported poorer quality childhood relationships with their fathers than mothers. T-tests comparing mean differences on the warmth and rejection EMBU sub-scales were conducted. The mean rating for paternal emotional warmth, 2.20 (SD = .77) was significantly lower than the mean rating for maternal emotional warmth, 2.82 (SD = .70), t (180) = -13.40, p < 001. Conversely, the mean rating for paternal rejection, 1.76 (SD = .62) was significantly higher than the mean rating for maternal rejection, 1.57 (SD = .48), t (180) = 4.80, p < 001. In summary, gay men recalled their fathers as being less emotionally warm and more rejecting than their mothers during childhood. Childhood Factors and Attachment 44 Research Question 3 This question addressed whether the associations between gender non-conformity and paternal rejection and the association between gender non-conformity and peer rejection were stronger than the association between gender non-conformity and maternal rejection. Femininity. The zero-order correlations between femininity and paternal, peer and maternal rejection were r_= -23, p_ < .01, r = .26, p < .001, and r = .22, p < .01, respectively. Tests of differences between dependent correlation coefficients revealed that there was no significant difference between these correlation coefficients. Thus, the association between paternal or peer rejection and femininity was not any stronger than the association between maternal rejection and femininity. Masculinity. The zero-order correlations between masculinity and paternal, peer and maternal rejection were r = -.13, g > .05, r = -.22, p < .01, and r = -.001, p > .05, respectively. That is, participants who recalled being more masculine were less likely to be rejected by their peers (but not fathers or mothers) in childhood. Or, said the other way, participants who recalled being less masculine were more likely to be rejected by their peers (but not fathers or mothers) in childhood. A test of differences between dependent correlation coefficients revealed that there was no significant difference between the masculinity and paternal rejection and the masculinity and maternal rejection correlation coefficient. However, there was a significant difference between the masculinity and peer rejection correlation coefficient and the masculinity and maternal rejection correlation coefficient, t (188) = -2.7, p < .01. This indicated that the association between masculinity and peer rejection was stronger than the association between masculinity and maternal rejection. Interestingly, the association between paternal warmth and masculinity was also significant, r = .23, p < .001, but the association between maternal warmth and masculinity was warm Childhood Factors and Attachment 45 not significant. Thus, although there was no significant association between paternal rejection and masculinity, participants who recalled being more masculine were more likely to recall relationships with their fathers in childhood. Or, said the other way, participants who recalled being less masculine were less likely to recall warm relationships with their fathers in childhood. Research Question 4 Research Question 4, building on the findings from Question 1 and 3, explored whether relationship quality (paternal, maternal and/or peer) mediated the relationship between gender non-conformity (femininity and masculinity) and anxiety and avoidance. Recall from Question 1, for a mediational model to be present, three relationships must hold true. Thus, in terms of this question, femininity or masculinity must correlate with anxiety or avoidance, relationship quality (with mothers, fathers and/or peers) must correlate with anxiety or avoidance, and femininity or masculinity must correlate with relationship quality (Barron & Kenny, 1986). Anxiety-femininity. The zero-order correlation between femininity and anxiety was, r = .20, p < .001, thus the first necessary association in the mediational model was significant. Zero-order correlations reported in Question 1 showed that paternal, maternal and peer relationships were each associated with anxiety, thus the second necessary condition was met. Finally, research question 3 also indicated that femininity was significantly correlated with all three forms of rejection (paternal, maternal and peer). Thus, there was sufficient reason to test whether paternal, maternal, and peer rejection mediated the relationship between gender non-conformity and anxiety. I explored whether paternal, maternal, and/or peer relationships mediated the relationship between femininity and anxiety using path analysis (depicted in Figure 3). Inspection of the path coefficients showed that rejection did mediate the relationship between femininity and anxiety as the direct path was not significant (Beta = .10, p = .16) when paternal, maternal, and peer Childhood Factors and Attachment 46 rejection were included in the model. Furthermore, as implied by the pattern of results, follow-up analyses showed that the mediational effect was driven by peer rejection (Beta = .30, p < .001). Anxiety-masculinity. The zero-order correlation between masculinity and anxiety was, r = -.21, p < .001, thus the first necessary association in the mediational model was significant. Zero-order correlations showed that paternal warmth, and peer rejection were each associated with anxiety, thus the second necessary condition was met. Finally, research question 3 also indicated that masculinity was significantly correlated with peer rejection and paternal warmth. Thus, there was sufficient reason to test whether peer rejection and paternal warmth mediated the relationship between masculinity and anxiety. Again, I explored the proposed mediational model using path analysis (depicted in Figure 4). Inspection of the path coefficients showed that peer rejection and paternal warmth mediated the relationship between masculinity and anxiety as the direct path was not significant (Beta = -.12, p = .09) when the two variables were included in the model. Follow-up analyses showed that the mediational effect was driven by peer rejection (Beta = .31, p < .001). Avoidance-femininity. Examination of the zero-order correlation between the predictor (femininity) and the criterion (avoidance) showed that this relationship was not significant (r = -. 12, p >. 05). Therefore, any test for mediation was untenable. Avoidance-masculinity. Again, examination of the zero-order correlation between the predictor (masculinity) and the criterion (avoidance) showed that this relationship was not significant (r = -.05, p >. 05). Therefore, any test for mediation was untenable. Alternate Path Models Because the models for femininity and masculinity were both fully recursive, a goodness-of-fit test was not appropriate. However, it was possible to test alternate mediational models. Childhood Factors and Attachment 47 Because it has, on occasion, been argued that sexual orientation and associated factors like gender non-conformity follow from parental rejection (e.g., Satinover, 1996), I tested whether femininity and masculinity each mediated the. associations between relationship quality (paternal, maternal, and peer) and the adult attachment dimensions (anxiety or avoidance). To isolate mediational effects, I conducted separate models for each significant measure of relationship quality. Alternate path model: Relationship quality to femininity to anxiety. The zero-order correlation between the predictor variables and the criterion (relationship quality and anxiety) were as follows: paternal rejection and anxiety, .28 (p < .01), maternal rejection and anxiety, .17 (p < .05), and peer rejection and anxiety, .39 (p < .01). The zero-order correlation between the mediator and the criterion (femininity and anxiety) was .20 (p < .01). Finally the zero-order correlation between the predictor variables and the mediator were as follows: paternal rejection and femininity, .23 (p < .01), maternal rejection and femininity, .22 (p < .01), and peer rejection and femininity, .26 (p < .01). Inspection of the path coefficients showed that the effects of the predictor variables on the criterion variable were not substantially reduced when the mediator was controlled (depicted in Figures 5). The Beta weight between paternal rejection and anxiety was .24 (p < .01). The Beta weight between maternal rejection and anxiety was .13 (p > .05). The Beta weight between peer rejection and anxiety was .35 (p < .001). Given that the effect of the mediator was twice as strong in the original model (between femininity, relationship quality, and anxiety) when compared to the alternate model, there was reason to opt for the former compared to the latter. Alternate model: Relationship quality to masculinity to anxiety. The zero-order correlation between the predictor variables and the criterion (relationship quality and anxiety) were as follows: paternal warmth and anxiety, -.23 (p < .01), and peer rejection and anxiety, .39 Childhood Factors and Attachment 48 (p < .01). The zero-order correlation between the mediator and the criterion (masculinity and anxiety) was -.21 (rj < .01). Finally the zero-order correlation between the predictor variables and the mediator were as follows: paternal warmth and masculinity, .23 (p < .01), and peer rejection and masculinity, -.22 (p < .01). Inspection of the path coefficients showed that the effects of the predictor variables on the criterion variable were not substantially reduced when the mediator was controlled (depicted in Figure 6). The Beta weight between paternal warmth and anxiety controlling for masculinity was -.20 (p < .01). The Beta weight between peer rejection and anxiety controlling for masculinity was .35 (p < .001). Given that the effect of the mediator was twice as strong in the original model (between masculinity, relationship quality, and anxiety) when compared to the alternate model, there was reason to opt for the former compared to the latter. Alternate model: Relationship quality to femininity or masculinity to avoidance. Examination of the zero-order correlation between the mediators (femininity and masculinity, respectively) and the criterion (avoidance) showed that these relationships were not significant: femininity and avoidance (r = -.12, p >. 05), masculinity and avoidance (r = -.05, p >. 05). Therefore, mediational tests of the alternate model were not conducted. Research Question 5 In Research Question 5,1 explored whether having a good relationship in one domain (i.e., either with mothers or fathers or peers) compensated for a poor relationship in another domain. Again, because no relationship other than that with peers was associated with avoidance, tests for compensation in terms of this attachment dimension were not possible. Therefore, analyses focused on relationship compensation and anxiety. Three hierarchical regressions were conducted to examine compensation among the following relationships: 1) mother and fathers, 2) mothers and peers, and 3) fathers and peers. Childhood Factors and Attachment 49 For each of the three hierarchical regressions, main effects were added in the first block and the interaction term was added in the second block. Interaction terms were created by calculating the product of the two zero-centred independent variables, thereby removing main effects. An interaction was significant if the R2Change for the second block was significantly different from the R generated in the first block. In the first hierarchical regression analysis with maternal rejection and paternal rejection, the main effect for maternal rejection and the interaction term were not significant; but the main effect of paternal rejection was significant (Beta = .26, t = 3.02, p < .01). This indicated that irrespective of whether mothers were rejecting or not, having a rejecting father contributed to anxiety. In the second hierarchical regression analysis with maternal rejection and peer rejection, the main effect of maternal rejection and the interaction term were not significant; but the main effect of peer rejection was significant (Beta = .37, t = 5.23, p < .001). This indicated that irrespective of whether mothers were rejecting or not, having rejecting peers contributed to anxiety. In the third hierarchical regression analysis with paternal rejection and peer rejection, both the main effect of paternal rejection (Beta = .18, t = 2.43, p < .05) and peer rejection (Beta = .32, t = 4.29, p < .001) were significant. In addition, the interaction approached significance (Beta = -. 11, t = -1.65 p =. 10). Given that the significance test for interactions using regression analysis is quite conservative, the observed interaction between paternal and peer rejection was further explored. Paternal and peer rejection were split into low and high groups at the median. The means (standard deviations in parentheses) for anxiety for each of the four groups were as follows: Low Paternal Rejection and Low Peer Rejection, 1.15 (SD = 50), High Paternal Rejection and Low Childhood Factors and Attachment 50 Peer Rejection, 2.26 (SD = 69), Low Paternal Rejection and High Peer Rejection, 1.97 (SD = .68), and High Paternal Rejection and High Peer Rejection, 2.86 SD = 50). This relationship is graphed in Figure 7. Anxiety was relatively high if relationship quality was poor in any one domain (i.e., either paternal or peer rejection was high). Only those who experience low paternal rejection and low peer rejection had low levels of anxiety. Thus, for this sample, relationships did not appear to function in a compensatory fashion. Secondary Research Question: Stigma Again, this study was not designed as a stigma analysis. However, because a relevant measure, internalized homophobia, was collected, I could explore whether the relationship between gender non-conformity and rejection (paternal, maternal, and/or peer) was mediated by self-stigma. In testing the mediational model, I first explored the necessary zero-order correlations. Results revealed that the predictor variable, gender non-conformity (both femininity and masculinity), did not correlate with the mediator, internalized homophobia (the measure of self-stigma). However, internalized homophobia did correlate with all three forms of rejection (internalized homophobia and paternal rejection, r = .27, p < . 0001; internalized homophobia and maternal rejection, r = .23, p < . 001; internalized homophobia and peer rejection, r = .31, g < . 0001). Since the first necessary condition was not met, further testing of the mediational model was not conducted. Additional Analyses Recall that participants who completed the follow-up component of the study differed significantly on two health-related variables when compared to 1) those individuals who completed the survey but either did not give permission to be re-contacted and/or 2) those individuals who gave permission to be re-contacted but did not actually complete the follow-up. Childhood Factors and Attachment 51 Specifically, differences were found in HIV status and degree of physical abuse ever experienced in a same-sex relationship. Because of these significant differences, I conducted additional analyses exploring whether HIV status and/or experience of physical abuse were related to any of the key variables. Results revealed that HIV status was not correlated with any key variables. However, degree of physical abuse was significantly associated with paternal rejection (r = .19, p < .05), maternal rejection (r = .33, p < .0001), peer rejection (r = .16, p < .05), and the two adult attachment dimensions of anxiety (r = .20, p < .01) and avoidance, (r = -.17, p < .05). Notwithstanding these findings, it is unlikely that degree of physical abuse influenced results: 1) degree of physical abuse was not related to gender non-conformity, a key variable in the model and, 2) the associations between variables did not appreciably change when abuse was controlled. Zero-order correlations between the two health-related variables and the key measures are presented in Table 4. Discussion Findings from each research question are summarized and discussed in turn. Although research questions 1, 4 and 5 were intended to explore both anxiety and avoidance, the ensuing summary and discussion will focus first on anxiety as avoidance was only related to one predictor variable, peer rejection. A separate discussion exploring why the avoidance dimension yielded fewer significant results will follow. Research Question 1 In this question, I explored whether the associations between both parental and peer relationships on anxiety were consistent with attachment theory. Specifically, I queried whether paternal, maternal and peer relationship quality independently predicted anxiety and whether quality of peer relationships partially mediated the association between parental rejection and anxiety. Childhood Factors and Attachment 52 Findings from this study were somewhat consistent with attachment theory. Results showed that paternal and peer rejection independently predicted anxiety but maternal rejection did not. This latter finding is somewhat contrary to the original framing of attachment theory in which mothers were seen as primary attachment figures (Bremerton, 1985). Thus one might expect mothers to have a significant impact on adult attachment. However, research is beginning to show that paternal influence uniquely and independently explains specific child and adult outcomes over and above that explained by maternal influence (Rohner, 1998). Most relevant is a study by Barnett, Marshall, and Pleck (1992) which found that when measures of both maternal and paternal relationship quality were considered simultaneously, only paternal relationship quality was related to adult son's anxiety and depression. This may be particularly true in the case of gay men. Isay (1990) emphasized the special importance of paternal relationships in gay male identity development (although he did not minimize the importance of mothers) and asserted that paternal rejection was an important reason why some gay men have difficulty forming loving and trusting relationships in adulthood. Two findings from the adult attachment research field may help explain the results suggesting a stronger influence of fathers over mothers. Collins and Read (1990) found in their heterosexual sample that recollection of the quality of childhood relationships with opposite-sex parents predicted romantic partner attachment style dimensions. Similarly, Shaver, Belsky, and Brennan (in press) found that heterosexual married women's feelings and expectations about male romantic partners were more strongly associated with their feelings and expectations about their fathers than their mothers. It seems that for heterosexuals, opposite-sex parents influence expectations about, and dynamics within, opposite-sex relationships, possibly through a relationship modeling process. Extending the logic to gay relationships, same-sex parents then may also influence expectations about, and dynamics within, same-sex relationships. Therefore, Childhood Factors and Attachment 53 poor relationships with fathers in childhood may impact on anxiety experienced by gay men in their adult romantic relationships. As previously stated, peer rejection in addition to paternal rejection independently predicted anxiety. This is consistent with other studies, which have shown that relationships with peers independently predict interpersonal and psychological adjustment in adulthood (Epstein, 1983; Bartholomew & Horowitz, 1991; Bagwell, Newcomb, & Bukowski, 1998). The significance of peers in the present study is also consistent with work by Kurdek which showed that for gay men, peers were a more important source of social support than family (Kurdek, 1988; Kurdek & Schmitt, 1987). The findings also indicated that peer rejection largely mediated the relationship between parental (paternal and maternal) rejection and anxiety. That is, relationships with peers accounted for some of the association between parental rejection and anxiety. This latter finding is consonant with a controversial theory by Judith Rich Harris (1995, 1998). She claimed that parents matter less than peers in the development of individual differences in personality—the more global domain under which individual differences in attachment is subsumed (Bartholomew & Shaver, 1998). Harris argued that parents (as a group) may influence children's peer groups, but it is the peer group that influences individual children. Thus, she suggested that peers mediate the influence of parents on personality. She did allow that children share genes with their parents and that heritability accounted for 40-50% of the variance in personality characteristics. Notwithstanding Harris' theory, the finding that peer rejection mediated the relationship between parental rejection and anxiety is consistent with attachment theory. Relationships with parents occur temporally before relationships with peers. According to attachment theory, poor quality relationships with parents can generalize to peers possibly because of working models. Childhood Factors and Attachment 54 Working models of attachment may underlie this transmission of poor relationships (from parents to peers) because working models are cognitive/emotional/behavioural filters that affect how individuals process relationship information as well as their expectations about self worth and treatment from others. However, it should be noted that data in this study regarding parent and peer relationships were collected at the same time. While it is reasonable to speculate that influence runs from parents to peers (and not in the other direction), I cannot draw causal conclusions with contemporaneous data. Moreover, paternal and maternal rejection combined only accounted for 17% of the variance in peer rejection indicating that other factors beyond parental treatment influence peer rejection. For example, peers might have rejected an individual because the individual violated a peer group norm to which parents were oblivious (e.g., the child fails to conform to fashion trends). If parents were oblivious to the peer group norm, it is not likely that this was an influential factor underlying their rejection. Research Question 2 In this second question I explored whether gay men reported poorer quality childhood relationships with their fathers than mothers. Consistent with the literature, results showed that fathers were seen as more rejecting than mothers were, and fathers were seen as less warm than mothers. Moreover, recall that gay men in this sample were more likely to report paternal and maternal rejection when compared to (primarily) heterosexual men in the Henderson (1998) sample. It should be noted that data regarding relationships with parents were retrospective and thus susceptible to memory distortions. It may be that the fathers (and mothers) of gay men were more rejecting in childhood, possibly in response to gender non-conforming behaviour. However, the rejection may have started later in life, when the gay men 'came out' to their Childhood Factors and Attachment 55 families (80% of the sample reported that most or all of their family knew that they were gay). It might be this current alienation that accounted for recollection of paternal (and to a lesser extent, maternal) rejection in childhood. Isay (1990) went one step further and argued that recollection of rejecting relationships with fathers could be defensive memory distortions arising from anxiety gay men feel about their early erotic attachment to their fathers. Research Question 3 In this question I addressed whether the relationship between gender non-conformity and paternal rejection and the relationship between gender non-conformity and peer rejection were stronger than the relationship between gender non-conformity and maternal rejection. Femininity. All three forms of relationship rejection (paternal, peer and maternal) were significantly associated with femininity and the respective correlation coefficients did not differ significantly. That is, the more effeminate a boy was the more likely he was to report having experienced maternal, paternal, and peer rejection. The finding that maternal rejection was positively associated with gender non-conformity runs contrary to findings in the literature (see Langlois & Downs, 1980; Seigleman, 1974; Sipova & Brzek, 1983). However, the magnitude of the effect for the observed association was small (r = .22). Therefore, previous studies may have had insufficient power to detect the association between maternal rejection and femininity. Masculinity. Only peer rejection was negatively associated with masculinity. That is, the more masculine a boy was, the less likely he was to be rejected by his peers. The strength of the association between masculinity and peer rejection was stronger than the association between masculinity and maternal rejection. Results also revealed that paternal warmth was positively associated with masculinity. It is not surprising that in childhood and adolescence, peer rejection was negatively associated with masculinity. Saghir and Robins (1973) report that many of their participants Childhood Factors and Attachment 56 recalled being teased during childhood because of their gender non-conformity. Moreover, Adler, Kless, and Adler (1992) report that pre-adolescent boys achieve high status on the basis of their athletic ability as well as coolness, toughness, social skills and success in cross-gender relationships. In looking at the masculine items in the BG(N)CS (e.g., As a child I imagined or wished I was a well-known sports figure), it becomes apparent that this scale indirectly assessed interest (and perhaps ability) in athletics. Since proficiency in this domain is crucial for the social success of boys, it follows that the more masculine a boy was, the less likely he was to be rejected. Boys' athletic interests and abilities may particularly affect fathers and peers. This might account for the finding that the association between masculinity and peer rejection was stronger than that between masculinity and maternal rejection and also that paternal warmth but not maternal warmth was associated with masculinity. That is, peers may have been less interested in befriending boys who did not share interests in athletic pursuits whereas this may not have been a relevant concern for mothers. Similarly, athletics may have been a domain, which fathers, more so than mothers sought to share with sons. Thus, fathers compared to mothers might have been more concerned with, and more affirming of, their sons' masculinity expressed in the form of athletic interest and ability. Research Question 4 Findings from the fourth question are consistent with the idea that relationship quality (paternal, and/or maternal, and/or peer) mediate the association between gender non-conformity and anxiety. Rejection by peers (and to a lesser extent, rejection by fathers) mediated the relationship between femininity and anxiety, and that rejection by peers (and to a lesser extent, paternal warmth) mediated the relationship between masculinity and anxiety. That is, gender non-Childhood Factors and Attachment 57 conformity (high femininity and low masculinity) may be associated with anxiety in close adult relationships primarily because it contributed to peer rejection, which, in turn, influenced anxiety. Because the association between peer rejection (and to a lesser extent, paternal warmth) and gender non-conformity has been addressed in Research Questions, further discussion of this particular finding is not necessary. A benefit of path analysis is that one can test opposing models. As noted, I did test an alternate model: relationship quality led to gender non-conformity (femininity and masculinity) which, in turn, led to anxiety. These tests showed that mediation was unlikely, as the effects of the predictors on the criterion were not substantially reduced when the mediator was controlled. Beyond statistical evidence, the alternate model(s) are theoretically implausible for two reasons. First, temporally, gender non-conforming behaviour appears as early as age 2 (Fagot, 1985; Huston, 1983). While children interact with their fathers before this age, they do not have much exposure to peers until later, after gender non-conforming behaviours are already manifest. Thus, it did not make sense to predict that peer rejection led to gender non-conformity. Second, although rejected children are known to act out aggressively or passively withdraw (e.g., Fraser, 1996; Williams & Gilmour, 1994), there is no theoretical reason or evidence to suggest that these children respond with gender non-conforming behaviour. However, as reported above, substantial evidence suggests that fathers and peers reject boys for gender non-conforming behaviour. Thus, the tested path model was chosen because it was theoretically grounded and psychologically meaningful. Research Question 5. I explored whether having a good relationship in one domain (i.e., either with mothers or fathers or peers) compensated for a poor relationship in another domain in terms of anxiety. Three hierarchical regressions were conducted to examine compensation among the following Childhood Factors and Attachment 58 relationships: 1) mother and fathers, 2) mothers and peers, and 3) fathers and peers. The first two hierarchical multiple regressions indicated both paternal and peer rejection independently predicted anxiety but maternal rejection did not. Results from the third hierarchical multiple regression (fathers and peers) were more novel. Here, both main effects of paternal and peer rejection were significant as was the interaction. Further inspection of the interaction revealed that anxiety was high for participants who experienced high levels of paternal rejection or high levels of peer rejection or high levels of both paternal and peer rejection. Only those participants who experienced low paternal rejection and low peer rejection had low levels of anxiety. Thus, a good quality relationship in one domain did not appear to compensate for a bad quality relationship in another domain. This is a particularly interesting finding because it suggests that there may be, to some degree, a rejection threshold. If a significant level of rejection is experienced, irrespective of the source (father or peer), anxiety may result. Moreover, it seems that the level of rejection can come from one source or accumulate across relationships since rejection was highest when both fathers and peers were rejecting. If I had used a more specific measure of peer relationships which explored opposite-sex and same-sex peer friendships separately, I may have been more likely to find compensation effects of opposite-sex friendships on gay men's anxiety. Bieber and Bieber (1979) found that gay men experienced difficulties with their same-sex friendships because they were not suited to the aggressive and combative ways typical of boys' peer group interactions. Adler et al., (1992), however, reported that emotional expression, intimacy and cooperation characterized girls' friendships. This suggests that girls might be more accepting of gender non-conforming (sensitive, non-aggressive, non-competitive) boys. Additionally, Isay (1990) claimed that it is not unusual for gay men to have close friendships with women throughout their lives. Childhood Factors and Attachment 59 Replication of Research Question 5 using a well-delineated measure of same-sex and opposite-sex peer relationship quality is necessary to more fully explore whether good peer relationships can compensate for poor paternal relationships. Secondary Research Question: Stigma Results showed that internalized homophobia, a measure of self-stigma, did not mediate the relationship between gender non-conformity and parental or peer rejection. This is because internalized homophobia was correlated with rejection but not gender non-conformity. To some degree, the failure to find a mediation relationship between these variables may be due to the limitations of the self-stigma measure. Internalized homophobia is only one aspect of stigma- a multi-dimensional construct. It is implied in this thesis that, to some degree, gender non-conformity violates social norms and therefore has the potential to be a source of stigmatization. This thesis, however, is not explicitly framed as a stigma analysis. If one wants to conduct an empirical investigation conceptualizing gender non-conformity as stigma, and thereby explicitly frame the study in terms of stigma, one must first operationalize and measure the multi-dimensional nature of stigma. Recall that stigma occurs at three levels: 1) the societal level, 2) the dyadic level, and 3) the individual level. This would involve then first measuring the social norms surrounding the gender non-conformity construct, which is not an easy task since people live in, and are affected by overlapping social worlds. Therefore, at the very onset, one must ascertain what constitutes the society in question (Stafford & Scott, 1986). Following that, one must then attempt to measure something as amorphous as social attitudes, all the while recognizing that even if one accurately measures attitudes, this does not necessarily make predicting subsequent behavior such as rejection any easier (Kraus, 1995). Childhood Factors and Attachment 60 Secondly, stigma is a relational construct. At the dyadic level, therefore, one must measure the extent to which individuals interacting with the marked person not only perceives the social norms but also associates the discrediting quality with the possessor (Jones et al., 1984). In the current example, this means that one would need to verify that parental and peer reactions to participants were a result of a social proscription against gender non-conformity. Such an endeavor would require that data be collected not only from participants but also from their fathers, mothers, and peers. From a purely practical point of view, collecting corroborating data, 20, 30 or 40 years after the fact is not possible. Parents may live in other cities, may have distant relationships with participants as a result of their sexual orientation, or may no longer be living. In fact, it could be argued that soliciting information from parents for cross-validation purposes in a study focusing on the associations between parent/childhood relationship quality and adult attachment would be unethical, as it could put participants at risk for further rejection. Similarly, most people lose contact with their childhood associates, especially if the relationship was characterized by rejection. Thus, if stigma was to be explicitly evoked as the mechanism underlying the association between gender non-conformity and rejection, the required data collection could become so onerous that most researchers would give up before they even started their investigation. Identity engulfment is central to the stigmatization dynamic. That is, for there to be sufficient evidence of stigmatization it would need to be shown that parents' and peers' impressions of participants were eclipsed by the latter's gender non-conformity (Jones et al., 1984). However, participants' reports of childhood relationships indicated that parents and peers impressions of participants were not totally engulfed or eclipsed by gender non-conformity. Recall that participants reported that both their fathers and mothers were occasionally warm. Childhood Factors and Attachment 61 They also reported having childhood and adolescent peer relationships that were more likely to be characterized as accepting rather than rejecting. Even if there is social consensus that a quality is discrediting, and even if others judge an individual because of the quality, one cannot conclude that full stigmatization is the end result. For stigma to be present, marked individuals must also perceive that they are being discredited because of the quality. Explicitly framing the associations between gender non-conformity and rejection in terms of stigma, therefore, requires evidence that participants internalized the stigma and that it affected their self-concept. In actuality, the opposite seemed to be true. Seventy percent of participants reported that they had not internalized negative social views regarding homosexuality. Additional Analyses Results revealed that HIV status was not correlated with any of the key variables. This indicates that HIV status did not have a direct bearing on any key findings. Degree of physical abuse was significantly associated with rejection (paternal, maternal, and peer), and the two adult attachment dimensions of anxiety and avoidance. However, degree of physical abuse did not have a significant influence on findings as it was not associated with gender non-conformity nor did the associations between key variables in the model change appreciably when degree of physical abuse was controlled. One cannot, therefore, conclude that the associations between gender-non-conformity, relationship quality, and anxiety were simply a function of HIV status or degree of physical abuse. Findings Related to Avoidance As previously reported, neither gender non-conformity (femininity and masculinity), paternal rejection, nor maternal rejection were associated with the avoidance attachment dimension (the only significant association was between peer rejection and avoidance). This Childhood Factors and Attachment 62 failure to find significant associations made it impossible to explore Research Questions 1,4 and 5 in terms of avoidance. One reason that associations with avoidance may not have been observed is because avoidance was a composite score. Recall, avoidance was derived from the four attachment prototypes by summing dismissing and fearful scores and subtracting secure and preoccupied scores. The end result is that this combining of prototypes may have obscured patterns of findings associated with the individual prototypes. Femininity. Inspection of the correlations between femininity and the individual prototypes showed that there was a negative association between the dismissing style and gender non-conformity (r = -.20), and a non-significant positive association between the fearful style and femininity (r = .04). Added together, these two resulted in a slight negative association. On the other side of the equation, there was a non-significant negative association between the secure style and femininity (r = -.08), and a significant positive association between the preoccupied style and femininity (r = .20). Added together, these two resulted in a slight positive association. Thus, one side of the equation reduced the effect of the other, rendering non-significant any association between femininity and avoidance. The negative association between dismissing and femininity was somewhat predictable. The dismissing style is an insecure form of attachment associated with self-competence (Brennan & Morris, 1997). It is adaptive for men in our society because it is consonant with societal expectations and norms for appropriate male behaviour. The adaptability of the dismissing style is partially evidenced by research which showed more men than women were categorized as predominantly dismissing (Bartholomew & Horowitz, 1991; Brennan, Shaver, Tobey, 1991). To some degree, success for boys in our society comes through athletic competence (Adler et al., 1992), which is probably inversely related to gender non-conformity. Therefore, given that athletic competence for boys is also probably related to self-competence, dismissing participants Childhood Factors and Attachment 63 who are concerned with self-competence may have also tended to be gender conformists (relatively low in femininity) during childhood. Alternatively, dismissing individuals are also prone to idealization (Bartholomew & Horowitz, 1991). It could be that these individuals reconstructed their memories of childhood to be consonant with societal expectations regarding masculinity. Perhaps the more dismissing an individual, the more likely he would be to downplay his childhood effeminate experiences and behaviours. The positive association between preoccupied and femininity was also predictable. Preoccupied individuals are overwhelmingly approach oriented. Irrespective of whether they were rejected or ostracized for their effeminate behaviours in childhood, these individuals would seek out others as a way to reduce felt anxiety. In fact, so intent are they to seek connection with others, that they are often oblivious to social cues suggesting inappropriate behaviour. Thus, data were consistent with the explanation that for a preoccupied individual, the greater his level of effeminate behaviour and concomitant rejection in childhood, the more likely he was to feel anxiety which would result in increased attempts to seek and maintain closeness with others. Masculinity. Inspection of the correlations between masculinity and the individual prototypes showed that there was a non-significant positive association between the dismissing style and masculinity (r = .12), and a significant negative association between the fearful style and masculinity (r = -. 15). Added together, these two resulted in a slight negative association. On the other side of the equation, there was a significant positive association between the secure style and masculinity (r = . 17), and a non-significant negative association between the preoccupied style and masculinity (r = -.09). Added together, these two resulted in a slight positive association. Thus, in taking both sides of the equations into consideration, a non-significant negative association between masculinity and avoidance was observed. Childhood Factors and Attachment 64 The positive association between security and masculinity may have been because the secure attachment style, like the dismissing style, is for the most part adaptive for men. It typically reflects a normative pattern of responding in childhood. As with the dismissing style, the secure style is associated with self-competence; but unlike the dismissing style, it is also associated with self-liking (Brennan & Morris, 1997). Again, it is plausible that the more stereotypical a boy was during childhood in terms of masculinity (and possible concomitant athletic skills), the more likely he was to experience validation and social success, which would, in turn, affirm his sense of self. The negative association between fearful and masculinity can be understood in terms of Bartholomew's conceptualization of the self- and other-models (Bartholomew, 1990; Bartholomew & Horowitz, 1991). Specifically, fearful individuals have both a negative sense of self and a negative sense of others. Thus, lower levels of masculinity in childhood might have lead to a more general sense of personal shortcomings in adulthood. With fearful individuals, this feeling of personal inadequacy is expressed in the avoidance of close relationships. Parental rejection. In terms of paternal rejection, inspection of correlations between this variable and the individual prototypes showed a non-significant negative association between dismissing and rejection (r = -.11), and a significant positive association between fearful and rejection (r = .17). There was also a significant negative association between secure and rejection (r = -.27), and a positive association between preoccupied and rejection (r = .16). Thus, one side of the equation reduced the effect of the other, rendering non-significant any association between paternal rejection and avoidance. In terms of the correlations between the individual prototypes and maternal rejection, there was a non-significant negative association between dismissing and rejection (r = -.01), and a non-significant positive association between fearful and rejection (r = .04). There was also a Childhood Factors and Attachment 65 negative association between secure and rejection (r = -.16) and a positive association between preoccupied and rejection (r = .22), again canceling each other and reducing to near zero any association between maternal rejection and avoidance. In contrast to the positive association between peer rejection and avoidance, it was not surprising that there was no association between parental rejection and avoidance. In dealing with rejecting peers as opposed to parents, individuals have considerably more latitude. Although it might be difficult, children can escape their peers and engage in solitary activities or possibly befriend an equally rejected peer. Avoidance with peers during childhood and adolescence then, could have been a successful coping style that was sufficiently reinforced to be carried on into adulthood. Conversely, dealing with rejecting parents can be more difficult because they are not as easy to avoid physically or psychologically, especially if the parents are intent on drawing the child out. Given the inherent power differential, the child is often at the mercy of his parents' demands. Moreover, at the end of the day, the child must eventually return to his family home. General Considerations, Strengths, Limitations and Conclusions General Considerations Personal developmental theories. It could be argued that the obtained pattern of results between variables might not reflect developmental reality so much as it reflects participants' present day theories of their developmental reality. Specifically, it could be reasoned that participants made certain attributions about their relationship anxiety that do not represent actual patterns. However, participants were never directly asked about relationship anxiety and may never have known this was a facet of the investigation; they were simply asked to tell the story of their relationship histories. Measurement of relationship anxiety was after-the-fact, based on expert codings of participants' narratives. Thus, it seems implausible that participants could put Childhood Factors and Attachment 66 forth their personal theories about a domain they were not even aware they were being asked about. Indeed, because participants were blind to the study's general hypotheses and data came from two different sources, self-report and expert ratings, there is less likelihood that the findings are simply an expression of participants' own theories. Moreover, while gender non-conformity is reported in the literature to be highly characteristic of gay men's childhood experiences, evidence also shows that discussion of gender non-conformity is not common in the gay community. Rofes (1994) stated, "Despite 25 years of gay liberation work..., there has been overwhelming silence about gay men's youthful experiences as sissies....To say sissies = gay male youth is considered offensive by many in the gay community" (p. 38). Thus, if as implied by Rofes, members of the gay community are intentionally silent about a particular aspect of gay male development, it seems unlikely that a large group of gay participants would systematically articulate a taboo personal theory. Finally, if participants were simply expressing their own personal theories, it is likely that both outcome measures, anxiety and avoidance, would be similarly influenced. That is, the findings would unfold in a clear and consistent manner, but this was not the case. In fact, no central findings related to the attachment dimension of avoidance were significant. The observed variation in findings suggests that factors beyond personal theories account for the given results. This is not a study about the causes of gay male sexual orientation. Researchers and theorists interested in the etiology of sexual orientation have sometimes looked at variables included in this study, specifically, gender non-conformity and parental rejection (e.g., Satinover, 1996). Notwithstanding the overlap of interest in these variables, it should be clear that this study was about the relationship between childhood socialization factors and adult attachment in gay men which is a necessary and worthy line of inquiry in its own right. This study was not intended or designed to predict or explain the causes of sexual orientation. To do this would Childhood Factors and Attachment 67 have required the inclusion of heterosexual men in the study. As such, this study cannot speak to the debate surrounding the nature or origins of gay male sexual preference. Insecure attachment in this sample. Interestingly, only 20% of this sample were rated as predominantly secure based on the history of attachment interviews. This is markedly lower than the 47%-51% security rates found in college student samples using the same methodology. It is also lower than the 59% security rate found in Mickelson et al.,'s (1997) nationally representative survey. Some of the difference between the Mickelson et al., (1997) and the present findings may be due to their use of self-report measures of attachment. However, additional corroboration of the low security levels in this sample come from self-report data that were collected, but not used, in this study (attachment interview ratings were the measure of choice used in this study because they show better reliability and validity—see Bartholomew, Henderson, & Marcia, in press). Comparable data from this sample showed that 38% of participants' self-reported being secure—the lowest level of self-reported security in any published studies of non-clinical samples. Even though findings were suggestive of heightened insecurity, it can not be concluded from this study that gay men are more insecure than heterosexual men as this study did not include a heterosexual comparison group. Recall that Ridge and Feeney (1998) looked at differences in relative frequency of attachment styles between gay and heterosexual men and found no significant difference. However, they used a self-report measure of attachment that was not necessarily valid in gay populations (Callander, 1999). Future research investigating sexual orientation differences in attachment, using the appropriate comparison group and sound methodology including expert ratings of attachment interviews, is still needed before any conclusions can be drawn. Any study of this nature would do well to consider the influence of Childhood Factors and Attachment 68 paternal and peer rejection on adult attachment, as well as societal pressures and internalized homophobia that might have a unique impact on gay men. Implications for policy. It is premature to draw conclusions for policy based on the findings of one basic research project. Notwithstanding this caveat, one could speculate about potential policy directions. Since findings from this study involve two social systems: family and peer, two different strategies would need to be considered. Families, however, are often closed systems that are not always available to the purview of social policy makers. Conversely, childhood interactions with peers often occur in public domain such as schools and social clubs (e.g., Boys Scouts) which are amenable to social pressures. Therefore, it seems changes could most easily be implemented in the latter system. Indeed, schools and social clubs are well suited to implement policies and practices that are affirming of all children irrespective of unconventional gender behaviour. There are a number of ways that implementation could occur. First, teachers and administrators might be made aware of, and sensitive to, bullying based on gender non-conformity. In addition to awareness, teachers could be encouraged to intervene and overtly disapprove of such bullying instead of turning a blind eye, which indicates tacit approval. Second, change could be made regarding the ways in which boys are affirmed. Currently, boys most often receive attention and praise for their athletic competence. However, a wide range of non-athletic, extracurricular activities could be made available and competence in these latter activities more readily affirmed. Finally, school gym classes could be designed to not draw attention to athletic competence. One way to do this would be to stop the practice of having students select one another for team membership, which invariably excludes and humiliates the non-athletic child. Instead, the focus of gym classes could be placed on more communal aspect of general health and fitness (Rofes, 1994). Childhood Factors and Attachment 69 Strengths Recruitment of a broad community sample was a notable strength in this study. Although the 49% response rate for the initial telephone survey was somewhat of a disappointment, the 65% response rate for the follow-up component was much higher than expected. This response rate can be directly compared to the 10% response rate (based on the number of originally surveyed participants who completed follow-up questionnaires and interviews) reported by Henderson (1998) who used a similar methodology. Albeit, Henderson's pool of original participants was much larger, 1200 compared to 300 in the present study. Further, she surveyed the Lower Vancouver Mainland, a broader geographic territory than the West End of Vancouver, which made conducting interviews more difficult. One significant factor contributing to the high response rate is the population that was studied, that is, gay men. Each man who was interviewed was asked why he agreed to say "yes" to the follow-up component of the study. The overwhelming response to this question was that participants felt that there was a societal need for objective, scientific inquiry into gay male relationships. Participants valued the goals of the research project and expressed confidence in the researchers' ability to objectively study and draw conclusions from the given information. Many participants conveyed to the researchers that they felt that the study was "theirs", that it "belonged to them" and that they had a responsibility as gay men to participate. Limitations A problem inherent in relationships research generally, but magnified in research with the gay community, is the lack of representativeness of the samples. Even though a community sample of gay men was recruited for this study using a random sampling methods (book-plus design), this did not mean that the sample was representative of all gay men. At best, these data could only be representative of openly gay men living in a well-known gay area of Vancouver. A Childhood Factors and Attachment 70 review of the demographic data indicated that participants in this study were culturally homogenous (the majority of British origin), well-educated, middle-class, urban men. This can be contrasted with Harry's (1990) probability sample of gay men in the United States which found that a substantial number of gay men were married, members of a racial minorities, lived in rural communities, and did not participate in a gay community. The overall response rate for the initial telephone survey was 49%. That is, in the initial telephone survey about half of all eligible respondents chose not to participate. It is not possible to know the various reasons underlying this self-selection process as the potential respondents declined to participate at the outset of the study, before any data were collected. It could be the case that factors salient to the issues under investigation were either minimized or maximized by self-selection, but the exact impact on the present findings remains unknown. What is known is that given the 49% response rate, random sampling methods did not necessarily result in a random sample. On a more positive note, Rosenthal and Rosnow (1975) have outlined ways in which participant volunteers differ from non-participant volunteers on a number of dimensions. They offer ten suggestions to reduce bias among volunteers, all of which were followed in the present study (see Appendix 8 for a complete list of their suggestions). There is some indication that respondents who complete the follow-up component of the study were more vulnerable or "at risk" than those respondents who only completed the survey component of the study. The former compared to the latter were more likely to be HIV positive and to have experienced a greater degree of physical abuse in same-sex relationships. Thus, some of the self-selection to participate in the follow-up study may be because the issues explored, such as adult close (romantic) relationships, were more relevant to vulnerable or "at risk" persons. Moreover, Schilder and Kort (1996) asserted that gender non-conforming boys are more likely to be targets of emotional, physical, or sexual abuse and that this may make them Childhood Factors and Attachment 71 i more vulnerable to suicide and high risk sexual behaviour leading to HIV infection in adulthood. Since participants who completed the study compared to those who did not were more likely to be HIV positive and to have experienced past abuse, it might be the case that gender non-conformity was also particularly prevalent in this sample. Thus, given that a certain amount of self-selection is evident, ability to generalize findings beyond this sample is reduced. Conclusions Little empirical research on same-sex relationships has been conducted to date. The research that is available is either descriptive, atheoretical, methodologically flawed, and/or has taken a deviance perspective (Allen & Demo, 1995; Deenen et al., 1995; Harry, 1990; Mohr, 1999; Sell & Petrulio, 1996). The field of Psychology, with its use of the scientific method and concomitant objectivity, is particularly well suited to empirically explore topics related to same-sex relationships. Unfortunately, the quantity and quality of psychological research on this topic has also been sorely lacking (Allen et al., 1995; Deenen et al., 1995). Many social psychological theories, especially those in the relationship research area, would lend themselves to the study of same-sex relationships. However, it is not sufficient to simply replicate opposite-sex relationship findings in same-sex populations. There is a real need to extend theories and incorporate factors specific and unique to gay relationships. In this thesis, I have taken attachment theory, a panhuman, ethological theory, and incorporated within it factors unique to gay male socialization. Thus, I took a validated theory, extended it to accommodate gay male childhood experiences and tested the fit between theory and adult attachment outcome measures. In summary, I looked at the associations between paternal, maternal, and peer relationship quality, childhood gender non-conformity (boyhood femininity and masculinity), and the adult attachment dimension of anxiety and avoidance. My findings were partially consistent with Childhood Factors and Attachment 72 attachment theory in that childhood relationship quality with fathers and peers (but not mothers) were important, independent predictors of attachment-related anxiety in adulthood. In fact, findings indicated that parental rejection might have influenced the way gay men experienced peer relationships in childhood, and this in turn, might have influenced current levels of anxiety in adult relationships. Moreover, gender non-conformity in childhood, a prominent concept in the gay male sexual orientation literature, was also a relevant factor in predicting adult relationship-related anxiety, possibly because it predisposed individuals to have poorer quality relationships with fathers and peers. Bearing in mind issues related generalizability, perhaps the most novel and significant conclusion that can be drawn from this study is that childhood factors such as gender non-conformity, and relationships with fathers and peers can be used to partially understand important features of gay men's subsequent adult attachment relationships. Beyond the novel and significant conclusions that can be drawn from this study, this study makes an important contribution to the field in other ways. With two known exceptions, there has been very little work done linking childhood factors such as gender non-conformity and parental and peer relationship quality with adult attachment in gay men. The first exception is Colgan's (1987) theoretical paper, which speculated that gender non-conformity was associated with childhood rejection contributing to over-attachment and over-separation in gay men. Recall, however, that Colgan (1987) never tested his hypothesis. Thus, this study is the first to empirically explore these associations. The second exception is the empirical study by Ridge and Feeney (1998) which looked at sexual orientation, recollected relationship quality with parents, and adult attachment. Notwithstanding the fact that Ridge and Feeney (1998) failed to include gender non-conformity in their analysis or include peer relationships as an additional socialization variable, their study was plagued by a number of serious limitations. These include use of a moderate sized Childhood Factors and Attachment 73 convenience sample of gay college students, and poor self-report measures of both adult attachment and parental relationship quality. The former limitation reduces generalizability beyond the sample studied, while the latter limitation reduces predictive power in finding hypothesized effects. In comparison, this project empirically explored a barely touched but enormously relevant aspect of gay men's lives using the best-known techniques available. These include a large community sample of gay men generated through random sampling, comprehensive self-report measures of recollected relationship quality with parents and peers (a 43-item scale and a 40-item scale, respectively), and a well-validated, thorough measure of adult attachment. Moreover, this thesis heeds the recommendations set out by both Sell and Petrulio (1996) and Deenen et al., (1995) for research conducted with gay men by using random sampling techniques, incorporating theory (attachment theory) from the general field of interpersonal relationships, and using correlational methods. This thesis, though, is only a first step and looks at a particular piece of the overall picture. Further research using sound theory and methods is necessary to truly understand gay male relationships. For example, collection of longitudinal data from prospective studies with gender non-conforming boys seems warranted. In deed, because much of the present findings were based on self-report and retrospective data, it would be useful to have corroborative information from siblings, parents, peers or teachers. Final thoughts. In a modern era where societal norms now encourage the increased role of fathers in child development, depending on the degree of concomitant paternal sensitivity, one might expect to see both the best and the worst outcomes for gender non-conforming boys. Results from this and other studies show that having a rejecting father in childhood is associated with negative aspects of adult relationships. Conversely, the affirming presence of a father in Childhood Factors and Attachment 74 childhood can contribute positively to trusting, warm, caring adult relationships. Hopefully, just as society has changed its perspective on the role of fathers in child development, there can be a similar change in societal tolerance of boyhood gender non-conformity. These two changes together, may have an overall beneficial influence on attachment security experienced by gay men. Childhood Factors and Attachment 75 References Adler, P. A., Kless, S. J., & Adler, P. (1992). Socialization to gender roles: Popularity among elementary school boys and girls. Sociology of Education, 65, 169-187. Ainlay, S. C, Coleman, L. M., & Becker, G. (1986). Stigma reconsidered. In S. C. Ainlay, G. Becker, & L. M. Coleman (Eds.), The dilemma of difference: A multi disciplinary view of stigma (pp. 1-13). New York: Plenum Press. Ainsworth, M.D.S.(1989). Attachments beyond infancy. American psychologist, 44, 709-716. Ainsworth, M.D.S., Blehar, M.C., Waters, E., & Wall, S. (1978). Patterns of attachment: A psychological study of the Strange Situation. Hillsdale, NJ: Lawerence Erlbaum. Allen, K. R., & Demo, D. H. (1995). The families of lesbian and gay men: A new frontier in family research. Journal of Marriage and the Family, 57, 111-127. Apperson, L. B., & McAdoo, W.G. Jr. (1968). Parental factors in the childhood of homosexuals. Journal of Abnormal Psychology, 73, 201-206. Armsden, G. C, & Greenberg, M. T. (1987). The inventory of parent and peer attachment: Individual differences and their relationship to psychological well-being in adolescence. Journal of Youth and Adolescence, 16, 427- 454. Bagwell, C. L., Newcomb, A. F., & Bukowski, W. M. (1998). Preadolescent friendship and peer rejection as predictors of adult adjustment. Child Development, 69, 140-153. Bailey, J. M., Kim, P. Y., Hills, A., Linsenmeier, J. A. W. (1997). Butch, femme, or straight-acting? Partner preferences of gay men and lesbians. Journal of Personality and Social Psychology, 73, 960-973. Bailey, J. M., Miller, J. S., & Willerman, L. (1993). Maternally rated childhood gender non-conformity in homosexuals and heterosexuals. Archives of Sexual Behavior, 22, 461-469. Childhood Factors and Attachment 76 Bailey, J. M., & Zucker, K. J. (1995). Childhood sex-typed behavior and sexual orientation: A conceptual analysis and quantitative review. Developmental Psychology. 3_1, 43-55. Barnett, R. C, Marshall, N. L., & Pleck, J. H. (1992). Adult son-parent relationships and the associations with sons' psychological distress. Journal of Family Issues, 13, 505-525. Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology. 51, 1173-1182. Bartholomew, K. (1990). Avoidance of intimacy: An attachment perspective. Journal of Social and Personal Relationships. 7, 147-178. Bartholomew, K., Henderson, A., & Marcia, J. (in press). Coded semi-structured interviews in social psychological research. In H. Reis & C. Judd (Eds.), Handbook of research methods in social psychology. Cambridge University Press. Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults: A test Of a four-category model. Journal of Personality and Social Psychology, 61, 226-244. Bartholomew, K., & Shaver, P. R. (1998). Methods of assessing adult attachment. In J. A. Simpson & W. S. Rholes (Eds.), Attachment theory and close relationships (pp. 25-45). New York: Guilford Press. Bell, A. P., Weinberg, M. S., & Hammersmith, S. K. (1981). Sexual preference: Its development in men and women. New York: Simon and Schuster. Bieber, I., & Bieber, T. B. (1979). Male sexuality. Canadian Journal of Psychiatry, 24, 409-421. Blumstein , P., & Schwartz, P. (1983). American couples: Money, work, sex. New York: Morrow. Childhood Factors and Attachment 77 Boswell, J. (1980). Christianity, social tolerance, and homosexuality. Chicago: University of Chicago Press. Bowlby, J. (1973). Attachment and loss: Vol. 2. Separation. New York: Basic Books. Bowlby, J. (1979). The making and breaking of affectional bonds. London: Tavistock. Bowlby, J. (1988). A secure base: Parent-child attachment and healthy human development. USA: Basic Books, Inc. Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult attachment: An integrative overview. In J. A. Simpson & W. S. Rholes (Eds.), Attachment theory and close relationships (pp. 46-76). New York: Guilford Press. Brennan, K. A., & Morris, K. A. (1997). Attachment styles, self-esteem, and patterns of seeking feedback from romantic partners. Personality and Social Psychology Bulletin, 23, 23-31. Brennan, K. A., Shaver, P. R., & Tobey, A. E. (1991). Attachment styles, gender, and parental problem drinking. Journal of Social and Personal Relationships, 8, 451-466. Bremerton, I. (1985). Attachment theory: Retrospect and prospect. Monographs of the Society for research in Child Development, 50 (1-2), Serial No. 209, 3-35. Buhrich, N., & McConaghy, N. (1978). Parental relationships during childhood in homosexuality, transvestitism and transsexualism. Australian and New Zealand Journal of Psychiatry, 12, 103-108. Callander, M. L. (1999). Validation of the four-category model of attachment in a sample of gay men. Unpublished Masters thesis, Simon Fraser University, Burnaby, British Columbia. Cameron, P., & Cameron, K. (1995). Does incest cause homosexuality? Psychological Reports, 76, 611-621. Childhood Factors and Attachment 78 Carter, D. B., & McCIoskey, L. A. (1983). Peers and the maintenance of sex-typed behavior: The development of children's conceptions of cross-gender behavior in their peers. Social Cognition, 2, 294-314. Colgan, P. (1987). Treatment of identity and intimacy issues in gay males, Journal of Homosexuality, 14, 101-123. Collins, N. L., & Read, S. L. (1990). Adult attachment, working models and relationship quality in dating couples. Journal of Personality and Social Psychology, 61, 644-663. Collins, N. L., & Read, S. L. (1994). Cognitive representations of attachment: The content and function of working models. In K. Bartholomew & D. Perlman (Eds.), Advances in personal relationships: Attachment processes in adulthood (Vol. 5, pp. 53-90). London, England: Jessica Kingsley. Cooper, C. R., & Cooper, R. G. (1992). Links between adolescents' relationships with their parents and peers: Models, evidence, and mechanisms. In R. D. Parke & G. W. Ladd (Eds.), Family-peer relationships: Modes of linkage (pp. 135-158). Hillsdale, NJ: Lawrence Erlbaum Associates. Cronbach, L. J., & Furby, L. (1970). How should we measure "change"—or should we? Psychological Bulletin, 74, 68-80. Deenen, A. A., Gijis, L., & van Naerssen, L. X. (1995). Thirty-five years of research into gay relationships. Journal of Psychology & Human Sexuality, 7, 19-39. DiPlacido, J. (1998). Minority stress among lesbian, gay men, and bisexuals: A consequence of heterosexism, homophobia, and stigmatization. In G. M. Herek (Ed.), Stigma and sexual orientation: Understanding prejudice against lesbians, gay men, and bisexuals (pp. 138-159). Thousand Oaks, California: Sage Publications, Inc. Childhood Factors and Attachment 79 Doll, L. S., Joy, D., Bartholow, B. N., Harrison, J. S., Bolan, G., Douglas, J. M., Saltzman, L. E., Moss, P. M., & Delgado, W. (1992). Self-reported childhood and adolescent sexual abuse among adult homosexual and bisexual men. Child Abuse and Neglect, 16, 855-864. New York: Free Press. Dutton, D. G. (1994). Behavioral and affective correlates of Borderline Personality Organization in wife assaulters. International Journal of Law and Psychiatry, 17, 265-279. Dutton, D. G., Saunders, K., Starzomski, A., & Bartholomew, K. (1994). Intimacy-anger and insecure attachment as precursors of abuse in intimate relationships. Journal of Applied Social Psychology, 24, 1367-1386. Dynes, W. R. (1987). Homosexuality: A research guide. New York: Garland. Elicker, J., Englund, M., & Sroufe, L. A. (1992). Predicting peer competence and peer relationships in childhood from early parent-child relationships. In R. D. Parke & G. W. Ladd (Eds.), Family-peer relationships: Modes of linkage (pp. 77-106). Hillsdale, NJ: Lawrence Erlbaum Associates. Epstein, S. (1983). The mother-father-peer scale. Unpublished manuscript, University of Massachusetts-Amherst. Evans, R. B. (1969). Childhood parental relationships of homosexual men. Journal of Consulting and Clinical Psychology, 33, 129-135. Fagot, B. I. (1985). Changes in thinking about early sex role development. Developmental Review, 5, 83-98. Feeney, J., & Noller, P. (1996). Adult attachment. Thousand Oaks, CA: Sage Publications, Inc. Fraley, R. C, & Davis, K. E. (1997). Attachment formation and transfer in young adults' close friendships and romantic relationships. Personal Relationships, 4, 131-144. Childhood Factors and Attachment 80 Fraley, R. C, & Waller, N. G. (1998). Adult attachment patterns: A test of the typological model. In J. A. Simpson & W. S. Rholes (Eds.), Attachment theory and close relationships (pp. 77-114). New York: Guilford Press. Fraser, M. W. (1996). Aggressive behavior in childhood and early adolescence: An ecological-developmental perspective on youth violence. Social Work, 41, 347-361. Freund, K., & Blanchard, R. (1983). Is the distant relationship of fathers and homosexual sons related to the sons' erotic preference for male partners, or to the sons' atypical gender identity, or to both? Journal of Homosexuality. 9, 7-25. Freund, K., Langevin, R., Zajac, Y., Steiner, B., & Zajac, A. (1974). Parent-child relations in transsexual and non-transsexual homosexual males. British Journal of Psychiatry. 124, 22-23. Gerslma, A., Emmelkamp, P. M. G., & Arrindell, W. A. (1990). Anxiety, depression and perception of early parenting: A meta-analysis. Clinical Psychology Review, 10, 251-277. Goffman, E. (1963). Stigma: Notes on the management of spoiled identity. Engelwood Cliffs, N. J.: Prentice-Hall Inc. Green, R. (1985). Gender identity in childhood and later sexual orientation: Follow-up of 78 males. American Journal of Psychiatry, 142, 339-341. Green, R. (1986). The "sissy boy syndrome" and the development of homosexuality. New Haven, CT: Yale University Press. Griffin, D. W., & Bartholomew, K. (1994). The metaphysics of measurement: The case of adult attachment. In K. Bartholomew & D. Perlman (Eds.), Advances in personal relationships: Attachment processes in adulthood (Vol. 5, pp. 17-52). London, England: Jessica Kingsley. Harris, J. R. (1995). Where is the child's environment? A group socialization theory of development. Psychological Review, 102, 458-489. Childhood Factors and Attachment 81 Harris, J. R. (1998). The nurture assumption: Why children turn out the way they do-parents matter less than you think and peers matter more. New York: Free Press. Harry, J. (1982). Gay children grown up: Gender culture and gender deviance. New York: Praeger. Harry, J. (1983). Defeminization and adult psychological well-being among male homosexuals. Archives of Sexual Behavior, 12, 1-19. Harry, J. (1989). Parental physical abuse and sexual orientation in males. Archives of Sexual Behavior, 18, 251-261. Harry, J., & Devall, W.B. (1978). The social organization of gay males. New York: Praeger. Hartup, W. H. (1979). The social worlds of childhood. American Psychologist, 34, 944-950. Hartup, W. H. (1983). Peer relations. In P. H. Mussen (Ed.), Handbook of child psychology (Vol. 4, pp. 103-196). New York: John Wiley & Sons. Hazan, C., & Shaver, P. (1987). Conceptualizing romantic love as an attachment process. Journal of Personality and Social Psychology, 52, 511-524. Hazan, C, & Zeifman, D. (1994). Sex and the psychological tether. In K. Bartholomew & D. Perlman (Eds.), Advances in personal relationships: Attachment processes in adulthood (Vol. 5, pp. 151-178). London, England: Jessica Kingsley Publishers, Ltd. Herek, G. M., & Capitanio, J. P. (1986). "Some of my best friends": Intergroup contact, concealable stigma, and heterosexuals' attitudes towards gay men and lesbians. Personality and Social Psychology Bulletin. 22, 412-424. Childhood Factors and Attachment 82 Hockenberry, S. L., & Billingham, R. E. (1987). Sexual orientation and boyhood gender non-conformity: Development of the Boyhood Gender Conformity Scale (BGCS). Archives of Sexual Behavior. 16, 475 492. Holtzen, D., Kenny, M. E., & Mahalik, J. R. (1995). Contributions of parental attachment to gay or lesbian disclosure to parents and dysfunctional cognitive processes. Journal of Counseling Psychology, 42, 350-355. Hooker, E. (1969). Parental relations and male homosexuality in patient and nonpatient samples. Journal of Consulting and Clinical Psychology. 33, 140-142. Hunter, F. T., & Youniss, J. (1982). Changes in functions of three relations during adolescence. Developmental Psychology, 18, 806-811. Huston, A. C. (1983). Sex-typing. In M. Hetherington (Ed.), Handbook of child Psychology: Vol. 4. Socialization, personality, and social development (pp. 387-467). New York: Wiley. Isay, R. A. (1989). Being homosexual: Gay men and their development. New York: Farrar Straus Giroux. Jones, E. E., Farina, A., Hastorf, A. H., Markus, H., Miller, D. T., & Scott, R. A. (1984). Social Stigma: The psychology of marked relationships. New York: W. H. Freeman and Company. Kelly, E. E., & Warshafsky, L. (1987, July). Partner abuse in gay male and lesbian couples. Paper presented at the Third National Conference for Family Violence Researchers. Durham, NH. Klem, L. (1995). Path analysis. In G. Laurence & P. R. Yarnold (Eds.), Reading and understanding multivariate statistics (pp. 65-97). Washington, DC: American Psychological Association. Childhood Factors and Attachment 83 Kraus, Stephen J. (1995). Attitudes and the prediction of behavior: A meta-analysis of the empirical literature. Personality and Social Psychology Bulletin, 21, 58-75 Kurdek, L. A. (1988). Perceived social support in gays and lesbians in cohabiting relationships. Journal of Personality and Social Psychology, 54, 504-509. Kurdek, L. A. (1995). Developmental changes in relationship quality in gay and lesbian cohabiting couples. Developmental Psychology. 31_, 86-94. Kurdek, L. A. (1997). The link between facets of neuroticism and dimensions of relationship commitment: Evidence from gay, lesbian, and heterosexual couples. Journal of Family Psychology, 1J_, 503-514. Kurdek, L. A., & Schmitt, J.P. (1986). Relationship quality of partners in heterosexual married, heterosexual cohabiting, and gay and lesbian relationships. Journal of Personality and Social Psychology. 51,711-720. Landolt, M. A., & Dutton, D. G. (1997). Power and personality: An analysis of gay male intimate abuse. Sex Roles, 37, 335-358. Langlois, J. H., & Downs, A. C. (1980). Mothers, fathers, and peers as socialization agents of sex-typed play behaviors in young children. Child Development, 51, 1237-1247. McWhirter, D. P., & Mattison, A. M. (1984). The male couple. Englewood Cliffs, NJ: Prentice-Hall. Meyer, I. H., & Dean, L. (1998). Internalized homophobia, intimacy, sexual behavior among gay and bisexual men. In G. M. Herek (Ed.), Stigma and sexual orientation: Understanding prejudice against lesbians, gay men, and bisexuals (pp. 160-186). Thousand Oaks, California: Sage Publications, Inc. Mickelson, K. D., Kessler, R. C, & Shaver, P. R. (1997). Adult attachment in a nationally representative sample. Journal of Personality and Social Psychology, 73, 1092-1106. Childhood Factors and Attachment 84 Milic, J. H., & Crowne, D. P. (1986). Recalled parent-child relations and need for approval of homosexual and heterosexual men. Archives of Sexual Behavior. 1_5, 239-246. Mohr, J. J. (1999). Same-sex romantic attachment. In J. Cassidy & P. R. Shaver (Eds), Handbook of attachment: Theory, research, and clinical applications, (pp. 378-394). New York: Guilford Press. Oldendick, R.W., Bishop, G.F., Sorenson, S.B., & Tuchfarber, A.J. (1988). A comparison of the Kish and last birthday methods of respondent selection in telephone surveys. Journal of Official Statistics, 4, 307-318. Patterson, C, Cohn, D., & Kao, B. (1989). Maternal warmth as a protective factor against risks associated with peer rejection among children. Development and Psvchopathology, I, 21-38. Peplau, L. A. (1993). Lesbian and Gay Relationships. In L.D. Garnets & D.C. Kimmel (Eds), Psychological perspectives on lesbian and gay male experiences, (pp. 395-419). New York: Columbia University Press Peplau, L. A., Garnets, L. D., Spalding, L. R., Conley, T. D., & Veniegas, R. C. (1998). A critique of Bern's "exotic becomes erotic" theory of sexual orientation. Psychological Review, 105, 387-394. Peplau, L. A., & Cochran, S. D. (1982). Value orientation in the intimate relationships of gay men. Journal of Homosexuality, 6, 1-19. Perris, C, Jacobsson, L., Lindstrom, H., von Knorring, L., & Perris, H. (1980). Development of a new inventory for assessing memories of parental rearing behaviors. Acta Psychiatrica Scandinavia, 61, 265-274. Phelan, J. E. (1996). Recollections of their fathers by homosexual and heterosexual men. Psychological Reports, 79, 1027-1034. Childhood Factors and Attachment 85 Pillard, R. C. (1990). The Kinsey Scale: Is it familial? In D. P. McWhirter, S. A. Sanders & J. M. Reinisch (Eds.), Homosexuality/heterosexuality: Concepts of sexual orientation (pp. 88-100). New York: Oxford University Press. Remafedi, G. (1987). Male homosexuality: The adolescent's perspective. Pediatrics, 79, 326-330. Ridge, S. R., & Feeney, J. A. (1998). Relationship history and relationship attitudes in gay males and lesbians: Attachment styles and gender differences. Australian and New Zealand Journal of Psychiatry. 32, 848-859. Roberts, C. W., Green, R., Williams, K., & Goodman, M. (1987). Boyhood gender identity development: A statistical contrast of two family groups. Developmental Psychology, 23. 544-557. Rofes, E. E. (1993/1994). Making our schools safe for sissies. The High School Journal, 77(1-2), 37-40. Rogosa, D. R. (1983). Demonstrating the reliability of the difference score in the measurement of change. Journal of Educational Measurement, 20, 335-343. Rohner, R. P. (1998). Father love and child development: History and current evidence. Current Direction in Psychological Science, 7,157-161. Ross, M. W. (1980). Retrospective distortion in homosexual research. Archives of Sexual Behavior, 9, 523-531. Ross, M. W., Campbell, R. L., & Clayter, J. R. (1982). New inventory of measurement of parental rearing patterns: An English form of the EMBU. Acta Psychiatrica Scandinavica, 66, 499-507. Childhood Factors and Attachment 86 Rotheram-Borus, M. J., Rosario, M., & Koopman, C. (1991). Minority youth at high risk: Gay males and runaways. In M.E. Colten & S. Gore (Eds.), Adolescent Stress: causes and consequences (pp. 181-200). New York: Aldine. Rutter, M. (1979). Protective factors in children's responses to stress and disadvantage. In M. W. Kent & J. E. Rolf (Eds.), Primary prevention and psychopathology. Vol 3. Social competence in children (pp.). Hanover, NH: University Press of New England. Rutter, M., & Quinton, D. (1984). Long-term follow-up of women institutionalized in childhood: Factors promoting good functioning in adult life. British Journal of Developmental Psychology, 2, 191-204. Saghir, M. T., & Robins, E. (1973). Male and female homosexuality: A comprehensive investigation. Baltimore, MD: Williams & Wilkins. Satinover, J. (1996). Homosexuality and the politics of truth. Grand Rapids Michigan: Hamewith Books. Schilder, A., & Kort, R. (1996). Childhood abuse and HIV vulnerability among gay men: Implications for HIV/AIDS work. The Canadian Health Psychologist, 4, 8-10. Sell, R. L., & Petrulio, C. (1996). Sampling, homosexuals, bisexuals, gays, and lesbians for public health research: A review of the literature from 1990-1992. Journal of Homosexuality, 30,31-47. Shulman, S., Elicker, J., & Sroufe, L. A. (1994). Stages of friendship growth in preadolescence as related to attachment history. Journal of Social and Personal Relationships, 11, 341-361. Siegelman, M. (1974). Parental background of male homosexuals and heterosexuals. Archives of Sexual Behavior, 3, 3-18. Childhood Factors and Attachment 87 Siman, M. L. (1977). Applications of a new model of peer group to naturally existing adolescent friendship groups. Child Development, 48, 270-274. Simpson, J. A., Rholes, W. S., & Phillips, D. (1996). Conflict in close relationships: An attachment perspective. Journal of Personality and Social Psychology. 71, 899-914. Sipova, I., & Brzek, A. (1983). Parental and interpersonal relationships of transsexual and masculine and feminine homosexual men. Journal of Homosexuality. 9, 75-85 Solomon, H. M. (1986). Stigma and western culture: A historical approach. In S. C. Ainlay, G. Becker, & L. M. Coleman (Eds.), The dilemma of difference: A multidisciplinary view of stigma (pp. 59-76). New York: Plenum Press. Sroufe, L. A., & Fleeson, J. (1986). Attachment and the construction of relationships. In W. Hartup & Z. Rubin (Eds.), Relationships within families (pp. 51-71). Hillsdale, NJ: Lawrence Erlbaum Associates. Stafford, M. C, & Scott, R. R. (1986). Stigma, deviance, and social control: Some conceptual issues. In S. C. Ainlay, G. Becker, & L. M. Coleman (Eds.), The dilemma of difference: A multidisciplinary view of stigma (pp. 77-91). New York: Plenum Press. Stacker, C. M. (1994). Children's perceptions of relationships with siblings, friends, and mothers: compensatory processes and links with adjustment. Journal of Child Psychology and Psychiatry, 35, 1447-1459. Tabachnick, B. G., & Fidell, L. S. (1989). Using multivariate statistics (2 n d ed.). New York: Harper Collins Publishers Inc. Thompson, N. L., Schwartz, D. M., McCandless, B. R., & Edwards, D. A. (1973). Parent child relationships and sexual identity in male and female homosexuals and heterosexuals. Journal of Consulting and Clinical Psychology. 41, 120 127. Childhood Factors and Attachment 88 Trinke, S. J., & Bartholomew, K. (1997). Hierarchies of attachment relationships in young adulthood. Journal of Social and Personal Relationships, 14, 603-625. van Aken, M. A. G., & Asendorpf, J. B. (1997). Support by parents, classmates, friends and siblings in preadolescence: Covariation and compensation across relationships. Journal of Social and Personal Relationships, 14, 79-93. Wagner, G., Serafini, J., Rabkin, J., Remien, R., & Williams, J. (1994). Integration of one's religion and homosexuality: A weapon against internalized homophobia? Journal of Homosexuality, 26, 91-110. Weiss, R. S. (1982). Attachment in adult life. In C. M. Parkes & J. Steveson-Hinde (Eds.), The place of attachment in human behavior (pp. 171-184). New York: Basic Books. Whitam, F. L. (1977). Childhood indicators of male homosexuality. Archives of Sexual Behavior. 6, 89-96. Whitman, F. L., & Mathy, R. M. (1986). Male homosexuality in four societies: Brazil, Guatemala, the Philippines, and the United States. New York: Praeger Publishers. Williams, B. T. R., & Gilmour, J. D. (1994). Sociometry and peer relationships. Journal of Child Psychology and Psychiatry and Allied Disciplines, 35, 997-1013. Zimmerman, D. W., & Williams, R. H. (1982). Gain scores in research can be highly reliable. Journal of Education Measurement, 19, 149-154. Zucker, K. J. (1990). Gender identity disorders in children: Clinical descriptions and natural history. In R. Blanchard & B. W. Steiner (Eds.), Clinical management of gender identity disorders in childhood and adults (pp. 1-23). Washington, DC: American Psychiatric Press. Childhood Factors and Attachment 89 Table 1 Demographic Characteristics of the Follow-up Participants (N =192) Demographic Characteristic Proportion Sexual Orientation Homosexual Bisexual Ethnic Background British/English/Scottish/Welsh/Irish Other European French Canadian Aboriginal/First Nations Other Canadian Latin/Central/South American Chinese/East Asian African Southeast Asian Not specified Education Some or all of a post-graduate education Some or all of a University education Some or all of a community college education Some or all of a high school education Some or all of a grade school education Employment Full time Part time Student Working at home without pay On leave (i.e. medical) Retired Unemployed Income $50,000 or more $40,000 to $49,999 $30,000 to $39,999 $20,000 to $29,999 $20,000 or less Refused 93.8% 6.3% 45.8% 27.6% 5.2% 1.0% 12.5% 2.1% 1.6% 1.0% .5% 2.6% 13.6% 39.6% 29.7% 15.1% 1.5% 77.1% 5.7% 1.0% 1.0% 4.2% 4.7% 6.3% 24.0% 16.1% 24.0% 18.2% 16.7% 1.0% Childhood Factors and Attachment 90 Table 2 Mean Scores for Self-report and Interview Measures Measure Minimum Maximum Mean SD EMBU Paternal Rejection* 1.00 3.75 1.76 .62 Maternal Rejection 1.00 3.33 1.57 .48 Paternal Warmth* 1.00 3.93 2.18 .77 Maternal Warmth 1.33 4.00 2.80 .71 BG(N)CS Boyhood Femininity 1.14 6.43 3.39 1.28 Boyhood Masculinity** 1.86 6.14 3.22 1.08 E-PEER Acceptance 1.00 4.80 3.10 .85 Rejection 1.00 5.00 2.78 1.00 IPA Trust 1.10 5.00 3.34 .86 Communication 1.00 4.88 2.83 .79 Alienation 1.00 4.71 2.68 .76 Total 1.28 4.68 3.18 .75 Peer Relationships Composite -2.00 2.45 .00 1.00 Internalized Homophobia 1.00 3.90 1.8 .63 HAI Secure 1.00 8.00 3.47 1.47 Fearful 1.00 8.00 4.01 1.86 Preoccupied 1.00 7.00 3.44 1.71 Dismissing 1.00 7.00 2.70 1.56 Anxiety -9.00 9.00 1.28 4.13 Avoidance -9.00 7.00 -.20 3.87 Note: N = 192 with two exceptions: * N = 181, **N = 191. o o 3 o o o » o —< o o o o o * o t-CN OS f -o o i/o OO UO ro o o o * © rO I CN VO .• — CN CN —' •3- ov — o OS VO ro CN 2 .• CN 173.2 E o. © o ro — c ro o E'ti I S " 0 0 0 0 VD VO CN ro CN CN o o o ro CN f*"> '-' m — ro CN • \ ! CN ro 0 0 ro CN ro CN CN ro ro vo ro <— O O O • O —' o o vo vo VO vo VO o o 0 0 0 0 o os rr-as CN VO CN 0 0 1 0 0 0 0 0 CN uo CN Os ro ro 0 0 CN CN ro ro CN Os CN CN VO CN VO CN CN ro «o ro E l £ ¥ p o TO ro ro .a 6 <L> 1*1 ro ro SO CN 0 0 CN ro os CN o vo o o CN ro CN CN o © CN OS ro ro VO CN CN ro ro ro O CN 0 0 . O 0 0 CN © CN O0 CN CN ro vo CN O CN O o ro Os CN — CN O © a ro •a 'o > < © -H Childhood Factors and Attachment 92 Table Four Correlations between Health-related and Key Variables Measure HIV Status Physical Abuse Paternal. Rejection .005 .19* N = 174 N=176 Maternal Rejection .01 .33** N = 184 N=186 Paternal Warmth .07 -.11 N=174 N = 176 Maternal Warmth .008 -.05 N = 184 N=186 .Femininity .07 .09 N = 184 N = 186 Masculinity .03 .-.01 N = 183 N = 185 E-PEER Acceptance -.01 -.12 N = 184 N = 186 E-PEER Rejection .03 .24** N = 184 N = 186 IPA Trust -.05 -.05 N=184 N = 186 IPA Communication -.004 -.04 N = 184 N = 186 IPA Alienation .02 .17* N = 184 N = 186 IPA Total -.03 -.08 N = 184 N = 186 Peer Composite .03 .16* N = .184 N = 186 Internalized .03 .09 Homophobia N = 184 N = 186 HAI Anxiety -.07 .20** N = 184 N = 186 HAI Avoidance -.09 -.17* N = 184 N = 186 ** Correlation is significant at the .01 level. * Correlation is significant at the .05 level. Figure 1. Two-Dimension Model of Attachment Childhood Factors and Attachment 93 Positive Model of Others/Closeness SECURE PREOCCUPIED Positive Positive Self-model Negative Self-model Model Positive Other-model Positive Other-model of Self Low Anxiety High Anxiety Closeness Closeness Low DISMISSING FEARFUL Anxiety Positive Self-model Negative Self-model Negative Other-model Negative Other-model Low Anxiety High Anxiety High Avoidance High Avoidance Negative Model of Self High Anxiety Negative Model of Others/High Avoidance Childhood Factors and Attachment 94 Figure 2. Path Model of Parental and Peer Rejection to Anxiety (bold line indicates mediation) .17* Paternal [Rejection .20' Maternal Rejectionl Peer Rejection .32**1 Relationship Anxiety -.02 ns Figure 3. Childhood Factors and Attachment 95 Path Model of Femininity to Rejection to Anxiety (bold line indicates mediation) .10 ns .16 ns Femininity Paternal Rejection .24* .17' .17' Maternal Rejection Peer Rejection 30** Relationship! Anxiety -.03 ns Childhood Factors and Attachment 96 Figure 4. Path Model of Masculinity to Warmth to Rejection to Anxiety (bold line indicates mediation) .12 ns 1 -.13 ns Masculinity .23 Paternal -.32** Peer Warmth Rejection .31**|Relationship| Anxiety -.15' Childhood Factors and Attachment 97 Figure 5. Alternate Path Model: Relationship Quality to Femininity to Anxiety (bold line indicates mediation) .24** Paternal Rejection Relationship Anxiety Maternal Relationship Rejection Anxiety .35*' Peer Relationship Rejection Anxiety .26 .20** Femininity Childhood Factors and Attachment 98 Figure 6. Alternate Path Model: Relationship Quality to Masculinity to Anxiety (bold line indicates mediation) -.20** Paternal Relationship Warmth Anxiety -.23** .35 Peer Relationship Rejection Anxiety .22** .21** Masculinity Childhood Factors and Attachment 99 Figure 7. Interaction of Paternal and Peer Rejection on Anxiety Low High Paternal Paternal Rejection Rejection Childhood Factors and Attachment 100 Appendix 1 E M B U Beside each statement please circle the number of the response listed below (1-4) that best describes how often the experience happened to you with your mother (or female guardian) and father (or male guardian) when you were growing up. If you had more than one mother/father figure, please answer for the persons who you feel played the most important role in your upbringing. 1 2 3 4 never occurred occasionally occurred often occurred always occurred Father or Mother or Guardian Guardian 1. My parent showed with words and gestures that he/she liked me. 1 2 3 4 1 2 3 4 2. My parent refused to speak to me for a long time. 1 2 3 4 1 2 3 4 3. My parent punished me even for small offenses. 1 2 3 4 1 2 3 4 4. I think that my parent wished I had been different in some way. 1 2 3 4 1 2 3 4 5. If I had done something wrong, I could apologize to my parent and make everything OK. 1 2 3 4 1 2 3 4 6. I was physically punished or scolded in the presence of others. 1 2 3 4 1 2 3 4 7. If things went badly for me, I felt my parent tried to comfort and encourage me. 1 2 3 4 1 2 3 4 8. My parent gave me more physical punishment than I deserved. 1 2 3 4 1 2 3 4 9. My parent would get angry if I didn't help at home when I was asked to. 1 2 3 4 1 2 3 4 10. I felt that it was difficult to approach my parent. 1 2 3 4 1 2 3 4 11. My parent would tell others what 1 had said or done so that I felt ashamed. 1 2 3 4 1 2 3 4 12. If 1 had a difficult task in front of me, I felt support from my parent. 1 2 3 4 1 2. 3 4 13. I was treated as the "black sheep" or "scapegoat" of the family. 1 2 3 4 1 2 3 4 14. My parent wished I had been like somebody else. 1 2 3 4 1 2 3 4 15. I felt my parent thought it was my fault when he/she was unhappy. 1 2 3 4 1 2 3 4 16. My parent showed me that he/she was fond of me. 1 2 3 4 1 2 3 4 Childhood Factors and Attachment 101 E M B U continued Father or Guardian Mother or Guardian 17. I think my parent respected my opinions. 1 2 3 4 1 2 3 4 18. I felt that my parent wanted to spend time with me. 1 2 3 4 1 2 3 4 19. I think my parent was mean and grudging toward me. 1 2 3 4 1 2 3 4 20. I think my parent tried to make my adolescence stimulating, interesting, and instructive. 1 2 3 4 1 2 3 4 21. My parent praised me. 1 2 3 4 1 2 3 4 22. When I was sad, I could seek comfort from my parent. 1 2 3 4 1 2 3 4 23. I was punished by my parent without having done anything. 1 2 3 4 1 2 3 4 24. My parent said he/she did not approve of my behaviour at home. 1 2 3 4 1 2 3 4 25. My parent criticized me in front of others. 1 2 3 4 1 2 3 4 26. My parent was abrupt with me. 1 2 3 4 1 2 3 4 27. My parent would punish me hard, even for little things. 1 2 3 4 1 2 3 4 28. My parent beat me for no reason. 1 2 3 4 1 2 3 4 29. My parent showed an interest in my own interests and hobbies. 1 2 3 4 1 2 3 4 30. I was beaten by my parent. 1 2 3 4 1 2 3 4 31. My parent treated me in such a way that I felt ashamed. 1 2 3 4 1 2 3 4 32. I felt that warmth and tenderness existed between me and my parent. 1 2 3 4 1 2 3 4 33. My parent would be angry with me without letting me know why. 1 2 3 4 1 2 3 4 34. My parent let me go to bed without food. 1 2 3 4 1 2 3 4 35. My parent respected the fact that I had other opinions than had he/she. 1 2 3 4 1 2 3 4 36. I felt that my parent was proud when I succeeded in something I had undertaken. 1 2 3 4 1 2 3 4 37. My parent hugged me. 1 2 3 4 1 2 3 4 38. My parent punished me for no reason. 1 2 3 4 1 2 3 4 39. My parent was severe with me. 1 2 3 4 1 2 3 4 40. I felt that my parent liked me. 1 2 3 4 1 2 3 4 Childhood Factors and Attachment 102 Appendix 2 MODIFIED BG(N)CS Below are a list of statements, please indicate how true each statement was of you as a CHILD. For example, put a "1" in the space provided if the statement was never or almost never true, and a "7" if it was always or almost always true. 1 2 3 4 5 6 7 Never or almost never true Sometimes Always or almost always true true 1. I felt like I was similar to or not very different from other boys my age. 2. I imagined or wished I were a policeman or a soldier. 3. I put on woman's' clothing, make-up, jewelry, etc. 4. I would imagine I was the male character (cowboy, detective, soldier, explorer) in stories I read or watched on TV. 5. I would imagine I was the female character (girl being saved, etc.) in the stories I read or watched on TV. 6. I preferred girls' games and toys (dolls, cooking, sewing etc.). 7. I liked to read adventure or sport stories. 8. I liked to read romantic stories. 9. I imagined or wished I were a well-known sports figure. 10. I imagined or wished I were a dancer or a model. 11. I preferred to be around older men (fathers, uncles, grandfather, coach, etc.). 12. I preferred being around older women (mother, aunts, grandmother, female teachers, etc.). 13. I looked to men and male peers to model my behavior and attitudes after. 14. I liked to engage in rough-and-tumble play. Childhood Factors and Attachment 103 Appendix 3 E-PEER The following statements describe your peer relationships in childhood and adolescence. Some of the statements refer specifically to your friends and some refer to other kids in general. Please read each of the items and rate how often the statement was true of your experience with your friends or other kids in your childhood through to your high-school years. 1 2 3 4 5 Almost never Seldom true Sometimes true Often true Almost always or or never true always true As a child or adolescent: 1. other kids liked to play or hang out with me. 1 2 3 4 5 2. other kids were usually friendly to me. 1 2 3 4 5 3. other kids often shared things with me. 1 2 3 4 5 4. other kids would usually stick up for me. 1 2 3 4 5 5. other kids liked to ask me to go along with them. 1 2 3 4 5 6. other kids were always criticizing me. 1 2 3 4 5 7. other kids often picked on me and teased me. 1 2 3 4 5 8. other kids wouldn't listen when I tried to say something. 1 2 3 4 5 9. other kids were often unfair to me. 1 2 3 4 5 10. other kids would often try to hurt my feelings. 1 2 3 4 5 Childhood Factors and Attachment 104 Appendix 4 IPA The following statements describe your peer relationships in childhood and adolescence. Some of the statements refer specifically to your friends and some refer to other kids in general. Please read each of the items and rate how often the statement was true of your experience with your friends or other kids in your childhood through to your high-school years. 1 2 3 4 5 Almost never or never true Seldom true Sometimes true Often true Almost always or always true As a child or adolescent: 1. my friends sensed when I was upset about something. 1 2 3 4 5 2. I liked to get my friends' point of view on things I was concerned 1 2 3 4 5 about. 3. when I discussed things with my friends they considered my point 1 2 3 4 5 of view. 4. talking over my problems with my friends made me feel ashamed or 1 2 3 4 5 foolish. 5. I wished I had different friends. 1 2 3 4 5 6. my friends understood me. 1 2 3 4 5 7. my friends encouraged me to talk about my difficulties. 1 2 3 4 5 8. my friends accepted me as I was. 1 2 3 4 5 9. other kids spread rumors about me. 1 2 3 4 5 10. I felt that I was not in touch with my friends as much as I would 1 2 3 4 5 have liked. 11. my friends didn't understand me. 1 2 3 4 5 12. I felt alone or apart when I was with my friends. 1 2 3 4 5 13. my friends listened to what I had to say. 1 2 3 4 5 14. I felt my friends were good friends. 1 2 3 4 5 15. my friends were fairly easy to talk to. 1 2 3 4 5 16. other kids threatened to hurt me in some way. 1 2 3 4 5 17. when I was angry about something, my friends tried to be 1 2 3 4 5 understanding. 18. other kids physically harmed me (e.g., hit, kicked, or beat me up) 1 2 3 4 5 without my provoking them. 19. my friends helped me to understand myself better. 1 2 3 4 5 20. my friends were concerned about my well being. 1 2 3 4 5 Childhood Factors and Attachment 105 IPA continued 1 2 3 4 5 Almost never or never true Seldom true Sometimes true Often true Almost always or always true As a child or adolescent: 21. other kids demanded money from me or stole my possessions. 1 2 3 4 5 22. I felt angry with my friends. 1 2 3 4 5 23. I could count on my friends when I needed to get something off my 1 2 3 4 5 chest. 24. I trusted my friends. 1 2 3 4 5 25. my friends respected my feelings. 1 2 3 4 5 26. I got upset a lot more than my friends knew about. 1 2 3 4 5 27. It seemed as if my friends were irritated with me for no reason. 1 2 3 4 5 28. other kids excluded me from groups. 1 2 3 4 5 29. I told my friends about my problems and troubles. - 1 2 3 4 5 30. If my friends knew something was bothering me, they would ask 1 2 3 4 5 me about it. Childhood Factors and Attachment 106 Appendix 5 IHS Below is a list of statements. Please indicate how strongly you agree or disagree with each of the statements by circling the appropriate number on the five-point scale below. For example, if you strongly disagree with the statement, circle "1" or if you strongly agree, circle "5". 1 2 3 4 5 Strongly disagree Disagree No opinion Agree Strongly agree 1. Male homosexuality is a natural expression of sexuality in human males. 2. I wish I were heterosexual. 3. When I am sexually attracted to another gay man, I do not mind if someone else knows how I feel. 4. Most problems that homosexuals have come from their status as an oppressed minority, not from their homosexuality per se. 5. Life as a homosexual is not as fulfilling as life as a heterosexual. 6. I am glad to be gay. 7. Whenever I think a lot about being gay, I feel critical about myself. 8. I am confident that my homosexuality does not make me inferior. 9. Whenever I think a lot about being gay, I feel depressed. 10. If it were possible, I would accept the opportunity to be completely heterosexual. 11. I wish I could become more sexually attracted to women. 12. 13. 14. 15. If there were a pill that could change my sexual orientation, I would take it. I would not give up being gay even if I could. Homosexuality is deviant. It would not bother me if I had children who were gay. 16. Being gay is a satisfactory and acceptable way of life for me. 17. If I were heterosexual, I would probably be happier. 18. Most gay people end up lonely and isolated. 19. For the most part, I do not care who knows I am gay. 20. I have no regrets about being gay. 2 2 2 2 2 2 2 2 2 2 2 2 2 2 .2 2 2 2 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 Childhood Factors and Attachment 107 Appendix 6 History of Attachment Interview GENERAL QUESTIONS: How old are you? What do you do for a living? What is your sexual orientation? HIV status? Review and verify relationship status. Relationship length. Do you have any children? (How many?) FAMILY RELATIONSHIPS Family Background * Please describe your family background. The kinds of things I'm interested in are: where you lived, how often you moved, what your parents did for a living, siblings, that sort of thing * Are your parents still alive? As appropriate: Did your parents ever separate or divorce? * Who lived in the household? * Were any other adults central in your upbringing? * Briefly describe what your parents' marital relationship was like when you were young. Was there much conflict? As appropriate: Did your parents' conflicts ever become physical? In front of you? Explore * Were they physically affectionate with one another in front of you? * Briefly describe what kind of a young child you were. Relationship with Parents * I'd like you to describe your relationship with your each of your parents as a child, going back as far as you can. * Which parent were you closest with? Why? If necessary, did you feel distant from either parent? * Was each of your parents affectionate? Describe how. If necessary, could you give me some adjectives describing your mother. If necessary, could you give some adjectives describing your father. If necessary, ask for specific memories or incidents to illustrate the adjectives for your mother, for your father. * Did either of your parents have a drug or alcohol problem? Explore: Which parent, alcoholism, binges, etc. Did it lead to anger or aggression? What effect did it have on family, on you as an individual? Upsets * If you were unhappy or upset as a child, what would you do? Example. How did your parents respond? * When you were ill or physically hurt? * When you were emotionally hurt? (teasing, conflicts with teachers or siblings) Childhood Factors and Attachment 108 Separations * If loss or separation within family, explore. * Do you remember the first time you were separated from your parents for any length of time? (e.g. camp, parents' holiday, hospitalization) Explore. If necessary: How about going to school for the first time? Or going to college? * As a child, did you ever get lost? How did you react (feelings)? How did your parents react? * Did you ever run away from home? Why? How did your parents react? Rejection * Did you ever feel rejected by your parents as a child? Describe. How did it feel? What did you do? * If not, did they ever hurt your feelings? * If not, did you ever feel that you'd disappointed your parents? * How did they show their disappointment? For instance, with regard to grades at school? * Did your parent realize she/he was rejecting (hurting your feelings, showing their disappointment towards) you? * Did you feel pushed by either parent? * Were your parents ever threatening - either jokingly or for discipline? * What did they do for discipline? If necessary, were you ever afraid of either parent? If necessary, how consistent was your mother in her parenting? Father? How predictable was each parent? * How often, intensely did each parent get angry at you? Explore. If necessary, how expressive was each parent? Range? * What are your feelings now about the discipline you received as a child? * Did you feel loved? Were they proud of you? How was that shown? * Did you feel that they understood you? Trauma * As a child or adolescent, did you have any sexual contact with an adult or older person? Explore: age? by whom? who did you tell? how did parents respond? If necessary: Did you consider that abusive or inappropriate? At the time? Now? * As a child or adolescent, did you ever attempt suicide? Explore. Changing Relationship with Parents * Did your relationship with each of your parents change during adolescence? * What is your relationship with your parents like now? If different, what brought about the change? * How often do you talk to them? Do you talk about personal concerns? Are there things that it would be hard to talk to them about? Have you talked to them about your sexual orientation? Explore. * Do you feel that they understand you? Childhood Factors and Attachment 109 Changes * What are your feelings now about the parenting you received as a child? If necessary: How would you have liked your parents to be different? * How do you think that your parents would have liked you to be different? Siblings * What was your relationship(s) like with your sibling(s) when you were growing up? * Siblings often get into conflicts. Were there any conflicts that you were involved in with siblings that caused physical or emotional harm? Explore. * What is your current relationship(s) like with your sibling(s)? If necessary, are you out to them? Any of them gay/bisexual? Effects * How do you think your experiences growing up with your family have influenced your relationships with people outside of the family? Childhood friendships * Did you have many close friends in childhood/adolescence? How did you get along with childhood peers in general? * Did you ever feel that you were 'different' in childhood or adolescence? How prevalent? Was this a concern? * When did you first become aware of a same sex attraction? * Did you experience any rejection, teasing, or forms of abuse from your peers as a child or adolescent? If appropriate, do you have any sense of why they did that? STOP AND CHANGE TAPE Childhood Factors and Attachment 110 PEER RELATIONSHIPS Now I'd like to move on and talk about your peer relationships but I'd first like to start with some general questions about how you see yourself in relation to other people. Personal Information * Compared to other people you know, how emotional would you say you are? Why do you say that? Expression of emotions? * How often do you cry? (If prompting, "once a day, every few days..,") * More often alone or with others? * If you do feel unhappy or upset about something, what are you likely to do? * Are you more likely to go to other people or do you tend to deal with upsets on your own? Friendships I'd like to turn now to talking about your friendships. * How many close friends do you have? * What does it mean to you to call someone a close friend? * Do you discuss personal matters with your close friends? Are there things you wouldn't talk about or that would be difficult to talk about? Example. Why? * With your friendships as a whole do you have a sense of who is more involved or invested? You or the other person? If discrepancy: How do you feel about that? * Have you ever had conflicts with your close friends? How do you handle it? What do you do? Have they ever been physical? If necessary: Have you ever had your feelings hurt by a close friend? Example Have there been times when you and F haven't talked to each other? * Have you ever felt torn between your friends and your romantic relationships? * What changes would you like to see in your friendships over time? * How do your current friendships compare with your childhood friendships? * When you meet new people do you think they will like you? * How confident are you about making new friends? * Would you consider yourself a generally shy person? * What impression do you think you make on other people? * What impression would you like to make? * To what degree are you out as a gay or bisexual man? Explore. Childhood Factors and Attachment 111 Romantic Relationships Now I'd like to spend some time talking about your romantic relationships. Relationship History * If currently in a romantic relationship: You said you were currently in a romantic relationship. Prior to your current relationship, how many relationships did you have that you would consider serious? * If not currently in rom rel: Have you been involved in a romantic relationship in the past? How many? * At what age did you first start dating or becoming involved in romantic relationships? Explore both same sex and opposite sex dating. * If not obviously sexually active: Have you ever been sexually active? * At what age did you first become sexually active? Clarify same and opposite sex partners. If participant is unclear what we mean by sexually active, "any genital contact between two people" | If no previous romantic relationships, move on to single questions" * Do you prefer to be in a monogamous or open relationship? * Has this changed over time across your relationships? Has this changed within your current or most recent relationship? * If open, how is it negotiated, does partner know, do you maintain boundaries, i.e. discussing it, rules, what to share, not share, etc. How do you think your partner feels about this arrangement? How do you feel about each others' activities? Explore. * Now I'd like to first talk briefly about your previous romantic relationships. Perhaps you could give me brief history of those romantic/sexual involvement's. The kinds of things I am interested in are how long you were together, how serious it was, what were the positives and the negatives of the relationship. If necessary, how did it end? Who initiated the ending? How was that for you? If necessary, explore times not in a relationship? * Do you see any patterns across your relationships? If necessary, do you have any issues with either trusting others too much, or having difficulties trusting others? * Was there any physical conflict in any of these relationships such as pushing, shoving, etc.? Can you briefly describe the incident. If necessary - context, impact on relationship, how did that make you feel? * Did you ever feel forced or pressured into sex in any of these relationships when you didn't want to? If yes, explore (abuse questions on page 16) If NOT currently in a relationship, go to singles questions. Childhood Factors and Attachment 112 Current Relationship * Briefly describe your relationship in terms of how serious it is, amount of time spent together, and whether or not you've considered future plans. * Can you tell me a little about what your relationship is like, what your partner is like? If necessary: What do you like about your relationship or your partner? What don't you like? * What do you think your partner likes most about you? And least? If necessary: What would they say? How do you know? Communication & Support * How comfortable are you discussing personal matters with your partner? What are some of the topics of conversation you avoid with your partner- because they're awkward to talk about or they lead to disagreements? * How does your partner respond when you would like help or support? * How does your partner respond when you're hurt or sick? * What about emotionally upset? What would you like your partner to do? Example. If necessary: Do you ever feel your partner is not responsive enough or too responsive? If necessary: Do you feel comfortable crying in front of your partner? If not, why? How does partner respond? * How well does your partner understand you? Love-worthiness & Trust * Have you ever felt rejected by your partner? Describe. If necessary: Have you ever had your feelings hurt by your partner? * Have you ever doubted that your partner loves or cares about you? How does P show it? * Have you ever felt neglected by your partner? * How does your current relationship compare with past ones? * Do you say "I love you" to each other? How often? One more often? Explore as necessary. Conflict Resolution * How often do you have disagreements or arguments? What about? What happens? Does it get resolved? How? As necessary: Do you ever have differences of opinion? i.e. Movies? Music? Do you ever wonder if your partner disagrees with you, but doesn't say anything? How long do you stay angry? Who initiates the arguments and the resolution? * Could you give me an example of a typical conflict and describe how it tends to go. Childhood Factors and Attachment 113 * How does this compare with past relationships? * Have you ever felt afraid of your partner? * How often are you and your partner mean to each other or critical? If necessary: Have the two of you had any issues about the amount of time you spend together? If necessary: How do the two of you go about making decisions? Is it mutual? * Have your conflicts ever become physical? (pushing, shoving, slapping, throwing things) If yes, explore (See abuse questions on page 16) If necessary, is there anything you or your partner have done that could be considered emotionally or psychologically abusive? Physical Relationship * How affectionate are the two of you within the relationship? Is one of you more so than the other? Ever an issue, in private or public? * Do you ever feel that your partner is not warm or affectionate enough? Or too affectionate? * How do feel about the sexual aspect of your relationship If necessary, how do think your partner feels about it? If necessary, has it changed over time? Is it more important for you or your partner? * Have you ever felt pressured by your partner into having sex or engaging in a sexual activity that you didn't feel comfortable with? Explore. * Have you ever pressured your partner into having sex or engaging in a sexual activity that he didn't feel comfortable with? If necessary: What changes would you like to see in your sexual relationship? (to make it more satisfying for you? your partner?) Separations * Have you and your partner ever been apart for any length of time? (e.g., holidays, business trips). Explore. If not, how would you feel if it did happen? * Have there been any separations or other involvements since you've been together? Explore. Mutuality * People in relationships commonly report that one partner seems more invested or committed to the relationship? Has this ever been the case in your relationship? If so, describe. * Some people feel concerned about becoming too dependent in a relationship. Is this a concern for you? For your partner? * How jealous or possessive is your partner? * If your partner is jealous, how does it make you feel? Childhood Factors and Attachment 114 * How about you? How jealous or possessive are you? Explore Regrets. Break-up * Have you ever thought about separating? When? Why? If yes, tell me your thoughts about it? * How difficult would it be for you to end this relationship? If necessary: Have you ever had regrets or doubts about having become involved with (married to) your partner? If appropriate: How much faith do you have that your relationship will last in the long term? * If it did break up, who do you think would be most likely to initiate a break up? Why? If necessary: If you and your partner ever did break up, how difficult do you think it would be for you? And for your partner? General Evaluation * How would you like to see your relationship change over time? If necessary: Any changes in the way you relate to your partner? Or how your partner relates to you? * If you could have the ideal relationship, how would it differ from your present relationship? * How does your current relationship compare with past ones? * If bisexual, how do your relationships with men compare to those with women? Childhood Factors and Attachment 115 RELATIONSHIPS IN GENERAL Now I'd like to ask you a few final questions concerning all your relationships - family, friends, romantic partners. * How many people do you think would be there for you no matter what? * Has that changed over time? * How connected do you feel socially? * Are you satisfied with your social support system? Changes? * How connected do you feel you are with the gay community? * What kinds of changes would you like to see in the way you relate to others? If necessary: What kinds of changes would you like to see in the way others relate to you? * Is there anything else about your social relationships that we haven't hit upon that seems important? Final Questions * How often have you felt rejected, stigmatized or discriminated against because of your sexual orientation? * Have you ever been publicly harassed, threatened or physically harmed because of your sexual orientation? Explore. * How did you feel about talking about the kinds of issues we've been talking about in this interview? Are these things that you've talked about with other people? What encouraged you to say yes to corning in to do this session? Thank you. Childhood Factors and Attachment 116 SINGLES' QUESTIONS Note about single questions. If person is not in a romantic relationship or dating, and would clearly like to be, ease off on a lot of the "why not" questions. Also any questions from the regular interview can be framed in a hypothetical way. For e.g.: if you were involved in a romantic relationship Again , be cautious about not making the person feel badly about not being involved. Dating Status Have you been dating anyone recently? Or have you dated anyone in the past IF YES: Could you give me a brief history of your dating experiences. If necessary, frequency, length, enjoyment, etc. * Would you prefer to see several people or would you rather have a steady relationship with one person? * How long do you see the same person before you consider yourself a couple? * What do you like about dating? What don't you like about dating? * What are you looking for when you date? (i.e., for fiin, to find partner, sex, etc.) * Was there any physical conflict in any of these relationships? If yes, explore (See abuse questions) * Do you see any patterns across your dating relationships IF NO: * Have you ever dated? And how do you feel about that? Can you tell me more about that? * Are there any particular reasons why you haven't dated much up to now (or lately)? * Do you ever meet people that seem interesting to you? What happens? * Do you ever have crushes on people? Explore. What happens? * Are you doing anything to pursue romantic relationships now? Explore. * What do you do? How does the other person respond? If not interested in dating * What about dating makes it uninteresting to you now? * Do you expect that to change in the future? Childhood Factors and Attachment 117 Feelings about being single * Do you ever feel envious of your friends' romantic relationships? * Do you ever feel left out? Explore. * Do you think they are ever envious of you? Explore * Do you experience any pressure to be in a romantic relationship? (from friends, parents) * Do you see any advantages to being single over being in a relationship? * How much time do you spend alone? Right amount? If appropriate: Would you consider living alone? * How often do you feel lonely? Explore Sex f If^ating: * How important a part does sex play in your dating relationships? If appropriate: How quickly do you become sexually involved with a dating partner? of sex in dating relationships? If not dating: * How important a part do you think sex should play in dating relationships? Conflict Resolution (for dating only) * How often have you had disagreements or arguments with people you are dating? Explore. If necessary, have you ever felt afraid of a dating partner? If necessary, how often are you and a dating partner mean to each other or critical? If necessary, in a dating relationships has there ever been issues of the amount of time you spend together? If necessary, have your conflicts ever become physical? (pushing, shoving, slapping, throwing things). If yes, explore (See abuse questions) Future * Are you interested in finding a steady relationship in the near future? How often do you think about it? If appropriate: Are you doing anything to pursue romantic relationships now? Explore. What? Why not? * If you did meet someone, do you feel that you'd be ready to make a serious commitment? * In general, how important is it to you to be in a romantic relationship? * In the long term, how important would it be to you to be in a romantic relationship? Childhood Factors and Attachment 118 * Do you ever worry about not finding someone to be with? What about that worries you? How realistic do you think that is? * What do you think you would miss by not being in a relationship? * What do you think you would gain? * What are your ideals for a romantic relationship? Kind of relationship, kind of person, etc. * Are there things you know you wouldn't want? RELATIONSHIPS IN GENERAL Now I'd like to ask you a few final questions concerning all your relationships - family, friends, romantic partners. * How many people do you think would be there for you no matter what? * Has that changed over time? * How connected do you feel socially? * Are you satisfied with your social support system? Changes? * How connected do you feel you are with the gay community? * What kinds of changes would you like to see in the way you relate to others? If necessary: What kinds of changes would you like to see in the way others relate to you? * Is there anything else about your social relationships that we haven't hit upon that seems important? Final Questions * How often have you felt rejected, stigmatized or discriminated against because of your sexual orientation? * Have you ever been publicly harassed, threatened or physically harmed because of your sexual orientation? Explore. * How did you feel about talking about the kinds of issues we've been talking about in this interview? Are these things that you've talked about with other people? What encouraged you to say yes to coming in to do this session? Thank you. Childhood Factors and Attachment 119 Appendix 7 History of Attachment Interview—Coding Form SUBJECT # DATE SEX INTERVIEWER . AGE CODER Family (1-9) Mother Father Acceptance Rejection of child/attachment . Neglect (1 -low, 9-high) Consistency/Predictability (1-none, 9-high) Emotional Expressivity (1-extreme reserve, 5-neither, 9-histrionic) Frequency of anger ' (I almost never, 3 once a year, 5 1-2/mo., 7 1/week, 9 every day) Intensity of anger • Pushed to achievement ' Role Reversal • Anger at parent • Idealization Proximity seeking (1-never, 9-always) Dominance (1-child, 5-ideal, 9-parent) Childhood closeness w/parent (1-9) Current closeness w/parent (1-9) Childhood Factors and Attachment 120 Overall quality (in childhood, 1-9) Other adult quality (1-9) Number of siblings Sister(s) #gay/bi gay/bi Overall quality of relationship(s) with siblings Childhood peers: Subject felt he was 'different' in childhood (1-9) OTHER Separation anxiety in childhood (1-9) (1-none, 5-average, 9-extreme) "Adolescent" rebellion (1-9) (1-none, 5-some, 9-extreme) Parental problem drinking Mother Father (l=No, 2=Potentially, 3=Yes) Any indication of abuse: Emotional (1-9) Physical (1-9) Sexual (1-9) Physical abuse (1-9) Mother —> Father Sibling —> Subject Father —> Mother Mother —> Sibling Mother —> Subject Father —> Sibling Father —> Subject Suicide attempt? Y/N Age Divorce (l=No, 2=Yes, simple, 3, Yes, complex) Notes PEER CODES Response when upset Crying pattern Frequency (1-9) Situation (1-9) (l almost never, 3 once a year, 5 1-2/mo., 7 1/week, 9 every day) Other adult impact Brother(s) # Childhood Factors and Attachment 121 FRIENDSHIPS Self-disclosure (1-9) Relative involvement (1-other, 9-self) Dominance (1-other, 9-self) ROMANTIC RELATIONSHIPS Length longest relationship attraction Length current relationship ss Level of involvement (1-9) ss Number serious relationships Self-disclosure (1 -9) Relative involvement (1-other, 9-self) Dominance (1-other, 9-self) Current physical abuse emotional abuse Overall physical abuse emotional abuse w/men Idealization (1-9) Quality of relationships (1-9) Number of close friends Age first aware of same sex Age began "dating" op Age began sex op w/women Idealization (1-9) Current quality (1-9) Overall quality (1-9) Subject as Receiver (l-9)_ Subject as Receiver (1-9) Subject as Receiver (l-9)_ Subject as Receiver (l-9)_ Perpetrator (1-9) Perpetrator (l-9)_ Perpetrator (l-9)_ Perpetrator (l-9)_ Time since last incident of abuse (physical/sexual) Number of abusive relationships with physical/sexual violence Comments (reasons and responses) Type of relationship (1, monogamous - 9, open) _ Comfort with arrangement (1, comfort - 9, disagree) _ OTHER SCALES (1-9) Overall adult proximity seeking (1-9) Emotional expressiveness Emotional dependence Care giving Warmth Elaboration (1 -9) Coherence (1-9) Overall Idealization (1-9) COUNTS (1-none, 5-some, 9-constant) Insistence on not remembering Inappropriate Laughter I don't knows (1 -9) STYLES (1-9) SECURE FEARFUL PREOCCUPIED DISMISSING FINAL NOTES (key points, evidence of change) Childhood Factors and Attachment 122 Jealousy Separation Anxiety Trust Self-confidence Childhood Factors and Attachment 123 Appendix 8 Suggestions for the Reduction of Volunteer Bias 1. Make the appeal for volunteers as interesting as possible, keeping in mind the nature of the target population. 2. Make the appeal for volunteers as non-threatening as possible so that potential volunteers will not be "put-off' by unwarranted fears of unfavorable evaluation. 3. Explicitly state the theoretical and practical importance of the research for which volunteering is requested. 4. Explicitly state in what way the target population is particularly relevant to the research being conducted and the responsibility of potential volunteers to participate in research that has the potential for benefiting others. 5. When possible, potential volunteers should be offered not only pay for participation but small courtesy gifts simply for taking the time to consider whether they will want to participate. 6. Have the request made by a person of high status as possible and preferably by a woman. 7. When possible, avoid research tasks that may be psychologically or biologically stressful. 8. When possible, communicate the normative nature of the volunteering response. 9. After a target population has been defined, an effort should be made to have someone known to that population make an appeal for volunteers. The request for volunteers itself may be more successful if a personalized appeal is made. 10. In situations where volunteering is regarded by the target population as normative, conditions of public commitment to volunteer may be more successful; where non-volunteering is regarded as normative, conditions of private commitment may be more successful. Rosenthal & Rosnow, 1975 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0089285/manifest

Comment

Related Items