_TH This is to certify that the dissertation entitled GENDER ROLE ORIENTATION, HOMOPHOBIC SELF-PRESENTATION, AND PERCEIVED KNOWLEDGE: PREDICTING SELF—DISCLOSIVE BEHAVIOR IN SAME-SEX MALE FRIENDSHIPS presented by JONATHAN MICHAEL BOWMAN has been accepted towards fulfillment of the requirements for the PHD. degree in COMMUNICATION % 777 W Major Professor’ s SW AUQUS‘I‘ 11 9005— U I Date MSU is an Affirmative Action/Equal Opportunity Institution LIBRARY Michigz: “ State University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 2/05 p:/ClRC/DateDuetindd-p.1 GENDER ROLE ORIENTATION, HOMOPHOBIC SELF-PRESENTATION, AND PERCEIVED KNOWLEDGE: PREDICTING SELF-DISCLOSIVE BEHAVIOR IN SAME-SEX MALE FRIENDSHIPS BY Jonathan Michael Bowman A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Communication 2005 ABSTRACT GENDER ROLE ORIENTATION, HOMOPHOBIC SELF-PRESENTATION, AND PERCEIVED KNOWLEDGE: PREDICTING SELF-DISCLOSIVE BEHAVIOR IN SAME-SEX MALE FRIENDSHIPS BY Jonathan Michael Bowman Much research has found that men are relatively reticent to engage in self- disclosure within their male friendships. The current study scrutinizes a few key factors which may lead to this finding by looking at relationships between gender role orientation and self-disclosive behavior. Also, the degree to which one perceives that his male friend has knowledge of him is discussed as a potential moderator of how homophobic self-presentation affects one’s self-disclosive behavior. Most significantly, the present research looks at the effects of self- disclosure on closeness within a relationship, using multiple conceptualizations of both. While the gender effects on self-disclosive behavior received mixed results and the current iteration of the homophobia scale failed to demonstrate reliable or valid findings, there was overwhelming support for the positive correlation between multiple measures of closeness (perceived closeness, interdependence, diversity of activity, and strength of influence) and multiple measures of participants’ self-reported self-disclosive behaviors (amount, control, breadth, and valence of disclosure). This link between various measures of both closeness and self-disclosive behaviors among established friends is well supported and is discussed as having wide-ranging implications for self- disclosure research. DEDICATION This work is dedicated to my father and mother, Michael and Sherri Bowman. Without their love and continued support, I would not have striven to reach the goals that they helped me believe were within my reach. I thank them for helping me to grow and mature in order that I might impact my world for my Lord and Savior Jesus Christ. ACKNOWLEDGEMENTS This work would not have been possible without the guidance and assistance of the author’s advisor, Dr. Gwen M. Wittenbaum, to whom a great deal of gratitude is directed. The author would like to thank Dr. Tim Levine, Dr. Kelly Morrison, and Dr. Stan Kaplowitz for their ideas and comments on all drafts of this work. The author is deeply indebted to Marge Barkman, Renee Strom, and Katie Klein for their assistance in maintaining the author’s emotional well-being in the office. The author would also like to acknowledge the work of some highly motivated undergraduate research assistants: Brian Willing, Megan Dowd, John Ryan, Keith Binder, Meredith Mulvaney, Allison Cherundolo, Erin Durkin, and Clare Robbins iv TABLE OF CONTENTS LIST OF FIGURES AND TABLES ........................................................... vii INTRODUCTION ................................................................................. 1 Gender Role Orientation and Self-Disclosure ..................................... 3 Concern with Appearing Homosexual ............................................... 5 Perceived Knowledge: A Potential Moderator ..................................... 7 Relational Closeness ..................................................................... 8 Hypotheses ................................................................................. 9 METHOD .......................................................................................... 12 Participants ............................................................................... 1 2 Friendship Behavior Questionnaire .................................................. 12 Procedure ................................................................................. 14 Introduction to the Study ...................................................... 14 Perceived Knowledge Induction ............................................. 14 Manipulation Check: Perceived Other-Knowledge of Self ..................... 14 Independent Variables .................................................................. 16 Masculine Gender Role Orientation ........................................ 16 Feminine Gender Role Orientation ......................................... 17 Concern with Appearing Homosexual ...................................... 17 Self-Disclosu re Items .................................................................... 1 8 Amount and Control of Disclosure ........................................... 18 Breadth of Disclosure ........................................................... 19 Valence of Disclosure ........................................................... 20 Closeness Items .......................................................................... 21 Perceived Closeness ............................................................ 21 Interdependence .................................................................. 21 Diversity of Activity ............................................................... 21 Strength of Influence ............................................................. 22 RESULTS .......................................................................................... 22 Descriptive Statistics .................................................................... 22 Hypothesis Tests ........................................................................ 22 Hypotheses 1a-1 b. .............................................................. 22 Hypothesis 2 ...................................................................... 25 Hypothesis 3 ...................................................................... 26 Hypotheses 4a-4d ............................................................... 27 DISCUSSION ..................................................................................... 29 Gender Role Orientation ............................................................... 29 Heterosexual Self-Presentation ...................................................... 32 Relational Closeness .................................................................... 34 Future Directions ........................................................................ 35 FOOTNOTES ..................................................................................... 38 APPENDICES .................................................................................... 39 REFERENCES ................................................................................... 63 vi LIST OF FIGURES AND TABLES Figure 1: Proposed Conceptual Model ...................................................... 56 Table 1: Descriptive Statistics for Main Factors (Predictor, Process, and Outcome Variables) Used in Correlations (n = 115) ....................................................... 57 Table 2: Correlations Between Independent Variables and Communication Process Variables (Self-Disclosure) (n = 115) ............................................. 58 Table 3: Standardized Beta Coefficients for Multiple Regressions Evaluating the Effects of Gender Role Orientation on Self-Disclosure Variables (n = 115) ................................................ 59 Table 4: Correlations Between One’s Concern with Appearing Homosexual and Self- Reported Self-Disclosure Behavior for More and Less Knowledge Conditions (n = 115) ....................................... 60 Table 5: Correlations Between Communication Process Variables (Self-Disclosure) and Outcome Variables (Closeness) (n = 115) ....................................... 61 Table 6: Standardized Beta Coefficients for Multiple Regressions Evaluating the Effects of Self- Disclosure on Closeness Variables (n = 115) ................................... 62 vii Gender Role Orientation, Homophobic Self-Presentation, and Perceived Knowledge: Predicting Self-Disclosive Behavior in Same-Sex Male Friendships “It is one thing for men to recognize that they need more intimate friendships, but it is quite another thing for men to behave in the ways necessary to develop those relationships.” (McGill, 1985, page 177) Although closeness has been defined and measured in various ways (see Berscheid, Snyder, & Omoto, 1989, for a review), lay views of what comprises close friendship include an open sharing of information about the self, known as self-disclosure. The nature of self-disclosure and its occurrence in intimate friendship is one of the most oft-studied variables in research on friendship and intimacy (Stewart, Cooper, Stewart, & Friedley, 2003). In fact, in their seminal typology of intimacy types, Schaefer and Olson (1981) described emotional intimacy, in part, as the act of self-disclosing. According to Pearce and Sharp (1973), self-disclosure is defined as occurring “when people voluntarily communicate information about themselves that other people are unlikely to know or discover from other sources.” In an extension of that definition, many researchers also believe that the communicated information must be personally private and that the discloser must believe that there are negative consequences were that information revealed to the general populace (See Miller & Steinberg, 1975 and Parks, 1982 for reviews). Using this extended definition of self-disclosure, much research has found that men are relatively reticent to engage in self-disclosure within their male friendships. Men are unlikely to discuss personal topics such as sadness and fears with male friends (Allen & Hacoun, 1976; Davidson and Duberrnan, 1982; Rubin, 1983). Consistently, men seem reticent to converse in a relational or personal nature in same-sex dyads, preferring to stick to topical discussions (Davidson & Duberrnan, 1982). As evidence of the robustness of this effect, meta-analytic research tends to suggest that, compared to women, men are relatively inhibited to self-disclose within their same-sex friendships (d = .31, Dindia & Allen, 1992). The current study scrutinizes a few key factors which may lead to this finding. This research looks at relationships between gender role orientation and self-disclosive behavior, empirically confronting the cultural stereotype that a masculine gender role orientation is negatively associated with self-disclosure and a feminine gender role orientation is positively associated with self- disclosure. This study also extends Bowman’s (2004) research on men and perceived appropriateness of self disclosure by predicting the self-reported friendship behaviors of men rather than their inferences about hypothetical other men. Also, the degree to which one perceives that his male friend has knowledge of him is discussed as a potential moderator of how homophobic self- presentation affects one’s self-disclosive behavior. As one’s time (and thus, knowledge) in a relationship increases, that individual may become less likely to allow homophobic self-presentation (i.e., the fear of being perceived as gay) to influence his self-disclosive behavior. Most significantly, the present research looks at the effects of self-disclosure on closeness within a relationship, with increasing self-disclosure positively correlated with increasing closeness within a relationship. In addition to looking at the effects of gender role orientation and homophobic self-presentation within male friendships, the present study examines the relationship between self-disclosure and closeness using multiple conceptualizations of both. Self-disclosure is seen as consisting of four key components: 1) amount of disclosure, 2) control of disclosure, 3) topical breadth of disclosure, and 4) valence of disclosure. Closeness also is operationalized using four elements: 1) one’s perceived closeness to their friend, 2) one’s perceived interdependence with their friend, 3) the strength of influence that the friend has over oneself, and 4) the diversity of activity in which the individual and their friend may engage. (See Figure 1 for the proposed conceptual model.) Gender Role Orientation and Self-Disclosure Sex and gender are the most oft-studied demographics with regard to their effect on self-disclosive behavior. Typically sex, gender, and gender roles are defined in terms of “those characteristics that actually differentiate the sexes, are stereotypically believed to differentiate the sexes, or are considered to be differentially desirable in the two sexes” (Lenney, 1991 ). ln elucidating the measurement of sex and gender roles, scholars had historically viewed masculinity and femininity as polar opposites on a single continuum, with Bern (1974) being among the first to question said conceptualization. Bern instead argued that individuals may in fact be able to possess both masculine features and feminine features simultaneously. Bem’s groundbreaking sex role inventory (Bem’s Sex Role Inventory, or BSRI, 1974) treated masculinity and femininity as independent dimensions, rather than a single continuum, and allowed for the creation of two new sex role categories: androgynous and undifferentiated. Individuals scoring equally high on the masculine and feminine dimensions of the BSRI represented androgyny, described as a “specific tendency to describe oneself in accordance with sex-typed standards of desirable behavior for men and women,” (Bern, 1974). Those individuals in the undifferentiated category scored low on both the masculine and feminine dimensions of the BSRI. As arguably the most oft-used measure of gender, the BSRI measures the degree to which an individual self-identifies with socially-desirable sex-typed characteristics, resulting in an index of one’s masculine, feminine, and androgynous or undifferentiated nature. Indeed, not only does one’s biological sex have an influence, where men are relatively reticent to disclose to other men as compared to women (Blieszner & Adams, 1992; Duck & Pittman, 1994) and are not expected to disclose as much or as well as women (Duran & Kelly, 1985; Jones & Brunner, 1984), but gender role orientation is also very influential for male self-disclosure, such that masculinity is negatively related to the intimacy of disclosure among men (Winstead, Derlega, & Wong, 1984). Historically, masculine gender roles have been shown to lead to low intimacy and vulnerability in friendships (Prager, 1995). In fact, there is practically no masculine socialization to engage in any form of expressive communication (Wood, 2000). Indeed, the act of expressing emotion and receiving other men’s expressions of emotion was actually associated with distress among men (Eisler, 1995). One wonders if these sex- and gender-based behavioral propensities have an effect on male behavior. Research has found that individuals are relatively invested in their cognitive generalizations of the self (e.g., gender role orientation) and behave according to those self-representations (Markus, 1977). Individuals who have a masculine self-schema (Le, a masculine cognitive generalization of the self) are able to process and enact stereotypically masculine information and behaviors more quickly than those individuals with a feminine self-schema (Markus, Crane, Bemstein, & Siladi, 1982). Therefore, those individuals who have cognitive representations of the self as masculine and heterosexual should desire to present the self consistently and be fluent in the successful portrayal of those behaviors. Additionally, in discussing gender role theory, Eagly, Wood, and Diekman (2000) claimed that performing "feminine” behaviors is seen as having a particularly negative impact on the gender role orientations of men when performed in the presence of other men. These social propensities become entrenched in men’s same-sex friendship behaviors and are seen as gender differences in self-disclosure and the creation of intimacy (Fehr, 2004). Concern with Appearing Homosexual One factor inhibiting male friendship behaviors among same-sex friends may be a concern with appearing homosexual. Phua (2002) claimed that oftentimes there is an inherent implication that there is a proper and normative way to act for heterosexual men which differs from homosexual men. Men are motivated to behave according to this norm. In a piece which spawned the field of heterosexism research, Rich (1980) argued that our culture has a “compulsory heterosexuality" in which there is a pervasive belief that being heterosexual is not only normative and laudable, but also prescriptive and obligatory (Rich 1980; Rubin 1975). This “presumption of heterosexuality" is entrenched to the point of heterosexism, and is believed to be a driving force for male behavior (Epstein & Johnson,1994;1998) The resulting concern with appearing homosexual may drive men’s self- disclosive activity. Bowman (2004) found support for this idea in a study examining men’s perceptions of disclosure appropriateness between two male friends. He had male college students read scenarios where a hypothetical male college student self-disclosed to another male friend. Participants evaluated the appropriateness of self-disclosure in scenarios that were both more and less normative for disclosure. Correlating Floyd’s (2000) homophobic beliefs scale with perceived appropriateness of self-disclosure revealed a negative relationship that held in both more and less normative contexts. Although Floyd (2000) claimed to measure general homophobic beliefs, these questions were developed to focus directly on the desire to not be seen as a homosexual male. This concern with appearing homosexual is a motivational characteristic and refers to the degree to which an individual fears that others may perceive him as being homosexual. The more these young men were concerned with appearing homosexual, the less they believed self-disclosure was appropriate between two male friends. Similarly, MacDonald and Games (1974) also found that men who rated highly on their Attitude Toward Homosexual Males Scale (ATHMS) measuring homophobia were likely to lack in intimacy behaviors such as sharing confidences and expressing love. Drawing from this research, a main barrier to male self-disclosure in same-sex friendships is a concern with appearing homosexual. Perceived Knowledge As A Potential Moderator An additional factor may come into the equation when dissecting the nature of one’s concern with appearing homosexual. This concern may be attenuated as friendship partners progress in their relationship. Early in the relationship, men may experience more relational uncertainty and therefore engage in tactics which serve to lessen that uncertainty. Some men may actively work to dispel others’ potential views of the self as homosexual, engaging in repudiative tactics which are used to disconfirrn a potential identity; other men may actively work to present the self as heterosexual, engaging in attributive tactics which are used to confirm a potential identity (Roth, Harris, & Snyder, 1988). As the relationship deepens, and individuals have greater knowledge and understanding of one another, those individuals may gain enough information and behavioral evidence to quell any potential thoughts that one is homosexual; thus, the degree to which one is concerned about appearing homosexual may be diminished. In this situation, as one’s perception of a friend’s knowledge of one’s self increases, the effect on self disclosure of the concern with appearing homosexual will decrease. Relational Closeness In addition to having increased relational knowledge, those relationships with more self-disclosure may, in fact, experience greater relational closeness. Because men are reticent to engage in self-disclosure unless they are very close to the other individual, these men may actually experience negative feelings due to the incongruency of engaging in a relationship that is disclosive but not close. As such, disclosure among men may, in fact, lead to increased perceptions of relational closeness within male friendships. Drawing upon and extending Bersheid, Snyder, and Omoto’s (1989) definition, in the present study closeness is conceptualized as the extent to which an individual within a relationship exhibits interdependence, liking, and mutual knowledge of the other individual. This interdependence is exhibited when individuals have a strong impact upon diverse activities of another individual for a long duration (Kelley et al., 1983). Additionally, as this relationship progresses towards friendship, the disclosure of intimate and personal information is more likely to happen. If men believe that disclosive relationships must, by definition, be close, then this relational closeness is expected to occur, in part through increased knowledge of the other individual. As such, these relationships are expected to be seen as close, as a result of this self-disclosure within a relationship (See Fehr, 2000, for a review). Hypotheses Multiple factors are seen as influential in the nature of friendship among men. Gender and homophobic self-presentation are expected to affect the likelihood that men are going to engage in disclosive conversation with their male friends. In addition, it is expected that there is a robust relationship between self- disclosure and closeness. The present study scrutinizes these independent factors, communication processes, and relational outcomes using multiple conceptualizations of both. As mentioned before, in general men are relatively reticent to disclose. As reviewed, the gender role orientations of these men are expected to have great explanatory power for the behaviors that they enact in their same-sex friendships. As each male participant’s degree of adherence to masculine gender roles possessed by an individual increases, the self-reported self-disclosure should decrease. Feminine gender role orientations are also stereotypically associated with disclosive, nurturing behavior (Bem, 1974). Consequently, as each male participant’s degree of adherence to feminine gender roles increases, the self-reported self-disclosure should also increase. The prior rationale suggests the following hypothesis: H1a: Male participant’s degree of adherence to masculine gender roles (one ’s masculine gender role orientation) will be negatively correlated with self-reported self-disclosive behaviors (amount, control, breadth, and valence) within same-sex male friendships. H 1b: Male participant’s degree of adherence to feminine gender roles (one ’s feminine gender role orientation) will be positively correlated with self-reported self-disclosive behaviors (amount, control, breadth, and valence) within same-sex male friendships. In addition to these effects of gender role orientation of self-disclosive behavior, prior research (Bowman, 2004) shows that as homophobic beliefs increase, the perceived appropriateness of that self disclosure decreases (r: - .30). Bowman assumed that heterosexual men’s inferences about other hypothetical men engaged in self-disclosure would match their own predispositions to self-disclose. By scrutinizing a slightly different perspective and looking at established friendships between men, the following proposed hypothesis tests that assumption: H2: Floyd’s (2000) scale assessing one ’5 concern for appearing homosexual will be negatively correlated with one ’s self-reported self-disclosive behaviors in male same-sex friendships. As the relationship among same-sex male friends deepens, those friends are expected to gain experience, knowledge, and history. These outcomes of a deepening friendship provide information that increasingly counteracts the otherwise-present concern that men have of appearing homosexual. The more that friends are mutually aware of each other’s personality, character, and sexuality, the less they should feel the need to worry about their friend’s superficial assessment of each other. Accordingly, as perceptions of the other‘s knowledge of self increases, the following interaction effect should occur: 10 H3: Perceived other’s knowledge of self moderates the negative association between one’s concem with appearing homosexual and self-reported self-disclosure (amount, control, breadth, and valence), such that a perception of more knowledge should demonstrate an attenuated negative correlation than a perception of less knowledge. Some scholars view self-disclosure as a primary means to the creation of feelings of closeness among same-sex friends, whereas other scholars view closeness among same-sex friends as created in one of two separate but equal ways, through either shared activity or self-disclosure (see Fehr, 2004 for a discussion). As aforementioned, this study adopts the theoretical perspective that self-disclosure is a primary means to the creation of feelings of closeness. Because of cultural views claiming that men would likely only disclose to close friends, men may uniquely create feelings of closeness by self-disclosing within their same sex friendship. Altemately, the argument could be made that disclosure is an outcome of relationships that are already close. Regardless of causality, the hypothesized relationship is likely to occur. Using this perspective, self-disclosure is expected to be positively associated with feelings of closeness in same-sex friendship. H4a: Self-reported self-disclosive behavior (amount, control, breadth, and valence of disclosure) will be positively correlated with relational closeness as measured by participants’ perceived closeness. 11 H4b: Self-reported self-disclosive behavior (amount, control, breadth, and valence of disclosure) will be positively correlated with relational closeness as measured by interdependence. H4c: Self-reported self-disclosive behavior (amount, control, breadth, and valence of disclosure) will be positively correlated with relational closeness as measured by diversity of shared activity. H4d: Self-reported self-disclosive behavior (amount, control, breadth, and valence of disclosure) will be positively correlated with relational closeness as measured by partner’s strength of influence. Method Participants The participants in the study were 115 heterosexual male undergraduate students at Michigan State University who participated in exchange for class credit in their introductory communication classes. Students ranged in age from 18 to over 25, with 50% of the students falling in ages 21-22. Participants also ranged from freshmen to senior level students, with the majority (66%) of students being juniors or seniors. Friendship Behavior Questionnaire Upon beginning the study, participants were asked to answer the Friendship Behavior Questionnaire (see Appendix A). First, the FBQ contained selected items from the Revised Self-Disclosure Scale (Wheeless & Grotz, 1976) and Jourard and Lasakow’s (1958) Self-Disclosure Scale, chosen to measure communicative depth, breadth, and valence. The closeness of the relationship through interdependence was measured using a modified version of Aron, Aron, 12 and Smollan’s (1992) Inclusion of the Other in Self scale, and also through items selected from the Berscheid, Snyder, and Omoto (1989) Relationship Closeness Inventory. Additionally, items were developed to measure closeness through the perceptions of other’s knowledge of self. Other items also assessed closeness among friends through liking, including some items from Rubin’s (1975) liking scale and McCroskey and McCain’s (1974) Social Attraction scale. Further questions determined the length of the friendship, the amount of time each individual spent together each week, and basic demographic information. Also embedded within the Friendship Behavior Questionnaire were questions which attempted to ascertain the gender role orientation of participants, and questions which attempted to determine one’s concern with appearing homosexual. Gender role orientation is often measured through the seminal and highly popular Bem’s Sex Role Inventory (1974). Bem’s inventory, though popular, has undergone significant revisions, particularly for issues of length. Because one of these revisions (Wheeless & Dierks-Stewart, 1981) has proven both reliable and significantly shorter in length (i.e., Aylor & Dainton, 2004; Reeder, 2003; Zhang, Norvilitis, & Jin, 2001 ), the questions used to determine the gender role orientation of participants were taken directly from the masculinity and femininity scales of the Revised Bem’s Sex Role Inventory (for a review of this version, see Wheeless & Dierks-Stewart, 1981) in an attempt to lessen participant fatigue. The questions related to the concern to not be seen as 13 homosexual were drawn directly from Floyd’s (2000) scale of homophobia. Procedure Introduction to the study. Participants volunteered for “a research study on friendship behaviors.” Upon arriving at the study, participants were assigned to sit in a moderate-sized classroom with tables and chairs arranged such that participants are unable to see one another’s writing surface. Participants then indicated their consent to participate in the study (Appendix B). Perceived Knowledge Induction. After consenting, participants received the “Friendship Behavior Questionnaire.” This questionnaire contained items to determine participants’ self-reported friendship behaviors. Participants were asked to report on one of two individuals: “Please think about a male friend of yours (non-relative) who knows you very well. Write their initials here: _ Please answer the following questions based on that relationship,” (more knowledge induction) or “Please think about a male friend of yours (non-relative) who does not know you very well. Write their initials here: __ Please answer the following questions based on that relationship,” (less knowledge induction). After completing the questionnaire, participants were debriefed, given credit for their participation, and invited to receive a copy of the results upon completion of the study. Manipulation Check: Perceived Other-Knowledge of Self Five questionnaire items tested the success of the manipulation of perceived other-knowledge of self: (a) “I feel like this person knows me very well,” (b) “This person has knowledge about me that other people likely do not know,” (0) “Based 14 on what this person has learned about me, they could probably predict my behavior in different situations,” (d) “He knows a lot about me,” and (e) “He knows me personally" rated on a scale from 1 (strongly disagree) to 9 (strongly agree). Because these six items were highly correlated, a composite individual measure was created by averaging the items (standardized Item 0 = .93). As expected, a one-way between groups analysis of variance using this composite measure found that participants given the more knowledge questionnaires rated their friends as having more knowledge about themselves (M = 7.22, SD = 1.41) than were those given the less knowledge questionnaires (M = 4.52, SD = 1.57), F (1, 114) = 94.62, p < .001, n2 = 0.46. Therefore, the induction of perceived other-knowledge of self was successful. Further proof of the distinction between these categories is evident in the relationship between friendship length and the degree to which participants described their male friends as having knowledge about them. A between groups one-way analysis of variance using a self-reported measure of friendship length in months found that participants given the more knowledge questionnaires had a longer history of friendship (M = 102.00, SD = 62.30) than were those given the less knowledge questionnaires (M = 47.63, SD = 42.87), F (1 , 112) = 28.69, p < .001, n2 = 0.21. The more a participant reported his friend as having knowledge about him, the longer that friendship had been in place. Also, the distinction between these categories is evident in the relationship between participants’ perceptions of the time spent with their male friends and the degree to which participants described those male friends as having 15 knowledge about them. A between groups one-way analysis of variance using a self-reported measure of average time spent together per week in minutes found that participants given the more knowledge questionnaires believed that they spent more time together on average (M = 284.90, SD = 333.46) than were those given the less knowledge questionnaires (M = 177.06, SD = 216.93), F (1 , 1 14) = 4.07, p < .05, n2 = 0.04. The more a participant reported his friend as having knowledge about him, the more he perceived that he spent time with that friend. Taken together, the participants in the more knowledge condition described a male friend who had more knowledge about participants, had a longer relational history with them, and spent more time with them as compared to those in the less knowledge condition. Independent Variables Masculine Gender Role Orientation. The questions used to determine the gender role orientation of participants were taken directly from the masculinity and femininity scales of the Revised Bem’s Sex Role Inventory (Wheeless & Dierks-Stewart, 1981 ). Participants were asked the degree to which they possessed certain behavioral characteristics, rated on a scale from 1 (never or almost never true) to 7 (always or almost always true), that were indicative of a masculine gender role orientation through ten questionnaire items: (a) “Acts as a leader,” (b) “Aggressive,” (c) “Assertive,” (d) “Competitive,” (e) “Dominant,” (f) “Forceful,” (9) “Has leadership qualities,” (h) “Independent,” (i) “Has strong personality,” and (j) “Willing to take a stand”. Because these ten items were highly correlated, a composite individual measure was created by averaging the 16 items (standardized Item 0 = .91 ). As expected, these masculine items held together as one factor in a confirmatory factor analysis (Hunter & Hamilton, 1988) and were seen as a distinct factor from the feminine items. Feminine Gender Role Orientation. Participants were also asked the degree to which they possessed certain behavioral characteristics, rated on a scale from 1 (never or almost never true) to 7 (always or almost always true), that were indicative of a feminine gender role orientation through ten questionnaire items: (a) “Compassionate,” (b) “Eager to soothe hurt feelings,” (c) “Friendly,” (d) “Gentle,” (6) “Helpful,” (f) “Sensitive to others’ needs,” (9) “Sincere,” (h) “Tender,” (i) “Understanding,” and (j) “Warm”. Because these ten items were highly correlated, a composite individual measure was created by averaging the items (standardized Item 0 = .93). As expected, these feminine items also held together as one factor in a confirmatory factor analysis (Hamilton & Hunter, 1988) and were seen as a distinct factor from the masculine items. Concern with Appearing Homosexual. The questions related to the concern to not be seen as homosexual were drawn directly from Floyd’s (2000) scale of homophobia. Floyd’s five items, rated on a likert-type scale from 1 (strongly disagree) to 9 (strongly agree), include (a) “I would be very upset if someone else thought I was gay,” (b) “I am careful not to do things that might make others think I am homosexual,” (c) “If someone questioned my sexual orientation, it would not bother me” (reverse scored), (d) “If a homosexual person began talking to me in public, I would be concerned about what other people might think,” and (e) “I would be very ashamed if someone I know thought I was 17 gay.” In Bowman’s (2004) research using this scale, these five items were embedded within a larger questionnaire that contained 45 additional filler items; this original embedding was intended to disguise the true nature of the questions. However, because of the large nature of the project and resulting fears of participant fatigue, these 5 items were presented together as one scale and were not embedded in a larger questionnaire. Unlike Bowman’s (2004) previous research, in this unembedded format these items did not hold together as one factor in a confirmatory factor analysis (Hamilton & Hunter, 1988) and exhibited a lower standardized item alpha value (or = .73) than reported in previous research (Bowman, 2004; Floyd, 2000). Accordingly, no results using this composite scale were found to be statistically significant at the p < .05 level. Self-Disclosure Items Amount and Control of Disclosure. Originally, amount of disclosure and control of disclosure were intended to have separate measures. Seven questionnaire items were intended to measure participant’s perceived amount of disclosure to their friends: (a) “I do not often talk about myself (reverse scored),” (b) “My statements of my feelings are usually brief (reverse scored),” (c) “I usually talk about myself for fairly long periods at a time,” (d) “My conversation lasts the least time when I am discussing myself (reverse scored),” (9) “I often talk about myself,” (f) “I often discuss my feelings about myself,” and (g) “I frequently express my personal beliefs and opinions,” rated on a scale from 1 (strongly disagree) to 7 (strongly agree). In addition, five questionnaire items were intended to measure participant’s perceived control of disclosure to their 18 friends: (a) “I intimately disclose who I really am, openly and fully in my conversation,” (b) “Once I get started, my self-disclosures last a long time,” (c) “I often disclose intimate, personal things about myself without hesitation,” (d) “I feel that I sometimes do not control my self-disclosure of personal or intimate things that | tell about myself,” and (9) “Once I get started, I intimately and fully reveal myself in my self-disclosures,” rated on a scale from 1 (strongly disagree) to 7 (strongly agree). A confirmatory factor analysis (Hunter & Hamilton, 1988) showed that a one-factor model better fit these twelve items than did the intended two-factor model, which separated both the amount of disclosure and control of disclosure. Indeed, with regard to the measure’s face validity, further scrutiny shows congruence of the items if combined to observe the same factor (henceforth referred to as amount of disclosure). Because these twelve items were highly correlated, a composite individual measure was created by averaging the items (standardized Item 0 = .85). As expected, these twelve items held together as one new factor in a confirmatory factor analysis (Hunter & Hamilton, 1988) and were seen as distinct from the other self disclosure items. Breadth of Disclosure. An additional measure which addressed the degree to which individuals engaged in self-disclosive behavior looked at the breadth of disclosure. Four questionnaire items tested the extent to which participants engaged in a broad range of topics of discussion with their friend: (a) “I talk about lots of different things with my friend,” (b) “I typically talk about one main subject with my friend,” (c) “My friend and I usually talk about a broad range of subjects,” and (d) “I usually have only one topic that I discuss with my friend” rated on a 19 scale from 1 (strongly disagree) to 7 (strongly agree). Because these four items were highly correlated, a composite individual measure was created by averaging the items (standardized Item 0 = .85). As expected, these items held together as one factor in a confirmatory factor analysis (Hunter & Hamilton, 1988) and were seen as distinct from the other self disclosure items. Valence of Disclosure. Because the willingness to share negative information is typically conceptualized as being more disclosive (see Miller, 1975 for a review), the valence of self-disclosure is expected to be an indicator of disclosive behavior. Eleven questionnaire items tested the extent to which participants were willing to self-disclose negative information as compared to positive information: (a) “I usually disclose positive things about myself (reverse scored),” (b) “On the whole, my disclosures about myself are more negative than positive,” (c) “I normally reveal “bad” feelings I have about myself,” (d) “I normally “express” my good feelings about myself (reverse scored),” (9) “I often reveal more undesirable things about myself than desirable things,” (I) “I usually disclose negative things about myself,” (9) “I will typically reveal information about myself that is about something bad happening in my life,” (h) “I will typically reveal information about myself that is about something good happening in my life (reverse scored),” (i) “I usually talk about happy things when I talk about myself with my friend (reverse scored),” (j) “I usually talk about the darker side of my life with my friend,” and (k) “On the whole, my disclosures about myself are more positive than negative (reverse scored)” rated on a scale from 1 (strongly disagree) to 7 (strongly agree). Because these eleven items were correlated, a 20 composite individual measure was created by averaging the items (standardized Item (1 = .67). As expected, these items held together as one factor in a confirmatory factor analysis (Hunter & Hamilton, 1988) and were seen as distinct from the other self disclosure items. Closeness Items Perceived Closeness. Three questionnaire items tested the extent to which participants rated their relationship with their friend as close: (a) “This person is a very good friend of mine,” (b) “This person and l are very close,” (c) “I consider this person to be a close friend,” rated on a scale from 1 (strongly disagree) to 9 (strongly agree). Because these three items were highly correlated, a composite individual measure was created by averaging the items (standardized Item (1 = .962). As expected, these items held together as one factor in a confirmatory factor analysis (Hunter & Hamilton, 1988). Interdependence. To measure interdependence, a modified version of Aron, Aron, & Smollan’s (1992) Inclusion of the Other in Self scale was included, showing increasingly overlapping circles as indicative of increasingly interdependent, “bonded” relationships (included in Appendix A). Diversity of Activity. The measure of diversity of activity consisted of a modified version (to reflect male friendships) of Berscheid, Snyder, & Omoto’s (1989) Relationship Closeness Inventory diversity scale (included in Appendix A). Participants received the prompt “In the past week, I did the following activities with my friend:” and then answered true or false to 22 items describing potential friendship activities (e.g., “watched TV or a video,” “ate a meal”). The number of 21 “true” responses was then summed to create a composite measure of the diversity of activities in which participants engaged with their male friend (standardized Item 0 = .90). Strength of Influence. This measure used portions of Berscheid, Snyder, & Omoto’s (1989) Relationship Closeness Inventory (included in Appendix A), with participants rating 27 items assessing the strength of their friends’ general influence on various aspects of life using a scale from 1 (strongly disagree) to 7 (strongly agree). These items (e.g. “This friend influences important things in my life,” and “This friend influences the basic values that I hold.”) were then summed to create a composite measure of the strength of influence held by participants’ friends (standardized Item 0 = .89). Results Descriptive Statistics Descriptive statistics for all predictor, process, and outcome variables that are used in the correlations in the following hypothesis tests may be found in Table 1. In general, variables had distributions which seemed to approach a unimodal normal curve. Exceptions to this normalcy included mostly negatively skewed distributions (Masculine Gender Role Orientation, Feminine Gender Role Orientation, and Breadth of Disclosure) with one positively skewed distribution (Diversity of Activity). Hypothesis Tests Hypotheses 1a-1b. The first series of hypotheses predicted that one’s gender role orientation would be significantly correlated with multiple elements of 22 participants’ self-reported self-disclosive behaviors (amount, breadth, and valence of disclosure), such that one’s degree of masculine orientation would be negatively correlated with self-disclosive behaviors and one’s degree of feminine orientation would be positively correlated with self-disclosive behaviors. These correlations are shown in the top two rows of Table 2. Hypothesis 1a - Amount of Disclosure. Using the aforementioned composite measure of one’s amount of disclosure, a correlation was employed to test the relationship between men’s self-reported amount of disclosure and their masculine gender role orientations. Counter to the predictions of Hypothesis 1a, a significant positive correlation emerged between men’s self-reported amount of disclosure and their masculine gender role orientation, r (1 13) = .21, p < .05. Hypothesis 1a — Breadth of Disclosure. Using the aforementioned composite measure of one’s breadth of disclosure, a correlation was employed to test the relationship between men’s breadth of disclosure and their masculine gender role orientations. Counter to predictions, no significant correlation emerged between masculine gender role orientation and the breadth of disclosure, r(113) = .15, n.s. Hypothesis 1a — Valence of Disclosure. Using the aforementioned composite measure of the valence of men’s disclosure, a correlation was employed to test the relationship between one’s willingness to share negative information and their masculine gender role orientations. As predicted, a negative correlation between a masculine gender orientation and one’s willingness to share negative information emerged, r (1 13) = -.20, p < .05. 23 One’s masculine gender role orientation was significantly related to one’s willingness to share negative information as predicted, but was not related in the expected manner to the amount, or breadth of self-disclosure. Therefore, the hypothesis (Hypothesis 1a) predicting the relationship between a masculine gender role orientation and one’s self-disclosive behaviors with male friends received mixed support. Hypothesis 1b — Amount of Disclosure. In a test of hypotheses 1b, the composite measure of one’s amount of disclosure was correlated with one’s feminine role orientation. As hypothesized, a significant positive correlation emerged between men’s self-reported amount of disclosure and their feminine gender role orientation, r(113) = .24, p < .01. Hypothesis 1b - Breadth of Disclosure. Breadth of disclosure was also expected to be affected by one’s feminine gender role orientation, such that an increase in levels of feminine gender role orientation were expected to be positively related to the breadth of disclosure. Counter to predictions, no significant correlation emerged, r(113) = .07, n.s. Hypothesis 1b — Valence of Disclosure. It was expected that a higher feminine gender role orientation would be positively correlated with one’s willingness to disclose negative information about the self. However, no significant correlation occurred between those two variables, r(113) = -.08, n.s. One’s feminine gender role orientation was significantly related to one’s amount of disclosure as expected, but no significant correlations emerged between the breadth and valence of disclosure. Therefore, hypothesis 1b, which 24 predicted a positive relationship between a feminine gender role orientation and one’s self-disclosive behaviors with male friends received mixed support. Hypotheses 1a-1b: Combined effects. The bivariate correlation tests between gender role orientation and self-disclosure ignore the strong positive correlation between masculine and feminine gender role orientations, r(113) = .63, p<.01 To examine the combined and independent effects of each gender role orientation on each of the three self-disclosure measures, a multiple regression analysis was performed. While the overall linear multiple regression analysis of the combined gender role orientations was statistically significant for amount of disclosure, F (3, 111) = 3.76, p < .05, the independent effects of masculine and feminine gender role orientation were not (see Table 3 for individual )6 values). It seems that the best predictor of amount of disclosure is the combination of both gender orientations. Linear analyses of the combined gender role orientation variables were not statistically significant for either tests of breadth of disclosure, F (3, 111) = 1.41, n.s., or valence of disclosure (F (3, 111) = 2.48, n.s.. Hypothesis 2. As discussed earlier and reported in other research (Bowman, 2004; Floyd, 2000), it was expected that a negative relationship would emerge between one’s concern with appearing homosexual and one’s self- reported self-disclosive behaviors. Unfortunately, the five items which comprised the homophobia scale were not embedded within a larger scale and subsequently did not hold together as one factor in a confirmatory factor analysis (Hamilton & Hunter, 1988) and exhibited a lower standardized item or than 25 presented in other research. Accordingly, no results using this composite scale were found to be significant. The correlation between one’s concern with appearing homosexual and amount of disclosure was not significant, r: -.07, n.s. The correlation between one’s concern with appearing homosexual and the breadth of disclosure was also not significant, r: -.08, n.s. Finally, the correlation between one’s concern with appearing homosexual and negative valence of self- disclosure was not significant, r: -.15, n.s. These correlations are shown in the third row of Table 2. Hypothesis 3. The third hypothesis predicted a potential moderator: participants’ perceptions of how much knowledge their friends have about each participant. It was hypothesized that a perception of more knowledge should demonstrate a lower negative correlation between concern with appearing homosexual and self-reported self-disclosive behaviors than would a perception of less knowledge. However, due to the failure of the scale that measured one’s concern with appearing homosexual, no negative correlation emerged. In addition, the null hypothesis tests for the difference between two correlations (Cohen, Cohen, West, & Aiken, 2003) failed to produce statistically significant differences between the more and less knowledge conditions when looking at one’s concern with appearing homosexual and the subsequent correlations with measures of the amount of disclosure z(114) = .45, n.s., breadth of disclosure 2 (114) = .77, n.s., and valence of disclosure 2(114) = .19, n.s. between these friends. Consequently, no moderating effect was found and Hypothesis 3 was not supported (see Table 4). 26 Hypotheses 4a-4d. The fourth series of hypotheses predicted that multiple measures of closeness (perceived closeness, interdependence, diversity of activity, and strength of influence) would be positively correlated with participants’ self-reported self-disclosive behaviors (amount, breadth, and valence of disclosure). These correlations are shown in Table 5. Hypothesis 4a - Perceived Closeness. Perceived closeness was expected to be positively related to all measures of self-disclosure. It was positively related to amount of disclosure (r (1 13) = .277, p < .005), and breadth of disclosure (r (113) = .409, p < .001 ). Counter to predictions, perceived closeness was not related to valence of disclosure (r (1 13) = -.058, n.s.). A linear multiple regression analysis with the independent variables of amount, breadth, and valence of disclosure entered simultaneously was performed to evaluate the effects of self- disclosive behavior on perceived closeness, and was statistically significant F (3, 111) = 9.85, p = .001, R2 = .210. Similar to the bivariate correlations reported above, a multiple regression analysis found that amount and breadth of disclosure were significant independent predictors of perceived closeness; surprisingly, the regression analysis showed valence of disclosure to be a predictor of perceived closeness in the opposite direction than what was expected. (see Table 6 for individual [3 values). Hypothesis 4b - Interdependence. Interdependence was expected to be positively related to all measures of self-disclosure. It was positively related to amount of disclosure (r(113) = .281, p < .005), and breadth of disclosure (r(113) = .413, p < .001). Counter to predictions, interdependence was not related to 27 valence of disclosure (r (1 13) = -.059, n.s.). A linear multiple regression analysis with the independent variables of amount, breadth, and valence of disclosure entered simultaneously was performed to evaluate the effects of self-disclosive behavior on interdependence, and was statistically significant F (3, 111) = 10.13, p = .001, I?2 = .215. As was found in the bivariate correlational analysis, both amount and breadth were significant independent predictors of interdependence. (See Table 6 for individual 6 values). Hypothesis 4c — Diversity of Shared Activity. Diversity of shared activity was expected to be positively related to all measures of self-disclosure. It was positively related to amount of disclosure (r (1 13) = .343, p < .001), and valence of disclosure (r (1 13) = .231, p < .05). Counter to predictions, diversity of shared activity was not related to breadth of disclosure (r(113) = .167, n.s.). A linear multiple regression analysis with the independent variables of amount, breadth, and valence of disclosure entered simultaneously was performed to evaluate the effects of self-disclosive behavior on diversity of shared activity, and was statistically significant F (3, 111) = 7.98, p = .001, R2 = .177. Similar to the bivariate correlations reported above, a multiple regression analysis found that amount of disclosure and valence of disclosure were significant independent predictors of diversity of shared activity, but counter to predictions breadth was not a significant predictor of diversity of shared activity. (see Table 6 for individual 8 values). Hypothesis 4d - Strength of Influence. Strength of influence was expected to be positively related to all measures of self-disclosure. It was positively related 28 to amount of disclosure (r(113) = .314, p < .001), and valence of disclosure (r (113) = .260, p < .005). Counter to predictions, strength of influence was not related to breadth of disclosure. (r(113) = -.025, n.s.). A linear multiple regression analysis with the independent variables of amount, breadth, and valence of disclosure entered simultaneously was performed to evaluate the effects of self-disclosive behavior on strength of influence, and was statistically significant F (3, 111) = 6.74, p = .001, 32 = .154. Similar to the bivariate correlations reported above, a multiple regression analysis found that amount of disclosure and valence of disclosure were significant independent predictors of strength of influence, but counter to predictions breadth was not a significant predictor of strength of influence. (see Table 6 for individual 6 values). Discussion The purpose of this study was to show the relationships between one’s gender role orientation, homophobic self-presentation, and self-reported self- disclosive behaviors. Additionally, the interrelationships between those self- disclosive behaviors were expected to be related to various measures of relational closeness between these established friends. Gender Role Orientation As noted earlier, gender effects on self-disclosive behavior received mixed results. Consistent with the body of literature implying that men are reticent to disclose negatively-valenced personal information about the self, an increasing masculine gender role orientation was associated with a decrease in the self- reported willingness to discuss negative information about the self. However, 29 one’s masculine gender role orientation was, surprisingly, not negatively related to the amount or breadth of self-disclosive behavior as expected. One potential explanation that was explored post hoc to explain this finding was the androgynous nature of the individuals. While a masculine role orientation was expected to be negatively related to the occurrence of disclosive or emotionally intimate behaviors among male friends, it has been suggested that androgynous individuals initiate and maintain satisfying relationships to a greater degree than do individuals with a masculine or feminine gender role orientation (lckes, 1985). As such, an individual with strong feminine and masculine characteristics (i.e., those individuals classified as having an androgynous role orientation) may be likely to engage in disclosure and friendship behaviors in a similar manner as was predicted for those with a high feminine gender role orientation; in essence, perhaps the presence of the feminine may “trump” the inhibiting nature of the masculine role orientation in certain categories. To test this idea, a measure of androgyny was created by summing the average of the ten masculinity items and the average of the ten femininity items of the Revised BSRI, and then subtracting the absolute difference between the two averages (i.e., [(average of masculine) + (average of feminine) ] - |(average of masculine) - (average of feminine)| ) to create a composite measure where the highest possible score (m = 14) represented an individual scoring as high as possible on both the masculine and feminine continua with no predilection towards either. The lowest possible score (m = 2) represented an individual with the most discrepant masculine and feminine scores or an individual scoring low 30 on both continua. As would be expected using these post hoc predictions, increasing androgyny was positively correlated with an individual’s amount of disclosure (r(113) = .26, p < .005), and perceptions of the other’s knowledge of the self (r (1 13) = .291, p < .005). Expected post hoc correlations between this new measure of androgyny and measures of breadth (r (1 13) = .132, n.s.) and valence of one’s disclosure (r (1 13) = .123, n.s.) did not emerge, indicating that androgyny may not ‘trump’ masculinity in the degree to which one discloses about a broad range of subjects or favors sharing mostly positively-valenced information about the self. Additionally, relational outcome variables like perceived closeness (r (1 13) = .336, p < .001) and interdependence (r (1 13) = .277, p < .005) were positively correlated with increasing androgyny as would be expected. However, this new measure of androgyny was not significantly related to outcome measures of strength of influence (r (1 13) = .029, n.s.) or diversity of activity (r (1 13) = .179, n.s.). These post hoc findings suggest that perhaps friendship behaviors are not easily categorized by masculine or feminine gender role orientations as previously conceptualized. Rather, one’s masculine and feminine gender role orientation have some influence on supporting and/or suppressing friendship behaviors and disclosure, but the combined effect of one’s gender role orientations (i.e., one’s degree of androgynous role orientation) may influence disclosure and friendship to a greater degree than was previously expected. Additionally, the gender role orientation of this group is unusually distinct, with high correlations between the masculine and feminine role orientations 31 (Table 2). As such, using this aforementioned composite measure, the current sample proved to be unusually androgynous, with 69% of participants falling above the midpoint of the scale (m = 9.52, s = 2.16). As such, both feminine and masculine gender role orientations produced parallel results for most of the self- disclosure items, despite a confirmatory factor analysis (Hunter & Hamilton, 1988) which showed the distinctive and separate nature of each scale. Heterosexual Self-Presentation A large portion of the study hinged on the expectation that the degree to which one desires to not be seen as a homosexual would have strong predictive ability in self-disclosure and friendships behaviors. As reported, the current iteration of the scale failed to demonstrate reliable or valid findings. The inability of the scale to display any significant results is both disappointing and intriguing, and worthy of scholarly scrutiny in itself. Several substantive differences may explain the current scale failure. First, Floyd’s (2000) initial use of the scale was in a context unrelated to friendship research. Studies have found that men are, in fact, reticent to even participate in such research (Lewis, Winstead, & Derlega, 1989). If the nature of the study alone was able to cause such a priming effect that activated generally homophobic tendencies, natural variations in heterosexual self-presentation would be attenuated. So, while Floyd’s original use of the scale did not prove unreliable or inappropriate, perhaps the research context that it was used in did not seem threatening to the individuals. As aforementioned, a second explanation for the failure of the scale may be due to the presentation of the heterosexual self-presentation scale. In past 32 friendship research using the scale, the author was able to embed the five items in a survey including 45 additional items which successfully disguised the true nature of the scale (Bowman, 2004).1 The current study did not preserve the author’s original embedding of said scale and, as such, may have created demand characteristics which influenced participants’ responses, resulting in a social desirability bias against revealing one’s true score of heterosexual self- presentation. Participants may have realized that the study examined heterosexual self-presentation and, as a result, answered in a more socially desirable manner. Additionally, while research has historically demonstrated the existence of homophobic tendencies in multiple forms (for a review see Morin & Garfinkle, 1978; see also Gentry, 1986; Millham, San Miguel, & Kellog, 1976; Weinberg, 1973), recent cultural shifts may have occurred which increase the perceived inappropriateness of homophobia. In fact, these ideas of increased social desirability biases led to Bowman’s (2004) original embedding of the questions to disguise the nature of the items. If cultural norms have increasingly prohibited the acceptability of espousing homophobic ideals, it is possible that similar cultural prohibitions may have arisen about consciously attempting to maintain a heterosexual image. Future research may want to look at the priming effects of such cultural prohibitions, where researchers may be able to induce the enactment of stereotypically feminine (or gay) behaviors by making salient one’s self-presentational biases; in an effort to present oneself in a socially desirable (non-homophobic) manner, individuals may be primed towards actually overestimating one’s amount of feminized behaviors. Subsequent research shall 33 need to explore the possibility of experimentally inducing this cognitive demand effect. Relational Closeness One of the most compelling findings elucidated in this project was the overwhelming support for the positive correlation between multiple measures of closeness (perceived closeness, interdependence, diversity of activity, and strength of influence) and multiple measures of participants’ self-reported self- disclosive behaviors (amount, breadth, and valence of disclosure). Indeed, of the sixteen bivariate correlations predicted, twelve were found to be statistically significant in the hypothesized direction. In addition, the four multiple regressions looking at the effects of disclosive behaviors on feelings of closeness produced significant R2 values, indicating that the amount of association between self- disclosure measures and each of the four dimensions of closeness may be a more useful measure of the relationships between these variables. While there may be other factors besides disclosure alone which lead to feelings of closeness among men, the combined correlational and multiple regression data indicates that the link between various measures of both closeness and self- disclosive behaviors among established friends has been well-supported in this study. Future research on this link between self-disclosure and closeness needs to be carried out in an experimentally controlled fashion to establish causality. Arguments can be made in either direction; the establishment of trust and rapport may cause closer friends to be more likely to engage in self-disclosure than 34 would less close individuals, but alternatively it is easily argued that individuals who engage in self-disclosure with one another may actually cause feelings of closeness to emerge or be reified by the act of self-disclosure itself. It may be useful to engage in research designed specifically to allow the causal processes to be clearly established. Future Directions The present research disentangles some of the interrelationships between masculinity, femininity, friendship behavior, and closeness within established friendships. Unlike much previous work on these concepts, the current study asked men to describe and detail information about specific current friendships, rather than having men simply describe what they believe generalized relationships among men look like. Additionally, the present research had a focus on ensuring that multiple measures and multiple conceptualizations of self- disclosure and various friendship behavior were included in the way that the project was carried out. Not only did individuals report on the amount of self- disclosure, but also the breadth and valence of that self-disclosive behavior. In addition, the current research used multiple conceptualizations of closeness to determine the multi-faceted ways that individuals conceive of what friendship looks like. Instead of simply determining the perceived closeness that individuals imagine exists between them and their friend, the research also looked at the diversity of activity that individuals engaged in, the degree of influence that the relationship exerts, and the interdependence among friends. 35 In scrutinizing friendship behavior, four key elements must be reviewed in order to gain an adequate picture of the nature of these important, universal relationships. It is important to look at how people conceptualize appropriate forms of friendship in general, how people believe they engage their own friends, and how people actually behave with their friends independent of what they think occurs. This three-pronged approach allows scrutiny of cognition, remembrance, perceived appropriateness, and actual behavior among male friends. Bowman (2004) assessed the perceived appropriateness of friendship behavior. Based on that work, the current study looked at how people believe they engage their own friends (remembrance and cognition) using a closed-ended approach. In the present study, participants were also asked some additional open-ended questions in an attempt to ascertain further remembrances and cognitions; these open-ended items were not analyzed in the present research, but in a future project they shall be coded, analyzed, and compared with various typologies of intimacy and friendship behaviors (i.e., Fehr, 2004), as well as correlated with appropriate closed-ended measures already used in the present research. (For example, measures items like “What do you typically do with this friend?” and “What types of things do you talk about with this friend?”, etc.) The next step in this line of research, currently under development, intends to look at actual observable behavioral elements of friendship interactions, correlating the degree to which individuals engage in such behavior with other elements of cognitive processes like participants’ degree of heterosexual self-presentation. Additionally, future research needs to be done which devises a useful and usable 36 measure of said heterosexual self-presentation, one which does not need to be embedded in such a unwieldy manner as the present form. Behavioral measures of heterosexual self-presentation and self-disclosive behavior need to be created, with coding schemes that produce both continuous and ogival measures of each. Self-disclosure has many effects on the nature of relationships. lndividuals’ self-disclosive behavior can cause discomfort or strengthen relationships, and lead to various cognitive and behavioral outcomes. Widening the scope of the already-broad field of self-disclosure research, further scrutiny of male self-disclosure is necessary to determine the complex interrelationships between men who reveal personal information about the self and the friends, family, and romantic partners who are on the receivingend of such information. 37 Footnotes 1 In fact, the population of individuals participating in Bowman’s (2004) research are expected to have no substantive differences from those participants in the current study. Participants were recruited from the same introductory communication courses at Michigan State University using the same recruitment procedures, approximately 12-15 months after the original study. 38 Appendix A: Friendship Behavior Questionnaire 39 Friendship Behavior Questionnaire Thank you for participating in this study! We are currently investigating the nature of friendships. Please think about a male friend of yours (non-relative) who knows you very well. Write their initials here: _ Please answer this entire packet of questionnaires based on that relationship. Again, please answer the following questions with regard to this particular person. Each survey is structured differently, so please read the directions carefully for each survey. PLEASE GO ON TO THE NEXT PAGE!!! 40 Friendship Behavior Questionnaire Thank you for participating in this study! We are currently investigating the nature of friendships. Please think about a male friend of yours (non-relative) who does NOT know you very well. Write their initials here: _ Please answer this entire packet of questionnaires based on that relationship. Again, please answer the following questions with regard to this particular person. Each survey is structured differently, so please read the directions carefully for each survey. PLEASE GO ON TO THE NEXT PAGE!!! 41 To begin, please answer the following questions as completely and honestly as possible. Your cooperation is appreciated! In this study, we are attempting to assess the nature of friendships. Everybody has their own definitions for friends, and that is fine with us as researchers. The following require only short answers. Please write your answers in the space provided. Write their initials again here: How long have you known this person? Below please indicate the number of years and/or months (for example, 3 years, 8 months) years months We would like you to estimate the amount of time you typically spend with this person during an average week. Think back over the past week and write in the average amount of time, per day, that you spent with X. If you did not spend any time with X, write 0 hour(s) 0 minutes. hour(s) minutes Compared with the "normal" amount of time you usually spend with this person, is this more than normal? Less than normal? Typical? Please Describe. PLEASE GO ON TO THE NEXT PAGE!!! 42 Instructions: Please mark the following statements to reflect how you communicate with the friend you mentioned before. Indicate the degree to which the following statements reflect how you communicate with this person by marking whether you (7) strongly agree, (6) agree, (5) moderately agree, (4) are undecided, (3) moderately disagree, (2) disagree, or (1) strongly disagree. Mark Your Answers by Filling in the Bubble on the Scantron that Corresponds with your Response, using the pencil provided. Ignore the numbers 8, 9, & 10 on the Scantron. Work quickly and just record your first impressions. 1 2 3 4 5 6 7 Strongly Disagree Moderately Undecided Moderately Agree Strongly Disagree Disagree Agree Agree 1) I do not often talk about myself 2) I usually disclose positive things about myself. 3) I intimately disclose who I really am, openly and fully in my conversation. 4) My statements of my feelings are usually brief. 5) On the whole, my disclosures about myself are more negative than positive. 6) Once I get started, my self-disclosures last a long time. 7) I usually talk about myself for fairly long periods at a time. 8) I normally reveal “bad” feelings I have about myself. 9) I often disclose intimate, personal things about myself without hesnaflon. 10) My conversation lasts the least time when I am discussing myself. 11) I normally “express” my good feelings about myself. 12) I feel that I sometimes do not control my self-disclosure of personal or intimate things that I tell about myself. 13) I often talk about myself. PLEASE GO ON TO THE NEXT PAGE! I! 43 Again, work quickly and just record your first impressions. 1 2 3 4 5 6 7 Strongly Disagree Moderately Undecided Moderately Agree Strongly Disagree Disagree Agree Agree 14) I often reveal more undesirable things about myself than desirable things. 15) Once I get started, I intimately and fully reveal myself in my self- disclosures 16) I often discuss my feelings about myself. 17) I frequently express my personal beliefs and opinions. 18) I talk about lots of different things with my friend. 19) I typically talk about one main subject with my friend. 20) My friend and I usually talk about a broad range of subjects. 21) I usually have only one topic that I discuss with my friend. 22) I usually disclose negative things about myself. 23) I will typically reveal information about myself that is about something bad happening in my life. 24) I will typically reveal information about myself that is about something good happening in my life. 25) I usually talk about happy things when I talk about myself with my fnend. 26) I usually talk about the darker side of my life with my friend. 27) On the whole, my disclosures about myself are more positive than negafive. PLEASE GO ON TO THE NEXT PAGE!!! 44 You have been answering questions about one of your male friends. Please answer the next question with that friend in mind: 28) Which of the following pictures best describes the closeness or bonding that you have with your friend? Fill in the appropriate number on your scantron sheet. OO 7 PLEASE GO ON TO THE NEXT PAGE!!! You have been answering questions about a specific person. With this person in mind, please respond to the following questions: The following is a list of different activities that people may engage in over the course of one week. If you have engaged in the listed activity with your friend in the past week, fill in the number “1” on your scantron. If you have not engaged in the listed activity, fill in the number “2” on your scantron. In the past week, I did the following activities with my friend: (1=TFIUE) (2 =FALSE) 29) Watched TV or a video 30) Ate a meal 31) Went to a grocery, department, book, hardware store, etc. 32) Went for a drive/walk 33) Threw a party 34) Attended class 35) Went on a trip (e.g., vacation, weekend, or day) 36) Cleaned house/apartment 37) Went to church/religious function 38) Worked on homework 39) Talked on the phone 40) Went to a movie/concert/play 41) Exercised / Participated in a sporting activity 42) Outdoor recreation (e.g., hunting, sailing, fishing) 43) Went to a bar 44) Visited family 45) Visited friends 46) Played cards/ video games 47) Attended a sporting event 48) Went to a party 49) Karaoke/ Played in a band PLEASE GO ON TO THE NEXT PAGE!!! 46 Instructions: Please mark the following statements to reflect the amount of influence your friend has on your thoughts, feelings, and behavior by marking whether you (7) strongly agree, (6) agree, (5) moderately agree, (4) are undecided, (3) moderately disagree, (2) disagree, or (1) strongly disagree. Mark Your Answers by Filling in the Bubble on the Scantron that Corresponds with your Response. Ignore the numbers 8, 9, & 10 on the Scantron. 1 2 3 4 5 6 7 Strongly Disagree Moderately Undecided Moderately Agree Strongly Disagree Disagree Agree Agree 50) This friend will influence my future financial security. 51) This friend does not influence everyday things in my life. 52) This friend influences important things in my life. 53) This friend influences which parties and other social events I attend. 54) This friend influences the extent to which I accept responsibilities in our friendship. 55) This friend does not influence how much time I spend doing household work. 56) This friend does not influence how I choose to spend my money. 57) This friend influences the way I feel about myself. 58) This friend does not influence my moods. 59) This friend influences the basic values that I hold. 60) This friend does not influence the opinions that l have of other important people in my life. 61) This friend does not influence when I see, and the amount of time I spend with, my family. 62) This friend influences when I see, and the amount of time I spend with, my other friends. 63) This friend does not influence which of my other friends I see. 64) This friend does not influence the type of career I have. 65) This friend influences or will influence how much time I devote to my career. PLEASE GO ON TO THE NEXT PAGE!!! 47 Again, Mark Your Answers by Filling in the Bubble on the Scantron that Corresponds with your Response. Ignore the numbers 8, 9, & 10 on the Scantron. 1 2 3 4 5 6 7 Strongly Disagree Moderately Undecided Moderately Agree Strongly Disagree Disagree Agree Agree 66) This friend does not influence my chances of getting a good job in the future. 67) This friend influences the way I feel about the future. 68) This friend does not have the capacity to influence how I act in various situations. 69) This friend influences and contributes to my overall happiness. 70) This friend does not influence my present financial security. 71) This friend influences how I spend my free time. 72) This friend influences when I see X and the amount of time the two of us spend together. 73) This friend does not influence how I dress. 74) This friend influences how my place looks (e.g., dorm room, apartment, house). 75) This friend does not influence where I live. 76) This friend influences what I watch on TV. PLEASE GO ON TO THE NEXT PAGE!!! 48 For the following questionnaire items, choose a number from the scale below that indicates the extent to which you agree with that statement. 1 2 3 4 5 6 7 8 9 Strongly Neither Agree Strongly Disagree Nor Disagree Agree Please answer the following questions by choosing a scale value that best represents your judgment. Mark your response on the scanner sheet by filling in the circle containing the number on which you decided next to the appropriate item number. Please ignore the #10 option on your scantron sheet. Be sure to mark you judgment ratings on your scanner sheet and NOT on this questionnaire. When you have finished these questions, please continue to the next page. ' 77) I feel like this person knows me very well. 78) This person has knowledge about me that other people likely do not know. 79) Based on what this person has Ieamed about me, they could probably predict my behavior in different situations. 80) This person is a very good friend of mine. 81) This person and I are very close. 82) I consider this person to be a close friend. 83) I like this person 84) I enjoy this person’s company 85) I think that this person is unusually well-adjusted. PLEASE GO ON TO THE NEXT PAGE!!! 49 1 2 3 4 5 6 7 8 9 Strongly Neither Agree Strongly Disagree Nor Disagree Agree 86) Most people would react very favorably to this person after a brief acquaintance. 87) I think that this person and I are quite similar to each other. 88) This person is one of the most likable people I know 89) This person is the sort of person whom I myself would like to be 90) It seems to me that it is very easy for this person to gain admiration 91) I think he is a friend of mine. 92) I like to have friendly chats with him. 93) We easily established a personal friendship with each other. 94) He fits into my circle of friends. 95) He is pleasant to be with. 96) I feel I know him personally. 97) I sometimes wish I were more like him. 98) He knows a lot about me. 99) He knows me personally. PLEASE GO ON TO THE NEXT PAGE!!! 50 The items below inquire about what kind of a person you think you are. Instructions: Please mark the following statements to reflect how much you think the following statements reflect YOUR personality. Mark Your Answers by Filling in the Bubble on the Scantron that Corresponds with your Response. Ignore the numbers 8, 9, & 10 on the Scantron. 1 2 3 4 5 6 7 Never or Always or Almost Never Almost Always True True 100) Has leadership qualities 101) Compassionate 102) Willing to take a stand 103) Eager to soothe hurt feelings 104) Independent 105) Friendly 106) Assertive 107) Gentle 108) Acts as a leader 109) Helpful 1 10) Aggressive 111) Sensitive to others’ needs 1 12) Competitive 113) Sincere 114) Dominant 115) Tender 116) Forceful 117) 118) Has Strong personality 119) Understanding Warm PLEASE GO ON TO THE NEXT PAGE!!! 51 For the following questionnaire items, choose a number from the scale below that indicates the extent to which you agree with that statement. 1 2 3 4 5 6 7 8 9 Strongly Neither Agree Strongly Disagree Nor Disagree Agree Please answer the following questions by choosing a scale value that best represents your judgment. Mark your response on the scanner sheet by filling in the circle containing the number on which you decided next to the appropriate item number. Please ignore the #10 option on your scantron sheet. Be sure to mark you judgment ratings on your scanner sheet and NOT on this questionnaire. When you have finished these questions, please continue to the next page. Again, mark your response on the scanner sheet and not on this questionnaire. Please ignore the # 10 option on the scanner sheet. 120) I would be very upset if someone else thought I was gay. 121) I am careful not to do things that might make others think I am homosexual. 122) If someone questioned my sexual orientation, it would not bother me. 123) If a homosexual person began talking to me in public, I would be concerned about what other people might think. 124) would be very ashamed if someone I know thought I was gay. PLEASE GO ON TO THE NEXT PAGE!!! 52 For the rest of the questionnaire, choose an answer and circle it on this sheet of paper. Please answer each question honestly. Your identity will in no way be associated with the answers you give on the survey questionnaire. Please feel free to ask the experimenter any questions. 125) What year in school are you? (Cirlce one) 1“ 2"" 3" 4’“ 5‘“+ 126) Where are you from? (Circle the number which corresponds to your answer) I 2 3 Michigan Another US State Another Country 127) What is your age? (Circle the number which corresponds to your answer) 18 I9 20 21 22 23 24 25+ 128) What is your sexual preference? 1 2 3 Heterosexual Homosexual Bisexual 129) What is your religious background? 1 2 3 4 5 6 Catholic Protestant Jewish Hindu/Muslim Eastern Tribal/Other 130) What is your race ? 1 2 3 4 5 6 White! Chicano/ Black/ Asian/ Native Middle European Latino African American Pacific Islander American Eastern YOU HAVE NOW COMPLETED THE FRIENDSHIP BEHAVIOR STUDY. PLEASE RETURN TO THE EXPERIMENTER’S DESK AT THIS TIME. 53 Appendix B: lnforrned Consent Form 54 Gender Role Orientation, Homophobic Self-Presentation, and Knowledge of Other: Predicting Self-Disclosive Behavior in Same-Sex Male Friendships - Informed Consent Welcome to the FRIENDSHIP BEHAVIOR study. During this session, you will be asked to answer open-ended questions about your friendships with other men. In addition, we have some multiple choice questions about those friends. The study also includes some anonymous demographic questions. Your identity will in no way be associated with the answers you give on the survey questionnaire. The questionnaires are anonymous, and will be kept on a password-protected computer hard drive in a locked office for two years before being destroyed, only to be viewed by the principle investigator and research assistants. If you participate in the study, your privacy will be protected to the maximum extent allowable by law. Full participation in this portion of the study will take 1 hour or less, and you will be given 1 hour of credit in your communication course. As described in your course syllabus, you may earn research credit by writing a short paper instead of participating in this study. Although participation in this study is not expected to produce discomfort or stress, please note that you may refuse to answer certain questions or withdraw from the experiment at any time without penalty. The experimentercan answer any questions you have about the study to help you choose whether to participate. Contact Jonathan Bowman (phone: 353-7252; office: 455 CAS) if you have any further questions or concerns regarding this study. Additionally, concerns about the rights of human participants in this study may be addressed to Peter Vasilenko, Chair of the University Committee on Research Involving Human Subjects, at (517) 355-2180. The office is located in 202 Olds Hall, and can be reached by e-mail at UCRIHS@msu.edu. Thank you, Jonathan Bowman If you have read the description of the research procedures involved in the FRIENDSHIP BEHAVIOR study and feel that the procedures have been explained to your satisfaction, please indicate your voluntary participation in the FRIENDSHIP BEHAVIOR study to receive course credit by completing the information below. Your Signature Today’s Date Your Printed Name 55 Figure 1: Proposed Conceptual Model Independent Communication Outcomes Variables Processes Gender Identity: \ Masculine - Closeness: Perceived + Closeness Gender Identity: \ Self-Disclosure Feminine -Amount *- Closeness: -Breadth T Interde ndence / -Valence pe Concern with _ Closeness: Appearing Strength of ‘ Homosexual Influence Closeness: Diversity of Perceptions of ACUVIIY Other’s Knowledge of Self — 56 Table 1 Descriptive Statistics for Main Factors (Predictor, Process, and Outcome Variables) Used in Correlations (n = 115) Standard Minimum Maximum Mean Deviation Hglsecghpeengzgg? 1.00 7.00 5.12 1.10 Ezgigcizrfigfiff 1 .00 7.00 5.09 1 .1 1 gfenggrfithaflgfe'f' 1 .00 8.20 4.89 1 .63 33323:; 1 .08 5.83 3.72 0.95 3:33:23, 2.00 7.00 5.38 1 .25 $35332: 3.27 6.27 4.83 0.64 33:37:55" 1 .00 9.00 6.16 2.10 Interdependence“ 1 .00 7.00 3.88 1 _92 ”WNW °I ACIIVM .00 20.00 5.37 5.24 Strength of Influence@ 29.00 140.00 85.58 24.42 ** Measured on a Likert-type scale from 1-9 * Measured on a Likert-type scale from 1-7 @Measured on a Likert-type scale from 1-7, with the scale responses summed # Measured with the total number of “true” responses summed 57 Table 2 Correlations Between Independent Variables and Communication Process Variables (Self-Disclosure) (n = 115) I. Masculine Role Orientation ll. Feminine Role Orientation III. Homophobic Self-Presentation IV. Perceived Other’s Knowledge V. Amount of Disclosure VI. Breadth of Disclosure Vll. Valence of Disclosure .632“r .026 .285" .210* .153 -.197* 1 -.O46 .270" .239” .072 -.078 III IV V VI VII 1 -.065 1 -.072 .346" 1 -.076 .449" .132 1 -.151 .032 .123 -.143 1 * p < .05; ** p < .01 (For descriptive statistics on each variable, see Table 1) 58 Table 3 Standardized Beta Coefficients for Multiple Regressions Evaluating the Effects of Gender Role Orientation on Self-Disclosure Variables (n = 115) Masculine Feminine Role Role Orientation Orientation 1. Amount of Disclosure 13 = .099 (3 = .176 2. Breadth of Disclosure B = .180 [3 = -.042 3. Valence of Disclosure B = -.246* B = .077 *p<.05; **p< .01; ***p< .001 59 Table 4 Correlations Between One ’3 Concern with Appearing Homosexual and Self- Reported Self-Disclosure Behavior for More and Less Knowledge Conditions (n = 115) More Knowledge Less Knowledge 1. Amount of Disclosure -.05 -.14 2. Breadth of Disclosure -.08 -.12 3. Valence of Disclosure -.08 +22 *p<.05;**p<.01 60 Table 5 Correlations Between Communication Process Variables (Self-Disclosure) and Outcome Variables (Closeness) (n = 115) I II III IV V VI VII l. Amount of 1 Disclosure ll. Breadth of Disclosure '183 1 III. Valence of Disclosure "093 "1 43 1 IV. Perceived .. u _ Closeness .277 .409 .058 1 V' '"terdependence .281 ** .413" -.059 .625" 1 VI. Diversity of .,, .. .. .. Activity .343 .167 .231 .261 .276 1 VII. Strength of n . up i... i. Influence .314 -.025 .260 .310 .283 .383 1 * p < .05; ** p < .01 (For descriptive statistics on each variable, see Table 1) 61 Table 6 Standardized Beta Coefficients for Multiple Regressions Evaluating the Effects of Self-Disclosure Behavior on Closeness Variables (n = 115) Amount Breadth Valence of of of Disclosure Disclosure Disclosure 1. Perceived B = .213“ B = .366*** B = -.296*** Closeness 2. Interdependence B = .216“ B = .370*** B = -.026 3. Diversity B = .296*** B = .145 B = .224* of Activity 4. Strength B = .302*** , = -.048 B = .225” of Influence *p<.05;**p<.01;***p<.001 62 REFERENCES Allen, J., & Haccoun, D. (1976). Sex differences in Emotionality: A multidimensional approach. Human Relations, 29, 711-722. Aron, A., Aron, E. N., Smollan, D. (1992). Inclusion of Other in the Self Scale and the structure of interpersonal closeness. Journal of Personality and Social Psychology, 63, 596-612. Aylor, B., & Dainton, M. (2004). Biological Sex and Psychological Gender as Predictors of Routine and Strategic Relational Maintenance. Sex Roles, 50 (9-10), 689-697. Bern, S. L. (1974). The measurement of psychological androgyny. Joumal of Consulting and Clinical Psychology, 42, 155-162. Berscheid, E., Snyder, M., & Omoto, A. M. (1989). The relationship closeness inventory: Assessing the closeness of interpersonal relationships. Interpersonal Relations and Group Processes, 57(5), 792-807. Blieszner, R., & Adams, R. (1992). Adult Friendships. Newbury Park, CA: Sage. Bowman, J. M. (2004). The Influence of Context Nonnativeness and Homophobic Beliefs on Perceived Appropriateness of Self-Disclosure in Male Heterosexual Friendships. Paper presented at the annual meeting of the lntemational Communication Association, New Orleans. Cohen, J., Cohen, R, West, S. G., Aiken, L. S. (2003). Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences. Mawah, NJ: Lawrence Erlbaum Associates. Davidson, J., & Duberrnan, L. (1982). Same-sex friendships: A gender comparison of dyads. Sex Roles, 8, 809-822. Dindia, K., & Allen, M. (1992). Sex Differences in Self-Disclosure: A Meta- Analysis. Psychological Bulletin, 112(1), 106-124. Duck & Pittman (1994). Social and Personal Relationships. In M. L. Knapp & Miller, G. R. (Eds.) Handbook of Interpersonal Communication (pp. 676- 695) Thousand Oaks, CA: Sage. Duran, R. L. & Kelly, L. (1985). An investigation into the cognitive domain of communicative competence. Communication Research Reports, 2, 112- 119. 63 Eagly, A. H., Wood, W., & Diekman, A. B. (2000). Social role theory of sex differences and similarities: A current appraisal. In T. Eckes & Trautner, H. M. (Eds) The Developmental Social Psychology of Gender. Mahway, N. J.: Lawrence Erlbaum Associates. Eisler, R. M. (1995). The relationship between masculine gender role stress and men’s health risk. In R. F. Levant & Pollack, W. S. (Eds.) A New Psychology of Men (pp. 207-225). New York: BasicBooks. Epstein, D. & Johnson, R. (1994). On the straight and narrow: the heterosexual presumption, homophobias, and schools. In D. Epstein, D. (Ed.). Challenging Lesbian and Gay Inequalities in Education. Buckingham: Open University Press. Epstein, D. & Johnson, R. (1998). Schooling Sexualities. Buckingham: Open University Press. Fehr, B. (2000). The life cycle of friendship. In C. Hendrick & Hendrick, S. S. (Eds.) Close Relationships: A Sourcebook. Thousand Oaks, CA: Sage. Fehr, B. (2004). A prototype model of intimacy interactions in same-sex friendships. In D. J. Mashek & Aron, A. (Eds). Handbook of Closeness and Intimacy. (pp. 9-26). Mahwah, NJ: Lawrence Erlbaum Associates, Inc. Floyd, K. (2000). Affectionate Same-Sex Touch: The Influence of Homophobia on Observers’ Perceptions. Journal of Social Psychology, 140(6), 774- 788. Gentry, C. S. (1986). Development of scales measuring social distance toward male and female homosexuals. Journal of Homosexuality, 13, (1 ), 75-82. Hunter, J.E. & Hamilton, MA. (1988) Confirmatory Factor Analysis. Lansing: Michigan State University. lckes, W. (1985). Sex role influences on compatibility in relationships. In W. Ickes (Ed.), Compatible and incompatible Relationships. New York: Springer-Verlag. Jones, T. S., & Brunner, C. C. (1984). The effects of self-disclosure and sex on perceptions of interpersonal communication competence. Women’s Studies in Communication, 7, 23-37. Jourard, S. M. & Lasakow, P. (1958). Some factors in self-disclosure. Journal of Abnormal and Social Psychology. 56, 91 -98. 64 Kelley, H. H., Berscheid, E., Christensen, A., Harvey, J. H., Huston, T. L., Levinger, G., McClintock, E., Peplau, L. A., & Peterson, D. R. (1983). Close Relationships. New York: Freeman Press. Lenney, E. (1991). Sex Roles: The Measurement of Masculinity, Femininity, and Androgyny. In J. P. Robinson, Shaver, P. R., & Wrightsman, L. S. (Eds.), Measures of Personality and Social Psychological Attitudes. San Diego: Harcourt Brace Jovanovich. Lewis, R. J., Winstead, B. A., & Derlega, V. J. (1989). Gender differences in volunteering for friendship research. Journal of Social Behavior & Personality, 4 (5), 623-632. MacDonald, A. P., Jr., & Games, R. G. (1974). Some characteristics of those who hold positive and negative attitudes toward homosexuals. Journal of Homosexuality, 1, 9-27. Markus, H. (1977). Self-schemata and processing information about the self. Journal of Personality and Social Psychology, 35 (2), 63-78. Markus, H., Crane, M., Bernstein, S., & Siladi, M. (1982). Self-schemas and gender. Journal of Personality and Social Psychology, 42 ( 1), 38-50. McCroskey, J. C., & McCain, TA. (1974). The measurement of interpersonal attraction. Speech Monographs, 41(3), 261 -266. McGill, M. E. (1985). The McGill Report on Male Intimacy. New York: Holt, Rinehart, and Winston. Miller, G. R., & Steinberg, M. (1975). Between people: A new analysis of interpersonal communication. Chicago: Science Research Associates. Millham, J., San Miguel, C. L., & Kellog, R. (1976). A factor-analytic conceptualization of attitudes toward male and female homosexuals. Journal of Homosexuality, 2 (1), 3-10. Morin, S. F., & Garfinkle, E. M. (1978). Male Homophobia. Journal of Social Issues, 34(1), 29-47. Parks, M. (1982). Ideology in interpersonal communication: Off the couch and into the world. In M. Burgoon (Ed.), Communication Yearbook 5 (pp.79- 107). New Brunswick, NJ: Transaction. Pearce, W. B., & Sharp, S. M. (1973). Self-disclosing communication. Journal of Communication, 23(4), 409-425. 65 Phua, V. C. (2002). Sex and sexuality in men’s personal advertisements. Men and Masculinities, 5(2), 178-191. Prager, K. J. (1995). The Psychology of Intimacy. New York: Guilford. Reeder, H. M. (2003). The effect of gender role orientation on same- and cross- sex friendship formation. Sex Roles, 49 (3-4), 143-152. Rich, A. (1980). Compulsory Heterosexuality and Lesbian Existence. Signs: Journal of Women in Culture and Society, 5, 631-660. Roth, Harris, & Snyder, 1988 Roth, D. L., Harris, R. N., & Snyder, C. R. (1988). An individual differences measure of attributive and repudiative tactics of favorable self- presentation. Journal of Social & Clinical Psychology, 6(2), 159-170. Rubin, G. (1975). The traffic in women: Notes on the ‘political economy’ of sex. In R. Reiter (Ed.) Toward an Anthropology of Women (pp.157-210). New York: Monthly Review Press. Rubin, L. (1983) Intimate Strangers. San Francisco: Harper & Row. Schaefer, M., & Olson, D. (1981). Assessing intimacy: The PAIR Inventory. Journal of Marital and Family Therapy, 7, 47-60. Stewart, L. P., Cooper, P. J., Stewart, A. D., & Friedley, S. A. (2003). Communication and Gender. Boston: Allyn & Bacon. Weinberg, G.(1973). Society and the healthy homosexual. New York: Anchor Books. Wheeless, V., & Dierks-Stewart, K. (1981). The psychometric properties of the Bem Sex Role Inventory: Questions concerning reliability and validity. Communication Quarterly, 28, 173-186. Wheeless, L. R. & Grotz, J. (1976). Conceptualization and measurement of reported self-disclosure. Human Communication Research, 2, 338-346. Winstead, B. A., Derlega, V. J., & Wong, P. T. (1984). Effects of sex-role orientation on behavioral self-disclosure. Journal of Research in Personality, 18(4), 541 -553. Wood, J. T. (2000). Gender and Personal Relationships. In C. Hendrick & Hendrick, S. S. (Eds.) Close Relationships: A Sourcebook. Thousand Oaks, CA: Sage. 66 Zhang, J., Norvilitis, J. M., & Jin, S. (2001). Measuring gender orientation with the Bern Sex Role Inventory in Chinese culture. Sex Roles, 44 (3-4), 237-251. 67 IIIIIIIIIllillilllilI