1 MULTIPLE MASCULINE SELVES, GENDER ROLE SELF CONCEPT, AND WELL BEING BY KIMBERLY DIANE JONES A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012
2 201 2 Kimberly Diane Jones
3 To Lance M y Ghost Writer
4 ACKNOWLEDGMENTS I would like to acknowledge my advisor, Dr Martin Heesacker, whose insistence on aski ng only those questions that are innovative and controversial has made my journey into psychological research an en joyable and rewarding one. I would also like to acknowledge my committee members, Dr. William Marsiglio, Dr. Greg Neimeyer, and Dr. James Sh epperd, whose time and wisdom have been greatly appreciated in the formation of this project Finally, I acknowledge Stephen Kadar, for his support and patience.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ........... 4 LIST OF TABLES ................................ ................................ ................................ ...................... 7 LIST OF FIGURES ................................ ................................ ................................ .................... 8 ABSTRACT ................................ ................................ ................................ ............................... 9 CHAP TER 1 INTRODUCTION ................................ ................................ ................................ ............. 11 Self Discrepancy Theory and Multiple Masculine Selves ................................ ................... 14 The Costs of Con flicting Gendered Selves ................................ ................................ ......... 15 Extending the Conflicting Standards Dilemma to the Study of Men ................................ ... 17 Ego Identity Development and the Gender Role Self Concept ................................ ............ 19 Reference Group Identity Dependence and Conflicting Standards ................................ ...... 22 Lack of Commitment: Fragmentation o r Specialization? ................................ .................... 23 2 PURPOSE AND SPECIFIC AIMS ................................ ................................ .................... 26 Specific Aim 1 ................................ ................................ ................................ ................... 26 Specific Aim 2 ................................ ................................ ................................ ................... 26 Specific Aim 3 ................................ ................................ ................................ ................... 26 3 METHODS ................................ ................................ ................................ ........................ 28 Participa nts and Procedure ................................ ................................ ................................ 28 Measures ................................ ................................ ................................ ............................ 29 Bem Sex Role Inventory (BSRI). ................................ ................................ ................... 29 Self Guide Discrepancy. ................................ ................................ ................................ 30 Positive Affect and Negative Affect Scale (PANAS). ................................ ..................... 31 Center for Epidemiologic Studies Depression Scale (CES D). ................................ ........ 31 The Satisfaction With Life Scale (SWLS). ................................ ................................ ..... 32 Rosenberg Self Esteem Scale (RSES). ................................ ................................ ........... 32 Self Concept Clarity Scale (SCC). ................................ ................................ .................. 33 Reference Group Identity Dependence Scale. ................................ ................................ 34 Demograp hics. ................................ ................................ ................................ ............... 35 4 RESULTS ................................ ................................ ................................ .......................... 36
6 5 DISCUSSION ................................ ................................ ................................ .................... 57 Review of Study Findings ................................ ................................ ................................ .. 57 Limitations of the Study ................................ ................................ ................................ ..... 64 Future Directions ................................ ................................ ................................ ................ 67 APPENDIX A S CALE ITEMS ................................ ................................ ................................ .................. 71 B SELF GUIDE DISCREPANCY PROMPT ................................ ................................ ........ 76 LIST OF REFERENCES ................................ ................................ ................................ .......... 77 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ..... 83
7 LIST OF TABLES Table page 4 1 Means, standard d eviations, a nd distribution statistics for study variables by g ender ...... 49 4 2 Means and standard deviations of average item level difference scores between c ont exts in Bem Sex Role Inventory r atings ................................ ................................ ... 50 4 3 Pearson correlation c oefficients for Well Being and Ge nder Discrepancy variables for m en ................................ ................................ ................................ .......................... 51 4 4 Pearson correlation c oefficients for Well Being and Ge nder Discrepancy variables for w omen ................................ ................................ ................................ ..................... 52 4 5 Hierarchical r egress ion analysis of the e ffects of Negative Affect, Satisfaction with Life, Self Esteem, and Depressive Symptomato logy on the No Reference Group s ubscale of the Reference Group Identity Dependence Scale ................................ .......... 53 4 6 Analyses of the moderation effects of Self Concept Clarity on the relationship between the No Reference Group s ubscale of the Reference Group Identity S tatus Scale and Self Esteem, Depressive Symptomatology, and Satisfaction with Life ........... 54
8 LIST OF FIGURES Figure page 4 1 Predicted graph of the r elati onship of No Reference Group s cores to Self Esteem, with Self Concept Clarity as a m oderator. ................................ ................................ ...... 55 4 2 Predicted graph of the r elationship of No Reference Group Scores to Satisfaction with Life, with Self Concept Clarity as a m oderator. ................................ ...................... 56
9 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requireme nts for the Doctorate of Philosophy MULTIPLE MASCULINE SELVES, GENDER ROLE SELF CONCEPT, AND WELL BEING By Kimberly Diane Jones December 20 12 Chair: Martin Heesacker Major : Psychology As times are changing for men situational and contextual messages and contingencies As such, d iffering social and spatial contexts and circumstances may now require different gendered repertoires or gendered selves from me n in ways that have received little attention in the empirical literature. This dissertation explores the idea through the lenses of self concept and ego identity as articulated by Higgins (1987) and by Marcia (1966). I t assess es whether a conflicting standards paradigm may be extended to our understa nding of men. Then it look s at how perceptions of discrepancy in messages about gender may relate to psychological well being, as well as how these discrepancies relate to the use of male reference groups I t then analyze s whether any relationships betwe en reference group identity dependence and well being are buffered by self concept clarity. I h ypothes ized that men would indeed report perceiving discrepancies in gendered trait expectations amongst self guides, and that these discrepancies would predict decrements to self esteem, satisfaction with life, and higher depressive symptomatology. Further, I predicted that perceptions of discrepancy would also predict feelings
10 of psychological connectedness to other men, and that lack of such connectedness wou ld also predict decrements to mental health outcomes. Finally, I hypothesized that this latter relationship would be moderated by levels of general self concept clarity. Findings indicate that men do in fact report significant discrepancies in social mes sages about both the preferred masculine and feminine traits they feel they are expected to possess. Further, self concept clarity is found to moderate the relationship between lack of a male reference group, self esteem and satisfaction with life. Some implications and limitations of these findings are discussed.
11 CHAPTER 1 INTRODUCTION How social messages about mas culinity an important area of inquiry for gender scholars as well as practicing psychologists. M ost of the research in the psychology of men thus far has focused on how traditional masculinity ideology, norms, and behaviors relate to negative psychological and physical health outcomes for men in American society However, recent shifts over the last few decades in gender relations at home and at work have changed the terrain in which men now experience themselves as gendered (e.g., Aarseth, 2007; Connell, 1995; Kimmel, 1987; Marsiglio, Amato, Day, & Lamb, 2000) Increasingly, traditional notions of masculinity no longer match economic and social realities, increasing the attractiveness and prevalence of more nontraditi onal enactments of masculinity During this time of transition from old to new guards, s ituational and contextual messages and conti bound to be multiple and inconsistent. As such, d iffering social and spatial contexts and circumstances may now require different gendered repert oires or gendered selves from me n in ways that have rece ived little attention in the empirical literature In this dissertation, I seek ed self concept s by determining whether men report diff ering expectations across various social settings in the gendered trait s they feel they a re expected to possess I explore whether the perception of such discre pancy relate s to the use of male reference groups concept Gender role self concept as it will be defined for this dissertati concept with regard to gender roles and includes gender related attributes, attitudes, and I investigate how perceptions of discrepancy may have a link to decrements in particularly in the form of depressive
12 symptomatology, satisfaction with life, and self esteem. I also look into how the use of reference concept may be rela ted to these aspects of well being Finally, I assess a possible mitigating fac tor in the relationship between a certain type of gender role self concept and well being outcomes. I begin by discussing some reasons that a study of gender ideology conflict, if we may so describe this notion of discrepancy, may be relevant to psychological study. Despite the essentialist and popular psychology arguments that would lead us to believe understand gender ideology and performa nce to be something that is socially constructed and contingent (e.g., Butler, 1988; Connell and Messerschmidt, 2005) Because these entities are socially constructed, they have been subject to redefinition and revision throughout time and in response to differ ing social contexts In most recent memory in the United States t he advancement of women during and after th e 1970s, and the movement of women into labor markets previously occupi ed almost exclusively by men has required, to varying degrees, a red efinition of what it means to be a man in American society (Kimmel, 1987). These societal changes have increased the incentives for men to engage in relational behaviors that violate the traditional masculinity (Levant, 1995) such as being mor e emotionally interdependent in relationships or sharing in the childcare Prior research on self perception theory ( Bem, 1967) and dissonance theory ( Festinger, 1957) suggest s th at as gendered behavior changes, they will also be more likely to re define their beliefs and attitudes regarding masculinity and feminini ty. This expectation of greater cooperation behavior may be even more prominent among you nger men who have only eve r known a world
13 large scale participation in the labor force (Singleton & Maher, 2004). Despite these new, more nontraditional mores that have come into effect and amplified over recent decades, there is little doubt that men are still subject to social pressure to conform to traditional gender repertoires as well (e.g., Korobov, 2005; Korobov & Thorne, 2006) s of being male (Goldberg, 1976) or more straightforwardly, the hazards of being a trad itional male, has predominated the psychology of men. Thus, in many ways, most of this research has further ed our knowledge about men only insofar as they conform to traditional hegemonic masculinity. Foc using only on traditional masculinity and its cor relat ions with diminished well being we may imperil research on gender more broadly, in that diversity and complexity amongst gendered construct ions are largely ignored ( Addis Mansfield, & Syzdek 2010; Ashmore, 1990; Hoffman, 2001 ; Lewin, 1984 ; 008b ) One possible source of complexity in gender self definitions at this point may simply be the presence of the multiplicity itself: if one perceives and feels press to choose from many different options of how to define oneself as a gendered self wh ich definition does one choose and will that definition be equally suited to all contexts ? As a burgeoning area of scholarly and empirical discourse, there are no doubt multiple ways to explore th e presence and sequelae of harboring conflicting message s about masculinity This dissertation will explore the idea through the lens es of self concept and ego identity as articulated by Higgins (1987) and by Marcia (1966) I then seek to combine these more general concepts of ego development with gender rol e self concept more specifically, and conjecture at how the former may inform the study of the latter. This discussion then leads to some overall well being.
14 Self Discrepancy Theory and Multiple Masculine Selves One clue to understanding how individuals may be negotiating the multiplicity of options around gender may be seen in self discrepancy theory which addresses the self concept more generally Higgins (1987 ) advanced self discrepancy theory to talk about how inconsistent or incompatible beliefs about oneself may lead to psychological discomfort, predicting that other negative emotional outcomes may also result when we experience these forms of internal disso nance. Higgins proposed the locus of this dissonance to be between self states or selves. The actual self ideal self is a representation of what self or other would like the pers on to ideally have, and typically includes such things as goals and hopes. Finally, the ought self represents what self or other believe a person ought to possess as attribute s and this is typically in the form of rules, injunctions, and obligations. Th e ideal and ought selves may be further distinguished by the sources from which they are most immediately maintained, with ought selves most often reflecting social pressures and mores, wh ereas ideal selves generally refer more to personal injunctions. Th ese domains and standpoints combine to form six self state representations: actual/own, actual/other, ideal/own, ideal/other, ought/own, and ought/other. The latter four are what Higgins considered standards or self guides (Higgins, Strauman, & Klein, 198 6) These self guides often contain shoulds, musts, and ideals of conduct and ability concept, or what a person perceives to be th eir actual conduct and level of abilit y at various tasks. Releva nt ideal/other, ought/other, and actual/other self guides may be a father, romantic partner, society in general, among others.
15 According to self self concept and other representations would lead to characteristic patterns of negative affect, including guilt, shame, dejection, agitation, and/or embarrassment (Higgins, 1987). For ex ample, if one possessed an ideal /other self guide that valued muscularity as an attribute but /own representation of self did not include this attribute, Higgins theorized that the resulting feeling may be shame or humiliation. Wh ereas outcomes has only received partial support (e.g., Tangey, Niedenthal, Covert & Hill Barlow, 1998), t he relationship between self discrepancies and negative outcomes more generally has a long standing basis in the theoretical and empirical literature (Bandura & Walters, 1963; Freud, 1917/1957; Rogers, 1961). In fact, h igher degre es of disparity in self representations have been associated with experiences of shame and guilt (Higgins, 1987 ; Tangn ey et al. 1998 ), negative affect, agitation, confusion, and dejection (Higgins, Klein, & Strauman, 1985; Stroink, 2004), and neuroticism (Pavot, Fujita, & Diener, 1997). The Costs of Conflicting Gendered Selves While the research on self discre pancies is clear er with respect to consequences when we look at the general self concept i t is unknown whether multiple and conflicting gender ed selves may be associated with psychological discomfort in some like manner As with many new areas of inquiry, qualitative studies of discrete groups of men experiencing such conflicts may give the first clues as to how to study the topic of gender messag e discrepancy by quantitative means. A few qualitative studies that will be touched upon here have helped elucidate some of the potential issues of concern One example may be seen in the therapeutic context for men an environment that traditionally ha s been more utilized by and associated with female consumers. Van Wagoner (2007) led him to conclude that his clients experienced polar ized
16 reactions to the demands of this group work in which open sharing of personal events and e motional expression are encouraged as these activities break with traditional expectations of masculine behavior He noted that the nontraditional cont ext of the groups seemed to lead some of the more traditional men to feel uncomfortable and to behave m ore passive ly in their exchanges. In contrast, those he identified as sensitive men evidenced more efforts to become increasingly other centered a nd reciprocal in their relationships. Van Wagoner noted that the latter group seemed to experience more inne r confl ict as they attempted to increase their comfort with expressiveness and interdependence in these relationships : [S]ensitive man can often verbalize a desire for emotional intimacy, and can verbalize feelings, yet the depth of feeling is somehow lack ing when one gets to know them. There is often a deep hunger for connection, and yet an inability to connect in an emot ionally intimate manner. (p. 49) Importantly however, Van Wagoner noted that the sensitive man does not represent a complete departure f rom the traditional man: Sensitive man is not the opposite of traditional man on such dimensions as competition and success, emotional expression and homophobia, but he is a man who has a budding understanding of the pitfalls of such gender role traps and is willing, at least on an intellectual level, to break free of traditional masculine gender roles. (p. 45) So while these sensitive men may have chosen to adopt more nontraditional actual ideal and ought gendered selves in response to changing times th ey appear to remain very much aware of and beholden to more traditional actual, ideal and ought gendered selves. According to Mahalik, Cournoyer, DeFranc, C herry & Napolitano (1998) more nontraditional men of this kind may struggle with feeling vulne rability and shame associated with expressing their emotional needs In seeking to avoid the use of more traditionally masculine defense mechanisms such as denial and projection, Mahalik et al. posited that more nontraditional men may struggle even more t o adapt when distressed Further s ome resear ch has
17 shown that men who experience confusion about what it means to be characteristically male are also more likely to experience social anxiety, low self esteem, anxiety, depression, diffused ego identity, a nd a lack of feeling of social connectedness (W ade & Brittan Powell, 2000 ; Wade & Gelso, 1998). However, there is also some suggestion that the link between confusion about masculinity and negative outcomes is mediated by nontraditionality in gender ideol ogy In other words, if one espouses more nontraditional beliefs about men, fewer of these negative psychological outcomes seem to result. P resumably this link exists because these men though unsure of what it means to be a man, are probably less likely to accept traditional masculine ris k taking, and avoiding help seeking for medical and psychological services ( Wade, 199 8). These studies provide some suggest ion that the gender role self concept may be related to broader well being. Extending the Conflicting Standards Dilemma to the Study of Men Though researchers have only recentl y begun to understand how men may navigate conflicting or confusing message s about masculinity, the notion is not without precedent in gender research. Stroink (2004) investigated the manifestation, and some sequelae, of these conflicting standards in a sample of women in her proposal of the Conflicting Standards Dilemma. Althou experiences. The Conflicting Standards Dilemma a 7) self discrepancy th eory, the observation repeated by gender scholars (e.g., Falk, 1998, Heriot, 1983) that multiple ways of constructing and enacting masculine and feminine behavior are salient to members of
18 contemporary Western culture. In response to this multiplicity, she argue d that individuals of either sex may lack a singular consistent basis upon w hich to define and understand their identit ies as either m e n or wom e n. Further, Stroink argue d may be a bewildering array of often mutually exclusive expectations and potential characteri stics Consistent with the basic notion advanced by Higgins (1987), s he posited that these messages are propagated by guides ehaviors and attributes th at are expected of them by specific entities These guides can be either specific or generalized. Specific guides refer to representations of specific others, such as a mother or partner. Generalized guides in contrast, refer to particular social groups and our conceptions of what these groups think are gender 1987; 1989) discrepancies in femi ninity self guides were predictive of negative psychological outcomes for women such as agitation, dejection, and confusion P resumably there may be a number of ways to process and respond to these discrepancies guides Therefore, Stroink su relation to these conflicting social messages about gender may elucidate an important aspect of ces in a post liberation age. For example, some women may align them self with a particular guide or consistent set of guides, while ignoring the expectation s of the rest Stroink also posited that some women may choose to adapt their behavior to the perceived expectat ions of each social situation or guide as it was salient resulting in inconsistent gendered behavior across contexts
19 (e.g., deferring in arguments with a romantic partner, while being competitive in a work environment). They may also simply fail to follo w the expectations of any specific guide, and amongst all the possibilities for action (p. 287) Despite these various possible coping styles, Stroink had reason to suspect that the perception of discrepancy amongst gui des would be associated with negative outcomes regardless of the coping style employed This prediction was based in part on self discrepancy t heory which posits that individuals experience a double approach avoidance conflict when they are motivated to live up to mutually contradictory ideals (Higgins, 1989 ; Van Hook & Higgins, 1988) Ego Identity Development and the Gender Role Self Concept We may be helped to understand how people resolve crises of gender rol e self concept by better understanding ho w people resolve identity crises more generally As Erikson (1968) proposed, people in late adolescence face the central task of defining a stable self concept amidst the multiplicity of possible identities available to them. the u se of coping styles to mitigate gender identity confusion is in many ways (1966) theories of identity development. Therefore, Marcia theory is briefly discussed here and then applied to the notion of the gender role self concept. Much like Stroink later theorized with her coping styles, Marcia postulated four ego identity statuses to refer to th e ways in whi ch people resolve identity crise s According to Marcia, the resolution of this stage may occur though ego diffusi on foreclosure moratorium or achievement These ego formations reflect varying levels of self exploration and commitment to a particular identity Self exploration refers to the extent to which someone has explored different ideological beliefs, prefe rences, and interpersonal roles. Commitment refers to the extent to which one has committed to a particular set of ideas and practices F oreclosure refers to a lack of exploration and premature commitment to an identity such as when a young man
20 adopts t he perspectives and practices of his father, without engaging in more serious exploration of alternatives. D iffusion refers to a lack of both self exploration and commitment and refers to those who are generally not concerned about their lack of directio n or identity incoherence. A chievement refers to the presence of both self exploration and commitment to an identity, and is thought to be the most autonomous and healthy way of resolving this stage in ego development. Finally, m oratorium refers to the p resence of self exploration in the absence of commitment someone who is amidst an exploration of their identity, but who has not yet settled on any particular one But how might these approaches to ident ity resolution inform the study of gender self conc ept resolution ? In an effort to e xtend the notion of ego formation to our understanding of Wade (1998) posited a theory of gender role self concep t development. Just as identity development occurs within a soci al and comparative context, Wade proposed that men also define their masculinity with reference to their male peers. They hypothesized that the way a man conceives of and relates himself to ot her men carries implications for how his own gender role self concept develops (W ade 1998). In this way, the Thereby, the diversity of masculinities we observe amongst men arises in part because men experience different characteristic types of psychologi cal relatedness to other males which is in turn related to ego identity development and to the development of masculinity ideologies in the form of the gender role self concept Th ese gender role self concepts a related att itudes and the quality of their experiences as gendered selves. Wade and Gelso (1998) found support for four characteristic types of psychological relatedness to other males in their validation of the Reference Group Identity Dependence Scale.
21 These types comprised the subscales labeled No Reference G roup Reference Group D ependent Reference Group N ondependent S imilarity and Reference Group N ondependent D iversity The same individual may score at varying levels on multiple of the subscale dimensions for this measure, and thus an individual will not always fit into a singular categorical type. Those who endorse content on the No Reference Group subscale indicate feelings of disconnectedness from other men and a lack of psychological relatedness to th em These men feel there are no other men like them and often possess gender role self concepts that are undefined and/or fragmented. The Reference Group Dependent sub scale taps into feelings of connectedness with a certain group of men with whom one ide ntifies, but not other groups of men These gender role self concepts are characterized by being externally defined, conformist, and often rigid. Reference Group Nondependency is defined by feelings of connectedness to all men, and is divided into two ty pes -Similarity and Diversity. The Similarity subscale assesses feelings of similarity with Corresponding gender self concepts are thought to be more internally d efined and flexible (Wade, 2008) As hypothesized the four reference group types have been shown to be related to eg o identity statuses. Wade and Gelso (1998) found that men who lacked a significant reference group in defining their male identit y ( No Reference G roup ) were also more likely to have a diffu se ego identity, to experience higher levels of depressive and anxiety symptoms, and to have lower self esteem. uncertaint concept as well. Reference G roup N ondependency was associated with achieved ego identity and unrelated to psychological outcomes indicat ing that feelings of relatedness to all men are most associated with a healthier type of ego development
22 (Kroger, 1993) The picture was slightly more complex for Reference Group D ependency, which was associated with both foreclosed and achieved ego identity This seems to indicate that defined reference gro up may occur with or without significant self exploration or consideration of alternatives Reference Group Identity Dependence and Conflicting Standards An as yet unexplored idea is whether reference group identification may relate in some way to perc eptions of conflicting standards around gender. The fact that self exploration has been correlated with reference group identification may provide some support for the idea (Wade and Gelso, 1998). If in fact there is some relation between perceptions of discrepancy about masculinity (i.e., conflicting standards) and feelings of psychological relatedness to other men, we may predict that the reference group identity statuses that are related to a lack of self exploration in ego identity ( i.e., No Reference Group or Reference Group D ependent) could be related to lower levels of conflicting standards in the social environment. This may occur because environments characterized by less conflicting standards may reduce opportunity or incentive for self explorat ion. Likewise we might expect that high levels of perceived discrepancy in messages about gender would be related to reference group statuses that acknowledge and embrace diverse groups of men (i.e., Reference Group Nondependent Similarity and Diversit y). This possibility will be explored in this study. reference group dependency theory Marcia (1966), (2004) conflicting standards dilemma all suggest that there is a significant relationship between social milieu has significant bearing on how the gender self concept develops as well, it makes sense to understand both the content and the level of heterogeneity within that mili eu In other
23 these influences. T he messages men hear about prefe rred masculine attributes may differ across contexts and create environments that may be confusing or difficult to navigate without choosing some strategy by which to sort out and make sense of these messages. Stroink (2004) argue d that the two most adapti ve ways of responding to the existence of conflicting standards are (a (b ) alignment, or relying on only one of many possible guides to det ndered conduct. These two coping styles presumably represent a more adaptive stance in that they manage multiplicity by in some way negat ing it, through commitment to a singu lar way of enacting gender. A central assumption of this ego identity based approach is then made visible that any form of compartmentalization or problematic incoherence in the individual. Similarly Wade & Gelso model also predicts that a more diffuse or undifferentiated identity would be associated with negative psychological outcomes. development study, which showed significant correlations betwee n No Reference Group scores and lower psychological well being. The findings and theory therefore suggest that, if men indeed perceive discrepant standards, that the best way to manage it is to commit to an identity that is intransient across contexts. Lack of Commitment: Fragmentation or Specialization? However, there is also some suggestion that a more uncommitted or diffuse identity may not always be related to negative psychological outcomes. A significant amount of literature exist s investigating a somewhat related notion: self differentiation
24 Neimeyer, & Brown, 2001, p. 396) Associations between levels of self differentia tion and psychological outcomes have been debated in both the theoretical and empirical literature, with mixed results (Bigler et al. 2001, Donahue, Robins, Roberts, & John, 1993; Gergen, 1991). On one side of the argument, self differentiation is thought to indicate a flexible and highly adaptive specialization that is responsive to what parts of the self may be most useful or expedient within particular contexts (Gergen, 1971; 1972 ; Goffman, 1959 ). For instance, the fa ther who is accommodating and nurtur ing when caring for a child, but uncompromising and assertive when negotiating the purchase price for a car is not necessarily evincing some lack of internal coherence, but rather readjusting to new contexts. Another side of the argument, in contrast, con tends that this inconsistency reflects not specializa tion, but self fragmentation ( Donahue et al., 1993 ). In support of this notion, Donahue and colleagues found that high self concept differentiation pred icted poor emotional adjustment: higher levels of depression and neuroticism and lo wer self esteem. However, Bigler and colleagues (2001) found that when levels of self concept clarity we re accounted for ( Campbell, Trapnell, Heine, Katz, Lavallee, & Lehman 1996 ), the negative psychological outcomes ass ociated with higher levels of self differentiation no longer obtain ed Self concept clarity beliefs are clearly, confidently defined and internally consistent, has been associated wi th lower neuroticism, higher self este em, low er negative affect, and low er depressive and anxiety symptom endorsement (Campbell et al., 1996). Their finding led Bigler and colleagues to conclude that self differentiation may not necessarily be associated with negative psychological outcomes w hen high differentiation occurs also within the context of a high degree of core level self concept stability.
25 To extrapolate from these results, the diminished well being that has been associated with the No Reference Group ascription may then be buffer ed by the degree to which one holds a clear self concept. In other words, it is possible that the No Reference Group type is comprised of two unique sets of individuals: those who lack self concept clarity and experience negative psychological outcomes, a nd those who have higher self concept clarity and do not experience as many negative outcomes. In other words, a non committ ed gender identity ma y not in and of itself be a less healthy way of adapting to situations characterized by conflicting gender rel ated demands. social roles may possess greater responsiveness t o environmental cues or greater resourcefulness in response to changing demands ( Gergen, 1971; 1972 ; Goffman, 1 959) thereby circumventing some of the negative psychological outcomes associated with lacking a male reference group In sum, evidence exists to suggest that the experience of conflicting social standards, gendered or otherwise, may be associated with negative psychological outcomes. Therefore, to explore this idea further t his dissertation assesses whether a conflicting standards paradigm may be extended to our understanding of men. It then look s at how perceptions of discrepancy in messages about gender may relate to psychological well being as well as how th ese discrepancies relate to the use of male reference groups It then analyze s w hether any relationships between reference group identity dependence and well being are buffer ed by self concep t clarity.
26 CHAPTER 2 PURPOSE AND SPECIFIC AIMS The present study seeks to investigate whether men report experiencing discrepancies in social messages about p referred gender related attributes and whether this perception is related to well being Furth er, this study seeks to elucidate the relationship between perception s of discrepancy in expected gendered attributes and a measure of male reference group dependence Finall y, various gender identity statuses will be analyzed for their relationship to ps ychological outco mes, and the potent ial buffering effect of self concept clarity on these relationship s will be assessed. Specific Aim 1 icting social standards regarding preferred masculinity, and evaluate how t hese perceptions may relate to psychological outcomes. Hypothesis 1a: Men will report significant discrepancies among self guides /contexts regarding preferred masculine traits. Hypothesis 1b: Higher rates of perceived discrepancy in masculine standards will be assoc iated with ( a) higher depressive symptomatology ( b) lower satisfaction with life, and ( c) lower self esteem. Specific Aim 2 To discern the relationship between perceptions of discrepanc y between self guides /contexts and reference group ident ity dependence Hypothesis 2 : ender identity is based in feelings of connection to all men will relate to perceptions of discrepancy such that greater feelings of connection will relate to greater perceptions of discrepancy Specific Aim 3 To explore the rel ationship between reference group identity dependence and psychological outcomes.
27 Hypothesis 3a : The extent to which me n lack a reference group in defining their gender identity will significantly predict negative psy chological outcomes, such that less identification will result in a) higher depressive symptomatology, b) lower satisfaction with life, and c) lower self esteem Hypothesis 3b : Self concept clarity will moderate the relationship between lack of reference group and negative psychological outcomes, such that those men possessing high self concept clarity will show a weaker association to negative psychological outcomes.
28 CHAPTER 3 METHODS Participants and Procedure 118 m ale and 311 female p arti cipants were recruited from undergraduate upper level psychology courses at a large southea stern public university and were compensated for their participation with a small amount of extra credit toward their course grade. Random responding was detected by embedding specific demand questions within each scale of more than 10 questions to this question female) were dropped from subsequent analyses for answering one or more these questions incorre ctly. Five participants were dropped from analysis due to missing values for entire scales. No other participants had mi ssing values, leaving 103 males and 280 females for subsequent analyses. Of the male participants, 63.1% (65) classified themselves a Also in the male sample, 47.6% (49) class ified themself as being in their first year of college, 19.4% (20) as second year, 18.4% (19) as third year, 13.6% (14) as fourth year, and 1% (1) as fifth year or more. Participants were given a web address for an online survey that they were able to com plete on their own ti me. They were asked to consent electronically to participate in the study after reading an informed consent description. Participants were then shown a prompt (see Appendix B ) that asked them to identify three relevant individuals, g roups, or contexts in their lif e that are most relevant to them in defining their gender. Suggestions for self guides in this others, although participants were also instructed that they may volunteer their own relevant
29 persons or contexts. After supplying these three guides, participants were then asked to complete three versions of a Bem Sex Role Inventory short form (BSRI; Bem, 1974 ) for each of their persona lly relevant self Please rate how much you feel [self guide inserted] believes you should possess each of the following attributes. Participants were then asked to complete the Positive Affect and Negative Affect Scale for affect at the time of participation (PANAS ; Watson, Clark, & Tellegen, 1988 ), Center for Epidemiological Studies Depression Scale (CES D ; Radloff, 1977 ), Satisfaction with Life Scale (SWLS; Diener, Emmons, Larsen, & Griffin, 1985 ) Rosenberg Self Estee m Scale (Rosenberg, 1965), Reference Group Identity Dependence Scale (RGIDS; Wade & Gelso, 1998), and Self Concept Clarity Scale (SCC; Campbell, Trapnell, Heine, Katz, Lavallee, & Lehman, 1996) All participants first by the PANAS, and then the remaining assessments were counterbalanced through computer randomization across participants Only men completed the RGIDS, as this assessment is specifically pointed at the experiences of men. Measure s Bem Sex Role Inventory ( BSRI) The BSRI (Bem, 1974) is a popular self report measure of sex role orientation. The short form of the BSRI contains 30 items measuring traditionally masculine and feminine traits The Femininity scale consists of 10 traits traditionally viewed as mor e desirable for a woman. The Masculinity scale consists of 10 traits traditionally viewed as more desirable for a man than for a woman. In the standard administration participants rate, on a 7 point scale from 1 = never or almost never true to 7 = always o r almost always true t he degree to which they feel each trait describes them as a person Sample items from the Masculinity scale include independent, competitive, and aggressive; sample items from the Femininity scale include compassionate, sym pathetic, and sensitive to the needs of others The remaining 10 items
30 were originally developed to measure social desirability, but are now generally used only as fil ler items. I nternal consistency reliabilities for the scales were reported by Bem (198 1) to range between .84 and .87 Following the methodology devised by Stroink (2004) to measure discrepancies between self guides in expected gen dered qualities, this study asked participants to complete three separate BSRIs wherein the views of various self guides are elicited from participants. The rating of multiple adjective lists (such as the BSRI) to determine self discrepancies has been utilized and validated by the work of Tangney, Niedenthal, Covert & Barlow (1998), who found this method to be e ssentially equivalent to qualitative methods of assessing these discrepancies. Following Stroink, directions for t he BSRI were alter ed ate how much you feel [ self guide inserted] believes you should possess e ach of the following attribute s. Scoring was done by summing ratings for the traditionally masculine and traditionally feminine adjectives separately in order to two create continuous variables for each guide Self G uide D iscrepancy Following the logic of Stroink (2004), BSRI Masc ulinity scale scores were summed separately for each of the three contexts, yielding a total of three BSRI Masculinity ratings. These values were divided by the total number of items in the scale (ten) to yield an average item rating for masculine traits. The absolute difference among the three possible parings of these item averages were then calculated (i.e., Context A C difference, A B difference, B C difference). These three scores were then averaged on the participant level to gi ve an estimate of av erage masculinity discrepancy (AMD) for each participant The same procedure was used to calculate femininity discrepancy scores yielding a total average femininity discrepancy score (AFD) for each participant. Higher AMD and AFD scores indicate a h igher degree of average difference between self guides expectations for preferred gendered behavior.
31 Positive Affect and Negative Affect S cale (PANAS) The PANAS (Watson, Clark, & Tellegen, 1988) is a 20 item self report measure of positive and negative af fect It consists of 10 adjectives for the negative affect ( NA ) dimension and 10 adjectives for the positive affect ( PA ) dimension. Participants were instructed to rate how descriptive with response alternative s ranging from 1 = very slightly or not at all to 5 = extremely on a 5 point scale. Ratings will be summed to provide separate total PA and NA scores. PANAS scale validity has been supported through correlations with general aspe cts of psychopathology (Huebner & De w, 1995), as well as other expressions of affect (Watson and Clark, 1984). This scale had an 8 week test retest alpha reliability of .85 for negative affect (NA) and .89 for positive affect (PA) in a sample of 660 psych ology undergraduate s from a southwestern university. This measure was in cluded in order to control for the possible effects of negative affect covariance between the various measures. for the negative affect subscale (NA) in this stud y was .84 Center for Epidemiologic Studies Depression Scale (CES D) The CES D (Radloff, 1977) is a 20 item scale that measures level s of depressive symptoms Each item is scored on a 4 point scale ranging from 0 = never to rarely (less than 1 day ) to 3 = most or all of the time (5 7 days ) previous week. The items on this scale form four subscales assessing: (1) depressed mood; (2) psychomotor retardation; (3) lack of well being; and (4) interpersonal diffi culties. Scores for items 4, 8, 12, and 16 are reversed before summing all items to yield a total score. Total scores can range from 0 to 60. Higher scores (both item and total scores) indicate more depressive symptoms. A score of 16 or higher has been us ed extensively as the cut off point for high depressive symptoms. Gatz and Hurwicz (1990) confirmed the original factor structure of the
32 CES D, and reported subscale reliabilities of 0.85, 0.75, 0.78 and 0.57 for depressed mood, psychomotor retardation, l ack of well being and interpersonal diffi culties, respectively. The CES D has been shown to be a reliable measure for assessing the number, types, and duration of depressive symptoms across racial, gender, and age categories (Knight, Williams, McGee & Olam an, 1997; Radloff, 1977; Roberts, Vernon, & Rhoades, 1989). High internal consistency ha s been reported with alpha coefficients ranging from .85 to .90 across studies for the full scale (Radloff, 1977). Concurrent validity by clinical and self report crit eria, as well as substantial evidence of construct validity 90 have been demonstrated ( Weissman, Prusoff, & Newberry, 1975 ). The for the overall sample in this study was .90 The Satisfac tion With Life Scale (SWLS) The SWLS (Diener, Emmons, Larsen, & Griffin, 1985) is a 5 item self report scale used to assess life satisfaction and as an indicator of psychological well being. It has been utilized with various ethnic minority groups and wit h various nationalities (Diener, Oishi, & Lucas, 2003). Participants will be asked to respond to items on a 7 point scale ranging from 1 = strongly disagree to 7 = strongly agree with higher scores indicating higher levels of life satisfaction. Construct validity was originally demonstrated with various other measures of well being, including the Affect Balance Scale and the well being subscale of the Differential Personality Questionnaire, as well as non significant correlation with the Marlowe Crowne so cial desirability scale. The two month test retest coefficient alpha was .87. The for the overall sample in this study was .89 Rosenberg Self Esteem Scale (RSES) The RSES is a 10 item self report scale devised by Rosenberg (1965) and is one of the most widely used measures of measure global self esteem. Participants respond to each item on
33 a 4 point scale from 1 = strongly disagree to 4 = strongly agree Half of the items are positively worded, while the other half is negatively word ed. Positive ly worded items in the measure negative ly worded Items written in a negative direction are reve rse coded and summed with positively worded items, yielding a score ranging from 10 to 40, with higher scores indicating higher global self esteem. The RS E S is the most c ommonly used measure of global self esteem and there is substantial evidence for its r eliability an d validity (e.g., Blascovich & Tomaka, 1991). The alpha for the overall sample in this study was .90 Se lf Concept Clarity S cale (SCC) The SCC scale is a 12 item self report scale developed by Campbell, Trapnell, Heine, Katz, Lav allee, & Lehman (1996) to measure self concept (e.g., perceived personal attributes) are clearly and confidently defined, internally consistent, and temporally point Likert between 12 and 60, w ith higher scores indicating high er self concept clarity. This scale had a 4 month test retest reliability of .79 and an average alpha coefficient of .86 in a sample of 155 Canadian undergraduate students. Construct validity of the SCC was supported by cor relations with measures of self concept clarity (Campbell et al., 1996). for the overa ll sample in this study was .90
34 Reference Group Identity Dependence Scale The RGIDS (Wade & Gelso, 1998) is a 30 male is dependent on a male reference group for his gender role self Pa rticipants rate items on a 6 point scale ranging from 1 = strongly disagree to 6 = strongly agree Item 24 only is reverse scored. The RGIDS postulates that three levels of ego id entity are associated with four levels of psychological relatedness to othe r males: No Reference Group (corresponding to a diffuse/undifferentiated ego), Reference Group Dependent (corresponding to a foreclosed/conformist ego), or Reference Group Nondependent which is subdivided into Similarity and Diversity subtypes (correspond ing to an achieved/integrated ego). Example items no reference group), I only feel relating to differen The No Reference Group other men. and feelings of connectedness with some men perceived as similar to oneself but not other men who are perceived as dissimilar. The Reference Group Nondependent status is divided into two types, and is characterized by psychological relatedness and feelings of connectedness with all m en The two types of Reference Group Nondependent are Similarity and Diversity. The Similarity sub scale assesses feel ings of similarity with all men, while t he Diversity sub appreciation of differences among men. Higher scores on t he scales indicate more endorsement of The subscales of the RGIDS were arrived at via factor analysis, and have been shown to have moderate internal consistency ranging from .70 to .78 (Wade & Gelso, 1998). There is also some evidence for the construct validity of this measure through correlations with subscales of the Gender Role Conflict Scale and measures of ego identity
35 development. The alpha coefficients for this study by subscale were as follows: No Reference .81 ); Reference Group Dependent .63 ); Reference Group Nondependent Total = .78 ). The subscales of the Reference Group Nondependent scale were Similarity ( .83 ) and .65 ). Demographics Information about gender identity was collected at the beginning of the survey, while information about race/ethnicity, and year in college were collected at the end.
36 CHAPTER 4 RESULTS The main objective of this study was to examine the presence of conflicting social messages abo ut preferred masculinity, and how these conflicts may relate to various measures of psychological health and relatedness to other males results are presented below For analysis of the initial hypotheses, two variables were created from the three BSRI context ratings that gave an estimate of a verage masculinity discrepancy and average femininity discrepancy. Discrepancy s cores were calculated on the participant level by computing the absolute difference in to tal BSRI masculinity scale scores for all possible pairings of ma sculinity self guides (e.g., average Context A masculinity item response minus average Context C masculinity item response ), yielding three masculinity di fference scores. The se three masculi nity guide to guide difference scores were then averaged to yield a total average masculinity discrepancy score (AMD) for each participant. The same procedure was used to calculate the three femininity difference scores and an averaged femininity discrepa ncy sco re (AFD) for each participant. Preliminary analyses of all dependent variables were conducted to determine the nature of distributions and any outliers present Skewness and kurtosis estimates of all dependent variables showed no significant devia nce from normality, with all values less than two ( Table 4.1) Outlie rs were identified by creating Z sc ores of all dependent variables and locating values equal to 3.29 or more standard deviations from the mean ( p. 77, Field, 2005). By this method, CES D (Depressive S ymptomatology) contained 1 outli er RSES (Self E steem) contained 1 outlier, PANn (Negative Affect ) contained 3 outliers, RGN (Reference Group Nondependent Total subscale) contained 1 outlier, and AMD (Average M ascul ine D iscrepancy) contain ed 1 outlier
37 The strategy used to correct for these outliers was to replace these cases with a Z s core equal to +/ 3.29 ( p. 79, Field, 2005). Specific Aim 1: icting social standards regarding preferred masculin ity, and evaluate how these perceptions may relate to psychological outcomes. Hypothesis 1a : Men will report significant discrepancies among self guid es /contexts regarding preferred masculine traits. Participants were given 21 examples of potential cont exts in which their gender may be most noticeable to them, but were also instructed that they may choose examples that were not on the list (see Appendix B) A total of only 14.6% of the male participants (15 males) selected one or more context that was n ot provided to them on the list of examples. Some examples of Of the contexts chosen verbatim or nearly verbatim from t he list of examples provided, just over half of all contexts selected came from the first five items on the list. The five es) Two analytic techniques were used to determine whether discrepancies among these contexts ( Table 4.2) First, two one sample t test s were cond ucted to test the null hypothesis that average masculine and feminine discrepancy scores differed from a test value of zero or no difference for the men in this sample The results of these tests indicated that Average Masculine Discrepancy was statisti cally different from zero, t (102) = 15.87, p < .001 and that Average Feminine D iscrep ancy also significantly differed from
38 zero, t (102) = 16.17, p <.001, indicating that men in this study were reporting differences in expected gendered t raits among self g uides/ contexts As a more rigorous test, and in line with the methodology used to determine significant discrepancy in Stroink (2004) two repeated measures ANOVA s were run to test f or discrepancies in expected masculine traits and feminine traits separa tely. To test for discrepancy among self guides regarding masculine traits, the thre e BSRI Masculinity scale average item scores from Contexts A, B, and C were entered as a three level within subject factor rather than the averaged AMD and AFD variables used in the t test analysis the assumption of sphericity was not violated 2 (2) = 5.69, NS The overall effect for context failed to reach significance F (2,204) = 1.94, NS p 2 = .02 A lthough the test may have been und e rpowered, the small partial eta squared m ay also indicate that context is not a significant factor in BSRI masculinity ratings for men However, consistent with Stroink (2004), context F (2,5 58 ) = 1 3.58 p < .001, p 2 = .05. Another repeated measures ANOVA was run to explore whether men reported significant discrepancies among contexts for feminine traits with the three BSRI Femininity scale average item scores from Contexts A, B, and C entered as a three level wi thin subject factor test indicated that the assumption of sphericity was not violated, 2 (2) = 2.55, NS The overall effect for context was significant, F(2,204) = 6.37, p = .002 p 2 = .06, indicating that femininity ratings differed a cross s elf guides/contexts. ratings of feminine traits, F(2,558) = 18.08 p < .001 p 2 = .06. These results were not further decomposed, given that such analysis would not be meaningful in this case (i.e., sele ction of particular self guides/contexts varied across participants).
39 Over all, these results may be seen as supportiv e of Hypothesis 1a that men perceive significant discrepancy in social messages about preferred gender related traits. Alt hough perception s of masculine discrepancy were significantly different from zero, this difference failed to ap pear in a repeated measures analysi s Somewhat surprisingly, however, both analyses show that men are report ing significant discrepancies regarding the feminine traits others expect them to possess. Hypothesis 1b : Higher rates of perceived discrepancy in masculine standards will be assoc iated with ( a) higher depressive symptomatology ( b) lower satisfaction with life, and ( c) lower self esteem. A series of Pear r correlation coefficients w ere calculat ed to examine the relationship s between perceptions of discrepancy ( Average Masculine D iscrepancy and Average Feminine D iscrepancy ) and all psychological well being variables (Depressive Symptomatology, Self Es teem, Satisfaction with Life, and Negative Affect ) for men and women separately ( Table s 4.3 and 4.4 ) though only the results for men are reported below Contrary to predictions, Average Masculine D iscrepancy did not significantly correlate with any of th e psychological health measures for men. Average Femini ne D iscrepancy for men however was significantly and positively correlated to Self E steem r (103) = .21, p = .03, indicating that greater discrepancy in expectations for feminine traits was related to greater self esteem Average Femin ine D iscrepancy also s ig nificantly correlated with Average M ascul ine D iscrepancy r (103) = .34, p < .001. Expected significant correlations were observed in the appropriate directions among all psych ological health me asur es (e.g., Satisfaction with L ife and Self E steem were positively correlated) Negative A ffect (assessing affect at the time of participa tion ) also significantly correla ted with well being variabl es in expected directions: Depressive S ymptomatology r (1 03)
40 = .43, p < .001 ; Self E steem r (103) = .37, p < .001; and Satisfaction with L ife r (103) = .25, p = .01. These results failed to suppor t Hypothesis 1b that masculine discrepancy would predict negativ e psychological outcomes. Femininity discrepancy was also unrelated to psychological outcomes with one exception: self esteem However, t his relationship was opposite of the expected direction, in that greater perceptions of discrepancy in expected feminine traits related to greater self esteem. Speci fic Aim 2: To discern the relationship between perceptions of discrepancy between self guides /contexts and reference group identity dependence Hypothesis 2 : connection to all men wil l relate to perceptions of discrep ancy such that greater feelings of connection will relate to greater perceptions of discrepancy. Pears ed to examine the relation ships of masculine discrepancy and feminine disc repancy to the four subscales of the Reference Group Identity Dependence Scale (RGIDS) for men ( Table 4.3 ). The subscales of the RGIDS showed correlations of similar direction and magnitude as Wade & Gelso (1998). Contrary to predictions, Average Masculi ne D iscrepancy and Average Feminine D iscrepancy did not significantly correlate with scores on any of the gender role identity statuses. Therefore, Hypothesis 2 was not supported, as the perception of discrepancy in self guides was not related to the Refer ence Group Nondependent Diversity subscale of the RGIDS. Specific Aim 3 : To explore the rel ationship between reference group identity dependence and psychological outcomes.
41 Hypothesis 3a: The extent to which me n lack a reference group in defining their gender identity will significantly predict negative psychological outcomes, such that less identification will result in a) higher depressive symptomatology b) lower sat isfaction with life, and c) lower self esteem A series of Pea oefficients were calculated to examine the relationship between the No Reference Group Identity (NRG) subscale of the Reference Group Identity Dependence Scale (RGIDS), and all of the psychological well being variables. Consistent with predictions, scores on the NRG subscale were associated with higher depressive symptomatology r (103) = 42 p < .001; lower self esteem r (103) = .45, p < .001, lower satisfaction with life r (103) = .23, p < .001 and higher negative affect, r (103) = .22, p = .03 Negat ive Affect and Satisfaction with Life are classified as small to medium size effects, while Depressive Symptomatology and Self Esteem could be classified as medium to large size effects (Cohen, 1992). To further ex plore the relationship of NRG scores to th e psychological well being variables, an inverted hierarchical regression analys is was run in which Depressive Symptomatology, Self E steem, Satisfaction with L ife and Negative A ffect were regressed onto NRG scores. The limitations and merits of this inve rsion of independent and dependent variables within this type of analysis are explored in the Discussion section The ratio of valid cases to number of independent variables (103:4, or 25.75:1) exceeded both the minimum (5:1) and preferred (15:1) ratio (F ield, 2005), indicating sufficient sample size for this analysis. The Durbin Watson statistic (2.07) for testing the presence of serial correlation among residuals was in th e acceptable range ( Durbin & Watson, 1951) indicating sufficient independence of errors. Tolerance values for all independent variables (from .44 to .71) exceeded .10, indicating that
42 multicollinearity wa s not a problem in this analysis (Field, 2005) The hierarchical regression model was run in two steps. N egative A ffect was entere d in to the first block to control for any possible negative affect covariance across subsequent variables entered into the model Next depressive symptomatology, self esteem, and satisfaction with life va riables were entered in the second block The fi rst step (Negative A ffect) produced a model with an R 2 of .05 F (1,101) = 5.21 p = .025 ( Table 4.5 ) indicating that this variable accounted for 5 % of the variance observed in NRG scores. The second step produced an overall m odel with an R 2 of .26 F (3 ,98) = 8 50 p < .001, indicating that Negative A ffect, Self Esteem, Depressive Symptomatology, and Satisfaction with L ife scores together accounted for 26% of the variance observed in NRG scores For the second model, the R square change statistic was si gnificant, F (3,98) = 8.50, p < .001 indicating that a significant amount of the variance in NRG sco res (21%) was accounted for with the well being variables even while controlling for the effects of negative affect covariance This constitutes a medium sized effect (Cohen, 1992). T wo of three well being variables contributed significantly to NRG : depressive symptomatology, = .32 t (98 ) = 2.43 p = .017 ); and self esteem, = .40 t (98) = 3.18 p < .002 ). Self Esteem had the largest standardized beta indicating it made the greatest independent contribution to the fit of the model. While S atisfaction with L ife (SWLS) was hypothesized to be related to NRG scores, it did not contribute significantly to the prediction of NRG when combined with the other predictor variables, = .21, t (98) = 1.75, NS despite a significant negative bivariate correlation between the two me asures This suggests that th e significant c orrelation between SWLS and NRG may most ly reflect shared variance of SWLS with the other target predictor variables and/or with negative affect
43 correlation coe fficients were also calculat ed betwe en each of the other subscales of the RGIDS a nd all well being outcomes ( Table 4.3). Although no specific predictions were made n o significant co rrelations were observ ed between psychological well being variables and the Reference Group Dependent and Ref erence Group No ndependent Similarity subscale scores H owever, there was a significant, small to medium sized cor relation between the Reference Group Nondependent Diversity subscale and self esteem r (103) = .24, p = .013 Some additional analyses of this hypothesis may be seen in the analyses of Hypothesis 3b. These results were partially supportive of Hypothesis 3a, that scores on the No Reference Group subscale of the RGIDS would be a significant predictor of negative psychological well being out comes. Self Esteem, Depressive Symptomatology, and Satisfaction with Life were all significantly correlated in the expected directions with No Reference Group endorsement. While no specific predictions were made about the relationships of the well being variables to the other subscales of the Reference Group Identity Dependence Scale, it was found that no significant correlations emerged excepting one. Self esteem showed a positive correlation with the Reference Group Nondependent Diversity subscale, i ndicating that men who expressed an appreciation for a diversity of male identities amongst men also reported higher self esteem. Hypothesis 3b : Self concept clarity will moderate the relationship between lack of reference group and negative psychologica l outcomes, such that those men possessing high self concept clarity will show a weaker associations to negative psychological outcomes. Hierarchical regression analyses were then run to test for a possible buffering moderation effect of Self Concept C lar ity on the relationship between No Reference Group scores and well being variables. A moderator variable alters the strength of a causal relationship, and indicates
44 the conditions under which a predictor variable has a relatively stronger or weaker effect upon an outcome variable. A buffering moderator reduces the strength of the relationship betwe en a predictor and outcome variable. In this case, it was predicted that participants with high No Reference Group scores who also possessed high Self Concept C larity would show less of an association between NRG and well being outcomes relative to those lower in Self Concept Clarity. In other words, high self concept clarity would serve to buffer the negative effects of lacking a reference group on well being. T he approach advo cated by Frazier, Tix, & Barron ( 2004) was used for this test of moderation, which suggests first standardizing predictor and moderator variables by creating Z scores of each and then creating a product term of those standardized variable s to represent the interaction. The predictor and moderator variables are entered first and the interaction te rm is entered separately A comparison between these models gives an indication of whether a significant moderation is present. Two exceptions to the standard regression interpretation should be kept in mind when evaluating the following results. First, as all variables entered into the analysis have been standardized (in the form of Z scores), the unstandardized B values should be interpreted, rather than the beta weights (Frazier et al., 2004). Second, these B coefficients reflect the conditional effects of each variable, in that they represent the contribution of that variable to the model when all other variables are held at a value of zero. Three separate hierarchical regressions were run using this method in which Self Esteem, Depressive Symptomatology and Satisfaction with Life served as the criterion vari able, while No Reference Group scores (NRG), Self Concept C larity (SCC), and th e interaction of NRG and SCC served as predictor variables. Consistent with analyses of Hypothesis 3a, Negative Affect was again included within the se models in the first step in order to control for negative affect
45 covariance. The Durbin Watson statistic s for the Self Esteem and Satisfaction with Life regression analyses were in the acceptable range (2.16 and 2.11, respectively), indicating sufficient independence of errors. There was however a violation of the indep endence of errors assumption for Depre ssive Sy mptomatology (1.12) indicating that the results obtained from this test should be interpreted cautiously due to some serial correlation in error terms Tolerance values for al l independent variables (from .77 to .94) indicated that multicollinear ity wa s not a problem in any analysis First, the possible moderating effects of SCC were examined with Self Esteem serving as the criterion variable ( Table 4.6 ) Three steps produced three separate models, though only the third model is discussed here as this model indicates both the conditional effects of th e predictors, as well as the presence of any moderation (Frazier et al., 2004) The third model including negative affect, NRG, SCC and the moderator product term (NRG x SCC) produced an R 2 of .4 9 F (4 ,98 ) = 23.20 p < .001 indicating that 49 % of the va riance in Self Esteem scores can be accounted for by No Reference Group scores, Self Concept Clarity and the interaction of these two variables NRG, = 23 t (102) = 2.82 p = .006 ; SCC scores, = .52 t (102) = 5 99 p < .001 ) ; as well as the moderator product term, = .19 t (102) = 2.28 p = .025 ) all contribute d significantly to Self Esteem Most importantly, this model represented a signi ficant increase in predictive power over the previous model which did not contain the moderator product term, with an R 2 change of .03, F change (1,98) = 5 21 p = .0 25 The negative B value for the product term suggests that the interaction of self concep t clarity and No Reference Group reduces the impact of No Reference Group endorsement on Self Esteem, and thereby moderates this relationship. The form of this relationship is discussed later in this section.
46 Next, the potential moderating effects of SCC were examined with Depressive Symptomatology serving as the criterion variable. The third model including negative affect, NRG, SCC and the moderator product term (NRG x SCC) produced an R 2 of .53 F (4,98) = 27.49 p < .001 indic ating that 53 % of the v a riance in Depressive Symptomatology scores can be accounted for by No Reference Group scores, Self Concept Clarity and the interaction of these two variables Both NRG, = .15 t (102) = 2.11 p = .037 and SCC scores, = .4 2 t (102) = 5 58 p < .001 ) contribute d significantly to Depressive Symptomatology. However, the moderator product term, = .07 t (102) = 1.03 NS ) failed to reach significance. Accordingly, th is model did not significantly increase predictive power over the previous model which did not contain the moderator product term, with an R 2 change of .01 F change (1,98) = 1.06 NS suggesting that self concept clarity does not moderate the effect of No Reference Group on Depressive Symptomatology. This result, as previously mentioned, must be interpreted cautiously, as the independence of errors assumption was violated in this particular analysis. Finally, the potential moderating effects of SCC were ex amined with Satisfaction with Life (SWLS) serving as the criterion variable. The third model including negative affect, NRG, SCC and the moderator product term produced an R 2 of .2 7 F ( 4 ,98 ) = 6.06 p < .001 indicating that 27 % of the variance in Satis faction with Life scores can be accounted for by No Reference Group scores, Self Concept Clarity and the interaction of these two variables NRG failed to contribute sig nifica ntly to Satisfaction with Life = 04 t (102) = 45 NS However SCC scores, = .45 t (102) = 4. 55 p < .001 ) as well as the moderator product term, = .18 t (102) = 4. 55 p < .001 ) contribute d significantly to S atisfaction with Life Most importantly, this model represented a significant increase in predictive power over the previous model which did
47 not contain the moderator product term, with an R 2 change of .03, F change (1,98) = 3.96, p = .05. The negative B value for the product term suggests that the interaction of self concept clarity and No Reference Group reduces the impact of No Reference Group endorsement on Satisfaction with Life, and thereby moderates this relationship To examine the form of these moderations, graphs of predicted va lues were created based on the r egressi on equations for Self Esteem and Satisfaction with Life (Aiken & West, 1991) The graphs represent the values of NRG and the outcome variables at one standard deviation above and below the mean value of NRG for men who scored low, average and high in self concept clarity ( Figure s 4.1 and 4.2 ) The pattern of the plotted interaction gives an indication of the conditions under which the moderation occurs. The moderation effect for self esteem was analyzed first. Visual inspection of the separate S CC groups appears to show a stronger association between NRG and self esteem for men high in SCC. Indeed, t he simple slopes of the these predicted relationship s indicated that NRG was actual l y more strongly associated with self esteem when self concept cl arity was high, B = .42, t (99) = 3.29, p = .001 than wh en SCC was average B = .23, t (99) = 2.75, p = .007, or low, B = .04, t (99) = .38, NS These results indicate that, while self concept clarity does moderate the relationship of No Reference Gr oup scores with self esteem and satisfaction with life, it does so in the opposite pattern of what was expected with high SCC actu ally exacerbating ra ther than buffering the effect of NRG on negative psychological health outcomes relative to participan ts with lower SCC It appears that the significant moderation found may a ctually express the buffering effects of low self concept clarity on negative outcomes, as there was no significant relationship between NRG and self esteem in this group. Although participants with high self concept clarity appear ed to be more sensitive to the neg ative sequelae of NRG endorsement they did also
48 appear to have higher levels of self esteem overall compared to participants with lower levels of self concept clarity Sec ond, the pattern of the interaction for Satisfaction with Life was assessed. Visual inspec tion appears to indicate no association between NRG and Satisfaction with Life for men at average levels of SCC ( Figure 4.2) However, there does appear to be a cro ssover interaction for men low and high in SCC, with participants high in SCC showing lower Satisfaction with Life as NRG increases, and participants low in SCC showing higher Satisfaction with Life as NRG increases Again, high SCC was still on average a ssociated with higher levels of Satisfaction with Life. Analysis of the simple slopes and their difference from zero, however, indicated that no slope was significant for high, B = .2 2, t (99) = 1.58 NS average B = 04 t (99) = 44 NS n or low SCC scorers B = 1 4, t (99) = 1.30 NS The presence of the moderation effect in previous analyses does however indicate that these slopes significantly differ from each other. Therefore, the associations between NRG and Satisfaction with Life in this analys is are most properly understood as patterns of relationships at various levels of SCC, rather than as standalone bivariate relationships.
49 Table 4 1 Means, Standard Deviations, and Distribution Statistics for Study Variables by Gender Mean SD SEM Skewness Kurtosis M en Average Masculine Di s crepancy 0.91 0.58 0.06 0.79 0.07 Average Feminine Discrepancy 1.43 0.89 0.09 0.53 0.56 Depressive Symptomatology Self Concept Clarity 19.38 39.17 8.77 8.91 0.86 0.88 0.93 0.04 0.46 0.72 Global Self Est eem 32.13 5.28 0.52 0.33 0.66 Satisfaction with Life 24.15 6.30 0.62 1.16 1.52 Positive Affect 29.76 6.59 0.65 0.14 0.35 Negative Affect 17.27 5.71 0.56 0.70 0.01 Reference Group Identity Scales No Reference Group 28.69 7.69 0.76 0.63 0. 21 Reference Group Dependent 25.31 5.06 0.50 0.18 0.75 Reference Group Nondependent Total 51.27 7.60 0.75 0.51 0.66 Reference Group Nondependent Diversity 30.05 4.31 0.42 0.37 0.73 Reference Group Nondependent Similarity 21.22 5.00 0.49 0.61 0.1 5 Women Average Masculine Di s crepancy 0.87 0.56 0.03 0.93 0.58 Average Feminine Discrepancy 1.35 0.93 0.06 0.67 0.37 Depressive Symptomatology Self Concept Clarity 20.60 39.36 10.07 9.92 0.60 0.59 0.92 0.11 0.50 0.69 Global Self Esteem 31 .46 5.36 0.32 0.47 0.01 Satisfaction with Life 24.69 6.93 0.41 0.73 0.22 Positive Affect 26.65 7.71 0.46 0.15 0.13 Negative Affect 16.47 6.06 0.36 1.36 2.04 Note. Male n = 103, Female n = 280.
50 Table 4 2 Means and Standard Deviations of Av erage Item Level Difference Scores Between Contexts in B em S ex R ole I nventory Ratings Masculinity Femininity Mean SD Mean SD M en Women Context A to B .81 .07 1.34 .12 Context A to C .93 .08 1.50 .13 Context B to C .98 .08 1.44 .11 Context A to B .84 .04 1.32 .07 Context A to C .90 .05 1.32 .08 Context B to C .88 .05 1.42 .07 Note. Averages based on absolute values of differences. Differe nce scores have a possible range of 0 t o 6 Male n = 103, Female n = 280.
Table 4 3 Pearso n Correlation Coefficients for Well Being and Gender Discrepancy Variables for Men 1 2 3 4 5 6 7 8 9 10 11 12 13 1. Depressive Symptomatology -2. Self Concept Clarity 0.59** -3. Global Self Esteem 0.64** 0.61** -4. Satisfaction with Life 0.60** 0.47** 0.65** -5. Positive Affect 0.19 0.22* 0.44** 0.29** -6. Negative Affect 0.53** 0.23* 0.37** 0.25* 0.08 -7. No Referenc e Group 0.42** 0.43** 0.45** 0.23* 0.16 0.22* -8. Reference Group Dependent 0.12 0.06 0.02 0.04 0.06 0.09 0.04 -9. Reference Group Nondependent Di versity 0.08 0.08 0.24* 0.01 0.14 0.14 0.21* 0.23* -10. Reference Group Nonde pendent Similarity 0.04 0.06 0.03 0 .00 0.04 0.01 0.47** 0.04 0.33** -11. Reference Group Nondependent Total 0.07 0.01 0.15 0.00 0.10 0.08 0.43** 0.16 0.78** 0.85** -12. Average Masculine Discrepancy 0.01 0.07 0.12 0.06 0.07 0.10 0.13 0.04 0.06 0.09 0.03 -13. Average Feminine Discrepancy 0.05 0.05 0.21* 0.10 0.00 0.10 0.16 0.03 0.10 0.02 0.04 0.34** -Note = significant at the .05 level; ** = significant at the .01 level
Table 4 4 Pearson Correlation Coefficients f or Well Being and Gender Discrepancy Variables for Women 1 2 3 4 5 6 7 8 1. Depressive Symptomatology -2. Self Concept Clarity 0.56 ** -3. Global Self Esteem 0.60 ** 0 .66 ** -4. Satisfaction with Life 0.56 ** 0.49 ** 0.67 ** -5. Positive Affect 0.31** 0.22 0.28 ** 0.36 ** -6. Negative Affect 0.50 ** 0.26* 0.23 ** 0.23 0.07 -7. Average Masculine Discrepancy 0.04 0.07 0.03 0.06 0.12* 0.06 -8. Average Feminine Discrepancy 0.04 0.06 0.01 0. 0 8 0.07 0 0 .32** -Note = significant at the .05 level; ** = significant at the .01 level
53 Table 4 5 Hierarchical Regression Analysis of the Effects of Neg ative Affect, Satisfaction with Life, Self Esteem, and Depressive Symptomatology on the No Reference Gro up Subscale of the Reference Group Identity Dependence Scale Variable B SE(B) R 2 R 2 change Step 1: Negative Affect .30 .13 .22* .05* Step 2: Model .26** .21 ** Negative Affect .05 .14 .04 Satisfaction with Life .26 .15 .21 Self Esteem .58 .18 .40* Depressive Symptomatology .28 .12 .32* Note Criterion variable: No Reference Group; p < .05, ** p <.001
54 Table 4 6 Analyse s of the Moderation Effects of Self Concept Clarity on the Relat ionship Between the No Reference Group S ubscale of the Reference Group Identity Status Scale and Self Esteem, Depressive Symptomatology, and Satisfaction with Life Variable B SE(B) R 2 R 2 change Criterion Variable: Self Esteem Ste p 1: Negative Affect .38** .10 .14** Step 2: Model .46** .32** Negative Affect .22* .08 No Reference Group .19* .08 Self Concept Clarity 51** .09 Step 3: Model .49** .03* Negative Affect .21* .08 No Reference Group .23* .08 Self Concept Clarity .52** .09 NRG x SCC .19* .08 Criterion Variable: Depressive Symptomatology Step 1: Negative Affect .49** .08 .28** Step 2: Model .52** .25** Negative Affect .37* .07 No Reference Group .14 .07 Self Concept Clarity .42** .08 Step 3: Model .53** .01 Negative Affect .37** .07 No Reference Group .15* .07 Self Concept Clarity .42** .08 NRG x SCC .07 .07 Criterion Variable: Satisfaction with Life Step 1: Negative Affect .25* .09 .06* Step 2: Mo del .24** .18** Negative Affect .15 .09 No Reference Group .01 .09 Self Concept Clarity .44** .10 Step 3: Model .27** .03* Negative Affect .14 .09 No Reference Group .04 .09 Self Concept Clarity .45** .10 NRG x SCC .18* .09 Note All predictor variables were analyzed as Z scores; NRG = No Reference Group; SCC = S elf Concept Clarity Total Score; p < .05, ** p <.001
55 Figure 4 1 Predicted Graph of the Rela tionship of No Reference Group Scores to Self Esteem, with Self Concept Clarity as a Moderator Note. Values are graphed as Z scores.
56 Figure 4 2 Predicted Graph of the Relationship of No Reference Group Scores to Satisfac tion with Life with Self Concept Clarity as a Moderator Note. Values are graphed as Z scores.
57 CHAPTER 5 DISCUSSION Review of Study Findings The purpose of this dissertation was to examine the presence and possible effects of conflicting standards of ma sculinity as experienced by men. Results were mostly supportive of the idea that men perceive significant discrepancy in social messages about preferred gender related traits though this discre pancy was reported most reliably with regard to expected femi nine, rather than masculine, traits Although perceptions of masculine discrepancy were significantly different from zero, this difference failed to appear in a repeated measures analysis though a power analysis predicted that an additional 60 male parti cipants would have produced a significant effect Whi le it is possible that the sample size was insufficient to detect a relatively small effect in masculine discrepancy for men, it cannot be concluded that the finding would have been significant given a la rger sample size However, another caveat should be added about the suitability of this analysis for testing this hypothesis In Stroink (2004), the self guides that were investigated were standardized across participants, meaning that all participants rated the same self guides as suggested by the instructions (e.g., all participants were asked to rate the perceptions of their mother as the self guide for Context A). In contrast, this study asked participants to select their own relevant self guides, and therefore a comparison of means between contexts (as with a repeated measures ANOVA) may collapse over significant differences that do in fact exist on the participant level. Therefore, t tests u s ing an estimate of discrepancy on the individual level (i.e., Average Masculine Discrepancy and Average Feminine Discrepancy) may be the preferable method in determining significant discrepancy given the nature of this methodology.
58 C ontrary to hypotheses, the perception of discrepancy among masculinity self g uides w as un related to any of the measures of well being assessed in this study Further, the re were no relationship s between perceptions of either feminine or masculine discrepancy and any dimension of the Reference Group Identity Dependence Scale Take n together, these results indicate that while men do report perceiving conflicting standards with regard to expected masculine attributes from relevant guides, these discrepancies are unrelated to the measures of well being assessed in this study. Discrep ancy also appears to be unrelated to perceptions of psychological relatedness to other men, and to perceptions of similarity and difference amongst men. Although these findings may reflect a truly null result, a few things may have impacted this result. T he method used to estimate discrepancy relied on three participant chosen data self guides, not the least imped ing of which may have been convenience. T o ensure better compli ance with instructions, participants received many examples of contexts and situations that could constitute a self g uide. Analysis of self guides chosen by participants shows that the majority of contexts selected came verbatim from the list of suggested contexts. Though these contexts may have held the most relevance for some of the participants who selected them, it cannot be counted out that having provided such a list in the instructions may have reduced the amount of cognitive effort that some parti cipants applied to the task. It is also possible that participants were inclined to select guides/contexts that were related for them in some way, thereby minimizing any discrepancy variance that might have existed among other, more unrelated guides. Par ticipants were given no instruction to intentionally avoid such similarity in guides, to avoid creating any demand characteristics. In fact, many of the most popular contexts selected in this study were
59 associated with more traditional masculine environme nts (e.g., at the gym, playing sports). Addi tionally, as participants were asked to rate only three contexts to reduce response fatigue, this may have been an insufficient number of guides to represent the true variability these individuals might have exp erienced across the hundreds of contexts experienced in everyday life. Therefore, a more robust estimate of discrepancy may have had more power to reveal possible relationships to well being or RGIDS measures. Unexpectedly, greater femininity discrepancy was related to greater self esteem. Though this finding may have been spurious, it may also indicate that men who enjoy greater latitude among significant contexts likely to restrict their expression to those behaviors and attributes deemed traditionally to a more positive self image in that, at least in some contexts, these men are less confine d by the rigid gender schemata that are associated with poorer psychological and health outcomes for men (Courtenay, 2000). This finding may also indicate that changes in social expectations of men have centered more on encouraging the adoption of more tr The results were partially sup portive of the hypothesis that scores on the No Reference Group subscale of the RGIDS would be a significant predictor of negative p sychological well being outcomes This result was consistent with previous demonstrat ions of a link between No Reference Gr oup endorsement and poorer psychological health ( W ade & Brittan Powell, 2000 ; Wade & Gelso, 1998 ). Self esteem, depressive symptomatology, and satisfaction with l ife were all significantly correlated in the expected directions with No Reference Group endo rsement.
60 While this may provide sufficient evidence of an association between No Reference Group endorsement and well being, the direction of causality, if any, is harder to determine statistically in a sample of this size When No Reference Group endors ement (NRG) wa s treated as an outcome variable within a regression analysis, both self esteem and depressi ve symptomatology we re predictive accounting for 21% of No Reference Group endorsement The use of a hierarchical regression analysis to explore th is hypothesis must be qualified for its relative contributions as well as its limitations. The hypothesis stated that a single independent variable (No Reference Group scores) could predict multiple psychological well being outcomes. However, NRG was ent ered as an outcome, rather than as a predictor variable. In other words, while this analysis has demonstrated how well certain well being outcomes predict NRG scores, one may not conclude that NRG scores predict the well being outcomes in the same way, as regression equations are not symmet rical in this manner. Despite this analysis may still be a useful one in that it helps to parse out the effects of negative affect at the time of participation, which was hypothesized to have accounted for some covarian ce among these well correlations are indeed unique and meaningful ones. Additionally, conducting this analysis gives a greater understanding of how these well being vari endorsement of content on the NRG subscale. Limitations to the regression analysis notwithstanding, the significant correlations between No Reference Group scores and lower well being deserve further discussion The items on this scale reflect the feeling that one is unlike all other males and that one lacks a feeling of connection or identification with other males. In lacking a specific male group with which to identify, Wade and Gelso (1998) conjectured tha t these men also lack any specific sense of male
61 identity. Such a lack of identification seems as t hough it may carry both costs and benefits however, given that identification a male reference group more often entails identification with the dominant ge nder ideology of traditional masculinity (Wade & Gelso, 1998). Holding more traditional views and favoring traditional norms of masculine conduct has been consistently 8 a ; Courtenay, 2000). Indeed, Wade and Gelso found that No Reference Group scores were unrelated to gender role conflict, a construct which measures negative psychological and interpersonal consequences of adherence to traditional masculine norms. The a uthors posited that no relation would exist between these constructs because high NRG endorsers would not have internalized or identified with traditional masculinity ideology enough to experience conflict as a result. Therefore, while lacking a sense of belongingness to a male reference group may carry costs associated with psychological or physical isolation from like others (Sherif & Sherif, 1969), lacking a reference group may also subvert some of the costs associated with alignment to more restrictive traditional norms. This raises the question of whether at least some high endorsers of items on the No Reference Group construct may lack a significant male reference group because they consciously or unconsciously eschew the expectations of traditional m asculinity yet feel at a loss for what other definition of masculinity to embrace. This idea would certainly require further exploration. Interestingly, the Reference Group Nondependent Diversity subscale of the RGIDS was positively correlated with sel f esteem indicating that men who expressed an appreciation for a diversity of male identities amongst men also reported higher self esteem. While it is possible that this relationship is ecologically valid, it is also possible that this significant relat ionship may reflect the high risk for Type I error due to inflated familywise error. Additionally, as the
62 reliability of this subscale in this sample was below acceptable levels ( ), this result should be interpreted cautiously. A similar relationship, however, was found by Wade & Gelso (1998) between another measure of self esteem and the other Reference Group Nondependent (RGN) subscale: Similarity. As both the Diversity and Similarity RGN subscales tap a construct of relatedness to other men that is more internally defined and associated with greater ego identity development, the association between these scales and self esteem is perhaps not too surprising. The final hyp othesis, that self concept clarity would moderate the relationship between No Reference Group endorsement and negative psychological outcomes was partially supported. Self concept clarity reduced the impact of NRG on self e steem and on s atisfaction with li fe, but did not significantly moderate the relationship for depressive symptomatology Lower self concept clarity was consistently associated with the lowest well being, regardless of NRG endorsement. The pattern of these relationships, however, was oppo site of what was expected. NRG was actually more strongly associated with self esteem and with satisfaction with life when self concept clarity was high such that there was more of a decrease in both well being outcomes as NRG endorsement increased So mewhat strangely, there was no association between NRG and self esteem when self concept clarity was low It appears tha t when self concept clarity is low, the presence or absence esteem. T his may make some sense in that when an individual lacks of a clear or consistent enough sense of who he is on a day to referent is to be found in some group, th ey are likely just as confused and inconsistent in their own self concept, and therefore possessing or lacking such a reference group may be
63 inconsequential esteem when there is no clear or consistent content in that social referent to help d The pattern of moderation for satisfaction with life was also difficult to decipher, given that the slopes for one standard deviation above, below, and at th e mean we re all non significant By more conservative standards (i.e., p < .01), this moderation interaction would have failed to reach significance, and so the possibility that this finding is spurious is not ruled out. However, visual analysis of the trends of these lines suggests that there was no relationship between NRG and satisfaction with life for those who were average in self concept clarity, while those high in self concept clarity saw more precipitous drops in satisfaction with life as NRG scores increased. Interestingly, participants low in self concept clarity t rended towards a boost in satisfaction with life as psychological relatedness to other males decreased. Perhaps t hose who identify with a reference group but remain unsure of themselves may be helped by abandon ing these external referents. In other words self concept, greater internal dissonance may lead to lower satisfaction with life, while the lack of a n external reference group may reduce this dissonance. These ideas would need t o be explored. It is important to note that a ll of these moderation results occurred within the context of a main effect for self concept clarity, such that higher levels of self concept clarity were associated with higher self esteem satisfaction with life, and lower depressive symptomatology when matched for levels of NRG Thus while self concept clarity may provide a general boost to well concept clarity does not buffer an individual from the costs of lacking a reference group otherwise. This pattern suggests that while lacking a male reference group may be more detrimental to those with high er
64 than lower self concept clarity, those with high self concept clarity still, on average, enjoy higher levels of self esteem and satisfaction with life regardless of the lack of a male reference group. The finding that some men possess high self concept clarity in the context of low feelings of relatedness to other males, is somewhat difficult to understand within the context of Reference Group Identity Dependence theory, which posits that the self concept is largely determined by these feelings of identification with a male reference group. In their original scale developme nt study, Wade and Gelso (1998) indicated that much of the variance in their construct was still categories, and this may explain the finding discussed above. Namely, a man who possesses high self concept clarity may possess this clarity within the context of some other significant reference group, such as a church or hobby group. Th is finding may a lso give some partial support to the idea previously suggested that at least some men who lack a significant reference group may be reject ing the prevailing traditional masculinity ideology, yet still perceive no suitable alternatives for identification in the gender terrain Indeed, in a social environment where gender proscriptions performances (Goffman, 1959). Limitations of the Study This study is limited by many of t he same issues as exist in all studies that seek to make inferences from self reported attitudes and perceptions to more externalized outcomes. In making i nferences from these results, one must place some modicum of t rust in the idea that the various manifestations of gender are sufficient ly dis cernable and articulable for respondents. A related limitation is that the construct is comprised of not only attributes as were focused on in this study, but also attitudes, self perceptions pre ferences, and behaviors, all of
65 which may have circumspect correlations to one another (Bartini, 2006). Therefore, one is limited in their ability to generalize from one domain of the construct to another. For instance, we may not assume that because som e men perceive discrepancies in expected gender attributes that they also act differently or per ceive themselves differently across dissimilar contexts No singular methodology can shed light on the totality of how gender manifests itself in and amongst i ts actors, as it remains a complex social phenomenon. In some ways, the use of a college setting to draw participants was ideal in that this population may still be grappling with the tasks of identity commitment (Erikson, 1968). In particular, more recen t literature has suggested that contemporary men of college age may experience an extended adolescence in which identity tasks are more drawn out ( Kimmel, 2008). If this be the case, one might expect the effects of any perceived discrepancies in expectati ons to be most consequential at this time, when identities may be more in flux. However, because the sampling for this study was not more cross sectional in nature, findings may be limited to college educated men in their late teens and early twenties. T he fact that certain contexts such as these men some of the most noticeable places where gender is salient may suggest that the set of normative developmental ta sks at various life stages may have much to do with perceptions of discrepancy. One might imagine that there may be considerably more homogeneity among these which may be more r eadily selected by men in their thirties and forties. In this way, one might suspect that perceptions of gender discrepancy may fluctuate as life roles change, and therefore age may be of considerable importance in studying these effects.
66 Further, a ltho ugh the racial/ethnic composition of this sample did roughly approximate the U.S. population statistics (U.S. Census Bureau, 2010), we would have reason to suspect that not only level of education al attainment, but also socioeconomic status of the family o f origin may be rather skewed in this sample though no specific data was collected to this effect An analysis of any potential cultural differences in the findings was not permitted, given the relatively small overall sample size. The r elation ship of ou r findings to (1987) self discrepancy theory is somewhat the attributes they felt others expected them to possess, i.e., in the language of Higgins According to self primary (actual/own) self concept and other selve s would lead to characteristic patterns of negative affect, including guilt, shame, dejection, agitation, and/or embarrassment (Higgins, 1987). Therefore, i n that this study did not investigate these particular distinctions amongst selves it may have missed some possible associations with negative outcomes for gender discrepancies between actual/own and the remain ing self states Perhaps this study failed to find associations between gender discrepancy and poorer psychological outcomes because the ought/other self is less central to the core self concept than is the actual/own self (Higgins, 1987), and thereby any dissonance that might exist there is insufficient to cause significant distress In other words, while men may perceive discrepancies amongst others regarding gender expectations, these discrepancies may be less internalized than discrepancies amongst on expectations around what constitutes masculinity. Further limitations of thi s study center around issues of analys is for moderation effects Aguinis (2004) has shown that mo deration effects are often small, and therefore power to detect
67 them is of ten quite low in most studies. Aguinis has suggested that researchers typically need a very large s ample size in order to have reasonable power to detect moderator effe cts utilizing a continuous variable There is also some suggestion that power is parti cularly low when tests of moderation utilize two continuous variables, as in the case of this study (McClelland & Judd, 1993). Although insufficient power to detect a moderation effects was not an issue for the analysis of self esteem and satisfaction wit h life as outcomes, the sample size may have been prohibitive in detecting moderation effects for depressive symptomatology. Future Directions I t is important to understand how conflicting external standards are being negotiated by men, and how these ne gotiations relate to subsequent mental health outcomes. Though initial suggestion from this study indicates that such discrepancies may not be related to well being outcomes, there is sufficient cause to continue to address a possible link via other outco me variables, and via different methodologies. Future studies should look at how discrepancies between self relate to unique constellations of outcomes In particular, discrepancies between gendered act ual own and other self states may be the most deeply internalized and therefore the most likely discrepancies to produce negative consequences. Additionally, an analysis of how participants responded to the prompt in this study suggests some possible direc tions for modification of methods in future studies. Participants seemed to favor contexts that were provided to them, and in particular, seemed to exercise some economy in selecting the initial examples provided to them in the prompt. Although this may end of crafting this list, randomization of the list order presentation in future studies may tease out whether there may be a potential confound of list ord er in this study. Further, the fact that
68 may have also led to differences in how well participants were able to respond to a prompt asking them to rate the ex prompt to make it better fit the nature of the contexts selected. Further, future studies coul d look at whether any characteristic differences appear between personal and more impersonal contexts in the extent to which men process or internalize concurrent gender expectations. For instance, might the accepted culture of the fraternity environment, for example, carry a narrative weight that is experienced more palpably by men than are the expectations of a given individual? In addition to these possible differences attributable to the nature of the self guide, o ne could also expect that there will be considerable variation in how individual men assign meaning and value to the perceived mores of any given guide will be relevant to explore in future stud ies, but also the valence of that context on the individual level. In that lack of a reference group has now been more consistently shown to correlate with diminished well being, further work should be done to elucidate this link and any other possible m ediators or moderators of this association. Some of the findings here indicate that the reasons for lacking a male reference group may be multiple and more complex than originally theorized by Wade (1998). Perhaps at least some men who lack a significant reference group may be rejecting the prevailing traditional masculinity ideology, yet still perceive no suitable alternatives for iden tification amongst male peers This segment of men may be increasingly aware of the costs associated with traditional ma sculinity (e.g., risk taking, emotional restrictiveness, lack of
69 intimacy), feel compelled or enticed to expand their roles as father s friends and partners that violate traditional codes of masculinity (e.g., Silverstein, Auerbach, & Levant, 2002; Single ton & Maher, 2004) and therefore lack a consistent definition of themselves as men in a way that translates in Exploratory qualitative studies might look at the conscious reasons men feel a sense of similarity to all, s ome, or no other men, and how these feelings may be related to gender ideologies more broadly. The link between refere nce group dependency and psycho l og ical health may also suggest new avenues by which counselors and psychologists may work with clients to treat mental illness and alleviate adjustment difficulties. It would also be interesting to investigate whether self concept clarity in the context of some non male reference group, such as a hobby group, may account for more or less variance in psych reference group. Further, how various psychological variables, such as lifestyle, values, and personality inte ract with the associations shown in this study needs further attention within the men and masc ulinities literature (McKelley & Rochlen 2010). Conclusion The findings in this study indicate that men indeed perceive discrepant standards around the masculin e, and even more strongly, the feminine attribut es they are expected to possess, though these discrepancies were not related to psychological outcomes or feeling s of psychological relatedness to other men. There was however a link between the lack of a significant male reference group and diminished well being, though this link was moderated by se lf concept clarity. Although overall only some support was found for the hypotheses in this study, the results obtained suggested new pathways of inquiry that may expand upon previous work on conflicting standards and reference group identity dependence. Further work will be
70 required to understand what impact, if any, the perception of discrepancies in gender
71 APPENDIX A SCALE ITEMS Bem Sex Role Inventory Directions Please rate how much you feel [ self guide inserted] believes you should possess each of the following attributes. 1. Defend my own beliefs 2. Affectionate 3. Conscientious 4. Independent 5. Sympathetic 6. Moody 7. Assertive 8. Sensitive to needs of others 9. Reliable 10. Stron g personality 11. Understanding 12. Jealous 13. Forceful 14. Compassionate 15. Truthful 16. Have leadership abilities 17. Eager to soothe hurt feelings 18. Secretive 19. Willing to take risks 20. Warm 21. Adaptable 22. Dominant 23. Tender 24. Conceit ed 25. Willing to take a stand 26. Love children 27. Tactful 28. Aggressive 29. Gentle 30. Conventional
72 Positive Affect and Negative Affect Scale This scale consists of a number of words that describe different feelings and emotions. Read each item and t hen mark the appropriate answer in the space next to that word. Indicate to what extent you feel this way RIGHT NOW Use the following scale to record your answers. very slightly or not at all a little moderately quite a bit extremely 1 2 3 4 5 1) interested 2) distressed 3) excited 4) upset 5) strong 6) guilty 7) scared 8) hostile 9) enthusiastic 10) proud 11) irritable 12) alert 13) ashamed 14) inspired 15) nervous 16) determined 17) attentive 18) jittery 19) active 20) afraid Satisfaction With Life Scale Below are five statements with which you may agree or disagree. Using the 1 7 scale below, indicate your agreement with each item by placing the appropriate number on the line preceding that item. Please be open and honest in your responding. The 7 point scale is 1 = strongly disagree, 2 = disagree, 3 = slightly disagree, 4 = neither agree nor disagree, 5 = slightly agree, 6 = agree, 7 = strongly agree. 1) In most wa ys my life is close to my ideal. 2) The conditions of my life are excellent. 3) I am satisfied with my life. 4) So far I have gotten the important things I want in life. 5) If I could live my life over, I would change almost nothing.
73 Self Concept Clarity Scale 1) My beliefs about myself often conflict with one another. 2) On one day I might have one opinion of myself and on another day I might have a different opinion. 3) I spend a lot of time wondering about what kind of person I really am. 4) Sometimes I feel that I am not r eally the person that I appear to be. 5) like. 6) I seldom experience conflict between the different aspects of my personality. 7) Sometimes I think I know other people be tter than I know myself. 8) My beliefs about myself seem to change very frequently. 9) If I were asked to describe my personality, my description might end up being different from one day to another day. 10) 11) In general, I have a clear sense of who I am and what I am. 12) what I want. Center for Epidemiologic Studies Depression Scale Instructions for questions: Below is a list of the ways you might have felt or behaved. Please tell me how often you have felt this way during the past week. Rarely or none of the time (<1 day) Some or a little of the time (1 2 days) Occasionally or a moderate amount of the time (3 4 days) Most or all of the time (5 7 days) 1. I was bothered by things that don't usually bother me. 2. I did not feel like eating; my appetite was poor. 3. I felt that I could not shake off the blues even with the help of my family or friend s. 4. I felt that I was just as good as other people. 5. I had trouble keeping my mind on what I was doing. 6. I felt depressed. 7. I felt everything I did was an effort. 8. I felt hopeful about the future. 9. I thought my life had been a failure. 10. I fe lt fearful. 11. My sleep was restless. 12. I was happy. 13. I talked less than usual. 14. I felt lonely. 15. People were unfriendly. 16. I enjoyed life. 17. I had crying spells. 18. I felt sad. 19. I felt that people disliked me. 20. I could not get "going ."
74 Rosenberg Self Esteem Scale Please mark the option that best represents how you feel: Strongly Disa gree (1 ) Disa gree (2 ) A gree (3 ) Strongly Agree (4 ) 1. I feel that I am a person of worth, at least on an equal plane with others. 2. I feel that I have a number of good qualities. 3. All in all, I am inclined to feel that I am a failure.* 4. I am able to do things as well as most other people. 5. I feel I do not have much to be proud of.* 6. I take a positive attitude toward myself. 7. On the whole, I a m satisfied with myself. 8. I wish I could have more respect for myself.* 9. I certainly feel useless at times.* 10. At times I think I am no good at all.* = reverse scored Reference Group Identity Dependence Scale Instructions: Below are a number of statements concerning how you may feel about yourself, other males, and your relationships with other males. Read each statement carefully and indicate the degree to which you agree or disagree by circling the corresponding number. There are no answers, and the only right ones are whatever you honestly feel or believe. It is important to answer every item. Use the following key to respond to each statement: 1 = Strongly disagree 3 = Probably Disagree 5 = Agree 2 = Disa gree 4 = Probably Agree 6 = Strongly agree No Reference Group 6) I find it difficult to describe who I am as a man. ected with any group of males. 8) I believe there are no other males who think the way I do about things. 11) Basically I am different from my male friends. 12) I have little in common with most other males. 15) I often wonder whether there are other men like myself. 16) with whom I identify. 19) I am not like most males. 23) Men are confusing to me. 29)
75 Reference Group Dependent 1) I only feel connected with a certain group of males. 4) Most of my social activities are centered aroun d a particular g roup of male friends. 5) My male friends and I all share the same perspective. 9) It is important that I share a particular commonalit y w ith a certain group of males. 14) Others might consider my friends and I a clique. 17) There are only certain ty pes of males with whom I relate. 27) I feel a common bond with my male friends but not so much with ot her males. Reference Group Nondepndent Diversity 2) I understand differences in men. 13) I feel comfortable relating to different types of males. 20) It does not matter to me whether my friends and I are all alike. 22) I find differences in men interesting. 24) I believe there is something wrong with guys who are very different from me, my male friends, and other males like me. ( reverse scored ) 25) I have different types of males as friends. 26) I feel connected with various types of males. Reference Group Nondepndent Similarity 3) I am similar in many ways to all males. 10) I share a common bond with all males. 18) Although males may differ in some ways, we a re essentially all the same. 21) To some degree I identify with all males. 28) Although I feel most similar to some males, I am also similar to all males. 30) I have much in common with most other males.
76 APPENDIX B SELF GUID E DISCREPANCY PROMPT Everyone experiences situations in which their gender is more meaningful and/or no ticeable to them. These situations or context s may consist of particular relationships with others, particular places, and/or social groups in which you may currently be involved. In the b ox below, please list the three most important and/or meaningful situations or contexts in which your gender is most apparent to you. The f ollowing are just some examples : with your mother with a romantic partner at the gym on a date participating in sports around women around men while shopping at a bar or club with your father at work with your brother at a religious institution or ceremony at school talking to your professor while doing household chores or activities at the park with male friends with female friends at your fraternity/sorority with your children Please select three situations that are most important or meaningfu l to you and enter them below. If you do not see an import ant situation/context for you abo ve, you may add your own below:
77 LIST OF REFERENCES Aarseth, H. (2009). From modernized masculinity to degendered lifestyle projects: Changes in 2005. Men and Masculinities 11(4), 424 440. doi:10.1177/10971 84X06298779 effects of gendered social learning in men. Psychology of Men & Masculinity 11(2), 77 90. doi:10.1037/a0018602. Aguinis, H. (2004). Moderated regression New York: Guilford. Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions Newbury Park, CA: Sage. Ashmore, R. (1990). Sex, gender, and the individual. In L.A Pervin (Ed.), Handbook of personality: Theory and r esearch (pp. 486 526). New York: Guilford Press Bandura, A., & Walters, R. H. (1963). Social learning and personality development. New York: Holt, Rinehart and Winston. Bartini, M. (2006). Gender role flexibility in early adolescence: Developmental chan ge in attitudes, self perceptions, and behaviors. Sex Roles 55(3 4), 233 245. doi:10.1007/s11199 006 9076 1 Bem, D. (1967). Self Perception: An alternative interpretation of cognitive dissonance phenomena. Psychological Review 74(3), 183 200. doi:10.103 7/h0024835. Bem, S. (1974). The measurement of psychological androgyny. Journal of Consulting and Clinical Psychology 42 (2), 155 162. doi:10.1037/h0036215. Bigler, M., Neimeyer, G., & Brown, E. (2001). The divided self revisited: Effects of self concep t clarity and self concept differentiation on psychological adjustment. Journal of Social and Clinical Psychology 20 (3), 396 415. doi:10.1521/jscp.20.3.396.22302. Blascovich, J., & Tomaka, J. (1991). Measures of self esteem. In J P. Robinson, P. R. Shav er, & L. S. Wrightsman (Eds.), Measures of personality and social ps ychological attitudes (Vol. 1, pp. 115 160). San Diego, CA: Academic Press. Butler, J. (1997). Performative Acts and Gender Constitution: An Essay in Phenomenology and Feminist Theory. In K. Conboy, N. Medina, S. Stanbury (Eds.) Writing on the Body: Female Embodiment and Feminist Theory (pp. 401 417). New York, NY: Columbia UP. Connell, R. W. (1995). Masculinities. Cambridge, UK: Polity. Connell, R., & Messerschmidt, J. (2005). Hegemon ic Masculinity: Rethinking the Concept. Gender & Society 19(6), 829 859. doi:10.1177/0891243205278639
78 Diener, E ., Emmons, R., Larsen, R., & Griffin, S. (1985). The Satisfaction With Life Scale. Journal of Personality Assessment 49 (1), 71 75. doi:10.1207 /s15327752jpa4901_13. Campbell, J., Trapnell, P., Heine, S., Katz, I., Lavallee, L., & Lehman, D. (1996). Self concept clarity: Measurement, personality correlates, and cultural boundaries. Journal of Personality and Social Psychology 70 (1), 141 156. doi :10.1037/0022 35184.108.40.206 Cohen, J. (1992). A power primer. Psychological Bulletin 112, 155 159. Courtenay, W. (2000). Constructions of masculinity and their influence on men's well being: A theory of gender and health. Social Science & Medicine 50(1 0), 1385 1401. doi:10.1016/S0277 9536(99)00390 1 Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The Satisfaction With Life Scale. Journal of Personality Assessment 49(1), 71 75. doi:10.1207/s15327752jpa4901_13 Diener, E., Oishi, S., & L ucas, R. E. (2003). Personality, culture, and subjective well being: Emotional and cognitive evaluations of life. Annual Review of Psychology 54403 425. doi:10.1146/annurev.psych.54.101601.145056 Donahue, E. M., Robins, R. W., Roberts, B. W., & John O. P (1993). The divided self: Concurrent and longitudinal effects of psychological adjustment and social roles on self concept differentiation. Journal of Personality and Social Psychology 64, 834 846. Watson, G. S., & Durbin, J. J. (1951). Exact tests of serial correlation using noncircular statistics. Annals of Mathematical Statistics 22446 451. doi:10.1214/aoms/1177729592 Erikson, E. H. (1968). Identity: youth and crisis. Oxford England: Norton & Co. Falk, G. (1998). Sex, gender, and social change: T he great revolution. Lanham, MD: University Press of America. Festinger, L. (1957). A theory of cognitive dissonance. Stanford University Press. Field, A. (2005). Discovering statistics using SPSS (2nd ed.). Thousand Oaks, CA US: Sage Publications, Inc. Frazier, P., Tix, A., & Barron, K. (2004). Testing Moderator and Mediator Effects in Counseling Psychology Research. Journal of Counseling Psychology 51 (1), 115 134. doi:10.1037/0022 0220.127.116.11. Freud, S. (1957). Mourning and melancholia. In J. Strac hey (Ed. and Trans.), The standard edition of the complete psychological works of Sigmund Freud (Vol. 14, pp. 243 258). London: Hogarth Press. (Original work published 1917)
79 Gatz, M., & Hurwicz, M. (1990). Are old people more depressed? Cross sectional da ta on Center for Epidemiological Studies Depression Scale factors. Psychology and Aging, 5(2), 284 290. doi:10.1037/0882 7918.104.22.1684 Gergen, K. J. (1991). The saturated self: Dilemmas of identity in contemporary life. New York: Basic Books. Gergen, K. J (1972 ). The healthy, happy human wears many masks. Psychology Today pp. 31 35, 64 66. Goffman, E. (1959). The presentation of self in everyday life. Garden City, NY: Doubleday. Goldberg, H. (1976). The hazards of being male: Surviving the myth of masc uline privilege. Oxford England: Nash. Heriot, J. (1983). The double bind: Healing the split. Women & Therapy, 2(2 3), 11 28. doi:10.1300/J015v02n02_03. Higgins, E. T, Klein, R., & Strauman, T. (1985). Self concept discrepancy theory: A psychological mod el for distinguishing among different aspects of depression and anxiety. Social Cognition, 3, 51 76. Higgins, E., Strauman, T., & Klein, R. (1986). Standards and the process of self evaluation: Multiple affects from multiple stages. Handbook of motivatio n and cognition: Foundations of social behavior (pp. 23 63). New York, NY US: Guilford Press. Higgins, E. T. (1987). Self discrepancy: A theory relating self and affect. Psychological Review, 94, 319 340. Higgins, E. T. (1989). Continuities and disconti nuities in self regulatory and self evaluative processes: A developmental theory relating self and affect. Journal of Personality, 57, 407 444. Hoffman, R. (2001). The measurement of masculinity and femininity: Historical perspective and implications for counseling. Journal of Counseling & Development 79(4), 472 485. Huebner, E., & Dew, T. (1995). Preliminary validation of the positive and negative affect schedule with adolescents Journal of Psychoeducational Assessment 13(3), 286 293. doi:10.1177/0734 28299501300307 Kimmel, M. (1987). Men's responses to feminism at the turn of the century. Gender & Society 1(3), 261 283. doi:10.1177/089124387001003003 Kimmel, M. S. (2008). Guyland: The perilous world where boys become men. New York: Harper
80 Knight, R. G., Williams, S., McGee, R., & Olaman, S. (1997). Psychometric properties of the Centre for Epidemiologic Studies Depression Scale (CES D) in a sample of women in middle life. Behaviour Research and Therapy, 35(4), 373 380. doi:10.1016/S0005 7967(96)00 107 6 Korobov, N. (2005). Ironizing masculinity: How adolescent boys negotiate hetero normative dilemmas in conversational interaction. 13, 225 246. Korobov, N., & Thorne, A. (2006). Intimacy and distancing: Young men's conv ersations about romantic relationships. Journal of Adolescent Research 21(1), 27 55. doi:10.1177/0743558405284035 Kroger, J. (Ed.) (1993). Discussions on ego identity Hillsdale, NJ: Lawrence Erlbaum. Levant, R. (1995). Toward the re construction of masc ulinity. A New Psychology of M en (pp. 229 251). New York, NY US: Basic Books. doi:10.1037/0893 3200.5.3 4.379 masculinity, 1. In M. Lewin (Ed.) In the shadow of the pa st: Psychology portrays the sexes (pp. 179 204). New York: Columbia University Press. Mahalik, J., Cournoyer, R., DeFranc, W., Cherry, M., & Napolitano, J. (1998). Men's gender role conflict and use of psychological defenses. Journal of Counseling Psycho logy 45(3), 247 255. doi:10.1037/0022 022.214.171.124 Marcia, J. (1966). Development and validation of ego identity status. Journal of Personality and Social Psychology 3 (5), 551 558. doi:10.1037/h0023281. Marsiglio, W., Amato, P. R., Day, R. D., & Lamb, M. E. (2000). Scholarship on fatherhood in Journal of Marriage and the Family 62, 1173 1191. McClelland, G. H., & Judd, C. M. (1993). Statistical difficulties of detecting interactions and moderator effects. Psychological Bulletin 114, 376 390. McKelley, R. A., & Rochlen, A. B. (2010). Conformity to masculine norms and preferences for therapy or executive coaching. Psychology of Men & Masculinity 11(1), 1 14. doi:10.1037/a0017224 O'Neil, J. (2008a). Summarizing 25 years of rese arch on men's gender role conflict using the Gender Role Conflict Scale: New research paradigms and clinical implications. Counseling Psychologis t, 36(3), 358 445. doi:10.1177/0011000008317057 O'Neil, J. (2008b). Complexity, contextualism, and multicultu ralism: Responses to the critiques and future directions for the gender role conflict research program. Counseling Psychologist 36(3), 469 476. doi:10.1177/0011000008314781
81 Pavot, W, Fujita, E, & Diener, E. (1997). The relations between self aspect congr uence, personality, and subjective well being. Personality and Individual Differences, 22, 183 191. Radloff, L. S. (1977). The CES D Scale: A self report depression scale for research in the general population. Applied Psychological Measurement 1(3), 385 401. doi:10.1177/014662167700100306 Roberts, R. E., Vernon, S. W., & Rhoades, H. M. (1989). Effects of language and ethnic status on reliability and validity of the Center for Epidemiologic Studies Depression Scale with psychiatric patients. Journal of N ervous and Mental Disease 177(10), 581 592. doi:10.1097/00005053 198910000 00001 Rogers, C. R. (1961). On becoming a person: A therapist's view of psychotherapy. Boston: Houghton Mifflin. Rosenberg, M. (1965). Society and adolescent self image. Princeto n, NJ: Princeton University. Sherif, M., & Sherif, C. W. (1969). Social psychology New York: Harper Row. Silverstein, L., Auerbach, C., & Levant, R. (2002). Contemporary fathers reconstructing masculinity: Clinical implications of gender role strain. Pr ofessional Psychology: Research and Practice 33(4), 361 369. doi:10.1037/0735 7028.33.4.361 Singleton, A., & Maher, J. (2004). The 'New Man' Is in the House: Young Men, Social Change, and Housework. The Journal of Men's Studies, 12(3), 227 240. doi:10.3149/jms.1203.227 Stroink, M. (2004). The Conflicting Standards Dilemma and Gender: A Mediating Model of Its Affective Implications and Coping Styles. The Journal of Social Psychology, 14 4(3), 273 292. doi:10.3200/SOCP.144.3.273 292. Tangney, J., Niedenthal, P., Covert, M., & Barlow, D. (1998). Are shame and guilt related to distinct self discrepancies? A test of Higgins's (1987) hypotheses. Journal of Personality and Social Psychology 75 (1), 256 268. doi:10.1037/0022 35126.96.36.1996. U.S.A. Quick Facts from the U.S. Census Bureau. (n.d.). Retrieved from http://quickfacts.census.gov/qfd/states/00000.html Van Hook, E., & Hig gins, E. T. (1988). Self related problems beyond the self concept: Motivational consequences of discrepant self guides. Journal of Personality and Social Psychology, 55, 625 633. Van Wagoner, S. (2007). Men, masculinity, and competition: Whither the new m an?. Envy, competition, and gender: Theory, clinical applications and group work (pp. 33 57). New York, NY US: Routledge/Taylor & Francis Group. Wade, J. (1998). Male reference group identity dependence: A theory of male identity. The Counseling Psycholo gist 26(3), 349 383. doi:10.1177/0011000098263001
82 Wade, J., & Brittan Powell, C. (2000). Male reference group identity dependence: Support for construct validity. Sex Roles 43 (5 6), 323 340. doi:10.1023/A:1026695209399. Wade, J., & Gelso, C. (1998). Re ference Group Identity Dependence Scale: A measure of male identity. The Counseling Psychologist 26(3), 384 412. doi:10.1177/0011000098263002 Watson, D., & Clark, L. A. (1984). Negative affectivity: The disposition to experience aversive emotional states Psychological Bulletin 96(3), 465 490. doi:10.1037/0033 2909.96.3.465 Watson, D., Clark, L., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psycho logy 54 (6), 1063 1070. doi:10.1037/0022 35188.8.131.523. Weissman, M. M., Prusoff, B., & Newberry P. (1975). Com parison of the CES D with standardized depression rating scales at three points in time. Technical Report, 1975, Yale University, Contract ASH 74 166, National Institute of Mental Health.
83 BIOGRAPHICAL SKETCH Despite a singular focus throughout her adult life on becoming a psychologist, journey to this point has been long and varied and many of the places she stopped or detoured alon g that path have their ghosts in the lines of th is dissertation. Kimberly began her education in psychology and sociology at the University of Florida, focusing on gender studies in the S ociology Department and studying under Dr. Peter Delaney in the are a of cognitive psychology. F ollowing her undergraduat es years, she studied the classic works of disciplines ranging from mathematics to religion worked as a case worker and behavior therapist with dev elop mentally disabled adults. She continued her pursuit of professional work in psych ology at Boston College, where she earned her enta l health c ounseling. In 2008, she returned to her alma mater to study under Dr. Martin Heesacker in the Acceptable Prejudices Lab, where she focused on the contextual responsiveness of gender ed expression from an i nterd isciplinary perspective. She received her Ph.D. from the University of Florida in the fall of 2012 and is currently completing a postdoctor al fellowship in geropsychology at the University of Colorado Aging Center in Colorado Springs, Colorado