Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Masculinity, sexual identity, and the life experiences of American adolescent men Easterbrook, Adam 2013

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

Item Metadata

Download

Media
24-ubc_2014_spring_easterbrook_adam.pdf [ 978.74kB ]
Metadata
JSON: 24-1.0165695.json
JSON-LD: 24-1.0165695-ld.json
RDF/XML (Pretty): 24-1.0165695-rdf.xml
RDF/JSON: 24-1.0165695-rdf.json
Turtle: 24-1.0165695-turtle.txt
N-Triples: 24-1.0165695-rdf-ntriples.txt
Original Record: 24-1.0165695-source.json
Full Text
24-1.0165695-fulltext.txt
Citation
24-1.0165695.ris

Full Text

       MASCULINITY, SEXUAL IDENTITY, AND THE LIFE EXPERIENCES OF AMERICAN ADOLESCENT MEN  by Adam Easterbrook   A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in The Faculty of Graduate and Postdoctoral Studies (Sociology)   THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) December 2013   ? Adam Easterbrook, 2013   ii  Abstract Demonstrating and defending one?s masculinity is an integral part of the adolescent male experience. Adolescent men earn their masculinity by engaging in performances of masculinity, which include concealing emotions, aggression, and athleticism. While men who give convincing performances of masculinity earn status and power, those who are unable to are victimized and ostracized. In this dissertation, I investigate the influence of masculinity on adolescent men?s lives through an analysis of the experiences of adolescents who participated in the National Adolescent Health study (ADD Health). My principle aim is to examine how performances of masculinity relate to important aspects of adolescents? lives. I examine the relationship between men?s performances of masculinity and their sexual identity. Many scholars posit that heterosexuals are more masculine than non-heterosexuals. I argue that these scholars fail to take into account that men?s use of homophobia to police masculinity biases people to erroneously perceive non-heterosexuals as less masculine than heterosexuals. The analysis of the ADD Health data confirms my argument; there is no difference between heterosexual and non-heterosexual men?s masculinity.  I also explore how adolescents? performances of masculinity and sexual identity affect their relationships with peers, life satisfaction, and risk-taking. I examine these life experiences because they strongly influence adolescents? development. The analyses indicate that concealing emotions does not strongly relate to adolescents? life experiences. Athleticism, however, leads to better peer relationships and greater life satisfaction, while aggression has a deleterious effect on peer relationships and life satisfaction as well as increases risk-taking. Sexual identity has almost no influence on life experiences.   iii   These findings have implications for future research. First, adolescent men?s performances of masculinity entail several different sets of behaviours that each uniquely influences life experiences. Consequently, researchers need to consider masculinity as a multidimensional construct. Second, there is no evidence that non-heterosexuals are less masculine than heterosexuals. Scholars exploring sexuality and masculinity must take into account how assumptions about sexuality and masculinity might be impacting their research. Third, performances of masculinity effect several aspects of adolescent men?s lives. Future research must focus on masculinity, and not just biological sex, when attempting to understand men?s experiences.       iv  Preface  Adam Easterbrook identified and designed the research program and analyzed all research data. The data used was from the National Study of Adolescent Health (ADD Health).       v  Table of Contents Abstract ............................................................................................................................................. ii Preface .............................................................................................................................................. iv Table of Contents ............................................................................................................................. v List of Tables ................................................................................................................................ viii List of Figures .................................................................................................................................. ix Acknowledgments ............................................................................................................................ x Chapter 1: Introduction .................................................................................................................. 1 Masculinity during Adolescence .................................................................................................... 3 Masculinity and Sexuality .............................................................................................................. 5 Influence of Masculinity and Sexual Identity on Men?s Life Experiences .................................... 9 Relationships with Peers ........................................................................................................... 10 Life Satisfaction ........................................................................................................................ 11 Risk-taking................................................................................................................................ 12 Adolescence, Masculinity, Sexual Identity, and Life Experiences .......................................... 13 Purpose of the Study .................................................................................................................... 14 Chapter 2: Literature Review and Hypotheses ........................................................................... 16 Conceptualizing and Defining Masculinity.................................................................................. 17 Masculinity and the Subordination of Women ......................................................................... 17 Masculinity and the Subordination of Men .............................................................................. 23 Masculinity during Adolescence is About the Guy Code ........................................................ 29 Sexuality and Masculinity ............................................................................................................ 35 The Influence of Masculinity and Sexuality on Men?s Life Experiences .................................... 41 Life Experiences ....................................................................................................................... 41 How Masculinity Influences Men?s Life Experiences ............................................................. 42 How Sexual Identity Influences Men?s Life Experiences ........................................................ 47   vi  Hypotheses: Influence of Masculinity and Sexuality on Men?s Life Experiences ................... 49 Summary ...................................................................................................................................... 59 Chapter 3: Methods ....................................................................................................................... 60 Study Dataset and Sample ............................................................................................................ 60 Operationalization of the Study Variables ................................................................................... 62 Independent Variables: Dimensions of Masculinity and Sexual Identity ................................ 62 Dependent Variables ................................................................................................................. 70 Control Variables ...................................................................................................................... 74 Analytic Strategy .......................................................................................................................... 76 Chapter 4: Results.......................................................................................................................... 79 Hypothesis 1.1: Sexual Identity and Masculinity......................................................................... 79 Life Experiences ........................................................................................................................... 79 Hypotheses 2.1.1 - 2.2.3: Peer Relationships ........................................................................... 79 Hypotheses 3.1.1 - 3.1.3: Life Satisfaction............................................................................... 85 Hypotheses 4.1.1 - 4.2.3: Risk-taking ...................................................................................... 87 Chapter 5: Conclusion ................................................................................................................... 93 Masculinity, Sexual Identity, and Life Experiences .................................................................... 93 Masculinity is about the Guy Code .......................................................................................... 94 Masculinity and Sexual Identity are not the Same ................................................................... 95 Masculinity and Sexual Impact Men?s Life Experiences ......................................................... 96 Implications of the Findings ....................................................................................................... 104 1. Masculinity Must Be Treated As Multidimensional .......................................................... 104 2. Sexual Identity and Masculinity Are Distinct Concepts .................................................... 106 3. Explorations of Men?s Life Experiences must Include Masculinity .................................. 107 Limitations ................................................................................................................................. 108 Future Research .......................................................................................................................... 111 Summary .................................................................................................................................... 113   vii  References ..................................................................................................................................... 115             viii  List of Tables TABLE 1: Descriptive Statistics for the Variables used to Measure the Dimensions of Masculinity (n=3147) ...................................................................................................................... 66 TABLE 2: Factor Analysis for the Dimensions of Masculinity (n=3147) ...................................... 67 TABLE 3: Descriptive Statistics for the Independent Variables: Dimensions of Masculinity and Sexual Identity (n=3147) ................................................................................................................. 68 TABLE 4: Descriptive Statistics for the Dependent Variables (n=3147)........................................ 72 TABLE 5: Descriptive Statistics for the Demographic and Control Variables (n=3147) ............... 76 TABLE 6: Unstandardized Slope Coefficients (Standard Errors) for the OLS Regression Models of Dimensions of Masculinity Regressed on Sexual Identity .......................................................... 80 TABLE 7: Log Odds (Standard Errors) for the Logistic Regression Models of Perception that Friends Care regressed on Dimensions of Masculinity ................................................................... 81 TABLE 8: Unstandardized Slope Coefficients (Standard Errors) for the OLS Regression Models of Social Desirability Regressed on Dimensions of Masculinity .................................................... 84 TABLE 9: Unstandardized Slope Coefficients (Standard Errors) for the OLS Regression Models of Life Satisfaction  Regressed on Dimensions of Masculinity ....................................................... 86 TABLE 10: Log Odds (Standard Errors) for the Logistic Regression Models of Desire to Take Risks Regressed on Dimensions of Masculinity .............................................................................. 88 TABLE 11: Log Odds (Standard Errors) for the Logistic Regression Models of Frequent Drunkenness Regressed on Dimension of Masculinity ................................................................... 91        ix  List of Figures Figure 1: Conceptual Models ........................................................................................................... 16        x  Acknowledgments I would like to thank Dr. Rima Wilkes, Dr. Richard Carpiano, and Dr. Nathan Lauster for always pushing me to do the best work possible. I also could not have finished without the support of my mom, who is always there for me regardless of what challenges she is facing ? I would not be where I am without your love and encouragement! Thank you to all my friends for offering both good advice and distractions when I was stressed. And finally, thank you Otavio for always encouraging me to be the best person I can be, both in school and in life. I can?t express how much I appreciate all the love and support you give me.        1  Chapter 1: Introduction  In 2012 Granville Island Brewing, a Vancouver based brewery, released pink raspberry ale. In order to promote the new ale, the company initiated a radio ad campaign. The ads featured men calling the pink ale ?man-genta? and described a man entering a party with the ale as a ?man-gentrance.? The company intended this ad campaign to be humourous, but it also exemplifies the importance of masculinity in men?s lives. These ads represent men?s need to be masculine and not feminine. To convince men to drink a pink-coloured beverage, Granville Island Brewing felt it necessary to justify the typically feminine colour by portraying masculine men drinking the ale and referring to its colour as man-genta. Men?s drive to demonstrate their masculinity influences more than just their drink and colour choices. Masculinity is a pervasive influence on men?s life experiences (Courtenay, 2001; Kimmel, 2008; Shoveller, Knight, Johnson, Oliffe, & Goldenberg, 2009).  Masculinity is defined here as the behaviours that men engage in to demonstrate their adherence to ideologies regarding what it means to be a man (Kimmel, 2008). Sociologists have often overlooked the influence of masculinity on men?s lives. That is, scholars have not adequately examined how the behaviours that men engage in to demonstrate their masculinity influence their life experiences. The lack of attention to men and masculinity has led scholars to call men ?the ignored gender? (Harris & Harper, 2010). Gender within the university setting is almost exclusively used in reference to women?s experiences. Most gender-focused university groups and classes, for instance, have as their aim protecting or empowering women (Harris & Harper, 2010). However, any attempt at improving women?s status that does not also address men?s experiences, is failing to take into account the relational nature of gender (Bannon & Correia, 2006; Connell, 2005).    2  The dearth of sociological research examining masculinity among men is particularly notable given the extensive feminist scholarship on femininity (Bannon & Carreia, 2006; Harper & Harris, 2010; Kimmel, 2008). Specifically, while feminist scholars have pushed for the study of women's histories and experiences, including re-examinations of femininity, the same cannot be said for men?s histories and experiences. An outcome of the focus on women's experiences and liberation is an increase in the roles that women can enact, which now ranges from athletes to stay-at-home mothers (Kimmel, 2008). In contrast, when asked what it means to be a man, men typically give a restrictive list of what men ought to be, such as being independent and strong.  Scholars must engage in knowledge generation in order to help liberate men from the restrictive ideologies surrounding masculinity, in much the same way that feminism has challenged notions of femininity.  When scholars examine men?s experiences, they often assume that all men are masculine. Specifically, researchers tend to equate sex with gender, thereby implying that men are a homologous and masculine group (Bottorff, Oliffe, & Kelly, 2012; Schrock & Schwalbe, 2009). Sex is biological and based on physical features, such as genitalia or chromosomes, which determine biological status as a man or a woman. Though the biologically based man/woman dichotomy is problematic, sex remains one of the most salient features individuals notice about others. In contrast, gender refers to the behaviours that men and women engage in that determine masculinity or femininity. Playing with dolls, for instance, indicates femininity while playing sports indicates masculinity. Within Western culture, it is typically assumed that there is a match between sex and gender in the sense that boys and men are masculine while girls and women are feminine. In reality, however, this matching does not always occur. Elton John and Dwayne Johnson are both men, yet they are not perceived to be equally masculine. Thus, when   3  researchers equate sex with gender they fail to take into account differences in men?s enactment of masculinity.  Several key areas within the literature on masculinity have not received adequate research attention. In the following sections, I will discuss three areas that require further study. First, researchers must develop a clear definition of masculinity that applies to the unique experiences of adolescents. Second, sexual identity is assumed to influence masculinity (Bailey & Zucker, 1995; Udry & Chantala, 2002; Whitam & Zent, 1984), but this taken-for-granted association needs to be critically re-examined. Third, the influence that masculinity and sexual identity have on men?s life experiences must be empirically investigated. Masculinity during Adolescence  Adolescence is one of the most salient periods of the life course regarding the enactment and enforcement of masculinity (Frosh, Phoenix, & Pattman, 2002; Kimmel, 2008; Mac an Ghaill, 1994; Morrell, 1998; Pascoe, 2007). High school operates as a ?boot camp? for masculinity since there are increased expectations regarding masculinity as well as severe repercussions for those who do not adequately demonstrate their masculinity (Kimmel, 2008). Men learn the expectations concerning what it means to be masculine through socialization into the ideologies of the ?Guy Code? (Kimmel, 2008). There are various ideologies within the Guy Code, such as the ideas that men should avoid femininity, be successful, have power, be self-reliant, be independent, and show aggression. Masculinity can be conceptualized as how close or far men?s behaviours are from the ideologies of the Guy Code (Kimmel, 2008). Specifically, masculinity refers to the practices and behaviours that men engage in to show their support of the ideologies of the Guy Code. The more of these behaviours that men exhibit, the more masculine they are. However, the behaviours through which men demonstrate their adherence to the Guy   4  Code are dependent on developmental context and thus the behaviours change over the life course. Adolescents demonstrate their adherence to these ideologies through behaviours such as playing sports or getting into fights, while adult men show their adherence through financial independence or watching sports (Brannon, 1976; Kimmel, 2004, 2008; Pascoe, 2007; Schrock & Schwalbe, 2009).  Many adolescents report that rather than supporting the Code because they believe it represents how men ought to act, they support the Guy Code out of fear of the numerous negative outcomes that being labelled as unmasculine leads to (Kimmel, 2004; 2008; Pascoe, 2007). Thus, an important means through which masculinity is encouraged is by monitoring and policing men?s performances of masculinity (Kimmel, 2004, 2008; Pascoe, 2007). Homophobic labels, such as calling a peer ?gay? or a ?fag,? are used to police masculinity through ostracizing and victimizing men who fail to conform to the Guy Code (Kimmel, 2008; Pascoe, 2007). The policing of masculinity therefore supports the Guy Code through empowering those who abide by it and disempowering those who do not.  Adolescent men experience anxiety and fear as to their own ability to be masculine (Kimmel, 2008). The expectations of the Guy Code are restrictive and unachievable since adolescent men cannot always be the top athlete or conceal their emotions. Moreover, these expectations, and the behaviours they encourage, lead adolescent men to isolate themselves when they need help the most, which increases feelings of distress (Brannon, 1976; Kimmel, 2004, 2008). Though the transition through adolescence and into adulthood is challenging, men are expected to deal with these stresses without support from others. Overall, the combination of the policing and monitoring of masculinity, restrictive and unachievable expectations, and the fear of being deemed unmasculine, make the transition from boyhood to manhood fraught with anxiety   5  and distress (Kimmel, 2004, 2008). Some young men compensate for their feelings of inadequacy by engaging in risky or immature behaviours, such as getting drunk, having casual sex, or playing videogames (Kimmel, 2008).  Along with the increased pressure to be masculine, adolescent men are also in the midst of important developmental tasks such as identity development (Erikson, 1963) and the creation of increasingly intense peer relationships (Parker, Rubin, Erath, Wojslawowicz, & Burkirk, 2006). It is likely that adolescents? performances of masculinity will have a profound impact on development since they will influence the foundations of their identity as well as their peer relationships. Adolescent men who are masculine, for instance, will likely develop more positive identities and have more successful peer relationships than those who are less masculine.     Though researchers are beginning to examine masculinity during adolescence and the transition into young adulthood (e.g., Kimmel, 2004, 2008; Pascoe, 2007), there is still very little integration with previous works on masculinity concerning how to conceptualize and define masculinity. Without a clear definition of masculinity, it is impossible to operationalize masculinity in a way that allows researchers to systematically study adolescent men?s performances of masculinity. In this dissertation, I integrate seminal works on masculinity in order to develop a definition, and ultimately a measure, of masculinity.  Masculinity and Sexuality Men?s sexual identity has a profound impact on their lives and is intimately connected with masculinity. Within mainstream culture, it is taken-for-granted that gay men are less masculine than heterosexual men. The flamboyant gay characters in blockbuster movies such as Mean Girls, As Good as it Gets, The Bride of Chucky, and My Best Friend?s Wedding all seem to prefer the company of women over men. In fact, many of the gay characters in movies and on   6  television are depicted more like girl-friends than as male friends. The idea that heterosexual men are more masculine than gay men has also been explored in academic research, which tends to find that gay men are, in fact, less masculine than their heterosexual counterparts. Gay men, compared to heterosexual men, report more childhood gender nonconformity (see review by Bailey & Zucker, 1995; Whitam & Zent, 1984). Male youth who report same-sex attraction are more feminine than those who report opposite-sex or both-sex attraction (Udry & Chantala, 2002) 1. The connection between masculinity and sexuality is so pervasive that parents assess their children?s level of gender nonconformity when unsure of their child?s sexual identity (D'Augelli, Grossman, & Starks, 2005).  There are methodological problems with the research that examines the relationship between masculinity and sexual identity. These studies often use retrospective accounts (e.g., Whitam & Zent, 1984), which can lead to recall biases. Given that gay men are assumed to be less masculine than heterosexual men, it is likely that gay men will be biased to recall their behaviours as more feminine due to the self-fulfilling prophecy (Gottschalk, 2003; Ross, 1980). Gay men?s tendency to assume that they are less masculine than heterosexual men is evident in their use of the colloquial term ?root.? A root is a feminine behaviour that gay men engaged in as a child that should have made it clear to others that they were (or would become) gay, such as playing with dolls or liking the colour pink. Studies that examine the relationship between sexuality and masculinity also often use samples in which respondents? sexuality is not clear.                                                               1 As measured based on questions that predict whether the respondent is male or female.   7  Udry and Chantala (2002), for instance, based their work on a sample that never self-identified as gay or bisexual. Instead, sexual identity was inferred by sexual attraction or partner choice. It is not clear if sexual attraction and partner choice accurately represent adolescents? own sexual identity. These methodological problems are compounded by researchers? (e.g., Udry & Chantala, 2002) failure to clearly define masculinity. Without an explicit definition of masculinity, it is unclear how masculinity is operationalized and whether there is comparability across cases or across studies.  Though the methodological problems inherent in these studies are a concern, the most pressing issue with this scholarship is theoretical. Most researchers have not addressed the socially constructed relationship between men?s performances of masculinity and sexual identity, including how this relationship is predicated on inequality. Specifically, heterosexuality is linked to masculinity while homosexuality is linked to being less masculine (and more feminine), which validates and empowers heterosexuals and marginalizes non-heterosexuals. The perceived connection between masculinity and sexual identity develops from what Pascoe (2007) calls ?the spectre of the fag.? By using homophobia as a way to police performances of masculinity (Kimmel, 2004, 2008; Pascoe, 2007), men come to fear homosexuality since it is seen as the antithesis of masculinity. Over time this leads to the spectre of the fag, which is the fear of engaging in any behaviours that will lead one to be deemed unmasculine and thus called a fag (see also Kimmel, 2004, 2008). The fear of this label is so strong that many men develop negative attitudes towards non-heterosexuals. Heterosexual men, for instance, tend to be much more hostile towards gay men than women are (Davies, 2004). This is a result of men?s need to both demonstrate (Connell, 1995; Kimmel, 2004) and defend (Kimmel, 2004; Wilkinson, 2004) their masculinity against the spectre of the fag.    8  The connecting of masculinity to sexual identity leads to the creation of stereotypic assumptions concerning the behaviours of heterosexual and non-heterosexual men. These assumptions are not representative of all gay or heterosexual men. However, they lead others to view highly effeminate gay men as the norm, while contrary examples of masculine gay men are ignored or minimized. The process of seeking affirming examples while ignoring contrary examples of stereotypes has been well documented via research on confirmatory hypothesis testing (e.g., Fiedler, Armruster, Nickel, Walther, & Asbeck, 1996; Snyder & Gangestad, 1981). Over time, confirmatory hypothesis testing leads to the belief that there is a biological or innate reason why gay men are less masculine. In fact, research that examines masculinity among different sexual identity groups often implicitly or explicitly assumes that differences are rooted in biological mechanisms (e.g., Bailey & Zucker, 1995; Bailey, 2003).  Research must critically re-examine the relationship between masculinity and sexual identity in a way that avoids many of these methodological and theoretical problems. The re-examination of this relationship can be accomplished through using sociological theory on masculinity and sexuality as a basis for research. Researchers must also move beyond simplistic comparisons of heterosexual and non-heterosexual men?s performances of masculinity. In particular, researchers should examine how masculinity and sexual identity interact and lead to different outcomes.  For instance, does being a less masculine heterosexual man have the same impact on peer relationships as being a less masculine gay man? In this dissertation, I aim to critically re-examine the relationship between masculinity and sexual identity and explore how masculinity and sexual identity influence adolescent men?s life experiences.     9  Influence of Masculinity and Sexual Identity on Men?s Life Experiences Though researchers are beginning to explore masculinity and sexuality during adolescence (e.g., Kimmel, 2004, 2008; Pascoe, 2007), there is still very little specific theory and research on the influence of masculinity and sexual identity on adolescent men?s life experiences. Research indicates that masculinity does influences men?s life experiences. Compared to masculine heterosexual men, less masculine heterosexuals are more likely to have health issues (Courtenay, 2001), eating pathology (Munren & Smolak, 1997), body/image dissatisfaction (Coates & Person, 1985), problems with peers (Coates & Person, 1985; Rekers, Bentler, Rosen, & Lovas, 1979), and depression, anxiety and distress (Harry, 1983; Kimmel, 2008). Concerning non-heterosexual men, less masculine gay and bisexual men are at risk for various negative outcomes such as bullying, victimization, parental rejection, emotional problems, and suicide (D'Augelli et al., 2005; Friedman, et al., 2006; Udry & Chantala, 2002). Overall, for both heterosexual and non-heterosexual men, being masculine increases positive life experiences while being less masculine increase negative life experiences.  The relationship between masculinity and life experiences is likely influenced by adolescents? sexual identity. Specifically, being gay or bisexual is associated with various negative outcomes, such as substance abuse, depression, isolation, and victimization (Banks, 2003). Moreover, homophobia is used to police adolescents? performances of masculinity, thereby illegitimating gay men?s performances of masculinity. Consequently, sexual identity may influence life experiences as well as moderate the relationship between masculinity and life experiences. This dissertation will explore the impact of masculinity and sexual identity on three life experiences: adolescents? relationships with their peers, life satisfaction, and risk-taking. Each of these are discussed below.    10  Relationships with Peers The first life experience that I examine is adolescents? relationships with their peers. In 1902, Charles Cooley suggested that peers provide a context for social and emotional development. Over a century later, researchers have confirmed that the development and maintenance of peer relationships is one of the most significant developmental tasks during adolescence (Parker, Rubin, Erath, Wojslawowicz, & Burkirk, 2006). This is because the experiences and skills gained while interacting with peers are a foundation for future relationships (Berndt, 1982; Furman, 1999). As such, adolescents who experience peer rejection are unable to develop the social skills needed for both platonic and romantic relationships since they cannot catch up to their peers? social skills development (Bierman, 2003; Schwartz, Dodge, & Coie, 1993). Over time, this leads rejected adolescents to develop negative internal understandings of both themselves and others, which negatively influences their developing sense of self and beliefs about the social world (Bierman, 2003; Kupersmidt, Coie, & Dodge, 1990).  Given that adolescents? experiences with peers influence the foundations of their future relationships, it is vital to understand what factors may increase or decrease adolescents? likelihood of being either accepted or rejected by their peers. Adolescents? performances of masculinity and their sexual identity will influence their peer relationships. Specifically, masculine adolescents will experience more acceptance while less masculine adolescents will experience more rejection. Heterosexual men who are less masculine, for instance, experience more peer abuse, isolation, and ostracism compared to their more masculine counterparts (Coates & Person, 1985; Kimmel, 2008; Rekers, Bentler, Rosen, & Lovas, 1977). Similarly, less masculine gay and bisexual men experience more bullying, victimization, and ostracism than   11  masculine gay and bisexual men (Friedman, et al., 2006; Remafedi, Farrow, & Deisher, 1991; Udry & Chantala, 2002). At the same time, sexual identity will influence the relationship between masculinity and peer relationships since gay and bisexual adolescents, compared to heterosexual adolescents, experience more peer rejection (Banks, 2003). Therefore, sexual identity may directly influence peer rejection as well as moderate the relationship between masculinity and sexuality such that those who are both unmasculine and non-heterosexual will experience the most negative outcomes (Pascoe, 2007). Life Satisfaction  The second life experience that I explore is adolescents? life satisfaction. Researchers have not directly examined how masculinity influences life satisfaction during adolescence and young adulthood among gay, bisexual, and heterosexual youth. However, it is likely that masculinity influences overall life satisfaction by affecting various facets of men?s lives. For both heterosexual and non-heterosexual men, being more masculine increases positive outcomes while being less masculine increases negative outcomes (e.g., Courtenay, 2000; D'Augelli et al., 2005; Kimmel, 2008; Pascoe, 2007). For instance, masculine men tend to be more socially desirable while less masculine men experience victimization and ostracism. The sheer range of outcomes associated with masculinity means that life satisfaction, which is a broad assessment of how fulfilled individuals are with their lives as a whole, is an ideal outcome to examine since it tests if masculinity is a pervasive influence on men?s lives. Given that gay and bisexual adolescents experience more negative life experiences than heterosexual adolescents do (Banks, 2003), it is expected that adolescents? sexual identity will positively influence life satisfaction and that sexual identity will moderate the relationship between masculinity and life satisfaction.    12  Risk-taking  The third life experience that I examine is adolescents? engagement in risk-taking. Masculinity may influence adolescents? life experiences through influencing their desire to engage in risk-taking. Courtenay (2001) attempts to explain men's higher rates of morbidity and mortality by arguing that the ways that men attempt to embody masculinity has a deleterious impact on their health. Specifically, he argues that the behaviours that men engage in when they attempt to demonstrate their masculinity are typically behaviours that increase health risks, such as binge drinking, over eating, or drug use. At the same time, the behaviours of women and less powerful men are rejected, which includes the rejection of many health-promoting behaviours such as going to the doctor or asking for help. The paradox is that the more unmasculine men are, the more likely they are to engage in risky behaviours since they are attempting to earn back their masculinity (Courtenay, 2001; Kimmel, 2008). That is, all men take risks but the further men are from the ideals of masculinity, the more risky behaviours that they will engage in as a means of earning their masculinity. Sexual identity relates to risk taking since gay and bisexual men?s masculinity is subordinate to heterosexual men?s masculinity (Connell, 1987; Courtenay, 2001). Given their lower status and power, gay and bisexual men, compared to heterosexual men, must engage in riskier behaviours in order to prove their masculinity.   In summary, adolescent men?s performances of masculinity and sexual identity will influence their life experiences. However, there is a lack of theory about the process through which performances of masculinity influences life experiences. I address the lack of theory through using existing theories about presentations of self (Goffman, 1959), labelling (Link et al., 1989), and stigma (Goffman, 1963; Link & Phelan, 2001) to develop a framework for understanding the processes through which masculinity influences life experiences among   13  adolescents with different sexual identities. I use this framework to derive hypotheses concerning how masculinity influences adolescents? relationships with peers, life satisfaction, and risk taking. Adolescence, Masculinity, Sexual Identity, and Life Experiences Adolescence and the transition into adulthood is an ideal period in which to examine how sexual identity influences performances of masculinity and life experiences. The most developmentally significant processes occurring during this period are identity development (Erikson, 1963) and the development of peer relationships (Parker et al., 2006). These developmental tasks are particularly challenging for gay and bisexual adolescents. During adolescence, gay and bisexual adolescents begin to accept and develop their non-heterosexual identities (Cass, 1984). Having to develop a non-heterosexual identity puts non-heterosexual adolescents at a disadvantage compared to heterosexual adolescents due to challenges with identity development, such as identity conflict and concealment (Cass, 1979, 1984; D'Augelli, 2006, Morrow, 1993). Compared to heterosexuals, gay and bisexual adolescents are also more likely to experience negative interactions with their peers, teachers, and parents (Banks, 2003; D'Augelli, Hershberger, & Pilkington, 1998; Gibson, 1989; Hammelman, 1993; Lindle, Walseman, & Carter, 2011; Mihalik, 1991; Remafedi, French, Story, Resnick & Plum, 1998; Savin-Williams, 2001; Sutton, 1994; Thompson & Johnson, 2003; Weston, 1995). Moreover, many gay and bisexual adolescents have very little support available to them since they are not out to their parents and have not entered the gay community (D?Augelli, 2006; Morrow, 1993). These challenges that gay and bisexuals experience coincide with experiencing increasingly greater demands concerning their own performances of masculinity (Kimmel, 2008). Due to the various unique challenges experienced by gay and bisexual adolescent men, adolescence is an   14  ideal life period in which to explore sexual identity and how it relates to adolescents? performances of masculinity and life experiences.  Purpose of the Study  In order to address the issues discussed above, this dissertation focuses on three interrelated areas. First, I develop a clear conceptualization and definition of masculinity during adolescence. Through reviewing and integrating seminal works on masculinity (e.g., Connell, 1987, 1996; Kimmel, 2004, 2008; Pascoe, 2007), I argue that masculinity can be defined as the behaviours men engage in while interacting with others that show their support for the ideologies of the Guy Code. During adolescence, support for the Guy Code is demonstrated through performances of masculinity that involve concealing emotions, aggression, and athleticism. Adolescents who engage in these behaviours are deemed to be masculine while those who do not engage in these behaviours are deemed to be unmasculine.  Second, I re-examine the taken-for-granted association between masculinity and sexual identity. By discussing sociological literature on sexuality and masculinity, I bring into question the assumption that heterosexuals are more masculine than non-heterosexuals. Sexual identity and masculinity appear to be connected is because homophobia is used as a tool to police men?s masculinity. By using homophobia to police masculinity, people develop biases concerning heterosexual and non-heterosexual men?s performances of masculinity. These biases then influence perceptions of the social world in a way that affirms the belief that heterosexuals are more masculine than non-heterosexuals. Through developing this argument, I will create and test a hypothesis concerning the relationship between men?s sexual identity and their performances of masculinity.    15  Third, I explore the influence that sexual identity and masculinity have on men?s life experiences. Though researchers are beginning to examine masculinity during adolescence and the transition into young adulthood (e.g., Kimmel, 2004, 2008; Pascoe, 2007), there is still very little specific theory as to how differences in the performance of masculinity may lead to different life experiences. A framework is required to explain how adolescents? adoption and enactment of the Guy Code influences their life experiences. I develop such a framework by using Goffman's (1959) theory of presentation of self as well as labelling theory (Link et al., 1989) and work on stigma (Goffman, 1963; Link & Phelan, 2001). From this framework, I formulate hypotheses concerning how masculinity and sexual identity influence adolescents? relationships with their peers, life satisfaction, and risk-taking.  Through systematically addressing these three interrelated goals, I will have defined masculinity as a multidimensional construct, re-examined the relationship between sexual identity and masculinity, and developed and empirically tested hypotheses concerning how masculinity and sexual identity influence adolescents? life experiences. This research builds upon the existing literature on masculinity through addressing several areas that require further investigation.    The following chapters will detail this endeavour. Chapter 2 develops and expands the three aims presented above. This includes the development of hypotheses concerning the relationship between sexuality and masculinity as well as the influence that masculinity and sexual identity have on adolescents? life experiences. Chapter 3 explains the methodology and analytic strategy that will be used to test these hypotheses. Chapter 4 details the results of the analyses. Chapter 5 is the concluding chapter, which will discuss the findings and their implications, acknowledge limitations of the current study, and present ideas for future research.      16  Chapter 2: Literature Review and Hypotheses The aim of this dissertation is to examine how performances of masculinity influence the lives of gay, bisexual, and heterosexual men during adolescence and the transition into adulthood. This chapter is divided into three sections. In the first section, I summarize and integrate works on masculinity in order to develop a clear conceptualization and definition of masculinity during adolescence.  In the second section, which is represented by conceptual model A in Figure 1, I explore the relationship between sexual identity and masculinity. Specifically, I use existing sociological literature to re-examine the taken-for-granted assumption that sexual identity influences masculinity. This discussion leads to the development of a hypothesis concerning the relationship between sexual identity and adolescent men?s performances of masculinity.  Figure 1: Conceptual Models   In the third section, I discuss the influence of masculinity and sexual identity on men?s life experiences. Following this discussion, I develop hypotheses concerning how masculinity and sexual identity affect adolescent men?s life experiences. These include predictions concerning both the direct (conceptual model B in Figure 1) and interactive (conceptual model C   17  in Figure 1) influence that masculinity and sexual identity have on life experiences. The focus of this section is on the independent variables rather than the dependent variables. That is, the main interest is on whether masculinity and sexual identity influence adolescent men?s life experiences, rather than the specific life experiences that are examined. However, the specific outcomes examined represent important life experiences during adolescence and are theorized to be affected by masculinity and sexual identity. Conceptualizing and Defining Masculinity In the following section, I conceptualize and define masculinity during adolescence. I discuss two broad research areas within the study of masculinity. The first research area focuses on how masculinity allows for the perpetuation of patriarchy through the subordination of women. This scholarship tends to equate being a man with masculinity, thereby assuming that the majority of men benefit from the power and status masculinity offers. The second research area focuses on how men?s attempts to demonstrate their masculinity lead to the subordination of less masculine men. After discussing these literatures, I proceed to develop a multidimensional definition of masculinity that links ideologies to behaviours.   Masculinity and the Subordination of Women Prior to the 1980s, much of the scholarship on masculinity was based on the sex role theory (Schrock & Schwalbe, 2009). This theory assumes that people internalize and enact norms about the appropriate roles for men and women (Schrock & Schwalbe, 2009). Thus, men and women?s behaviours are a product of norms concerning what are, and what are not, acceptable roles. Since these sex roles represent how people ought to behave, they are believed to be both complementary and necessary according to the sex role theory. Specifically, men and women behave as they do in order to support one another. When women enact the role of stay-at-  18  home mother, men are able to enact the role of breadwinner. This clearly demarcates what each member of couples ought to do in order for the institution of the family to function. These roles are held up as ideals, whereby deviations from them are seen as unhealthy or as a risk to the stability of social functioning.  Many researchers argue against the sex role theory since it uses functionalist arguments to explain men and women?s sex roles (Carrigan, Connell, & Lee, 1985; Connell, 1987). Specifically, sex role theory does not take into account the power dynamics that exist both between, and within, the sexes, and how these power differences influence behaviours. Sex role theory assumes that men and women?s respective roles and behaviours are natural and functional, rather than a product of inequality. As such, sex role theory fails to take into account the various ways that social structure creates and maintains inequality between the sexes.  Sociologist Raewyn Connell?s (1987) social constructionist theory of hegemonic masculinity is a response to the criticisms of sex role theory. According to the theory of hegemonic masculinity, gender refers to the practices and behaviours that define men and women as masculine or feminine. Gender is a verb because it is enacted through practices, rather than fixed or passively internalized through norms (Connell, 1987, 1996). In particular, the theory argues that men act in ways that are masculine because this gives them power via patriarchy. Men?s enactment of masculinity therefore influences social structures, such as the family, in a way that allows for the reproduction of patriarchy through granting men power and status over women. Thus, men enact the role of breadwinner to gain power and status over women, and consequently maintain patriarchy. By viewing gender as enacted, the theory of hegemonic masculinity is able to take power into account when examining masculinity.    19  The idea that power is a key element in the enactment of masculinity is the cornerstone of the theory of hegemonic masculinity (Connell, 1987, 1995). Specifically, dominance and power are attained through hegemonic masculinity, defined by Connell as ?the configuration of gender practice which embodies the currently accepted answer to the problem of the legitimacy of patriarchy, which guarantees (or is taken to guarantee) the dominant position of men and the subordination of women? (Connell, 1995, p. 77). Stated another way, hegemonic masculinity is defined as the means through which men obtain and maintain power and status over others, particularly women, through the support and propagation of patriarchy as natural and normative (Carrigan, Connell, & Lee, 1985; Connell, 1987). Connell?s use of the term hegemony is important, since it refers to the maintenance of power by a group over others through consent and persuasion, rather than force per se (Steer, 2001).  Most men are unable to fully embody hegemonic masculinity, yet the hegemonic ideal allows men to maintain power through being supported and promoted as a cultural ideal (Edley & Wetherell, 1997). Though not all men are heterosexual, for example, the hegemonic ideal of heterosexuality remains strong within society and offers a template from which some non-heterosexual couples organize their relationships. Lesbians embodying roles such as ?butch? versus ?fem,? for example, represent more masculine versus more traditionally feminine roles. Therefore, hegemonic masculinity maintains its legitimacy through being the taken-for-granted ideal. Most men, however, cannot enact hegemonic masculinity. Instead, they support its ideals by engaging in ?complicit masculinity.? Complicit masculinity allows men to benefit from patriarchy through supporting the ideals of hegemonic masculinity without having to fully embody it (Connell, 1987). Watching sports and action movies, for instance, supports the hegemonic ideals of strength and power without having to be strong or powerful. Therefore,   20  most men support hegemonic masculinity without fully enacting it, which grants men power through normalizing patriarchy and inequality between men and women.  The masculinities of men who do not embody hegemonic or complicit masculinity are called ?subordinate masculinities? (Connell, 1987). These masculinities are less legitimate than other masculinities and therefore have comparatively less power and status. However, subordinate masculinities still have more power and status than femininity. Gay men are an example of a group of men who have a subordinate masculinity. The subordination of gay men?s masculinity occurs through the marginalization and discrimination of gay men (Banks, 2003; Carrigan, Connell, & Lee, 1985; Kimmel, 2008). This subordination occurs because same-sex desire between men challenges heterosexuality, which is problematic since heterosexuality is vital for the reproduction of patriarchy. Moreover, according to modern hegemonic masculinity, women are sexual objects (Carrigan, Connell, & Lee, 1985). Sex between men challenges this view through sexualizing and eroticizing the male body while desexualizing the female body. The subordination and illegitimization of gay men?s masculinity sustains heterosexuality and patriarchy.  Connell?s argument that relationships are dependent on, and explained by, hegemonic masculinity and the support of patriarchy, is one of her major contributions to the study of masculinity (Demetriou, 2001). Specifically, the inclusion of power in theorizing about masculinity was an important move away from the essentialist notions of gender found in the sex role theory. In fact, Connell's work on hegemonic masculinity spurred an increase in the study of masculinity, often in relation to how masculinity subordinates women and supports patriarchy.  Though Connell?s theory is an important contribution to the study of masculinity, there are several criticisms of hegemonic masculinity as a concept. Researchers studying hegemonic   21  masculinity tend to equate masculinity with sex (i.e., being a biological man) and then focus on how masculinity (being a man) relates to various outcomes (Schrock & Schwalbe, 2009). This research, and hegemonic masculinity as a concept, takes the category of men for granted, rather than viewing it as emerging from practices that are embedded with meaning2 (Schrock & Schwalbe, 2009). Put another way, much of the work on hegemonic masculinity takes for granted that most men have power and status and therefore equates being a man with having power.  The criticism that hegemonic masculinity equates masculinity with being a man has lead researchers to argue for the use of interactionist frameworks when theorizing about masculinity (Schwalbe, 2005). These scholars posit that men earn their masculinity through interacting with others in ways that support the masculine ideal. As such, the focus should be on the processes through which men subordinate others through their practices and interactions (Schrock & Schwalbe, 2009). The practices that define masculinity change and develop over time and context. Moreover, factors related to both the self and the audience influences the practices that men engage in to demonstrate their masculinity. In contemporary society, the essential element of masculinity is the ability to exert control over the self and others (Johnson, 2005). Consequently, masculine practices are attempts to claim privilege, elicit deference, and resist exploitation (Schrock & Schwalbe, 2009). Thus, unlike hegemonic masculinity, which links                                                               2 Connell and Messerschmidt (2005) acknowledge that the attempt to explain masculinities based on the dominance of men over women is too simplistic.    22  masculinity to men, Schrock and Schwalbe argue that the male and female body types are irrelevant except insofar as a male body operates as a symbolic asset for performances of masculinity. The idea that the body and masculinity are separate allows scholars to move from a focus on biological sex to a focus ?on what males do to create, maintain, and claim membership in a dominant gender group? (pp. 281).  The argument that gender and sex are different is critical. Having a male body should not be equated with being masculine. Even though being a man can help legitimate performances of masculinity, it does not necessarily guarantee masculinity. The realization that body and masculinity are separate allows for the development of a more agential understanding of masculinity than is found in both sex role theory and the theory of hegemonic masculinity. Moreover, it brings masculinity from a macro focus on societal structures to a micro level focus on interactions. The change from macro to micro allows researchers to specifically measure and examine masculinity through men?s interactions and performances. The theory of hegemonic masculinity has also received criticism for its argument that there are multiple masculinities (Schrock & Schwalbe, 2009). In particular, research on hegemonic masculinity has moved from developing a theoretical understanding of masculinity and how it relates to power and inequality, to cataloguing the various ways different groups of men perform masculinity. While thinking about masculinities, rather than masculinity, allows for the examination of the various ways that men can enact masculinity, it is problematic to ignore what these practices, no matter how different, have in common (Schrock & Schwalbe, 2009). As Schrock and Schwalbe (2009) argue, the commonality of these practices lies in men?s attempts to exert and resist control. Thus, in creating countless categories of masculinities scholars are missing the fact that these practices are being used to maintain control over both men and   23  women. Focusing on the practices that all men engage in to create inequality allows for the development of theories of masculinity in general, including how masculinity influences the lives and life experiences of men (Jefferys, 2005; McCarry, 2006; Schrock & Schwalbe, 2009).  In the next section, I discuss an emerging literature that addresses the concern that researchers should return to understanding masculinity as developing through men?s interactions with others. Specifically, the focus is on men?s performances of masculinity, including how masculinity empowers some men while simultaneously disempowering others. Attention shifts from how masculinity leads to the subordination of women to how performances of masculinity lead to the subordination of men.  Masculinity and the Subordination of Men The research discussed thus far focuses on how masculinity allows men to have power and control over others, especially women (Carrigan et al., 1985; Connell, 1987; Schrock & Schwalbe, 2009). This ignores the fact that masculinity leads to anxiety, distress, and the subordination of many men (Seidler, 2002). An emerging literature spurred on by sociologists Michael Kimmel (2004; 2008) and Cheri Jo Pascoe (2007), examines how men?s performances of masculinity lead to differences in status and power between men. In his book Guyland, Kimmel (2008) interviews over 400 American men and some women between the ages of 16 and 26. His research explores how masculinity influences the transition from adolescence into adulthood. Kimmel (2004) has also extensively theorized and researched about how masculinity is often policed using homophobia. In her book, Dude, You?re A Fag, Pascoe (2007) discusses her ethnographic study of adolescent men in an American high school. Her focus is on how adolescent men demonstrate and maintain their masculinity among their peers, including the role that homophobia plays in evaluating men?s performances of masculinity.   24  A consistent finding in this research is that men fear being emasculated or deemed unmasculine by other men (Kimmel, 2004, 2008; Pascoe, 2007). An important part of the emasculation process involves using ?inferior men? - typically gay men - to shame men who are perceived to be unmasculine (Kimmel, 2004; 2008; Pascoe, 2007). As such, men often police one another's performances of masculinity using homophobic insults, such as ?fag.? The policing of masculinity leads men to be on guard concerning both their own and others? performances of masculinity. The policing and enforcement of masculinity leads to the paradox whereby even though men have power, they often claim that they do not feel powerful. This occurs because men fear they are not measuring up to other men?s expectations of masculinity (Kaufman, 1993; Kimmel, 2004). Kimmel uses an analogy of a chauffeur to explain this point. He argues that the chauffer appears to have power by dressing well, driving the vehicle, and knowing where he is going. That is, the chauffer is engaging in the behaviours that would lead others to assume he has power and status. The reality, however, is that he is just taking orders from someone else who is, often, a man. Therefore, men seem to have power, yet are actually just following what expectations of them concerning their performances of masculinity. In reality, few men actually gain significant power and status, with most men competing amongst one another for power. Inequalities between men based on their performances of masculinity have not received adequate academic attention. The women?s movement, for instance, has increased the range of acceptable behaviours that women can engage in while interacting with others. In contrast, such a movement has not occurred for men concerning more flexibility in what are acceptable performances of masculinity (Kimmel, 2008). Kimmel (2008) argues that compared to women, men's lives are structured around relationships of power. Power, defined here, is the ability to influence the behaviours of other people through either force or hegemony. Very few men   25  actually meet the rules of manhood and very few men have significant power. As a result, men strive for more power, rather than support feminists? efforts to make power more equitable. Moreover, men experience many restrictions on the behaviours that are deemed acceptable, thus they must often limit their own choices so that they can appear to be masculine. Unfortunately, few researchers have examined how the drive to be masculine creates inequality among men (notable exceptions are Pascoe, 2007 and Kimmel, 2008). Thus, while scholars have focused on trying to empower women, the challenges that men face at the hands of other men has not been adequately addressed. Researchers must examine the process through which men attempt to gain power, or limit others? access to power, through interactions with one another (Kimmel, 2004, 2008).   To understand the process through which masculinity creates inequality among men, it is important to examine the practices in which men engage in order to demonstrate their masculinity. These practices are based on a restrictive set of beliefs and expectations concerning how men ought to act (Kimmel, 2008). The ideologies dictating how men ought to behave are often accepted because they are entrenched within cultural beliefs (Reeser, 2010) and there are severe repercussions for rejecting them (Kimmel, 2008; Pascoe, 2007). Moreover, these ideologies are linked to power and status in that conforming to these beliefs leads to more power while not conforming leads to a loss of power and status (Resser, 2010; Kimmel, 2004, 2008). Ideologies related to masculinity (and heterosexuality) also influence social structures, such as marriage (Heath, 2009), and thus are quite resistant to change (Heath, 2003). It is important to understand the ideologies that are guiding men?s practices. Merely looking at practices, though important, is akin to looking at the symptoms of a disease rather than examining its cause.   26  Without understanding the underlying root cause, changing behaviours will not change the underlying ideologies.   The ideologies guiding men?s masculinity develop from the ?Guy Code,? which is the basis for socialization concerning what men ought, or ought not, do (Kimmel, 2008). The Code is very restrictive. Kimmel (2008), for instance, asked women what it means to be a woman and found that women listed various potential social roles, such as mother or athlete. In stark contrast, when he asked men what it means to be a man, they listed a set of restrictive behaviours and beliefs, such as not showing weakness or avoiding being emotional. He argues that together these restrictive beliefs constitute the Guy Code and, specifically, the ?collection of attitudes, values, and traits that together compose what it means to be a man. ... The criteria that will be used to evaluate whether any particular guy measures up? (pp. 45). Interestingly, even transmen (people born as biological women who have transitioned to men) experience pressure to behave in masculine ways by colleagues (Schilt & Connell, 2007). This demonstrates people?s belief that to be a real man one must act in masculine ways.      The modern Guy Code is similar to the four dimensions of masculinity developed by Brannon (1976) nearly 40 years ago (Kimmel, 2008). According to research on the male sex role, four categories of attitudes and behaviours demonstrate masculinity. The first category is ?no sissy stuff,? which involves avoiding feminine characteristics that would lead to being labelled ?a sissy? or ?gay?. From this, masculinity is a rejection of femininity and behaviours related to it. The second category is ?the big wheel,? which represents men?s need to demonstrate power, wealth, and status in order to be masculine. Next is ?the sturdy oak,? which embodies men?s need to be confident, independent and self-reliant. This category develops from the idea that, during a crisis, men are able to remain calm and respond appropriately. The fourth category is ?give ?em   27  hell,? which includes aggression, daring and violence: men should do what they want and take risks, regardless of what others think. Another way of labelling these four categories is: concealing emotions, success, independence, and toughness (Springer & Mouzon, 2011).   Men endorse these beliefs of the Guy Code in order to be positively evaluated by other men, such as their fathers, friends, and coaches (Kimmel, 2008). The ideologies around the Guy Code develop in childhood through socialization and then are policed by other men throughout men?s lives (Kimmel, 2004; 2008). From childhood on, men have every part of their selves evaluated, from mannerism to movement. Any behaviour that contradicts the ideologies of the Guy Code is insulted and ridiculed in order to correct the behaviour. As such, men evaluate other men?s performances of masculinity and compare the performances to the ideologies of the Guy Code. Adherence to the Code is not necessarily about men trying to dominate women. Men are trying to impress other men. In fact, other men have the power to dictate who is, and is not, masculine. Kimmel?s argument, in contrast to previous research (e.g., Connell, 1987), proposes that performances of masculinity are based on inequalities between men, with men attempting to emasculate other men in order to gain more power.  Homophobia plays a central role in the policing of masculinity and heterosexuality (Kimmel, 2004, 2008; Korobov, 2004; Pascoe, 2007; Plodel & Fartacek, 2009). Less masculine men are often publically emasculated, typically through being called ?gay? or ?fag.? In fact, these remain the most common and offensive insults men use against one another (Kimmel, 2008; Pascoe, 2007). Interestingly, however, these insults are not about sexuality per se, but rather they are about masculinity, or a lack thereof (Pascoe, 2007). However, the linking of masculine behaviours with heterosexuality and less masculine behaviors with homosexuality scares men into acting masculine. This creates what Pascoe (2007) calls ?the spectre of the fag.? Specifically,   28  the association of non-heterosexuality with being less masculine teaches men to fear homosexuality since it is supposedly related to unmasculinity (Pascoe, 2007). Real men must avoid all things associated with homosexuals; therefore, heterosexual men must dress casually, not listen to women, and not care about art and literature since these are all behaviours that gay men are stereotypically believe to engage in (Kimmel, 2008). Similarly, the military?s ?Don?t ask, don?t tell, don?t pursue? philosophy is a means of policing masculinity through maintaining the heterosexual masculine ideal through distancing itself from non-heterosexuality (Britton & Williams, 1995). The policing of masculinity through the association of being less masculine with homosexuality enforces masculinity through creating fear of deviating from the Guy Code.  Overall, men?s adherence to the Guy Code ?demands a lot - that boys and young men shut down emotionally, that they suppress compassion, and inflate ambition. And it extracts compliance with coercion and fear? (Kimmel, 2008, p. 55). At the same time, conforming to the Guy Code brings promises of power and status, though these rarely, if ever, actually come to fruition. Therefore, adherence to the Guy Code creates tension and conflict between men, with few men actually obtaining the power and status masculinity is expected to offer. Over time, many men feel anger, resentment, and shame for not earning the power and status that they were promised (Kimmel, 2008).   Kimmel and Pascoe?s work frame masculinity as the behaviours men engage in to show their support of the Guy Code. Their research leads to the conceptualizing of masculinity as the internalization of ideals that are enacted through behaviours. Men who are the most masculine, according to this research, should tend to have better outcomes than those who the least masculine. Similar to West and Zimmerman?s idea of ?doing gender? (West & Zimmerman, 1987), I argue that men are ?doing masculinity? through their everyday interactions with others.   29  This scholarship is a vital move away from previous research, which tended to assume that masculinity benefited the majority of men. Focusing on interactions also answers Schrock and Schwalbe?s (2009) call for researchers to move from cataloging how men demonstrate masculinity to exploring how men?s performances of masculinity are related to power differences among men.  This work, particularly Kimmel?s research, provides little discussion of how men may challenge these ideologies. For instance, do men from disadvantaged groups, such as gay and bisexual men, challenge these ideologies or do they attempt to follow them? Kimmel (2008) himself notes that his research focuses on the experiences of white, middle-class men who are university educated. As such, it is important to examine a broader population of men to ascertain the generalizability of Kimmel?s findings. To address this concern, this dissertation will use the above theoretical and empirical works to develop and test hypotheses about the influence that masculinity has on adolescent men?s life experiences.   Masculinity during Adolescence is About the Guy Code In this section, I will develop my definition of masculinity during adolescence. Researchers have found that school is a critical site in which gender identities are formed (Frosh, Phoenix, & Pattman, 2002; Mac an Ghaill, 1994; Morrell, 1998), including increasingly greater expectations on men concerning performances of masculinity (Mac an Ghaill, 1994; Kimmel, 2008; Morrell, 1998; Frosh, Phoenix, & Pattman, 2002). This occurs because the potential rewards and punishments for being, or not being, masculine become more powerful than they were during childhood. Adolescents who are more masculine, for instance, are popular while adolescents who are less masculine are ostracized and victimized (Connell, 2002; Pascoe, 2007). As such, adolescents are strongly encouraged to demonstrate their masculinity in the ways   30  dictated by the Guy Code. While attempting to prove their masculinity during high school, these men must also deal with the stresses and challenges of transitioning into adolescence. At the same time, adolescents receive less support from others since the expectation is that men should be independent and calmly deal with life?s challenges (Brannon, 1976). Thus, when young men are feeling significant pressure to be masculine, they are given less support and taught to avoid seeking help or admitting weakness. These problems are compounded by the unrealistic and unachievable expectation of the Guy Code, which lead many men to feel inadequate and inferior (Kimmel, 2008).   Masculinity during adolescence is encouraged through the policing of masculinity (Kimmel, 2008; Pascoe, 2007). Policing of other men?s masculinity during high school can manifest itself in various ways, such as bullying, hazing, and predation of those men who engage in behaviours that are deemed to be unmasculine (Kimmel, 2008; Pascoe, 2007). The enforcement of masculinity is so pervasive that Kimmel (2008) calls high school a boot camp for masculinity. As mentioned, the policing of masculinity typically involves the use of homophobic labels to emasculate and disempower men. The consequences of being labelled as unmasculine are severe, often leading to taunting, violence, sexual assault, ostracism, bullying, dropping out of school, and assaults (Kimmel, 2008; Pascoe, 2007). These experiences lead to various negative outcomes, including depression, loneliness, isolation, suicide, or even school shootings (Kimmel, 2008).  While adolescents are experiencing the challenging and isolating transition into adolescence, two important developmental tasks must occur ? identity development (Erikson, 1963) and the development of peer relationships (Parker et al., 2006). The coinciding of these developmental tasks with the need to demonstrate masculinity means that while youth are   31  navigating the expectations of masculinity, they are also in the process of forming their own identity and learning how to develop and maintain relationships with peers. It is likely that adolescents? performances of masculinity will influence both their developing sense of self as well as their ability to develop peer relationships. Specifically, those who are unable to demonstrate their masculinity will likely have both their identity as a worthy man and their ability to form peer relationships challenged. A definition, and theory, of masculinity during adolescence must take into account the importance of masculinity on all areas of men?s lives, including identity development and relationships with others.   Adolescent Masculinity Defined Though researchers, such as Pascoe (2007) and Kimmel (2008), are beginning to explore the importance of masculinity during adolescence, there is still no clear definition of what masculinity during adolescence entails. I argue that masculinity refers to the behaviours that men engage in that show their support of the ideologies of the Guy Code. Thus, men engage in performances of masculinity while interacting with others (Connell, 1987; Kimmel, 2008; Schrock & Schwalbe, 2009), which leads them to be deemed as more or less masculine. However, three important elements of my definition should be highlighted. First, masculinity is about more than men subordinating women. During adolescence, men often must compete with one another in order to gain power and status through engaging in convincing performances of masculinity. Second, masculinity is not one set of behaviours, but rather a product of several groups of behaviours that, taken together, encompass overall masculinity. Third, men?s level of masculinity varies according to their performances of masculinity.      32  Subordination of men  Much of the work on masculinity (e.g., Connell, 1987; Schrock & Schwalbe, 2009) has tended to focus on how masculinity is a means through which men can subordinate women. Though performances of masculinity can support patriarchy and the subordination of women, I argue that researchers need to examine the ways in which performances of masculinity also influence men?s status and power relative to other men (Kimmel, 2004, 2008; Pascoe, 2007). I focus on how differences in adolescent men?s performances of masculinity may lead to different life experiences. As mentioned, masculinity is earned through practices that demonstrate adherence to the Guy Code (Kimmel, 2008), thus not all men will be equally masculine. Men will judge other men?s performances of masculinity based on how close their behaviours fit with the ideologies of the Guy Code. When men?s performances are deemed inadequate, they will be labelled as unmasculine, which will lead to a loss of power and status relative to other men. Therefore, men?s performances of masculinity will influence their life experiences relative to other men.  Sets of Behaviors I argue, based on previous research, that there are three groups of behaviours that adolescent men engage in to demonstrate their masculinity.  Concealment of emotions. The first set of behaviours focuses on the concealment of emotions (Kimmel, 2008; Pascoe, 2007). From early in life, men learn that expressing so called feminine emotions, such as worrying or crying, is not masculine and will lead them to be viewed as unmanly. Being emotional embodies the ultimate in unmanly since it is associated with being stereotypically gay (Kimmel, 2008). In fact, during her observations, Pascoe (2007) found that   33  mocking portrayals of this stereotype were often performed by youth as a means of policing other men deemed to be unmasculine. Aggression. The second set of behaviours that demonstrate masculinity during adolescence centers on using physical or verbal aggression with others (Kimmel, 2004, 2008; Schrock & Schwalbe, 2009). Specifically, physical and verbal aggression are often used by adolescents to assert power with particular others or within certain spaces (Harper & Harris, 2010; Kimmel, 2008; Pascoe, 2007; Schrock & Schwalbe, 2009). It is through physical and verbal aggression that young men demonstrate to other men that they are able to defend themselves and others around them (e.g., girlfriends). Pascoe (2007), for instance, found that many adolescent men use shows of aggression, such as acting loud and rowdy or challenging other men to fights, to prove their own masculinity or challenge another adolescent?s masculinity. Even for youth who are bullied, masculinity is related to defending oneself through physical and verbal aggression (Grossman, 2002).  Athleticism. The third set of behaviours focuses on how athleticism increases adolescent men?s masculinity. Playing sports tends to be male dominated, with boys more likely to play sports than girls (Sport Canada, 1998). Men also tend to play contact sports that involve greater risks and chances for injury (Emery, Meeuwisse, & McAllister, 2006; Kujala et al., 1995). Sports are an important means through which men can construct their masculinity around other men (Messner, 1990). During high school, men are keenly aware of how being in sports increases their masculinity (Pascoe, 2007). In fact, the most athletic adolescents tend to be the most masculine and have more status and power relative to those who are not athletic (Pascoe, 2007). In contrast, adolescent men who do not play sports often have their masculinity challenged (Kimmel, 2008). The competitive nature of sports can lead adolescents to develop a conditional   34  sense of self-worth since their masculinity is dependent on their success or failure, which may decrease adolescents? desire to develop intimacy with others (Messner, 1990).      In summary, I argue that adolescent men demonstrate their masculinity through three sets of behaviours: concealing emotions, aggression, and athleticism. The combined influence of these sets of behaviours determines how masculine adolescents will be perceived to be by their peers. This means that to examine adolescent masculinity, it is important to look at each group of behaviours.  Masculinity varies according to performances of masculinity Up to this point, adolescents? performances have been discussed as if they are either masculine or unmasculine. Researchers often equate sex and gender, thereby implying that men are a homologous and masculine group (Bottorff, Oliffe, & Kelly, 2012; Schrock & Schwalbe, 2009). However, there are degrees of masculinity ranging from very masculine to not masculine at all. Adolescent men who are the most masculine give performances that best support the Guy Code and thus have the most positive life experiences, such as being popular. On the other hand, when adolescent men are unable to give performances of masculinity that adequately support the Guy Code, they will be deemed to be less masculine and have more negative life experiences. Most adolescent men will fall between these two groups. That is, most adolescents occupy the middle range, being neither very masculine nor very unmasculine. These middle-range adolescents? may receive some teasing from peers as a means of reinforcing the need to be masculine, but this will be less severe than the teasing that occurs when men are labelled as very unmasculine.        35  Sexuality and Masculinity As discussed in the previous section, sexuality is linked to men?s performances of masculinity. It was argued that homophobia is used to police masculinity and that heterosexual adolescents are assumed to be more masculine than non-heterosexual adolescents. Stereotypes and research tend to support this taken-for-granted association (e.g., Bailey & Zucker, 1995; Udry & Chantala, 2002; Whitam & Zent, 1984). In reality, however, heterosexuals and non-heterosexuals are equally masculine. Specifically, by re-examining sexual identity and masculinity, I will argue that there is no reason to assume that sexual identity influences masculinity during adolescence. All men, regardless of their sexual identity, become socialized into the ideologies of the Guy Code. Societal and personal biases lead to the ascription of less masculinity to gay and bisexual men irrespective of their actual performances of masculinity. As such, when actual performances of masculinity are examined, there will be no differences between adolescents based on sexual identity.  It is important to note that research indicates that gay and bisexual men do not reject traditional notions of masculinity. Gay men, like heterosexual men, tend to idealize masculine and muscular bodies (Hennen, 2005), sexual risk-taking and activity (Green & Halkitis, 2006), and masculine fashions (Mosher et al., 2006). Moreover, gay men use stereotypical perceptions of masculinity (e.g., being emotionally restrictive, competitive, and muscular) and femininity (e.g., being emotional, passive, and small framed) when evaluating their own as well as others? masculinity (Sanchez, Greenburg, Liu, & Vilain, 2009). In fact, gay men tend to desire masculinity in themselves and others (Sanchez, Greenburg, Liu, & Vilain, 2009) and tend to avoid effeminate gay men (Bergling, 2001). This prior research supports the proposition that gay   36  and bisexual men engage in practices of masculinity that are similar to heterosexual men, thereby offering evidence that gay men do not redefine masculinity.   To understand why researchers and laypeople argue that gay and bisexual men are less masculine, it is important to differentiate performances of masculinity from ascribed masculinity. Performances of masculinity are exactly what the previously discussed definition entails - the performances adolescent men engage in as a means to demonstrate their masculinity that include behaviours related to concealing emotions, aggression, and athleticism. Ascribed masculinity, in contrast, concerns the characteristics that endow men with more or less masculinity regardless of their actual practices. Examples of ascribed masculinity are body type (tall perceived as masculine and short as feminine) and voice tone (low pitched perceived as masculine and high pitched as feminine). The ascription of masculinity occurs even though these men are not engaging in any behaviours related to performances of masculinity. The characteristics that lead men to be ascribed as more or less masculine are fluid and change depending on time and context. Ascribed masculinity has a profound impact on men since it advantages some men?s performances of masculinity while simultaneously disadvantaging others.  The distinction between performances of masculinity and ascribed masculinity is important since, compared to heterosexuals, gay and bisexual men are ascribed less masculinity irrespective of their actual performances of masculinity (Kimmel, 2008; Pascoe, 2007). The ascription of less masculinity to gay and bisexual men leads people to assume gay and bisexual men are less masculine. Performances of masculinity can still increase or decrease gay and bisexual men?s masculinity as perceived by others, but non-heterosexuals are given a lower baseline of masculinity. Pascoe (2007), for instance, found that gay adolescents who were deemed to be masculine were given more status and received less victimization than those who   37  were less masculine. However, gay and bisexual men?s ascribed masculinity is so low that performances of masculinity must be very strong in order for them to counteract their ascribed masculinity, thereby becoming labelled as masculine. As such, even though gay and bisexual men?s performances of masculinity are actually equal to that of their heterosexual male peers, they are perceived to be less masculine due to their lower ascribed masculinity.   The argument that gay and bisexual men are ascribed less masculinity than heterosexual men leads to an important question: Why ascribe less masculinity to gay and bisexual men irrespective of their performances of masculinity? I argue that this is a result of adolescents? adoption of the Guy Code (Kimmel, 2008) and what Pascoe (2007) calls ?the spectre of the fag.? Men often use homophobia as a way to police performances of masculinity and heterosexuality when other men are not supporting the ideologies of the Guy Code (Kimmel, 2004, 2008; Korobov, 2004; Pascoe, 2007; Plodel & Fartacek, 2009). Homophobia is often used to police masculinity. Over time, this leads to the spectre of the fag, which is the fear of engaging in any behaviours that will lead one to be labelled as gay and thus deemed to be unmasculine (see also Kimmel (2004; 2008)). The spectre of the fag teaches adolescents that masculinity is related to heterosexuality and that unmasculinity is related to homosexuality. This consequently leads men to fear homosexuality and behaviours (assumed to be) associated with it. The fear of the label fag is so strong that many men develop negative attitudes towards non-heterosexuals (Kimmel, 2008; Pascoe, 2007). In fact, hostility towards gay men is often a means for both demonstrating (Connell, 1995; Davies, 2004; Kimmel, 2004) and defending (Kimmel, 2004; Wilkinson, 2004) heterosexual men?s masculinity against the spectre of the fag. In summary, the use of homophobia as a means of policing masculinity and heterosexuality leads to the development of negative perceptions of gay and bisexual men. Over time, gay and bisexual men are ascribed less   38  masculinity, even though their performances of masculinity may be similar to heterosexuals? performances of masculinity.  The ascription of less masculinity to gay and bisexual men encourages the development of biases concerning the way people experience the social world. Specifically, the expectation that gay and bisexual men are less masculine leads people to focus on those who affirm this stereotype, such as gay men who act in ways that are considered feminine, while simultaneously rejecting contradictory images, such as gay men who act in masculine ways. This process of seeking affirming examples while ignoring contrary examples of stereotypes has been well documented in studies of confirmatory hypothesis testing (e.g., Fiedler et al., 1996; Snyder & Gangestad, 1981). Over time, these perceptions become so ingrained that they become truisms, which are linked to biological mechanisms. Ultimately, both heterosexual and non-heterosexual men come to perceive gay and bisexual men as less masculine regardless of their actual behaviours. This likely even influences gay and bisexual men?s recall of their own masculinity, given that they will assume that since they are gay they should have been less masculine than their heterosexual peers. Gay men, for instance, refer to some effeminate behaviour as the ?root? of their homosexuality. This implies that their failure to perform masculinity adequately should have given away that they are, or would become, gay.  Empirical research, in contrast, tends to support the assumption that gay and bisexual men are less masculine than heterosexual men. Gay men tend to retrospectively report more childhood gender nonconformity than heterosexual men (see review by Bailey & Zucker, 1995; Ploderl & Fartacek, 2006; Whitam & Zent, 1984). When examining sexual attraction, those who report only same-sex attraction are significantly more feminine (as measured using questions that predict whether the respondent is male or female) than those who report opposite-sex or bisexual   39  attraction (Udry & Chantala, 2002). Moreover, non-heterosexuals are much more likely to report gender identity disorder, which may indicate greater gender non-conformity (Bailey & Zucker, 1995). Finally, twin studies have found that twins rate non-heterosexual co-twins as less gender conforming even when the sexual identity of the co-twin is unknown (Bailey, Dunne, & Martin, 2000). This research offers empirical support for the assumption that gay and bisexual men are less masculine than heterosexual men, or at least give performances of masculinity that are deemed by others to be less masculine.    If  non-heterosexuals? performances of masculinity are similar to heterosexuals? performances, why do researchers (e.g., Bailey & Zucker, 1995; Ploderl & Fartacek, 2006; Udry & Chantala, 2002; Whitam & Zent, 1984) tend to find that gay men are less masculine than their heterosexual counterparts? There are various conceptual and methodological problems with this previous research. First, many studies use retrospective accounts of masculinity, which are prone to recall biases and stereotypes concerning the link between sexuality and masculinity. Retrospective accounts are problematic since gay men, compared to heterosexual men, are likely to recall themselves as less masculine due to their lower ascribed masculinity. Other researchers have made similar arguments about the biased recall of childhood masculinity among gay men (Gottschalk, 2003; Ross, 1980). However, the argument that gay men are biased in their recall of their childhood masculinity has not been empirically validated (Bailey & Zucker, 1995; Bailey, Dunn, & Martin, 2000; Zucker, 2005). Second, studies often fail to clearly define and operationalize masculinity, thereby making comparisons between heterosexuals and non-heterosexuals challenging. Third, studies use samples that do not self-identify as gay or bisexual (e.g., Udry & Chantala, 2002), which means research participants are not actually defining their own sexual identities. Rather, it is inferred based on sexual attraction. Inferring that a man who   40  has been attracted to another man is gay or bisexual is akin to the one-drop rule when used to determine race.  Fourth, and most problematic, is that many of these studies are explicitly (e.g., Bailey, 2003; Bailey & Zucker, 1995) and implicitly (e.g., Bailey, Dunn, & Martin, 2000; Ploderl & Fartacek, 2006; Udry & Chantala, 2002; Zucker, 2005) predicated on the assumption that differences in masculinity between heterosexuals and non-heterosexuals have a biological basis. The problem with biological theories is that they do not take into account the social construction of masculinity and sexuality. What is perceived to be masculine is culturally and historically dependent. Victorian fashion, for instance, was quite flamboyant in terms of modern standards, yet Victorian-era men who wore these garments were perceived to be masculine even though modern era men who dress flamboyantly are perceived to be feminine. By assuming that masculinity and sexuality are biologically linked, results from these studies are questionable. Overall, previous research is biased, both methodologically and theoretically, in a way that makes it likely that differences in heterosexual and non-heterosexual men?s performances of masculinity will be found even though actual performances of masculinity are likely similar.  In summary, there is no reason to expect that sexual identity will influence masculinity among adolescents since all adolescent men are socialized into the Guy Code and are expected to be masculine. Instead, it is various biases, which develop from the adoption of the Guy Code and the spectre of the fag, which lead to the ascription of less masculinity to gay and bisexual men irrespective of their actual performances of masculinity. Adolescent men, regardless of sexual identity, should have similar performances of masculinity.  Before stating my hypothesis concerning the relationship between sexual identity and masculinity, I should note that the literature on sexuality and masculinity conceptualizes sexual   41  identity as influencing masculinity (e.g., Bailey & Zucker, 1995; Ploderl & Fartacek, 2006; Udry & Chantala, 2002; Whitam & Zent, 1984). Therefore, in the current study, sexual identity will be the independent variable and masculinity will be the dependent variable. That is, based on the previous literature, I will treat sexual identity as a potential influence on men?s performances of masculinity when examining the relationship between sexuality and masculinity.  Based on the above arguments, it can be expected that:  Hypothesis 1.1. During adolescence, sexual identity is unrelated to masculinity. The Influence of Masculinity and Sexuality on Men?s Life Experiences  The following section discusses how masculinity and sexual identity, as separate constructs, influence adolescent men?s life experiences. The focus will be on the direct influence of masculinity and sexual identity on life experiences as well as on how sexual identity moderates the relationship between masculinity and life experiences. I begin this section by discussing the importance of life experiences in the life course of men. Next, I consider the process through which men?s performances of masculinity may influence their life experiences. Following this, I explore how sexual identity affects men?s life experiences. I will then conclude by discussing the impact of masculinity and sexual identity on three important life experiences: peer relationships, life satisfaction, and risk-taking. Life Experiences According to the life course perspective (Elder, Johnson, & Crosnoe, 2003), the life transitions and trajectories of individuals affect their life experiences, which then, in cyclic fashion, affects future life transition and trajectories. Important influences on life course trajectories are adolescents? performances of masculinity and their sexual identity, since both will either facilitate or hinder the ability to transition though life stages (Easterbrook, 2009;   42  Kimmel, 2008). Masculine heterosexual adolescent men, for example, are more likely to be popular and to have girlfriends (Pascoe, 2007). This allows them to better develop the social skills needed to maintain and develop  platonic and romantic relationships compared to their less masculine and/or non-heterosexual counterparts. Over time, those with the most beneficial qualities, such as being masculine and heterosexual, will experience more positive life experiences. These positive life experiences will allow for more beneficial life transitions and trajectories, which will further increase the chances of positive life experiences occurring. In contrast, those with less desirable qualities, such as being unmasculine and/or non-heterosexual, will experience more negative life experiences and thus their transitions and trajectories will be more detrimental.  How Masculinity Influences Men?s Life Experiences  In this section, a framework for understanding the process through which masculinity influences men?s life experiences is developed. I propose that, as men interact with others, they engage in performances of self in which they attempt to demonstrate their masculinity through behaviours that support the Guy Code. When performances allow men to earn masculinity they are labelled as masculine, thereby earning power and status. In contrast, when performances of masculinity come into question, men are labelled as unmasculine and become stigmatized, leading to a loss of power and status. Performances of masculinity thus increase or decrease men?s power and status relative to other men, which will in turn influence men?s life experiences.  Performances of Masculinity and Presentation of Self  Adolescents earn masculinity through their interactions with others (Connell, 1987; Kimmel, 2004, 2008; Schrock & Schwalbe, 2009). These social interactions can be understood   43  as performances that occur within structures that dictate how men ought to act based on the ideologies of the Guy Code. Goffman?s (1959) theory of presentation of self is a useful framework for exploring how men?s performances of masculinity influence their life experiences. According to the theory of presentation of self, social interactions are performances wherein people, as social actors, can play or perform various roles for others, the audience. Social actors use social cues that are received from other people during an interaction to decide how to manage or express information in that particular social context (Goffman, 1959). Performances often involve the use of information management, so that only information that supports the role that actors want the audience to know are revealed. An important part of information management is information control, which occurs when actors carefully monitor the information given to the audience. This control of information is vital for social actors because it allows actors to avoid revealing any details that could jeopardize the legitimacy of the performance as well as the current interaction and future relationship (Goffman, 1959). An adolescent gay man, in an effort to hide his homosexual identity from others, may agree with fellow students that a female classmate is sexually attractive in order to legitimate his outward social performance as heterosexual and hence masculine.  Goffman?s work on presentation of self is useful for understanding how the Guy Code influences adolescents? performances with others. Given that the Guy Code is an integral part of socialization, adolescents understand the importance of demonstrating their masculinity through their performances with others. Adolescents? knowledge of the Guy Code, including the consequences for not following it, also means that they will understand that there are repercussions for being unmasculine, such as victimization and ostracism. During interactions with others, adolescent men will therefore engage in information control so that they will appear   44  masculine. Exaggerating sexual prowess or concealing feminine emotions, for instance, will help adolescent men earn their masculinity. However, masculinity is a limited resource and not all men can be labelled as masculine. If all men were masculine, it would not confer power or status. Furthermore, the means through which adolescents can demonstrate their masculinity (such as through sexual prowess or athleticism) are limited, thus only the most masculine men are selected for sports teams or have sex with the most desirable female classmates. Men often police one another's performances of masculinity in an attempt to inflate their own masculinity while emasculating others (Kimmel, 2004, 2008). Due to the competition to be masculine, some men?s performances of masculinity will not measure up to the ideals. Once this occurs, their performances will be judged as illegitimately masculine, which typically means emasculation through being labelled as a fag (Kimmel, 2004, 2008; Pascoe, 2007).  Labelling Theory and Stigma  Once adolescents? performances of masculinity come into question by others, they are labelled as unmasculine. The modified labelling theory is a useful theory for examining the process through which being labelled as unmasculine may influence men?s life experiences (Link et al., 1989). The modified labelling theory was developed in order to explain how labelling influences the negative outcomes experienced by people who have been diagnosed with psychiatric disorders. Central to the theory is the idea that people internalize society's perceptions of various phenomena (Mead, 1934), which then influences their own perceptions of themselves. Specifically, people internalize societal perceptions of what it is to be labelled with a psychiatric disorder before they themselves have been labelled. Therefore, being labelled becomes associated with the negative views that have been internalized. This leads individuals who are labelled with a disorder to see themselves as devalued. The combined impact of the   45  label, as well as others? reaction to it, leads to outcomes such as withdrawal, discrimination, or rejection (Link et al., 1989). A major contribution of the modified labelling theory is the argument that it is the label and the societal expectations attached to it, rather than the disorder, that creates many of the negative outcomes associated with mental illness. In a longitudinal study on the effects of stigma and diagnosis of mental disorder (Link, Strueing, Rahav, Phelan, & Nuttbrock, 1997), it was found that the label of mentally ill has a negative impact on well-being even after treatment. Therefore, once labelled, the stigma continues to have a negative impact even after the symptoms of the initial problem have been treated.  Though developed in order to understand the experiences of those who have been diagnosed with a mental illness, the modified labelling theory is also very useful for understanding the experiences of adolescent men who are labelled as unmasculine. Men are socialized to adopt the ideologies of the Guy Code (Kimmel, 2008). This means that they understand what is expected of them concerning performances of masculinity, including that unmasculinity is devalued socially. In fact, the fear of being labelled as unmasculine is so strong that many adolescents support the Guy Code out of fear, rather than a belief in the Guy Code as a mantra (Kimmel, 2008). Adolescents who are labelled as unmasculine have internalized societal perceptions of those who are unmasculine and are thus aware of how others will perceive them. The combined impact of the label, as well as others? reaction to it, leads to various negative outcomes, such as withdrawal and rejection (Link et al., 1989). These outcomes are particularly problematic during adolescence since it is during this period that adolescents are developing their identities and forming increasingly intimate relationships with peers. Given the important developmental tasks occurring during adolescence, it is likely that adolescents deemed to be unmasculine will have their identity development challenged (e.g., leading them to ponder their   46  own masculinity) and find it difficult to develop and maintain positive peer relationships since peers may avoid the labelled adolescent.   An important part of the labelling process is the stigmatization of the labelled individual. Goffman (1963) defined stigma as an attribute, reputation or behaviour that is discrediting and thus causes others to classify the stigmatized as undesirable and discredited. The person goes from being a whole person to one who is tainted. The stigmatizing attribute then overshadows and taints other unrelated characteristics. Link and Phelan (2001) argue that stigmas develop through differential access to power and status. Specifically, the stigmatization process occurs though linking certain characteristics to negative attributes. People or groups place individuals into categories depending on these characteristics, creating an ?us? versus ?them? situation. This leads to status loss and discrimination for the stigmatized and status gains and power for the stigmatizers.   Once adolescents are labelled as unmasculine, they are stigmatized, thereby making attempts to earn masculinity through their performances much more challenging. The stigmatizing label also means that other unrelated behaviours are scrutinized as others look to further reinforce the label of unmasculine. A young boy who brings a doll to school, for instance, may thus receive the stigmatizing label of unmasculine. The stigma of the behaviours that are labelled as unmasculine becomes the defining feature of the boy and other parts of him will come into question, such as how he walks or talks. The stigmatizing process is also an important means through which the Guy Code and the behaviours attached to it are reinforced. Specifically, the fear of being stigmatized as unmanly becomes so strong that men are coerced into supporting it (Kimmel, 2008). Thus, consistent with Link and colleagues? (1989) theory, the stigma creates a   47  power imbalance such that those who are the most masculine can maintain their power by creating a fear of being stigmatized as unmasculine.        In summary, I argue that from an early age, boys are socialized into the ideologies of the Guy Code. A boy?s socialization into the Guy Code includes learning that those who do not act in accordance with the Code are not masculine and thus are not real men. As boys enter high school, there is an increase in the pressure to be masculine, which is enforced through the policing of others? performances of masculinity. As such, adolescent men come to learn the importance of giving convincing performances of masculinity, which are enacted through engaging in information control while interacting with others. The problem is that masculinity is a limited resource and consequently not all adolescent men can be masculine. This creates competition to be masculine, thereby encouraging men to police one another?s masculinity. Once adolescents? masculinity is repeatedly questioned they become labelled as unmasculine and stigmatized. The stigma has a powerful impact on adolescents since they have internalized negative perceptions of what it means to be unmasculine, including the consequences for being unmasculine. The label of unmasculine taints the self so that it becomes one of the most salient features others will focus on, thus making the labelled adolescent less desirable to his peers. The stigmatizing process is particularly problematic given that this is all occurring during a period of the life course in which identity formation and relationships with peers are salient developmental tasks.  How Sexual Identity Influences Men?s Life Experiences Having just discussed how masculinity influences men?s life experiences, I will now turn to exploring how sexual identity influences life experiences. Gay and bisexual adolescents, compared to heterosexual adolescents, experience various negative life experiences due to   48  marginalization, stigmatization, and victimization (Banks, 2003; Lindley et al., 2011; Mihalik, 1991; Savin-Williams, 2001; Sutton, 1994; Thompson & Johnson, 2003; Weston, 1995). Non-heterosexuals, for instance, are targets of both overt and covert forms of harassment and discrimination, and, compared to heterosexuals, experience higher rates of drug and alcohol abuse, depression, stress, loneliness, and suicide (Banks, 2003; Lindley et al., 2011; Mihalik, 1991; Sutton, 1994; Weston, 1995). Gay and bisexual adolescents are also at an increased risk of physical and verbal abuse and rejection from family due to negative family reactions (D'Augelli, Hershberger, & Pilkington, 1998). Familial rejection increases the likelihood that adolescents will run away, feel isolated, become depressed, attempt suicide, or engage in high-risk behaviours (Gibson, 1989; Hammelman, 1993; Remafedi, French, Story, Resnick & Plum, 1998; Dube, & Savin-Williams, 1994).  Sexual identity also influences life experiences through the internalization of the negative label associated with being non-heterosexual. Based on the modified labelling theory, gay and bisexual men internalize negative societal perceptions of what it is to be non-heterosexual before they themselves are aware of their non-heterosexual identity (Cass, 1979; 1984). Similar to discussed previously regarding masculinity, once adolescents come to develop their identity as gay or bisexual they become aware that they are devalued socially, which leads many non-heterosexual adolescents to devalue themselves and hide their true identity through engaging in information management (D'Augelli, 2006, Morrow, 1993). For adolescent men, the experience of devaluation is particularly challenging given the socially constructed link between sexuality and masculinity. That is, an identity as gay or bisexual is indicative of being less masculine.   Adolescents? sexual identity will influence life experiences in two ways. First, sexual identity will directly influence life experiences through the stigmatization and marginalization   49  that non-heterosexuals experience. Second, adolescent men?s sexual identity will influence life experiences through moderating the relationship between performances of masculinity and life experiences. Therefore, I argue that sexual identity will have a direct influence on adolescent men?s life experiences as well as influence the impact that masculinity has on life experiences. Hypotheses: Influence of Masculinity and Sexuality on Men?s Life Experiences  Having discussed a framework for understanding how masculinity and sexual identity influence men?s life experiences, I now develop hypotheses about the influence that masculinity and sexuality have on three important life experiences: peer relationships, life satisfaction, and risk-taking.  Peer Relationships  At the start of the 20th century, sociologist Charles Cooley (1902) argued that peers provide a context in which social and emotional development and adjustment can take place. Specifically, Cooley was one of the first scholars to suggest that children experience themselves through the responses of their peers. As children interact with their friends, they learn about themselves through how others respond. For instance, through playing games with peers, children learn to understand others? perspectives in relationship to their own perspective. Peer interactions become a means through which children come to develop their sense of self. In the tradition of Cooley, recent scholarship has demonstrated the importance of friendships for creating support networks, validating the self, increasing feelings of self-worth, and helping children develop reliability, affection, intimacy, empathy, and perspective taking (Furman & Robbins, 1985; Hartup, 1993; Parker et al., 1995). Moreover, adolescents? experiences with friendships influence development throughout the life course (Sternberg & Silverberg, 1986) because these early relationships serve as models from which intimate and romantic adult   50  relationships will be based (Berndt, 1982; Furman, 1999). As such, the ability to develop and maintain peer relationships is one of the most important adolescent developmental tasks.  In contrast, adolescents who are rejected by their peers are less likely to develop the social skills needed for both platonic and romantic relationships (Bierman, 2003; Schwartz, Dodge, & Coie, 1993). Peer rejection creates negative internal understandings of both the self and others through affecting adolescents? developing sense of self and beliefs about the social world (Bierman, 2003; Kupersmidt et al., 1990). Adolescents who are rejected by peers experience various problems, such as dropping out of school, aggression, disruptive behaviours, withdrawal, delinquency, poor adjustment, loneliness, low self-esteem, and anxiety (e.g., Asher, Hymel, & Renshaw, 1984; Bierman, 2003; Coie et al., 1990; Kupersmidt et al., 1990; Leary, Kock, & Hechenbleiker, 2001; Olweus, 1993; Parker et al., 1995). Due to these negative outcomes, rejected adolescents are unable to catch up to their peers? level of social skills development, which has a detrimental effect on their ability to develop future relationships (Bierman, 2003; Schwartz, Dodge, & Coie, 1993). The impact of peer rejection is particularly problematic during adolescence since these youth are within the tight knit and closed system of the school. Specifically, within the context of school, the negative consequences of peer rejection are intensified since youth cannot seek alternate contexts to avoid further rejection. Adolescents must interact with their peers daily and cannot easily attend a different school. In fact, many rejected gay adolescents drop out of school since they are unable to escape the stigmatizing label (Kupersmidt, Coie & Dodge, 1990; Ollendick, Weist, Borden, & Greene, 1992).  Given the important role that peer relationships play in the lives of adolescents, it is necessary to examine what factors influence the likelihood of adolescents being either accepted or rejected by their peers. For adolescent men, performances of masculinity will influence their   51  relationships with their peers by increasing or decreasing their desirability as a friend, or partner. Previous research indicates that masculinity is related to adolescents? peer relationships, though direct comparisons of adolescents from different sexual identity groups are scarce. Heterosexual adolescents who give the most convincing performances of masculinity, for instance, tend to have more friends and be more popular than less masculine heterosexual adolescents (Pascoe, 2007; Udry & Chantala, 2002). In contrast, heterosexual men who are less masculine experience more peer abuse, isolation, and ostracism (Coates & Person, 1985; Green, 1976; Kimmel, 2008; Rekers et al., 1979) as well as report more stressful interactions with peers (Washburn-Ormachea, Hillman, & Sawilowsky, 2004). Being unmasculine is also associated with poorer interpersonal adjustment, lower relationship satisfaction, and less participation in leisure activities during adolescence and adulthood (Aube & Koestner, 1992). For non-heterosexual adolescents, being unmasculine is associated with bullying, victimization, and rejection (Fagot, 1977; Friedman, et al., 2006; Landolt et al., 2004; Pascoe, 2007; Udry & Chantala, 2002). Unmasculine gay men, compared to masculine gay men, are also more likely to be loners, teased while playing sports, and to not fight back when being bullied (Grossman, 2002). This research indicates that for both heterosexual and non-heterosexual adolescents, being unmasculine is associated with more problematic peer relationships.    The means through which adolescents? performances of masculinity influence their peer relationships is not well understood. Based on the framework previously developed, I argue that adolescent men who give convincing performances of masculinity will earn the power and status needed to develop and maintain peer relationships. However, when adolescent men?s performances of masculinity are not accepted they will be labelled and stigmatized. This leads them to become members of a devalued group, which will increase the likelihood they will be   52  further rejected by their peers. Moreover, even peers who would have potentially befriended a less masculine peer may distance themselves due to the fear of becoming labelled and stigmatized through association (Kimmel, 2008). Heterosexual adolescents, for instance, may fear interacting with a peer deemed unmasculine because it may lead others to question their own performances of masculinity as well as their heterosexuality. At the same time, much like Cooley (1902) argues, rejected adolescents will feel increasingly undesirable due to the negative interactions they experience with their peers. When performances of masculinity come into question, adolescents become labelled and stigmatized, leading them to feel, and be, less desirable to their peers.  Sexual identity will influence the relationship between masculinity and peer relationships. Compared to heterosexuals, gay and bisexual adolescents are disadvantaged in their peer relationships due to the labelling and stigma associated with being non-heterosexual. Gay and bisexual adolescents, for instance, experience severe rejection from their peers (CDC & MDE, 1997; Henning-Stout, James, & Macintosh, 2000; Kosciw, Greytak, & Diaz, 2009; Ploderl & Fartacek, 2009), which has a negative impact on psychological well-being (Thurlow, 2001). The rejection non-heterosexuals experience leads to internalization and problems such as alienation, isolation, and difficulties with school (Roffman, 2000; Thompson & Johnson, 2003). This occurs because once labeled as gay or bisexual adolescents are stigmatized, thereby making them less desirable to their peers. Peers, particularly male peers, also fear non-heterosexual adolescents due to the spectre of the fag, since being non-heterosexual is believed to be the antithesis of masculinity. Many gay and bisexual youth cope with their fear of being rejected and stigmatized by hiding their identity until after graduation (D'Augelli, 2006, Morrow, 1993). However, even if adolescents withhold their non-heterosexual identity from their peers, they are   53  likely to have negative peer interaction due to their own perceptions of being devalued as well as the stress associated with having to hide their identity. Overall, gay and bisexual adolescents? experiences of alienation, rejection, and identity concealment likely disadvantages their ability to develop and maintain relationships with peers. As such, sexual identity will directly influence peer relationships as well as moderate the relationship between performances of masculinity and peer relationships.   Two specific peer relationship outcomes will be examined: adolescents? perceptions that their friends care about them and adolescents? feelings of being socially desirable. The perception that friends care measure examines adolescents? own perceptions concerning how much their friends care about them. In contrast, the social desirability measure examines adolescents? sense of being desirable to others, not just their friends.    Based on the above premises, the present study will test two sets of hypotheses concerning adolescents? peer relationships. For adolescent men: Hypothesis 2.1.1 Performances of masculinity are positively associated with the feeling that their friends care about them. Hypothesis 2.1.2 Heterosexuals will be more likely to feel that their friends care about them than non-heterosexuals. Hypothesis 2.1.3 Sexual identity will moderate the relationship between performances of masculinity and perceptions that friends care.  Hypothesis 2.2.1 Performances of masculinity are positively associated with feeling socially desirable.     54  Hypothesis 2.2.2 Heterosexuals will be more likely to feel socially desirable than non-heterosexuals. Hypothesis 2.2.3 Sexual identity will moderate the relationship between performances of masculinity and feeling socially desirable.   Life Satisfaction Life satisfaction is a broad assessment of how satisfied adolescents feel about their lives, thus it represents overall experiences, rather than a specific outcome. Given that masculinity and sexual identity influence various facets of men?s life experiences, it is likely that both will influence adolescents? life satisfaction. Specifically, based on the framework developed above, adolescents? performances of masculinity influence life satisfaction through differential access to power and status. When performances of masculinity are accepted, adolescents gain power and status, which increase the likelihood of positive life experiences occurring. Masculine adolescents, for example, are more likely to have a girlfriend or be popular (Pascoe, 2007). In contrast, when performances of masculinity are rejected, adolescents are labelled and stigmatized, thus leading them to be denied access to power and status. Those labelled as unmasculine are likely to experience more negative life experiences than those who are labelled as masculine. As such, differential access to power and status based on performances of masculinity will influence overall life satisfaction via their influence on life experiences.  There is evidence that being unmasculine is associated with more negative life experiences than being masculine. For heterosexual men, being less masculine is related to negative health outcomes (Courtenay, 2001), eating pathology (Munren & Smolak, 1997), body/image dissatisfaction (Coates & Person, 1985), peer abuse/isolation (Coates & Person,   55  1985; Rekers et al., 1979), psychological wellbeing (Carver, Yunger, & Perry, 2003; Harry, 1983; Kimmel, 2008; Yunger, Carver, & Perry, 2004), and suicide ideation (Reinherz et al, 1995). Moreover, childhood gender nonconformity is associated with psychological strain and negative feelings into midlife (Aube & Koestner, 1992). For gay adolescents, those who report same-sex only attractions are at risk for various negative outcomes such as victimization and emotional problems (Udry & Chantala, 2002). Compared to masculine gay and bisexual men, those who are unmasculine are also more likely to experience bullying, victimization, and have an increased risk of suicide (Friedman, et al., 2006; Ploderl & Fartacek, 2006). Moreover, non-heterosexual children who are unmasculine experience more negative parental comments and reactions (D'Augelli et al., 2005; Landolt et al., 2004), including a greater likelihood of being abused (Corliss, Cochran, & Mays, 2002; Harry, 1989).  Sexual identity will influence adolescents? life satisfaction as well as the relationship between performances of masculinity and life satisfaction. The label and stigma associated with being non-heterosexual influences life satisfaction by disempowering and marginalizing non-heterosexuals compared to heterosexuals. Gay and bisexual adolescents, compared to their heterosexual peers, are much more likely to have negative life experience such as victimization, rejection, dropping out of school, job termination, housing discrimination, and suicide (e.g., Banks, 2003; CDC & MPE, 1997; Henning-Stout, James, & Macintosh, 2000; Kosciw, Greytak, & Diaz, 2009; Ploderl & Fartacek, 2009; Sutton, 1994). The various negative outcomes associated with being non-heterosexual lead gay and bisexual men to experience more negative life experiences and thus feel a lower sense of life satisfaction. As such, even when performances of masculinity are similar, it would be expected that gay and bisexual adolescents will experience lower life satisfaction.    56  Based on these premises, the present study hypothesizes that for adolescent men: Hypothesis 3.1.1 Performances of masculinity are positively associated with life satisfaction.   Hypothesis 3.1.2 Heterosexuals will report greater life satisfaction than non-heterosexuals.   Hypothesis 3.1.3 Sexual identity will moderate the relationship between performances of masculinity and life satisfaction.    Risk-Taking The final life experience that I examine is adolescents? risk-taking. Courtenay (2001) explains men's higher rates of morbidity and mortality by using Connell?s (1987) work on hegemonic masculinity as a framework. Courtenay argues that men demonstrate their masculinity through power relationships, with the amount of risk behaviours men engage in linked to their masculinity relative to other men. This means that even though men are taught that being masculine involves taking health risks (Springer & Mouzon, 2011), the amount of risk depends on men?s level of masculinity relative to their peers. Those who are closer to the hegemonic ideal will take risks as a means to maintain their power and status relative to other men. However, those who belong to subordinate masculinities will take risks as a means of earning back their masculinity and hence their status and power. The risks these men engage in may be even more dangerous than the risks taken by those closer to the hegemonic ideal. This is because those with subordinate masculinities are marginalized due to being labelled and stigmatized as unmasculine, therefore they must engage in even more risky behaviours to earn their masculinity. Courtenay argues, for instance, that gay men engage in many sexual health   57  risks, such as unprotected sex with multiple partners, as a means of proving their masculinity relative to heterosexual men.  Courtenay?s definition of masculinity is based on Connell?s (1987) work on hegemonic masculinity. As discussed, Connell defines hegemonic masculinity as the means through which men obtain and maintain power and status over others, particularly women, through the support and propagation of patriarchy as natural and normative. My definition, in contrast, more specifically defines masculinity as the behaviours that men engage in to show their support for the ideologies of the Guy Code. Though there are differences between Connell?s definition and my own, I argue that Courtenay?s scholarship is amenable to my definition of masculinity as well as the framework I developed concerning how masculinity and sexual identity influence life experiences. Based on Courtenay?s work, I predict that men labelled as masculine will engage in risk-taking behaviours as a means of demonstrating their masculinity and thereby maintaining their power and status. However, those labelled and stigmatized as unmasculine will engage in increasingly risky behaviours as a means of earning back their masculinity. In fact, given the importance of both masculinity and peer relationships during adolescence, adolescents will have a strong desire to prove their masculinity. Specifically, the negative stigma associated with being unmasculine are so strong that the benefits of engaging in risky behaviours may outweigh the potential risks. Sexual identity plays a pivotal role in risk taking since gay and bisexual adolescents are ascribed less masculinity than their heterosexual peers. As such, gay and bisexual adolescents will have to engage in the greatest amounts of risk to prove their masculinity relative to heterosexuals. Peers do not have to know the sexual identity of adolescents for there to be an increased desire to take risks, since non-heterosexual adolescents are aware that they belong to a devalued and stigmatized group.     58    Risk-taking is assessed in this study in two ways: desire to take risks (attitudinal) and engagement in actual risks (behavioural). The reason for using two assessments of risk taking is that attitudes do not always represent actual behaviours. Specifically, La Piere?s (1934) seminal study found that even though almost all the hotels and motels he contacted stated that they would not give a Chinese couple a room or let them eat in the restaurant, when the couples did arrive they did receive a room and service. Based on this study, researchers began to question the link between attitudes and actual behaviours (Wicker, 1969). The link between attitudes and behaviours is complex and influenced by factors such as the strength of the attitude (Kraus, 1995). Given that attitudes may not directly predict behaviours, it is advantageous to examine both attitudes about risk-taking as well as actual risk-taking. During adolescence, I argue that an important risk factor to examine is frequency of being drunk (Richer, Leppin, & Gabhainn, 2006). Specifically, overuse of alcohol is related to alcohol abuse/misuse (Christiansen, Smith, Roehling, & Goldman, 1989), health problems (Richter, Leppin, & Gabhainn, 2006) and engaging in other dangerous and illegal behaviours, such as unplanned sex, injuries, and assaults (Baer, 1993; Wechsler et al., 1994).     Based on these premises, I hypothesize that for adolescent men: Hypothesis 4.1.1 Performances of masculinity are negatively associated with the desire to take risks.   Hypothesis 4.1.2 Heterosexuals will be less likely to report desiring to take risks than non-heterosexuals.   Hypothesis 4.1.3 Sexual identity will moderate the relationship between performances of masculinity and the desire to take risks.      59  Hypothesis 4.2.1 Performances of masculinity are negatively associated with frequent drunkenness.   Hypothesis 4.2.2 Heterosexuals will be less likely to report frequent drunkenness than non-heterosexuals.   Hypothesis 4.2.3 Sexual identity will moderate the relationship between performances of masculinity and frequent drunkenness.   Summary In summary, this dissertation aims to test three conceptual models, which were shown in figure 1 at the beginning of this chapter. First, I examine the relationship between sexual identity and masculinity. Specifically, the null hypothesis that there is no relationship between adolescents? sexual identity and their performances of masculinity is tested. Next, I explore the direct effect of masculinity and sexual identity on life experiences. Finally, I test the moderating effect of sexual identity on the relationship between masculinity and life experiences. The following chapter will detail the methods that will be used to test these hypotheses.     60  Chapter 3: Methods  In this chapter, I outline the methods used to test the hypotheses concerning masculinity, sexual identity, and life experiences among adolescent men. First, I discuss the study dataset and sample. Second, I operationalize the variables, including the independent variables, dependent variables, and control variables. Third, I detail the general analytic strategy for testing my hypotheses. Study Dataset and Sample  The data that are used for this study are from the public use section of the first three waves of the National Longitudinal Study of Adolescent Health (ADD Health). ADD Health is an American survey that focuses on adolescence and the transition into adulthood. The survey contains questions on a wide range of topics, including peer relationships, health, sexuality, family context, and psychological well-being. ADD Health is ideal for this study since it is one of the most comprehensive and utilized longitudinal datasets on adolescents and their transition into adulthood (e.g., Lindley, Walsemann, & Carter, 2011; Russell, Seif, & Trung, 2001; Udry & Chantala, 2002, 2004). The ADD Health dataset surveyed adolescents from 132 American high schools. The schools represented a stratified sample based on region, school size, and racial composition. The dataset oversampled groups such as Black adolescents from well-educated families and students with disabilities. The oversamples allow for the examination of minority groups that would otherwise be too small to provide meaningful results. Data were collected over multiple waves through means such as interviews, surveys, and biological tests. The first wave of data was collected between 1994 and 1995 and sampled students in grades 7 to 12. Wave two was collected one year later and included all wave one respondents who were still attending high   61  school. All respondents who graduated high school after wave one were excluded in wave two. Wave three was collected 6 years after wave one (2001 to 2002) from all respondents who were in wave one and could be contacted by wave three. Accordingly, wave three included those who had graduated high school between wave one and wave two. The adolescent respondents completed these surveys themselves in their school in wave one and in their homes in waves two and three.  Due to the small sample size of adolescents who identify as gay or bisexual, I took measures to avoid sample loss due to missing data. The primary cause of missing data is that wave two does not include respondents who were in grade 12 in wave one, although many of these respondents were included in wave three. These respondents appear to be missing at random in that their missingness is unrelated to any of the variables under examination. As such, I used multiple imputations to obtain estimates for the missing values using the Amelia II program (Graham 2009; Honaker & King 2010; Rubin 1987; Schafer, 1997). The use of multiple imputations, compared to listwise deletion, has been shown to be less biased3 (Honaker, King & Blackwell, 2011). When using multiple imputations, missing values are created over multiple datasets and filled with a distribution of values that reflect the uncertainty concerning the missing values (Honaker, King, & Blackwell, 2011). The combined estimates of the coefficients and                                                               3 Analyses were also run using listwise deletion and results were similar to those obtained using multiple imputations. However, results using multiple imputations did appear to be more conservative, thus they were used in the current study.   62  model fit statistics were produced with the SAS 9.3 Minianalyse procedure and are based on five imputed datasets.  The public use data from the self-administered surveys has a total sample size of 6,504 adolescents. The current study focuses on men and thus all women were removed from the study sample. Data was missing for most variables examined, though the majority of the missing data is in waves two and three. Specifically, wave two variables had, on average, 26.8% missing data and wave three had, on average, 29.2% missing data. The wave two missing data, as mentioned, is mostly due to the exclusion of all wave one grade 12 students. The wave three missing data, in contrast, seems to be due to sample attrition. The final sample, once data was imputed, is 3,147 adolescent men. All analyses are based on the imputed data.   Operationalization of the Study Variables Independent Variables: Dimensions of Masculinity and Sexual Identity Two independent variables require operationalization: masculinity and sexuality identity. I begin this section by detailing how the multiple dimensions of adolescents? performances of masculinity are measured. Following this, I discuss how adolescents? sexual identity is determined.  Dimensions of Masculinity As outlined in the previous chapter, I focus on several dimensions of masculinity, rather than an overarching measure of masculinity. Specifically, I argued that three dimensions of masculinity are most salient during adolescence: concealing emotions, aggression, and athleticism. These three dimensions correspond with the dimensions of masculinity found in adulthood (Brannon, 1976; Springer & Mouzon, 2011), though I conceptualize and measure them in ways that better represent adolescent men?s experiences (e.g., Kimmel, 2004, 2008;   63  Pascoe, 2007; Schrock & Schwalbe, 2009). First, concealing emotions embodies avoidance of feminine emotions, such as worrying or crying, because they are viewed as unmanly (Brannon, 1976; Kimmel, 2008; Pascoe, 2007; Springer & Mouzon, 2011). In fact, being emotional embodies the antithesis of masculinity since it is associated with being stereotypically gay (Kimmel, 2008). Second, aggression is an important means of demonstrating masculinity (Brannon, 1976; Kimmel, 2004, 2008; Schrock & Schwalbe, 2009; Springer & Mouzon, 2011). In particular, physical and verbal aggression are used by adolescent men to assert power with particular others or within certain spaces (Harper & Harris, 2010; Kimmel, 2008; Pascoe, 2007; Schrock & Schwalbe, 2009). Third, an important component of masculinity is success (Brannon, 1976; Springer & Mouzon, 2011), which during adolescence centers on athleticism (Messner, 1990; Kimmel, 2008; Pascoe, 2007).  The ADD Health dataset does not include items that were designed to measure adolescents? performances of the above dimensions of masculinity, thus, I had to develop a measure of masculinity for each of the three dimensions. To accomplish this, ADD Health survey questions that tap each of the three dimensions of masculinity were selected based on two criteria. First, the questions had to be asked in both wave one and wave two. Second, the questions had to assess the dimensions of masculinity in developmentally appropriate ways. That is, the questions had to measure performances of masculinity based on behaviours that adolescents engage in. Upon reviewing the ADD Health questions, 14 potential items were selected based on these criteria. The 14 items were factor analyzed to determine if they approximated the three dimensions of masculinity. I used three criteria to select the factors. First, the items had to load at .5 or higher. Second, the items needed to load much higher in one factor than any other factor.   64  Third, the items had to load into the same factors in wave one and wave two. I removed items that failed to meet these criteria and then factor analysed the remaining items again. After running several factor analyses in this way, I selected eight items to measure the dimensions of masculinity.  The first dimension, concealing emotions, is assessed using three questions that ask respondents, ?In the last month, how often (were you moody/did you cry a lot/were you fearful)?? Responses are ?never,? ?just a few times,? ?about once a week,? ?almost every day,? and ?every day.? These responses were reverse coded from 0=every day to 4=never. These questions do not explicitly ask about the purposeful concealment of emotions. However, I argue that adolescent men who are masculine will be less likely to openly report feeling these ?feminine? emotions than adolescents who are less masculine. Respondents' responses to these items were summed to get an overall score of concealing emotions. Higher scores indicate greater concealment of emotions.  The second dimension is aggression. Aggression is measured using two questions which ask respondents, ?How often were you loud, rowdy, or unruly in a public place?? and ?How often did you get into a serious physical fight?? in the last year. Possible responses are ?never,? ?1 or 2 times,? ?3 or 4 times,? and ?5 or more times.? Responses were coded from 0=never to 3=5 or more times. Importantly, it is unknown why these aggressive acts, particularly the fighting, took place. I do not know, for instance, if the adolescent is the instigator or the victim. However, aggression, even when used to defend oneself, increases masculinity (Grossman, 2002). As such, I argue that aggressiveness, regardless of the cause, will increase adolescents? masculinity. I summed responses to get an overall aggression score, with higher scores indicate greater aggression.    65  The third dimension is athleticism and it is assessed using three questions. The first two questions ask respondents ?During the past week, how many times did you play an active sport, such as baseball, softball, soccer, swimming, or football?? and ?During the past week, how many times did you exercise, such as jogging, walking, karate, jumping rope, gymnastics or dancing?? Responses are ?never,? ?just a few times,? ?about once a week,? ?almost every day,? and ?every day.? These are coded as 0=never to 4=every day. The third question that assesses athleticism has respondents answer how strongly they agree or disagree with the statement, ?You are physically fit.? Possible responses to this questions are ?strongly agree,? ?agree,? ?neither agree nor disagree,? ?disagree,? ?strongly disagree,? which are coded as 0=strongly disagree to 4=strongly agree. Respondents? answers to these questions will indicate their overall athleticism. Scores on these variables were summed, with higher scores indicate more athleticism. Table 1 gives the descriptive statistics for the variables used to create each of the dimensions of masculinity.  Table 2 shows the results of the factor analyses, which were run using varimax rotation. These factor analyses include the eight items selected to measure the dimensions of masculinity. As shown, these items loaded into three factors that approximate the three dimensions of masculinity. In both waves, the same three factors emerged thus supporting the assertion that there are three dimensions of masculinity. The correlations between the items in each scale were generally low to moderate. Each item is therefore tapping different parts of the masculinity construct. The correlations between the factors are also low, which further confirms that each dimension is measuring a different factor of masculinity.      66  TABLE 1: Descriptive Statistics for the Variables used to Measure the Dimensions of Masculinity (n=3147)  Mean Standard Deviation Actual Range Wave 1       Moodiness 2.89 .96 0-4    Crying 3.84 .42 0-4    Fearful 3.57 .66 0-4    Fight .61 .87 0-3    Rowdy .71 .92 0-3    Exercise 1.59 1.09 0-4    Sports 1.70 1.13 0-4    Fit 3.10 .84 0-4 Wave 2       Moodiness 3.08 .98 0-4    Crying 3.78 .48 1-4    Fearful 3.51 .66 0-4    Fight .37 .66 0-3    Rowdy .65 .92 0-3    Exercise 1.58 1.08 0-4    Sports 1.68 1.27 0-4    Fit 3.11 .88 0-4  The Conbach?s alphas for each factor are also shown in table 2. These alphas are low, indicating that there is relatively low internal consistency between the items within each factor. In fact, these alphas are lower than those obtained in other studies of masculinity. Springer and Mouzon's (2011) masculinity scale has an alpha of .61 and Pleck, Sonenstein and Ku?s (1993) masculinity measure has an alpha of .56. My low alphas may mean that the items within each factor are not measuring the same concept or construct. However, the low values obtained in this study are partially explained by the fact that there are only two or three questions per factor and that within some of the factors the format of the questions is different. These problems mean my alphas will be underestimates of the actual reliability of the factors (Tavakol & Dennick, 2011).       67  TABLE 2: Factor Analysis for the Dimensions of Masculinity (n=3147)    Concealing Emotions  Athleticism Aggression Wave 1       Moodiness .60 .10 -.33    Crying .74 .02 .02    Fearful .78 .01 .03    Fight .04 .05 .78    Rowdy -.13 -.02 .75    Exercise -.04 .72 .05    Sports -.03 .76 .02    Fit .19 .62 -.06 Wave 2       Moodiness .54 .09 -.15    Crying .71 -.08 .02    Fearful .73 .03 -.02    Fight .01 .03 .80    Rowdy -.14 .03 .77    Exercise -.16 .69 .00    Sports .03 .69 -.01    Fit .20 .68 .01 Cronbach's alpha       Wave 1 .54 .41 .49    Wave 2 .40 .39 .45  Table 3 provides the descriptive statistics for the dimensions of masculinity measures. Adolescents, on average, show high levels of concealing emotions, low levels of aggression, and moderate levels of athleticism. For all dimensions, scores do not differ greatly from wave one to wave two, which is expected given that the waves are only one year apart. Nevertheless, this does indicate stability in masculinity over time.  The concealing emotions and aggression scales are skewed. Skewness is not problematic as long as the distribution of the residuals in all the models approximates a normal distribution. When the distribution of residuals were examined for the dimensions of masculinity, they tended to be normally distributed. Nevertheless, to ensure the skewness of the variables does not affect the results, I also recoded the masculinity dimensions as three categories (low, medium, and high) and then re-ran all analyses using these three categories measure of masculinity. Results   68  replicated what was found when I used the scale variables. This offers further evidence that the skewness of the variables does not influence the results of this study. All analyses reported here are based on the scale versions of the masculinity measures.  TABLE 3: Descriptive Statistics for the Independent Variables: Dimensions of Masculinity and Sexual Identity (n=3147)   Percent Scale Mean Standard Deviation Actual Range Masculinity Measures:       Scale Measures       Wave 1          Concealing Emotions  10.30 1.50 0 - 12      Aggression  1.33 1.41 0 - 6      Athleticism  6.39 2.17 0 - 10   Wave 2          Concealing Emotions  10.34 1.33 3 - 12      Aggression  1.08 1.18 0 - 6      Athleticism  6.29 2.09 0 - 10   Change scores (W1 - W2)         Concealing Emotions  .041 1.622 -12 - 12     Aggression  -.250 1.436 -6 - 6     Athleticism  -.099 2.090 -10 - 10 Sexual Identity:          Heterosexual 95.1%         Gay/Bisexual 4.9%          Given that there are two waves of data for each dimension of masculinity, change scores were created as a means of examining the impact of changes in adolescents? performances of masculinity on their life experiences. Change scores allow for an examination of how changes in adolescents? performances of masculinity relate to their life experiences. Specifically, the use of change scores offers a further test of the hypotheses by investigating how changes in performances of masculinity, regardless of baseline masculinity, influence adolescent men?s life experiences. The change scores were created by subtracting wave one scores from wave two scores (W2 ? W1). Positive scores indicate an increase in adolescents? performances of masculinity while negative scores indicate a decrease in their performances of masculinity. As   69  shown in Table 3, from wave one to wave two, adolescents show a slight increase in concealing emotions (.041), while both aggression (-.250) and athleticism (-.099) decrease. Thus, from wave one to wave two, adolescents tended to increase their concealment of feminine emotions and decrease their levels of aggression and athleticism.  Sexual Identity The second independent variable requiring operationalization is sexual identity. Previous studies examining GLB and heterosexual adolescents using the ADD Health dataset (e.g., Russell, Seif, & Trung, 2001; Udry & Chantala, 2002) have often selected samples based on respondents' romantic relationships or their indication of who they are attracted to (males, females, or both). In contrast, the current study uses a measure of self-identification to determine adolescents? sexual identity. Specifically, in wave three respondents were asked to ?Please choose the description that best fits how you think about yourself? with responses being ?100% heterosexual (straight),? ?Mostly heterosexual/somewhat attract to people of own sex,? ?Bisexual-attracted to men and women equally,? ?Mostly homosexual/somewhat attracted to the opposite sex,? and ?100% homosexual (gay).? There is a sixth category, which includes adolescents who are not attracted to anyone, that was excluded from the sample (n=25). The straight category was collapsed with mostly heterosexual because minor attraction to the same-sex does not negate one from being basically heterosexuals. Specifically, I wanted to avoid categorizing men as non-heterosexuals if they demonstrated slight attraction to the same sex. Given the small size of non-heterosexuals, respondents who indicate that they are bisexual, mostly homosexual, or homosexual are treated as one category. As shown in Table 3, the majority of the sample identifies as heterosexual (n=2993).     70  Dependent Variables  The waves of the ADD Health dataset vary slightly from one another due to the addition of new questions and/or the removal of old questions. As a result, some of the questions used to measure the dependent variables were only asked in certain waves. This means that each set of hypotheses is analyzed using different waves of data depending on the data available for the variables being examined. Sexual identity is the independent variable and masculinity is the dependent variable for the first hypothesis, which examines the relationship between sexual identity and masculinity. For all other hypotheses, sexual identity and masculinity are treated as independent variables.  Sexual Identity and Masculinity 1.1. Masculinity: Masculinity is measured as the three dimensions of masculinity discussed above: concealing emotions, aggression, and athleticism. Sexual identity is measured using the two category measure (heterosexual vs. gay/bisexual) that was discussed above.  Masculinity, Sexual Identity and Life Experiences Peer Relationships were measuring with two items. The first item examines adolescents' perception that their friends care about them while the second measure assesses adolescents' feelings of being socially desirable.  2.1. Friends care: This wave one and wave two measure asks respondents ?How much do you feel that your friends care about you?? Possible responses are 1=?not at all,? 2=?very little,? 3=?somewhat,? 4=?quite a bit,? and 5=?very much.? This variable is skewed, with most respondents indicating that their friends care about them. This skew was expected since most adolescence should indicate that their friends care about them. To compensate for the skew, I created a binary variable for friends care. Responses indicating that friends care ?not at all?, ?very   71  little?, or ?somewhat? are coded as 0 and the responses of those who rate their friends as caring ?quite a bit? or ?very much? are coded as 1.  2.2. Social desirability: Social desirability is assessed using the scores of adolescents? responses to four questions that were asked in waves one and two. The first two questions ask respondents, ?How often was each of the following true: You felt that people disked you? and ?How often was each of the following true during the last week: People were unfriendly to you.? Responses are 0=?never/rarely,? 1=?sometimes,? 2=?a lot of the time,? and 3=?most/all of the time.? The next two questions asked respondents how strongly they agree or disagree with the statements, ?You feel socially accepted? and ?You feel loved and wanted.? Responses are 1=?strongly agree,? 2=?agree,? 3=?neither agree nor disagree,? 4=?disagree,? and 5=?strongly disagree.? The social desirability scale was created by summing adolescents' scores. The scores were then reverse coded so that higher scores indicated feeling more socially desirable. The alpha for combining these items is .75 in wave one and .71 in wave two.  Table 4 provides the descriptive statistics for the peer relationship variables. In both wave one and wave two, the majority of adolescents report that their friends care about them. This, again, is expected given that adolescents choose their friends and would typically only be friends with someone who cared about them. Adolescents also report feeling socially desirable. Though this variable is skewed, very few adolescents report scores below 9. This indicates that most adolescents feel socially desirable, while very few feel socially undesirable. In general, adolescents tend to feel that their friends care about them and that they are socially desirable.      72  TABLE 4: Descriptive Statistics for the Dependent Variables (n=3147)  Percent Mean (SD) Actual Range    Friends care W1 (vs care less) 81.3      Friends care W2 (vs care less) 79.3      Social Desirability W1  11.70 (1.86) 2-14    Social Desirability W2  11.71 (1.77) 1-14    Life Satisfaction W3  16.02 (2.68) 0-20    Risk Taking W2 (Yes) 60.0      Risk Taking W3 (Yes) 62.6      Drunkenness W2 (3+ times) 24.7      Drunkenness W3 (3+ times) 41.7   SD = Standard Deviation     Life Satisfaction 3.1. Life satisfaction: Life satisfaction is assessed using adolescents? responses to five items asked in wave three, which corresponds to the transition into young adulthood. The first question asks respondents ?How satisfied are you with your life as a whole?? with the responses of 1=?very satisfied,? 2=?satisfied,? 3=?neither satisfied nor dissatisfied,? 4=?dissatisfied,? and 5=?very dissatisfied.? The next four items ask respondents how strongly they agree or disagree (1=?strongly agree,? 2=?agree,? 3=?neither agree nor disagree,? 4=?disagree,? and 5=?strongly disagree?) that they ?have many good qualities,? ?have a lot to be proud of,? ?like yourself just the way you are,? and ?feel that you are doing things just about right.? Scores were calculated by summing respondents? responses across all items. The alpha for this scale is .80. Table 4 provides the descriptive statistics for the life satisfaction measure. Adolescents? generally indicate that they have a relatively high sense of life satisfaction. Again, this variable is skewed, though very few respondents have a score below 12, thus indicating that most respondents do not feel dissatisfied with their lives.       73  Risk-Taking is assessed with respect to attitudes and behaviours.  4.1. Desire to take risks: In waves two and three respondents are asked how strongly they agree or disagree (1=?strongly agree,? 2=?agree,? 3=?neither agree nor disagree,? 4=?disagree,? and 5=?strongly disagree?) to the statement, ?You like to take risks.? Given that my interest is in desiring to take risks, this variable was recoded as 0 =neutral/disagree and 1=agree.  4.2. Frequent Drunkenness: In waves two and three respondents were asked, ?Over the past 12 months, on how many days have you gotten drunk or ?very, very high? on alcohol?? Responses are 1=?every day/almost every day,? 2=?3-5 days/week,? 3=?1 or 2 days/week,? 4=?2 or 3 days/month,? 5=?once a month or less (3-12 times in past 12 months),? 6=?1 or 2 days in past 12 months,? and 7=?never.? In wave three, the responses were reverse ordered. Given that this variable is highly skewed, it was recoded into a binary variable. Specifically, those who rarely get drunk (scores of 6 or 7 which represents never getting drunk or getting drunk 1 or 2 days in the past year) are coded as 0 while those who more frequently get drunk (scores of 1 to 5 which represents getting drunk 3 or more times in the last year) are coded as 1. Previous research (Richer, Leppin, & Gabhainn, 2006) has labelled frequent adolescent drunkenness as having been drunk 2 or more times ever. The current study, in contrast, uses the cut-off of having been drunk 3 or more times in a year as an indication of frequent drunkenness. This cut-off was chosen because the adolescents in the current study are older, and thus are more likely to get drunk (Richer, Leppin, & Gabhainn, 2006) Table 4 provides the descriptive statistics for the risk-taking measures. Over half of the adolescent men report that they desire to take risks, though a much lower percent engage in actual risk taking as measured by frequent drunkenness. This may indicate that attitudes about taking risks do not necessarily correspond to actual risk-taking. This supports my earlier   74  assertion that beliefs do not necessarily translate into actual behaviours. The descriptive statistics also indicate that drunkenness increases dramatically from wave one to wave two, which may have occurred because the average age of the sample increased by one year.   Control Variables  Also included in all models are a series of demographic and other control variables.   Demographics: Birth year, race, maternal and paternal education, and household income are controlled for in all models. Race is coded as 0=?non-white? and 1=?white.? The binary coding of race was done because of the small sample size of the non-heterosexual adolescents. Maternal and paternal education are coded as 0=?less than a B.A.? and 1=?B.A. or more.? Household income is a scale variable indicating parents? self-reported annual household income.  Urbanicity: Contextual data from wave one codes respondents? level of urbanicity, with 0=?not completely urban? and 1=?completely urban.? Urbanicity influences the life experiences of adolescents. For instance, GLB individuals are more accepted in urban areas than they are in less urban areas (e.g., Snively, Kreuger, Stetch, Watt, & Chadha, 2004).  Depression: It is important to control for adolescents? feelings of depression given that depression is related to sexual identity, masculinity, and life experiences. For instance, gay and bisexual adolescents, compared to heterosexual adolescents, report higher rates of depression (Fergusson, Horwood, & Beautrais, 1999; Hatzenbuehler, McLaughlin, & Nolen-Hoeksema, 2008; Safren & Heimberg, 1999). Consequently, it is important to determine whether differences between heterosexual and non-heterosexual adolescents? life experiences are due to their performances of masculinity rather than differences in their feelings of depression. Moreover, there may be some overlap between the concealing emotions dimensions of masculinity and feelings of depression. In particular, concealing emotions is likely also tapping adolescents?   75  emotional stability or depression. That is, respondents who are less masculine show more ?feminine? emotions, such as crying or worrying, which are also symptoms of depression. In fact, the correlations between concealing emotions and depression at both waves range from .3 to .5. Given these potential issues, it is important to control for depression.    The ADD health dataset includes the CES-D, which is a well-validated measure of feelings of depression (Radloff, 1977). Importantly, this scale was developed for use in research on the general population and it is not a clinical tool. The CES-D has 20 items, which examine several dimensions of depression. Respondents are asked, ?How often was each of the following true? to statements such as ?You were bothered by things that usually don?t bother you,? ?You felt that you were just as good as other people,? and ?You felt people disliked you? over the past week. Responses are 0=?never/rarely,? 1=?sometimes,? 2=?a lot of the time,? and 3=?most/all of the time.? A depression symptom score is calculated by summing responses.  Table 5 provides the summary statistics for all the demographic and control variables. Adolescents were born between 1975 and 1983, with a mean birth year of 1979. Most adolescents are White and just over half live in more urban areas. Concerning parental education, about a third of respondents? mothers and fathers have a B.A. or more. The average household income is about $50,000. The mean depression score is 10.19 out of 60, indicating a low level of depression symptomology. Weights and Clustering: ADD Health (Chantala, 2006) recommends using weights and cluster variables to take into account the stratified sampling procedures used to collect the data. Specifically, the 132 school chosen did not have equal probability of selection. Moreover, certain groups (such as educated Blacks and adolescents with disabilities) were purposefully over- sampled. As such, the use of weights and cluster variables allow analyses to be unbiased   76  TABLE 5: Descriptive Statistics for the Demographic and Control Variables (n=3147)  Percent Mean (SD) Actual Range    Birth year  1979 1975-1983    White (vs. other) 66.7      More Urban (vs. less Urban) W1 52.9      Mother > B.A. (vs. less than B.A.) W1 30.8      Father  > B.A. (vs. less than B.A.) W1 33.2      Income W1  50,303 (54,698) 0-999,000    Depression Scale W1  10.03 (6.72) 0-48    Depression Scale W2  10.19 (6.43) 0-56 SD = Standard Deviation     estimates of population parameters and standard errors. There are separate weights for each wave, and it is recommended to use the weight from the latest wave in the current analysis (Chantala, 2006). For example, if examining wave one masculinity and wave two peer relationships, the wave two weight should be used.  Analytic Strategy  All data were analyzed with multiple linear or binary logistic regression models using the SAS 9.3 program. When using OLS regressions, the unstandardized slopes and standard errors were reported. For the logistic regressions, I reported the log odds and standard errors. P-values under .05 were reported as significant while p-values under .10 were reported as marginally significant. All analyses included the sampling weights and cluster variables as well as all the control variables.  Hypothesis 1.1, which focuses on the influence of sexual identity on masculinity, was tested using one set of models for each dimension of masculinity. Specifically, the models examined the influence of sexual identity on each dimension of masculinity in both wave one and wave two as well as on adolescents? masculinity change scores.    77   The remaining hypotheses (hypotheses 2.1.1 to 4.2.3) focused on the influence of sexual identity and masculinity on adolescents? life experiences. These hypotheses were tested using three different sets of models. In the first set of models, the direct relationship that sexual identity and the dimensions of masculinity have on life experiences was tested. When there were multiple waves of data for an outcome, models tested the impact of each wave of masculinity on each wave of the life experience being examined4. Sexual identity is only available in wave three, thus it is always wave three sexual identity that is being examined. These first sets of models tested how adolescents? sexual identity and performances of masculinity at each wave influence their life experiences.  In the second set of models, the relationship between the masculinity change scores and life experiences was examined. This, in contrast to the first set of models, examined how changes in adolescents? performances of masculinity were related to life experiences.  The third set of models tested for moderation by adding the interaction terms. Interaction terms were developed to test if adolescents? sexual identity moderates the relationship between masculinity and life experiences. Specifically, testing interactions allows for an examination of how the relationship between each dimension of masculinity and life experiences is moderated by adolescents? sexual identity as heterosexual or gay/bisexual. Three interactions terms were created for each wave so that each factor of masculinity could be examined. The data was                                                               4 Models were also run that separately test the relationship that sexual identity and masculinity have on life experiences. Results follow the same pattern as is found when both masculinity and sexual identity are in the same model.   78  imputed after the interaction terms are created to avoid biasing the regression parameters (von Hippel, 2009).      79  Chapter 4: Results In chapter 2, I developed several sets of hypotheses concerning adolescent men?s masculinity, sexual identity, and life experiences. Chapter 3 detailed how these hypotheses were tested. This chapter discusses the results from the models testing each set of hypotheses. I begin by summarizing the results for the models testing the relationship between sexual identity and masculinity. Next, I review the findings for the models examining the influence that masculinity and sexual identity have on adolescent men?s life experiences.  Hypothesis 1.1: Sexual Identity and Masculinity I predicted that sexual identity will not influence men?s performances of masculinity. Specifically, hypothesis 1.1 states that there is no relationship between adolescents? sexual identity and their performances of masculinity. Table 6 shows the results of the models testing this hypothesis. As shown in model 6, being heterosexual is associated with greater athleticism in wave two (B = .437). This indicates that, compared to non-heterosexual men, heterosexual men score almost a half point higher in athleticism. However, this relationship is only marginally significant (p < .10). There are no other significant relationships between adolescents? sexual identity and their performances of masculinity. In addition, adolescents? sexual identity is unrelated to their masculinity change scores. Overall, these findings indicate that sexual identity and masculinity are distinct concepts.  Life Experiences Hypotheses 2.1.1 - 2.2.3: Peer Relationships  This section focuses on the hypotheses concerning adolescents? peer relationships. I begin by discussing the results of the models examining adolescents? perceptions that their friends care. This will be followed by the results of the models examining social desirability.       80   Wave 1 Wave 2 Change Scores  Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Model 9  No Emotion Aggression Athleticism No Emotion Aggression Athleticism No Emotion Aggression Athleticism Sexual Identity             Heterosexual (vs.                                    Gay/Bisexual) .080(.221) .047 (.189) .299 (.247) .004 (.186) .001 (.159) .437? (.253) -.076 (.273) -.046 (.228) .137 (.252) Demographics             Birth year -.059** (.018) .039* (.016) .133** (.024) -.003 (.025) -.038* (.018) .209** (.037) .056* (.027) -.077** (.022) .076? (.037)    White (vs. other) -.477** (.066) .227** (.062) -.408** (.104) -.192* (.074) .127* (.061) -.159 (.125) .285** (.072) -.100 (.075) .249* (.115)    More Urban (vs. < Urban) .042 (.063) .192** (.057) .092 (.080) -.111? (.065) .120 (.072) .169 (.113) -.152* (.070) -.073 (.076) .077 (.120)    Mom  > BA (vs. < than BA) -.050 (.075) -.049 (.084) .200 (.114) -.107 (.088) -.040 (.065) .046 (.113) -.057 (.091) .009 (.086) -.154 (.130)    Dad  > BA (vs. < than              BA) -.104 (.089) -.081 (.092) .297* (.130) -.068 (.092) -.002 (.104) .078 (.127) .036 (.091) .079 (.127) -.219 (.134)    Income -.000 (.001) .000 (.000) .000 (.001) -.001 (.001) .000 (.001) .003* (.001) -.000 (.000) -.000 (.001) .003* (.001)    Depression Scale -.098** (.005) .042** (.005) -.065** (007) -.057** (.005) .020** (.004) -.052** (.009) .041** (.006) -.022** (.005) .013 (.009) ? p ? .10; * p ? .05; ** p ? .01  Depression scale wave is the same as the masculinity wave being examined.             TABLE 6: Unstandardized Slope Coefficients (Standard Errors) for the OLS Regression Models of Dimensions of Masculinity Regressed on Sexual Identity   81    TABLE 7: Log Odds (Standard Errors) for the Logistic Regression Models of Perception that Friends Care regressed on Dimensions of Masculinity  Friends Care W1 Friends Care W2  Model 1 Model 2  Model 3 Model 4 Model 5 Model 6 Dimensions of Masculinity       Wave 1          Concealing Emotions .007 (.041) -.015 (.119) -.050 (.046)       Aggression .018 (.039) .152 (.227) -.003 (.039)       Athleticism .110* (.033) .129 (.190) .033 (.027)    Wave 2          Concealing Emotions    -.031 (.051) -.062  (.089)     Aggression    .003 (.048) -.084 (.208)     Athleticism    .030 (.026) .137 (.140)  Change Scores          Concealing emotions      .013 (.038)    Aggression      -.008 (.033)    Athleticism      .000 (.029) Sexual Identity          Gay/Bisexual -.027 (.341)  -.178 (.290) -.178 (.297)  -.163 (.293) Interaction Terms          Conc. Em. X Heterosexual  .023 (.117)   .032 (.081)     Aggression X Heterosexual  -.142 (.246)   .093 (.199)     Athleticism X Heterosexual  -.020 (.195)   -.113 (.137)  Demographics          Birth year -.052 (.031) -.053 (.031) .021 (.040) .012 (.042) .012 (.041) .018 (.041)    White (vs other) .630** (116) .634** (.117) .207 (.163) .199 (.161) .200 (.122) .196 (.163)    More Urban (vs < Urban) .036 (.116) .036 (.116) -.108 (.140) -.095 (.141) -.093 (.140) -.084 (.140)    Mom > BA (vs < than BA) -.024 (.150) -.024 (.150) .235 (.193) .229 (.189) .228 (.190) .236 (.192)    Dad  > BA (vs < than BA) .302 (.201) .303 (.200) .346 (.175) .259 (.180) .259 (.181) .263 (.180)    Income .002 (.001) .002 (.001) .003 (.002) .003 (.002) .003 (.002) .002 (.002)    Depression Scale -.047** (.010) -.047** (.010) -.025** (.009) -.034** (.010) -.034** (.010) -.033** (.009)    Friends Care W1   1.055** (.144) 1.049** (.147) 1.050** (.147) 1.065** (.148) ? p ? .10; * p ? .05; ** p ? .01 Depression scale wave is the same as the masculinity wave          82    Friends Care   The results for the models testing the hypotheses concerning the influence that masculinity and sexual identity have on adolescents? perceptions that their friends care are shown in Table 7. Hypothesis 2.1.1 predicts that masculinity is positively associated with adolescents? perceptions that their friends care. The only significant relationship found in these models is model 1. Athleticism in wave one is positively associated with perceiving that friends care in wave one (Log odds = .110). This corresponds to an odds ratio of 1.12, meaning that athletic adolescents have 1.12 times higher odds of reporting that their friends care about them. Specifically, adolescents who are high in athleticism are more likely to feel that their friends care about them in wave one. There are no other statistically significant association between the dimensions of masculinity and perception that fiends care. Overall, these findings do not support the hypothesis that adolescents? performances of masculinity are associated with perceiving that their friends care about them.   Hypothesis 2.1.2 states that heterosexual adolescents are more likely to feel that their friends care about them than non-heterosexual adolescents. Results, however, indicate that sexuality is not significantly related to perceptions that friends care.   Hypothesis 2.1.3 predicts that sexual identity moderates the relationship between performances of masculinity and the perception that friends care. None of the interaction terms for masculinity and sexual identity are significantly related to perceptions that friends care. Thus, sexual identity does not significantly moderate the relationship between dimensions of masculinity and perceptions that friends care.        83    Social Desirability Next, I discuss the models testing the hypotheses concerning the influence that masculinity and sexual identity have on adolescents? perceptions of being socially desirable. Table 8 shows results for these models. Hypothesis 2.2.1 states that masculine adolescents will feel more socially desirable. As shown in model 1, concealing emotions in wave one is positively associated with social desirability in wave one (B = .058). This indicates that a one-point increase on the concealing emotions scale corresponds to a .058 increase in social desirability. Thus, the more adolescents conceal their emotions, the more social desirable they feel in wave one. As shown in models 1, 3, and 4, athleticism in wave one is positively associated with feeling socially desirable in wave one (B = .113) and wave two (B = .062), while athleticism in wave two is positively associated with feeling socially desirable in wave two (B = .089). A one-point increase in athleticism corresponds to between a .062 and .113 increase in social desirability. These results indicate that athleticism increase adolescents? feelings of being social desirable. The athleticism change score, as shown in model 6, is also positively associated with social desirability (B = .045). A one-point increase in the athleticism change score corresponds to a .045 increase in social desirability, thus indicating that becoming more masculine from wave one to wave two is related to feeling more socially desirable in wave two. In contrast, aggression in both wave one (B = -.057) and wave two (B = -.059) is negatively associated with social desirability in wave two, as shown in models 3 and 4. Specifically, a one-point increase in aggression corresponds to almost a .6 decrease in social desirability. Adolescents who are high in aggression tend to feel less socially desirable in wave two, though not in wave one. Overall, these findings offer mixed support for the hypothesis that masculinity and social     84    TABLE 8: Unstandardized Slope Coefficients (Standard Errors) for the OLS Regression Models of Social Desirability Regressed on Dimensions of Masculinity  Social Desirability W1 Social Desirability W2  Model 1 Model 2 Model 3 Model 4 Model 5  Model 6 Dimensions of Masculinity       Wave 1          Concealing Emotions  .058* (.027) .062 (.061) .046 (.034)       Aggression  -.024 (.021) .010 (.106) -.057* (.028)       Athleticism  .113** (.016) .105 (.088) .062** (.017)    Wave 2          Concealing Emotions     .022 (.032) -.025 (.052)     Aggression    -.059* (.026) -.151 (.171)     Athleticism     .089** (.017) .187** (.076)  Change Scores          Concealing emotions      .028 (.024)    Aggression      -.025 (.024)    Athleticism      .045* (.018) Sexual Identity          Gay/Bisexual .020 (.205)  -.074 (.162) -.034 (.140)  -.071 (.138) Interaction Terms          Conc. Em. X Heterosexual  -.003 (.055)   .049 (.043)     Aggression X Heterosexual  -.035 (.114)   .097 (.171)     Athleticism X Heterosexual  .008 (.089)   -.103 (.076)  Demographics          Birth year .016 (.022) .016 (.022) .066* (.024) .002 (.023) .003 (.022) .014 (.024)    White (vs other) .202** (.074) .201** (.073) -.041 (.082) .142* (.064) -.143* (.064) .184** (.069)    More Urban (vs < Urban) .063 (.068) .062 (.068) .082 (.079) .019 (.077) -.019 (.075) .010 (.075)    Mom > BA (vs < than BA) .090 (.068) .089 (.068) .102 (.114) .153 (.097) -.152 (.099) .140 (.096)    Dad  > BA (vs < than BA) .104* (.089) .103 (.090) -.012 (.076) .011 (.070) -.009 (.069) -.008 (.071)    Income .000 (.001) .000 (.001) .001? (.002) .001 (.001) .001 (.001) .001 (.001)    Depression Scale -.154** (.006) -.154** (.006) -.036** (.008) -.126** (.007) -.126** (.007) -.132** (.007)    Social Desirability W1   .337** (.028) .259** (.023) .258** (.023) .283** (.022) ? p ? .10; * p ? .05; ** p ? .01 Depression scale wave is the same as the masculinity wave         85    desirability are positively related. Although athleticism and concealing emotions increase feelings of social desirability, aggression actually decreases social desirability. Hypothesis 2.2.2 predicts that heterosexual adolescents are more likely to feel socially desirable than non-heterosexual adolescents. Results indicate that there is no difference in adolescents? perception of being socially desirable based on sexual identity.  Hypothesis 2.2.3 predicts that adolescents? sexual identity moderates the relationship between masculinity and socially desirability. None of the interaction terms for masculinity and sexual identity are significantly related to social desirability. Sexual identity does not moderate the relationship between dimensions of masculinity and social desirability.    Hypotheses 3.1.1 - 3.1.3: Life Satisfaction  Table 9 shows the results of the models testing the hypotheses concerning the influence that masculinity and sexual identity have on adolescents? life satisfaction. Hypothesis 3.1.1 states that masculinity is positively associated with overall life satisfaction. Aggression in both wave one (B = -.128) and wave two (B = -.126) is negatively associated with life satisfaction, as shown in models 1 and 3. A one-point increase in aggression corresponds to about a .130 decrease in life satisfaction. This indicates that as aggression increase, life satisfaction decreases. In contrast, as shown in models 1 and 3, athleticism in both wave one (B = .143) and wave two (B = .124) is positively associated with life satisfaction. This indicates that a one-point increase in athleticism corresponds to between a .124 and .143 increase in life satisfaction, thus indicating that adolescents who are high in athleticism report more life satisfaction. These findings both refute and support the prediction that performances of masculinity will be positively associated with life satisfaction. Specifically, aggression is related to lower life satisfaction while athleticism is related to higher life satisfaction.     86    TABLE 9: Unstandardized Slope Coefficients (Standard Errors) for the OLS Regression Models of Life Satisfaction  Regressed on Dimensions of Masculinity  Model 1 Model 2  Model 3 Model 4 Model 5 Dimensions of Masculinity      Wave 1         Concealing Emotions  .047 (.046) -.049 (.120)       Aggression -.128* (.050) -.111 (.177)       Athleticism  .143** (.028) .189 (.160)    Wave 2         Concealing Emotions    .035 (.042) -.059 (.109)     Aggression   -.126** (.047) -.140 (.197)     Athleticism    .124** (.033) .179 (.133)  Change Scores         Concealing emotions     -.017 (.035)    Aggression     .013 (.047)    Athleticism     -.010 (.031) Sexual Identity         Gay/Bisexual -.772? (.387)  -.736? (.375)  -.788? (.384) Interaction Terms         Conc. Em. X Heterosexual  .103 (.119)  .098 (.101)     Aggression X Heterosexual  -.017 (.177)  .015 (.210)     Athleticism X Heterosexual  -.049 (.161)  -.058 (.138)  Demographics         Birth year -.017 (.043) -.017 (.043) -.058 (.041) -.057 (.041) -.028 (.039)    White (vs other) .109 (.184) .107 (.186) .227 (.173) .228 (.174) .269 (.171)    More Urban (vs < Urban) .073 (.158) .073 (.159) .037 (.147) .037 (.147) .028 (.156)    Mom > BA (vs < than BA) .082 (.138) .083 (.139) .082 (.143) .081 (.144) .082 (.147)    Dad  > BA (vs < than BA) .023 (.135) .023 (.136) -.014 (.131) -.015 (.132) -.019 (.131)    Income -.001 (.001) -.001 (.001) -.001 (.001) -.001 (.001) -.000 (.001)    Depression Scale -.057** (.010) -.057** (.010) -.091** (.012) -.091** (.012) -.106** (.010) ? p ? .10; * p ? .05; ** p ? .01 Depression scale wave is the same as the Masculinity wave being examined    87     Hypothesis 3.1.2 predicts that heterosexuals will report higher life satisfaction than non-heterosexuals. In all models, non-heterosexual did report lower life satisfaction, though results were only marginally significant (p < .10).   Hypothesis 3.1.3 states that sexual identity moderates the relationship between performances of masculinity and life satisfaction. None of the interaction terms for masculinity and sexual identity are related to life satisfaction, indicating that sexual identity does not significantly moderate the relationship between adolescents? performances of masculinity and their life satisfaction.    Hypotheses 4.1.1 - 4.2.3: Risk-taking  This section focuses on the hypotheses concerning risk-taking. First, I will discuss the results of the models examining adolescents? desire to take risks. Second, the results of the models examining actual risk behaviours will be detailed. Desire to Take Risks   The results for models testing the hypotheses concerning the influence that masculinity and sexual identity have on adolescents? desire to engage in risk-taking are shown in Table 10. Hypothesis 4.1.1 predicts that masculinity is negatively associated with the desire to engage in risk-taking. Aggression in wave one, as shown in models 1 and 2, is positively associated with risk-taking in wave two (Log Odds = .156) and wave three (Log Odds = .105). As shown in model 5, aggression in wave two is positively associated with risk-taking in wave two (Log Odds = .181). Thus being high in aggression corresponds to between a 1.11 and 1.20 greater odds of desiring to take risks. As shown in model 6, aggression in wave two is also positively related to risk taking in wave three (Log Odds = .073), although this association is only marginally      88 TABLE 10: Log Odds (Standard Errors) for the Logistic Regression Models of Desire to Take Risks Regressed on Dimensions of Masculinity  Risk Taking W2 Risk Taking W3  Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Dimensions of Masculinity         Wave 1            Concealing Emotions .016 (.047)    -.006 (.039)       Aggression  .156** (.043)    .105* (.040)       Athleticism  .076** (.023)    .051?  (.027)    Wave 2            Concealing Emotions   .048 (.039) .024 (.082)   .004 (.050) -.009 (.079)     Aggression  .181** (.037) .129 (.177)   .073? (.044) -.074 (.199)     Athleticism   .077** (.023) .106 (.138)   .056* (.023) .047 (.090)  Change Scores            Concealing Emotions    .026 (.032)    .006 (.035)    Aggression    -.010 (.034)    -.040 (.031)    Athleticism    .000 (.024)    .004 (.028) Sexual Identity            Gay/Bisexual .157 (.298) .153 (.292)  .188 (.286) .173 (.231) .169 (.237)  .189 (.230) Interaction Terms            Conc. Em. X Heterosexual   .026 (.075)    .011 (.054)     Aggression X Heterosexual   .056 (.179)    -.001 (.191)     Athleticism X Heterosexual   -.029 (.139)    .009 (..092)  Demographics            Birth year .078* (.032) .092* (.035) .093* (.035) .098** (.034) .091* (.033) .094** (.034) .094** (.034) .099** (.032)    White (vs. other) .520** (.108) .536** (.116) .538** (.117) .521** (.110) .187 (.108) .192 (.121) .191 (.120) .183 (.118)    More Urban (vs. < Urban) .215? (.120) .209? (.118) .211? (.118) .235? (.121) -.048 (.100) -.047 (.100) -.047 (.100) -.033 (.100)    Mom > BA (vs. < than BA) .099 (.112) .125 (.111) .123 (.110) .118 (.113) -.040 (.129) -.033 (.135) -.033 (.135) -.032 (.130)    Dad  > BA (vs. < than BA) -.280* (.115) -.274* (.117) -.272* (.117) -.266* (.117) -.110 (.149) -.106 (.148) -.106 (.147) -.095 (.150)    Income .000 (.001) .000 (.001) .000 (.001) .000 (.001) .001 (.001) .001 (.001) .001 (.001) .001 (.001)    Depression Scale .021* (.010) .033 (.009) .033** (.009) .030** (.009) .001 (.008) .005 (.009) .005 (.009) .004 (.009)    Risk Taking W2     .842** (.103) .848** (.107) .848** (.107) .881** (.103)  ?p ? .10; * p ? .05; ** p ? .01  Depression scale wave is the same as the masculinity wave     89  significant. This represents a 1.08 greater odds of desiring to take risks. Similarly, as shown in models 1, 2, and 6, athleticism in wave one is positively associated with risk taking in wave two (Log Odds = .076) while athleticism in wave two is positively associated with risk taking in waves two (Log Odds = .077) and three (Log Odds = .056). This means that athleticism increases the odds of taking risk by between 1.06 and 1.08 times. As shown in model 5, athleticism in wave one is positively associated with risk taking in wave three (Log Odds = .051), although this relationship is only marginally significant. This corresponds to an odds ratio of 1.05. These results indicate that as aggression and athleticism increase, adolescents are more likely to indicate that they desire to take risks. Aggression, however, seems to have a stronger influence on adolescents? desire to take risks. Concealing emotions, in contrast, does not influence adolescents? desire to take risks. These findings do not support the hypothesis that there is a negative relationship between performances of masculinity and adolescents? desire to take risks. In contrast, results indicate that adolescents? desire to take risks is positively associated with aggression and athleticism.   Hypothesis 4.1.2 predicts that heterosexuals will report a lower desire to engage in risk-taking than non-heterosexuals. The results, however, indicate that sexual identity is not related to the desire to engage in risk-taking.   Hypothesis 4.1.3 states that adolescents? sexual identity moderates the relationship between masculinity and the desire to engage in risk-taking. None of the interaction terms for masculinity and sexual identity are significantly related to risk-taking. These results indicate that sexual identity does not moderate the relationship between dimensions of masculinity and risk-taking.      90  Engaging in Risk Behaviours The results of models testing the hypotheses concerning the influence that masculinity and sexual identity have on adolescents? engagement in frequent drunkenness are shown in Table 11. Hypothesis 4.2.1 states that masculinity is negatively associated with drunkenness. Aggression in wave one, as shown in models 1 and 2, is positively associated with frequent drunkenness in wave two (Log Odds = .330) and wave three (Log Odds = .074). These indicate that aggression, respectively, corresponds to 1.39 and 1.08 greater odds of engaging in frequent drunkenness. Similarly, wave two aggression is positively associated with frequent drunkenness in waves two (Log Odds = .298) and three (Log Odds = .164), as shown in models 5 and 6. These findings indicate that aggression increases the odds of engaging in frequent drunkenness by 1.35 and 1.18 times. Together, these results indicate that the higher adolescents? aggression, the more likely they are to report frequent drunkenness. As shown in model 4, the aggression change score, however, is negatively associated with frequent drunkenness in wave two (Log Odds = -.079). This corresponds to an odds ratio of .92, thereby indicating that adolescents high in aggression are .08 times less likely to engage in frequent drunkenness. This indicates increasing aggression from wave one to wave two is associated with less frequent drunkenness in wave two. As shown in model 5, athleticism in wave one is positively related to frequent drunkenness in wave three (Log Odds = .050), although this result is only marginally significant (p < .10). This shows that athletic adolescents have 1.05 higher odds of engaging in frequent drunkenness. Concealing emotions is not related to frequent drunkenness. These results do not support the hypothesis that masculinity is negatively associated with drunkenness. In contrast, the results indicate that adolescents? aggression is positively associated with frequent drunkenness.       91 TABLE 11: Log Odds (Standard Errors) for the Logistic Regression Models of Frequent Drunkenness Regressed on Dimension of Masculinity  Frequent Drunkenness W2 Frequent Drunkenness W3  Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Dimensions of Masculinity         Wave 1            Concealing Emotions  -.012 (.045)    -.063 (.040)       Aggression .330** (.036)    .074* (.037)       Athleticism  .008 (.026)    .050? (.056)    Wave 2            Concealing Emotions   .006 (.044) -.008 (.096)   -.076? (.041) -.072 (.079)     Aggression  .298** (.040) .214 (.254)   .164** (.049) .077 (.196)     Athleticism   -.026 (.029) -.025 (.140)   .010 (.022) .062 (.097)  Change Scores            Concealing Emotions    .026 (.035)    -.016 (.029)    Aggression    -.079* (.032)    .047 (.032)    Athleticism    -.034 (.031)    -.039 (.024) Sexual Identity            Gay/Bisexual .259 (.272) .284 (.376)  .271 (.351) -.267 (.290) -.257 (.290)  -.238 (.251) Interaction Terms            Conc. Em. X Heterosexual   .016 (.087)    -.005 (.060)     Aggression X Heterosexual   .088 (.271)    .091 (.184)     Athleticism X Heterosexual   -.002 (.137)    -.054 (.102)  Demographics            Birth year -.310** (.050) -.267** (.049) -.268** (.049) -.279** (.045) .111** (.031) .127** (.032) .127** (.032) .142** (.032)    White (vs. other) .602** (.156) .651** (.151) .650** (.153) .682** (.149) .642** (.168) .644** (.161) .645** (.163) .706** (.166)    More Urban (vs. < Urban) -.031 (.133) -.005 (.140) -.007 (.139) .018 (.125) -.043 (.123) -.058 (.124) -.058 (.124) -.034 (.120)    Mom > BA (vs. < than BA) -.034 (.168) -.036 (.160) -.037 (.160) -.048 (.161) .217 (.123) .225 (.138) .225 (.138) .228? (.132)    Dad  > BA (vs. < than BA) -.076 (.176) -.095 (.176) -.096 (.178) -.100 (.178) .105 (.143) .118 (.142) .119 (.143) .117 (.144)    Income .001 (.001) .001 (.001) .001 (.001) .002? (.001) .004* (.001) .004* (.001) .004* (.002) .004** (.001)    Depression Scale .007 (.009) .016? (.008) .016? (.008) .026** (.008) -.014? (.009) -.016* (.007) -.016* (.008) .009 (.009)    Drunk W2     .771** (.145) .750** (.146) .749** (.146) .812** (.132) ? p ? .10; * p ? .05; ** p ? .01 Depression scale wave is the same as the masculinity wave       92  Hypothesis 4.2.2 predicts that heterosexuals will be less likely to report frequent drunkenness than non-heterosexuals. Sexual identity is unrelated to drunkenness, thus these results do not support this hypothesis.  Hypothesis 4.2.3 predicts that adolescents? sexual identity moderates the relationship between masculinity and drunkenness. None of the interaction terms for masculinity and sexual identity are significantly related to frequent drunkenness. This indicates that sexual identity does not moderate the relationship between dimensions of masculinity and frequent drunkenness.         93  Chapter 5: Conclusion Men?s drive to demonstrate and defend their masculinity influences their interactions with others, the choices they make, and their life experiences. Given that masculinity is a pervasive influence on men?s lives (Courtenay, 2000; Harper & Harris, 2010; Kimmel, 2008; Pascoe, 2007; Shoveller et al., 2009) it is surprising that sociologists have overlooked masculinity?s influence on men?s lives to such as a degree that men have been called the ignored gender (Harper & Harris, 2010). Feminist scholars have explored and re-examined women?s experiences and femininity, yet men?s experiences and masculinity have been largely neglected. When masculinity is examined, it is often assumed that all men are masculine and that masculinity benefits all men (Bottorff, Oliffe, & Kelly, 2012). This ignores the variation that exists in men?s performances of masculinity as well as fails to take into account that masculinity is both a positive and a negative influence on men?s life experiences. Overall, researchers have often oversimplified the diversity of men?s experiences. This dissertation addresses these problems by exploring adolescent men?s masculinity, sexual identity, and life experiences.   Masculinity, Sexual Identity, and Life Experiences I add to the literature on masculinity in three ways. First, I use existing literature to argue that adolescent men earn their masculinity through performances of masculinity that show their support for the ideologies of the Guy Code. The three most salient sets of behaviours that demonstrate masculinity during adolescence are concealing emotions, aggression, and athleticism. Second, I argue that adolescents? sexual identity will be unrelated to their performances of masculinity. The taken-for-granted assumption that non-heterosexuals are less masculine than heterosexuals are has developed because homophobia is used to police masculinity. Third, I argue that masculinity and sexual identity will influence adolescents? life     94  experiences. This included examining both the direct impact of masculinity and sexual identity on life experiences as well as how sexual identity moderates the relationship between masculinity and life experiences. I elaborate on these contributions to the literature on masculinity below. Masculinity is about the Guy Code To being with, I define masculinity during adolescence. Adolescence is an ideal period in which to examine masculinity given that high school is a critical site for the development of gender identities (Frosh, Phoenix, & Pattman, 2002; Mac an Ghaill, 1994; Morrell, 1998) and corresponds with greater expectations concerning performances of masculinity (Mac an Ghaill, 1994; Kimmel, 2008; Morrell, 1998; Frosh et al., 2002). Moreover, it is during adolescence that individuals form their own identity (Erikson, 1963) and develop increasingly intimate relationships with peers (Parker et al., 2006). Adolescent men?s performances of masculinity affect the foundations of identity development as well as their peer relationships (Kimmel, 2008; Pascoe, 2007). Given the pervasive influence of masculinity during adolescence, it is a critical period for examining masculinity and its impact on men?s life experiences.  Masculinity during adolescence refers to how much adolescents? performances of masculinity conform to the ideologies of the Guy Code (Connell, 1987; Kimmel, 2008; Schrock & Schwalbe, 2009). These performances lead adolescents to appear as either more or less masculine. There are three key elements to my definition of masculinity. First, masculinity is often about men competing with one another in order to earn their masculinity. Accordingly, the focus should not be solely on how masculinity is used to subordinate women. Second, masculinity is not one set of behaviours, but rather a product of several groups of behaviours that, taken together, encompass overall masculinity (e.g., Barron, 1976; Kimmel, 2008; Schrock     95  & Schwalbe, 2009). During adolescence, the three most salient sets of behaviour that demonstrate masculinity are concealing emotions, aggression, and athleticism. Third, adolescent men vary considerably regarding how masculine their performances are. That is, not all men are masculine since masculinity must be earned through men?s interactions with others.  Masculinity and Sexual Identity are not the Same  Next, I re-examine and test the taken-for-granted association between sexual identity and masculinity. According to stereotypes and research (e.g., Bailey & Zucker, 1995; Udry & Chantala, 2002; Whitam & Zent, 1984), non-heterosexual men are less masculine than heterosexual men. I argue that heterosexuals and non-heterosexuals are equally masculine. There is no reason that sexual identity should influence masculinity during adolescence. All men, irrespective of their sexual identity, become socialized into the ideologies of the Guy Code and are expected to act in masculine ways. The use of homophobia in the policing of masculinity leads people to ascribe less masculinity to non-heterosexual men, regardless of their actual performances of masculinity. The assumption that non-heterosexuals are less masculine than heterosexuals leads to biases in the perceptions of gay and bisexual men?s masculinity (Gottschalk, 2003; Ross, 1980). Previous researchers have not taken the socially constructed link between sexual identity and masculinity into account, thus results are biased in a way that portrays non-heterosexuals as less masculine than they are (Gottschalk, 2003; Ross, 1980).   The current study finds that there are no differences in heterosexual and non-heterosexual men?s performances of masculinity. Adolescent men, regardless of their own sexual identity, give similar performances of masculinity. This finding contradicts previous research (e.g., Bailey & Zucker, 1995; Udry & Chantala, 2002; Whitam & Zent, 1984) as well as taken-for-granted assumptions regarding gay and bisexual men?s performances of masculinity. Due to the     96  pervasive influence of masculinity in men?s lives, most men, irrespective of their sexual identity, conform to the expectations of the Guy Code. The spectre of the fag, however, leads to the ascription of less masculinity to non-heterosexuals. Once non-heterosexuals are ascribed as less masculine, people?s perceptions are biased in a way that affirms heterosexual?s masculinity while at the same time non-heterosexual men?s masculinity is rejected or ignored. As such, scholars must endevour to challenge these biases through basing their research on sociological theory that challenges the homophobic stereotype that gay and bisexual adolescents are less masculine than their heterosexual counterparts.  Masculinity and Sexual Impact Men?s Life Experiences Lastly, I tested the impact that masculinity and sexual identity have on adolescent men?s life experiences. To accomplish this, I developed a framework that sets out how masculinity and sexual identity influences men?s life experiences. With reference to the effect of masculinity, I argue that since masculinity is a limited resource and thus not all adolescent men can be masculine (Kimmel, 2008). This means that adolescents must police one another?s performances of masculinity as a means of earning and defending their masculinity relative to their peers (Kimmel, 2008; Pascoe, 2007). Adolescents who are able to demonstrate their masculinity gain status and power. In contrast, adolescents? who are unable to demonstrate their masculinity are labelled as unmasculine through homophobic labels. These labels lead adolescents to be stigmatized and devalued by both themselves and their peers, thereby decreasing their status and power relative to their masculine peers. The power and status adolescents gain or lose based on their performances of masculinity will influence their life experiences. In particular, those who are labelled as masculine will have more positive life experiences while those who are labelled as unmasculine will have more negative life experiences.      97  Regarding the influence of sexual identity on life experiences, gay and bisexual adolescents, compared to heterosexual adolescents, report more marginalization, discrimination, and victimization due to the stigma associated with being non-heterosexual (e.g., Banks, 2003; Lindley et al., 2011). This has a negative influence on non-heterosexual adolescents? life experiences. Moreover, while adolescents develop their identity as gay or bisexual they know that their non-heterosexual identity is devalued socially. This leads many non-heterosexual adolescents to devalue themselves and hide their identity through engaging in information management (Goffman, 1959; Morrow, 1993). For gay and bisexual adolescent men, the experiences of devaluation are particularly challenging given that an identity as non-heterosexual leads to the ascription of less masculinity. That is, developing an identity as gay or bisexual is associated with being, de facto, less masculine. Non-heterosexuals become doubly marginalized because of both their sexual identity and their lower ascribed masculinity. Overall, gay and bisexual adolescents, compared to their heterosexual peers, experience more negative life experiences.   Masculinity and sexual identity influence various facets of adolescent men?s lives. However, I focused on three life experiences that are particularly salient during adolescence. The first life experience I examined was adolescents? relationships with their peers since the development and maintenance of peer relationships is a major developmental task (Parker, Rubin, Erath, Wojslawowicz, & Burkirk, 2006). The second life experience I explored was life satisfaction. Testing life satisfaction allowed me to examine how pervasive the influence of masculinity and sexual identity are on adolescents? lives. That is, if masculinity and sexual identity are a pervasive influence on men?s lives, then they should influence overall life satisfaction. The third life experience I focused on was risk-taking, which was assessed as both     98  attitudes towards taking risks as well as actual risk-taking behaviours. It was important to examine both attitudes and behaviours since attitudes do not always correspond with actual behaviours (Kraus, 1995; La Piere, 1934; Wicker, 1969). I detail the influence of masculinity and sexual identity on each of these life experiences below.  Peer Relationships  Adolescents? ability to create and maintain peer relationships is vital for their psychological and social development. Performances of masculinity influence adolescent men?s ability to sustain friendships by increasing their desirability among peers (Friedman, et al., 2006; Kimmel, 2008; Pascoe, 2007; Udry & Chantala, 2002). In contrast, those who are labelled as less masculine are more likely to experience problematic peer relationships (Pascoe, 2007). Problematic peer relationships challenge adolescents? ability to develop the social skills needed for future platonic and romantic relationships (e.g., Bierman, 2003; Leary, Kock, & Hechenbleiker, 2001). Similarly, adolescents? sexual identity influences peer relationships through making adolescents? appear as more or less desirable to their peers. Non-heterosexual adolescents, for instance, tend to experience more problematic peer relationships (Banks, 2003) and often must cope with their fear of being rejected by hiding their identity until after graduation (D'Augelli, 2006, Morrow, 1993). Even when adolescents withhold their non-heterosexual identity from peers, they experience negative peer interaction due to their own perception of being devalued as well as the stress associated with having to hide their identity. Based on these arguments, I hypothesized that adolescents? masculinity and sexual identity will influence their peer relationships. Specifically, I predicted that masculinity positively influences adolescents? relationships with their peers. I also hypothesized that heterosexuals will report     99  more positive peer relationships than non-heterosexuals and that sexual identity will moderate the relationship between masculinity and peer relationships.  Adolescents? performances of masculinity do influence their relationships with their peers, though not always in the directions predicted. Athleticism, and concealing emotions to a lesser degree, is associated with adolescents? perception of more positive relationships with others. In contrast, aggression negatively influences adolescents? relationships with their peers. Thus, while athleticism (and concealing emotions to a much lesser degree) is beneficial to adolescents? relationships with their peers, aggression is detrimental. This indicates that masculinity is both a positive and a negative influence on men?s life experiences (Kimmel, 2008). Athleticism may be beneficial because it gives adolescents the opportunity to earn status and power among peers. As such, athleticism offers many chances for adolescent men to raise their own reputations and social standings relative to their peers. Athleticism seems to be one of the most important means through which adolescents earn positive approval from peers (Kimmel, 2008; Pascoe, 2007) For instance, the most popular adolescent men, for instance, are often on sports teams (Pascoe, 2007). This may explain why athleticism is the dimension of masculinity that is the most consistently related to better peer relationships. In contrast, aggression may hinder adolescents? ability to develop positive reputations due to peers? avoidance of those who engage in physical and verbal aggression. That is, adolescents who get into altercations with others are likely to be ostracized. When youth use aggression they are more likely to be rejected by their peers, particularly when the aggression is disruptive and does not serve a purpose, such as being used to defend oneself (Bierman, 2003; Perry et al., 1992). It is surprising that concealing emotions does not influence peer relationships. Based on previous research on masculinity, it seemed likely that adolescent men who do not conceal their emotions     100  should be more likely to be rejected by their peers. However, this was not found in the current study. I will discuss why this may be in the limitations section. Adolescents, regardless of their sexual identity, have similar perceptions of their relationships with their peers. This is somewhat surprising given the extensive literature documenting gay and bisexual adolescents? negative experiences with their peers (e.g., Banks, 2003). However, it is unknown if non-heterosexual adolescents? are out to their peers, which would influence how positive or negative their experiences with peers are. Many adolescents, for instance, withhold their sexual identity from their peers as a way to avoid being rejected (D'Augelli, 2006, Morrow, 1993). Adolescents? sexual identity also did not moderate the relationship between masculinity and peer relationships. This indicates that masculinity has a similar influence on adolescents peer relationships irrespective of their sexual identity. This, again, may be due to the fact adolescents are not out to their classmates.   Life Satisfaction Life satisfaction assesses how satisfied adolescents feel about their lives. The sheer range of outcomes associated with masculinity and sexual identity (e.g., Courtenay, 2001; Kimmel, 2008; Landolt et al., 2004; Munren & Smolak, 1997; Yunger, Carver, & Perry, 2004) means that life satisfaction is an ideal outcome to examine. That is, testing the influence of masculinity and sexual identity on life satisfaction determines if masculinity and sexuality are, indeed, pervasive influences on men?s lives. Masculinity and heterosexuality are related to power and status while unmasculinity and non-heterosexuality are related to stigmatization. Therefore, masculine and heterosexual adolescents will experiences more positive life experiences than less masculine and non-heterosexual adolescents. Based on these arguments, I predicted that adolescents? performances of masculinity will positively influence their life satisfaction. In addition, I     101  hypothesized that heterosexuals will report higher levels of life satisfaction and that sexual identity will moderate the relationship between masculinity and life satisfaction.   Adolescents? athleticism is positively associated with life satisfaction. The more athletic adolescent men are, the higher their life satisfaction. This, like the results for peer relationships, demonstrates the beneficial nature of athleticism (Kimmel, 2008; Pascoe, 2007). Specifically, they replicate the finding that athleticism is a positive influence on adolescent men?s life experiences. Again, this may be because athleticism offers opportunities to increase social status through increasing one?s reputation and creating opportunities to interact with other popular peers. Aggression, on the other hand, is negatively associated with life satisfaction. The higher adolescents? aggression, the less satisfied they are with their lives. The findings that aggression decreases life satisfaction may stem from adolescents? motivations for engaging in physical and verbal aggression. Aggressive adolescents may be less satisfied with their lives, and they externalized these feelings through aggression. Furthermore, this aggression may lead to negative outcomes, such as less positive peer relationships, which will then further increase adolescents? aggressive tendencies. Surprisingly, concealing emotions was unrelated to life satisfaction. As mentioned, I will discuss why this may be in a later section on limitations. Overall, these findings indicate that while the athleticism dimension of masculinity increases life satisfaction, the aggression dimension of masculinity decreases it. As found with adolescents? peer relationships, these findings indicate that masculinity is both beneficial and detrimental to men?s life experiences.   As expected, heterosexual adolescents reported higher levels of life satisfaction that non-heterosexual adolescents, though these results were only marginally significant. This finding affirms the challenges that gay and bisexual adolescents experience (e.g., Banks, 2003),     102  particularly given that non-heterosexuals are indicating lower life satisfaction while in late adolescence and young adulthood. Moreover, non-heterosexuals? lower life satisfaction was found even while controlling for adolescents? performances of masculinity and feelings of depression, thus indicating that lower life satisfaction may occur independently of both these variables. Given the importance of this period for future well-being and the development of a positive identity, it is discouraging that non-heterosexuals are already feeling dissatisfied compared to their heterosexual counterparts.  Sexual identity did not moderate the association between masculinity and life satisfaction, thereby indicating that sexual identity does not influence the impact that masculinity has of life satisfaction. This means that masculinity influences adolescents? life satisfaction similarly regardless of their specific sexual identity.  Risk-taking Courtenay (2001) argues that men?s performances of masculinity will influence their desire to engage in risk-taking. Masculine men engage in risk-taking as a means of earning and defending their masculinity. In contrast, when men are laballed as less masculine, they engage in risk-taking as a means of combating the label of unmasculine so that they can earn back the power and status associated with being masculine. However, combating the label of unmasculine is very difficult, meaning that less masculine men must engage in increasingly risky behaviors (i.e., masculine behaviours) in order to earn back their masculinity. Thus, Courtenay (2001) predicts a negative relationship between masculinity and risk-taking, such that those deemed the least masculine will engage in the most risk-taking. Based on this work I developed hypotheses concerning adolescents? risk-taking. In particular, I hypothesized that adolescents? performances of masculinity will be negatively associated risk-taking. Moreover, I predicted that heterosexual     103  adolescents will have lower risk-taking than non-heterosexuals and that sexual identity will moderate the relationships between masculinity and risk-taking. Risk-taking was measured as adolescents? desire to take risks as well as their engagement in actual risk behaviours Courtenay?s (2001) arguments were not supported. The findings from this study indicate that masculinity actually increases adolescents? risk-taking. In particular, athleticism increases adolescents? desire to take risks yet it is unrelated to frequent drunkenness. Aggression, in contrast, increases both adolescents? desire to take risks as well as their engagement in frequent drunkenness. These results indicate that the athleticism and aggression dimensions of masculinity increase adolescents? desire to take risks, though only aggression is related to actual risk taking. Therefore, contrary to Courtenay, I find that masculinity increases adolescent men?s risk-taking. These findings shows that those who are less masculine are not compensating by desiring to, or actually engaging in, more risky behaviours. Adolescents? sexual identity also does not influence risk taking, nor does sexual identity moderate the relationship between masculinity and risk taking. This indicates that non-heterosexual men, compared to heterosexual men, are not taking more risks as a means of combating their lower ascribed masculinity.  Interestingly, however, adolescents who increased their aggression from wave one to wave two were less likely to engage in frequent drunkenness. That is, regardless of their actual level of aggression, increasing aggression from wave one to wave two leads to lower risk-taking behaviours. One explanation may be that the increase in aggression represents effective aggression (Bierman, 2003; Perry et al., 1992). Effective aggression is used to control/coerce others or to avoid being controlled. This sort of aggression is not related to negative outcomes since it serves a purpose and can actually increase social desirability. The increased social desirability may counter-act the need to engage in risk-taking.     104  Similar to the previous results, the findings from these hypotheses indicate that athleticism and aggression differently influence adolescent men?s life experiences. Athleticism increases adolescents? desire to take risks but not actual risk behaviours, while aggression increases both the desire to take risks as well as actual risk-taking. This indicates that athleticism, but not aggression, has protective factors in that even though it increases the desire to take risks it protects adolescents against actual risk-taking that could have long-term negative consequences. This may occur because athleticism increases status and power while aggression increases ostracism and rejection (Bierman, 2003; Pascoe, 2007; Perry et al., 1992). As such, athleticism may increase adolescents? desire to take risks, but the expression of this desire is only about defending and earning masculinity. As such, it does not increase actual risk-taking. In contrast, aggression increases both the desire to take risks and actual risk-taking because those who are high in aggression are desperately attempting to earn the respect and admiration of their peers so that they can be accepted.  Implications of the Findings There are several broader implications from the current study. First, masculinity is a multidimensional concept and should be measured as such. Second, sexual identity and masculinity are unrelated to one another. Third, masculinity is not synonymous with being a man, thus sociologists must take into account differences in men?s performances of masculinity when examining men?s life experiences.   1. Masculinity Must Be Treated As Multidimensional This study demonstrated the importance of conceptualizing performances of masculinity as several different dimensions. That is, adolescent men demonstrate their masculinity by engaging in sets of behaviours, each of which have a different influence on life experiences.     105  Athleticism is beneficial to men?s life experiences while aggression is detrimental. Athleticism is related to more positive perceptions of peer relationships and to higher ratings of life satisfaction. In fact, in reference to social desirability, becoming more athletic further increased adolescents? feelings of desirability. Particularly interesting is that even though athleticism was related to an increased desire to take risks, it was not related to actual risk behaviours. As discussed, athletic adolescents may desire to take risks as a means to demonstrate their masculinity, yet avoid engaging is overly risky behaviours. In this sense, athleticism is actually a protective factor in relation to adolescents? engagement in certain risk behaviours. In contrast, aggression is detrimental to adolescents? life experiences. Aggression negatively influenced peer relationships and life satisfaction, and increased both the desire to engage in risk-taking as well as actual risk-taking behaviour.  The differences found concerning the influence that athleticism and aggression have on adolescents? life experiences are remarkable since they indicate that each dimension of masculinity has a unique effect on men?s live experiences (Kimmel, 2008; Harper & Harris, 2010). The implication of this is that researchers exploring men and masculinity must take into account the various dimensions of masculinity. Problematic, however, is that many scholars do not actually do this (e.g., Udry & Chantala, 2002). Results that are based on one-dimensional measures of masculinity distort the relationship between masculinity and life experiences. Additionally, given that there are multiple dimensions of masculinity, it is likely that some men are more masculine in certain domains and less masculine in others. Conceivably, as shown by the positive influence of athleticism and negative influence of aggression, being more masculine in different dimensions may lead to different life experiences. Men who demonstrate their masculinity through athleticism, for instance, will have much better life experiences than men     106  who demonstrate their masculinity through aggression. Overall, researchers must take the multiple dimensions of masculinity into account when examining men?s experiences.  2. Sexual Identity and Masculinity Are Distinct Concepts Throughout this dissertation, I have argued that there is no reason that gay and bisexual men should be less masculine than their heterosexual peers. I directly tested if sexual identity and masculinity are unrelated as well as tested if adolescents? sexual identity moderates the relationship between masculinity and life experiences. Thus, I examined both the direct relationship between sexual identity and masculinity as well as the indirect influence that sexual identity has on the relationship between masculinity and various outcomes.  Adolescents? sexual identity does not influence their performances of masculinity. Moreover, sexual identity does not moderate the relationships between masculinity and life experiences. Overall, sexual identity, at least during adolescence, does not influence masculinity and its impact on life experiences. Therefore, the current study challenges the taken-for-granted assumption that sexual identity influences masculinity. In contrast, this study supports the argument that the relationship between sexual identity and masculinity is socially constructed and predicated on the marginalization of non-heterosexuals through homophobia. All men, regardless of sexuality, ascribe to the ideology of the Guy Code and tend to prize masculinity (e.g., Hennen, 2005; Mosher et al., 2006).   The implication of these findings is that scholars should critically re-examining the taken-for-granted association between masculinity and sexual identity, rather than perpetuate it through research that is methodologically and theoretically flawed. Future research should explore the commonalities between men?s performances of masculinity, rather than attempt to find differences that are based in stereotypes rather than theory (Schrock & Schwalbe, 2009).     107  Challenging and re-examining the relationship between sexual identity and masculinity may help combat homophobia. Specifically, educating adolescents about the spurious link between masculinity and sexual identity may be an important means of both challenging homophobia and easing men?s fears regarding their own performances of masculinity. By becoming aware that gay and bisexual men are equally masculine, adolescents may be less inclined to use homophobia in the policing of masculinity. This study adds to a developing literature (e.g., Kimmel, 2008; Pascoe, 2007) that explores the influence that homophobia has on men?s performances of masculinity.    3. Explorations of Men?s Life Experiences must Include Masculinity Researchers have recently begun to theorize about how differences in men?s performances of masculinity influence their life experiences (e.g., Courtenay, 2001; Kimmel, 2004, 2008; Pascoe, 2007). However, few sociological studies have empirically examined how differences in men?s performances of masculinity are related to their life experiences. Instead, when scholars explore the experiences of men they often assume that all men are masculine. That is, they erroneously equate sex with gender (Bottorff, Oliffe, & Kelly, 2012; Schrock & Schwalbe, 2009). As shown in this study, differences in adolescent men?s performances of masculinity lead to different life experiences. By failing to take into account differences in adolescent men?s performances of masculinity an important influence on men?s life experiences would have been overlooked. The implication of my research is that when researchers examine men?s life experiences the focus should be on performances of masculinity rather than sex as a biological man. Researchers must focus on differences between men, rather than comparisons between men and women. Comparisons between men and women oversimplify men?s (and women?s) experiences     108  (Courtenay, 2001). Men?s performances of masculinity, for instance, influence their likelihood of seeking preventative care (Springer & Mouzon, 2011). Specifically, masculinity is associated with not receiving preventative care and negates the benefits that socioeconomic status has on seeking preventative care. Future research must focus on masculinity, and not just biological sex, when attempting to understand men?s life experiences. Limitations  Like many studies that explore the experiences of non-heterosexuals (e.g., Russell, Seif, & Truong, 2001), the sample size of gay and bisexual adolescents is quite small. A small sample size increases the chances of making type II errors ? indicating a failure to reject the null hypothesis. This, however also means that any significant results are found under more conservative conditions. Moreover, given that this is a representative sample of adolescents and that this area remains under-researched, the small sample size should not deter needed contributions.  Ideally, non-heterosexuals would be separated into two groups (gay adolescents and bisexual adolescents) rather than be grouped into one category. However, due to the small sample size of the non-heterosexual adolescents, this was not possible. The combining of all non-heterosexual adolescents into one group has the potential to affect the findings. Specifically, if gay men are indeed less masculine than heterosexual and bisexual men, having bisexuals in the same category may bias results in a way that minimizes gay men?s lower masculinity. This, however, is not the case since gay men far outnumbered bisexual men. Therefore, even if bisexuals are more masculine than gay men, this would not have changed the findings dramatically. In addition, as a precaution, I ran analyses with each of these groups individually and results followed a similar pattern.      109  The sample of non-heterosexual men may not be representative of all gay and bisexual adolescent men. Specifically, my non-heterosexual sample may be biased because it includes those who are the most comfortable with their sexual identity. The question I use to ascertain sexual identity asks respondents to acknowledge their sexual orientation. Not all men will be comfortable telling a researcher their true sexual identity, thus those who indicate their non-heterosexual identity may be more comfortable with their identity than those who chose to withhold it. Moreover, men vary concerning the age that they come to accept their non-heterosexual identity (Cass, 1979, 1984; Morrow, 1993). Since the men are in late adolescence and early adulthood when asked to indicate their sexual identity, it is possible that I have a selection bias towards those adolescents who have accepted their non-heterosexual identity earliest. As such, my measure of sexual identity may be selecting adolescents who are most comfortable, and hence confident, with their sexual identity.  The dimension of masculinity termed ?concealing emotions? may tap into more than just adolescents? concealment of emotions. Specifically, the questions in this dimension measure respondents? experience of negative affective states, such as being moody, sad, or fearful. It is likely, as I argued, that masculine respondents rate themselves lower on these factors due to their desire to appear masculine. However, it would have been more advantageous if the questions more precisely measured the purposeful concealment of emotion. This could be accomplished by asking adolescents if they actively attempt to conceal emotions such as sadness or fear from others. I accounted for the potential of this measure to be tapping affective states (rather than concealing emotions) by controlling for depression in all analyses. Although the results did control for symptoms of depression, the problems associated with this particular dimension of     110  masculinity may explain why the relationship between concealing emotions and life experiences was inconsistent.  Another limitation is that the Cronbach?s alphas obtained for the dimensions of masculinity are quite low. This indicates that the internal consistency of each dimension of masculinity is low, and thus my measures of the dimensions of masculinity may not be reliably assessing a single construct. However, this may also be due to the low number of items in each dimension as well as the fact that items within the same dimension use different scales. Furthermore, factor analysis consistently revealed the same three dimensions in both wave one and wave two. This is strong evidence that even though the alphas are low, these items are indeed measuring three dimensions of masculinity. Finally, the data and theories used in this dissertation are based on the experiences of American adolescents. Moreover, the ADD Health data was obtained from adolescents in the 1990s and early 2000s. Context will influence the relationships between masculinity, sexual identity, and life experiences during adolescence. Consequently, the findings from this dissertation may not represent the experiences of adolescent men in other contexts (e.g., contemporary Canada).  Though there are limitations in the current study, there are also several strengths that should be acknowledged. First, this research is based on theory, rather than taken-for-granted assumptions regarding men, masculinity, and sexual identity. Specifically, I assimilated emerging theories and scholarship concerning masculinity and sexual identity during adolescence (e.g., Courtenay, 2001; Kimmel, 2008; Pascoe, 2007). From this literature, I defined masculinity as a multidimensional construct, examined the relationship between masculinity and sexual identity, and explored how masculinity and sexual identity influence adolescent men?s life     111  experiences. By using theory as the basis for my empirical research, I was able to challenge various assumptions, such as that all men are masculine, that masculinity is universally beneficial to all men, and that non-heterosexual men are less masculine than heterosexual men. Overall, the use of theory to frame my empirical research allowed for a more detailed understanding of masculinity on its impact on men?s lives.  Second, I was able extend Kimmel (2008) and Pascoe?s (2007) research by demonstrating that their theories apply to a representative sample of adolescent men. This is an important addition to the literature on masculinity since these qualitative studies were based on non-representative samples of men. Specifically, Kimmel?s (2008) interviews were conducted on young adult men from white, middle class and educated backgrounds while Pascoe?s (2007) ethnographic study was conducted using a school in a more rural and lower socioeconomic area. These works contributed greatly to the development of theory regarding masculinity, but the utility of these theories for understanding the experiences of men in general was unclear. However, by using this emerging literature as the basis for this study I demonstrated that their findings are generalizable. As argued by Ragin (1994), an important component of social research is the use of both qualitative and quantitative studies to develop and test theories. Specifically, he argues that qualitative studies are useful to generate theories, which can then be used to generate and test hypotheses in quantitative research. This study attempted to accomplish the latter of these goals. Future Research In this section, I offer three suggestions for future research. First, future research should further explore how performances of masculinity influence adolescents? relationships with their peers. In particular, my measures focused on the quality of relationships, rather than on objective     112  measures of the relationships. Subjective measures are difficult to compare since rejected adolescents may have fewer expectations of their friends than popular adolescents (Parker & Asher, 1993). Masculine adolescents, for instance, may rate their friendships similarly to those who are labelled as unmasculine since each group is using different standards to rate their friendships. As such, both groups will give similar subjective evaluations of their friends even through there may be significant differences in their actual relationships. Future researchers should attempt to use objective measures of peer relationships, such as the number of friends or the sex composition of friends. These measures would allow researchers to examine both the density and composition of adolescents? peer networks as well as how these relate to adolescents? performances of masculinity. Measures such as these have been used in studies examining the influence of depression on adolescents? peer relationships (Hatzenbuehler, McLaughlin, & Xuan, 2012) and have proven to be quite useful.  Second, this study focused on performances of masculinity, and thus examinations of femininity were absent. Masculinity and femininity are not necessarily opposite sides of the same scale. It is important for future researchers to examine how performances of femininity influence adolescent men?s life experiences as well. The creation of a measure of femininity would be the best way to accomplish this. By creating a measure of femininity, adolescents? performances of both masculinity and femininity can be measured. It is conceivable that adolescents can be both masculine and feminine, or high in one and low in the other. Differences in adolescents? performances of masculinity and femininity will likely lead to different life experiences. Research that examines both dimensions at once appears to be quite rare, so research in this area is needed.      113  Third, as I discussed previously, the data and theories used in this dissertation are based on the experiences of American adolescents who are now in their thirties. The findings from this dissertation may not represent the experiences of adolescent men in other contexts. Canadian scholars have been examining masculinity (e.g., Lafrance as cited in Nebenzahl, 2012; MacDonald, 2012; Synnott, 2001, 2009) as well as the experiences of GLBT individuals (e.g., Adam, 2004; Brennan, Ross, Dobinson, Beldhuizen, & Steele, 2010; Fetner, Elafros, Bortolin, & Drechsler, 2012; Fumia, 2010; Green, 2008, 2011). Though many of the findings are similar to those found in American research, there are important differences. Canadians, for instance, are become increasingly more accepting of non-heterosexuals and tend to be more accepting than Americans (Andersen & Fetner, 2008). This coincides with the development of school programs such as gay-straight alliances (mygas.ca) as well as alternative schools designed to help at risk GLBT youth (such as the Triangle program at OASIS Alternative School in Toronto). Researchers have explored these ?safe spaces? (e.g., Fetner et al., 2012) and have found that gay-straight alliances help GLBT youth through increasing safety, acceptance, support and academic achievement (Heck, Flentje, & Cochran, 2011; Lee, 2002; Walls, Kane, & Wisneski, 2010). It is likely that the more positive GLBT climate in Canada influences the relationship between masculinity and sexual identity. Given that these differences exist, future research should examine the generalizability of these findings to contexts outside of the United States.  Summary In this dissertation, I use existing literature to argue that men engage in performances of masculinity that entail behaviours related to the concealment of emotions, shows of aggression, and demonstrations of athleticism. Using this definition as a framework, I was able to develop a multidimensional measure of masculinity that allowed me to differentiate adolescent men?s     114  performances of masculinity. Next, I argued, and found, that sexual identity does not influence adolescents? performances of masculinity. Finally, I tested the influence that masculinity and sexual identity have on men?s life experiences. The findings indicate that each dimension of masculinity uniquely influences adolescents? peer relationships, life satisfaction, and risk-taking. Athleticism, and concealing emotions to a lesser degree, is beneficial to adolescents while aggression is detrimental. Sexual identity does not influence adolescents? peer relationships and does not moderate the relationship between masculinity and life experiences.   This study offers several implications for future research that examines men and masculinity. First, adolescents men? performances of masculinity must be measured as a multidimensional construct lest important relationships will be overlooked. Second, the taken-for-granted argument that sexual identity influences masculinity is fast and likely rooted in homophobia. Researchers must be aware of this if they are to avoid perpetuating this stereotype through research that is methodologically and theoretically flawed.  Third, differences in men?s performances of masculinity must be taken into account when researchers explore men?s life experiences. Including measures of masculinity avoids conflating sex with gender by acknowledging that not all men are masculine.         115  References Abue, J., & Koestner, R. (1992). Gender characteristics and adjustment: A longitudinal study. Journal of Personality and Social Psychology, 63, 485-493. Adam, B. (2004). Care, intimacy, and same-sex partnership in the 21st century. Current Sociology, 52(2), 265-279. Andersen, R., & Fetner, T. (2008). Cohort Differences in Tolerance of Homosexuality: Attitudinal Change in Canada and the United States, 1981-2000. Public Opinion Quarterly, 72, 311-330. Asher, S. R., Hymel, S., & Renshaw, P. D. (1984). Loneliness in children. Child Development, 55, 1456-1464.  Baer, J. (1993). Etiology and secondary prevention of alcohol problems with young adults. In J. Baer, G. Marlatt, & R. McMahon, (Eds.), Addictive behaviours across the lifespan: Prevention, Treatment, and policy issues (pp. 111-137). Newburg Park, CA: Sage.  Bailey, J., & Zucker, K. (1995). Childhood sex-typed behaviour and sexual orientation: A conceptual analysis and quantitative review. Developmental Psychology, 31, 43-55. Bailey, J., Dunne, M., & Martin, N. (2000). Genetic and environmental influences on sexual orientation and its correlates in an Australian twin sample. Journal of Personality and Social Psychology, 78, 524-536. Banks, R. (2003). The cost of homophobia: Literature review on the human impact of homophobia in Canada. Saskatoon, SK: Gay and Lesbian Health Services. Bannon, I., & Correia, M. (2006). Introduction. In I. Bannon & M. Correia (Eds), The other half of gender: Men?s Issues in Development (pp. xvii ? xxiv). Washington, DC: The World Bank.     116  Bergling, T. (2001). Sissyphobia: Gay men and effeminate behaviour. Binghamton, NY: Southern Tier Editions.  Berndt, T. J. (1982). The features and effects of friendship in early adolescence. Child Development, 53, 1447-1460. Bierman, K. L. (2003). Peer Rejection: Developmental Processes and Intervention Stategies. New York: The Guilford Press. Bottorff, J., Oliffe, J., & Kelly, M. (2012). The gender(s) in the room. Qualitative Health Research, 22(4), 425-440. Brannon, R. (1976). The male sex role: Our culture's blueprint for manhood, what it's done for us lately. In D. David and R. Brannon (Eds.), The forty-nine percent majority: The male sex role (pp. 1-49). Reading, MA: Addison-Wesley.  Brennan, D, Ross, L, Dobinson, C., Veldhuizen, S., & Steele, L. (2010). Men's Sexual Orientation and Health in Canada. Canadian Journal of Public Health, 101(3), 255-258. Britton, D., & Williams, C. (1995). ?Don?t ask, don?t tell, don?t pursue?: Military policy and the construction of heterosexual masculinity. Journal of Homosexuality, 30(1), 1-21. Carrigan, T., Connell, R., & Lee, J. (1985). Towards a new sociology of masculinity. In M. Kaufman (Ed.), Beyond patriarchy: Essays by men on pleasure, power, and change (pp. 156-186). Toronto: Oxford University Press.  Carver, P., Yunger, J., & Perry, D. (2003). Gender identity and adjustment in middle childhood. Sex Roles, 49, 95-109. Cass, V. C. (1979). Homosexual identity formation: A theoretical model. Journal of Homosexuality, 4, 219?235. Cass, V. C. (1984). Homosexual identity formation: Testing a theoretical model. Journal of Sex     117  Research, 20, 143?167. Center for Disease Control and the Massachusetts Department of Education, 1997. Chantala, K. (2006). Guidelines for analyzing Add Health data. Carolina Population Center. University of North Carolina at Chapel Hill. Christiansen, B., Smith, G., Roehling, P., & Goldman, M. (1989). Using alcohol expectancies to predict adolescent drinking behaviour after one year. Journal of Consulting and Clinical Psychology, 57(1), 93-99. Coates, S., & Person, E. S. (1985). Extreme boyhood femininity: Isolated behaviour or pervasive disorder? Journal of the American Academy of Child Psychiatry, 24, 702?709. Connell, R. (1987) Gender and power: Society, the person and sexual politics. Cambridge: Polity. Connell, R. (1995). Masculinities. Sydney: Allen & Unwin.  Connell, R. (1996). Politics of changing men. Arena, 6, 56 Connell, R. (2005). Masculinities: Second Edition. Los Angeles, CA: University of California Press.  Connell, R., & Messerschmidt, J. (2005). Hegemonic masculinity: Rethinking the concept. Gender and Society, 19, 829-859. Cooley, C. H. (1902). Human nature and the social order. New York: Schocken.  Corliss, H., Cochran, S., & Mays, V. (2002). Reports of parental maltreatment during childhood in a United States populatio0n-based survey of homosexual, bisexual, and heterosexual adults. Child Abuse and Neglect, 26, 1165-1178. Courtenay, W. (2001). Constructions of masculinity and their influence on men's well-being: A theory of gender and health. Social Science and Medicine, 50, 1385-1401.     118  Craver, P. Yunger, J., & Perry, D. (2003). Gender identity and adjustment in middle childhood. Sex Roles, 49, 95-109. D?Augelli, A. (2006). Developmental and contextual factors and mental health among lesbian, gay, and bisexual youths. In A. E. Omoto & H. M. Kurtzman (Eds.), Sexual orientation and mental health: Examining identity and development in lesbian, gay, and bisexual people (pp. 37-53). Washington, DC: APA Books. D'Augelli, A., Grossman, A., & Starks, M. (2005). Parents' awareness of lesbian, gay, and bisexual youths' sexual orientation. Journal of Marriage and Family, 67, 474-482. D'Augelli, A., Hershberger, S., & Pilkington, B. (1998). Lesbian, gay, and bisexual youth and their families: Disclosure of sexual orientation and its consequences. American Journal of Orthopsychiatry, 68(3), 361-371. Davies, M. (2004). Correlates of negative attitudes towards gay men: Sexism, male role norms, and male sexuality. Journal of Sex Research, 41, 259-266. Demetriou, D. (2001). Connell's concept of hegemonic masculinity: A critique. Theory and Society, 30, 337-361. Easterbrook, A. (2009). Rethinking families over the life course development perspective: Including the lives of same-sex families. Sociology Compass, 3(6), 1000-1016. Edley, N., & Wetherell, M. (1997). Jockeying for position: The construction of masculine identities. Journal for the Theory of Social Behaviour, 20, 203-217. Elder, G., Johnson, M., & Crosnoe, R. (2003). The emergence and development of life course theory. In J. Mortimer & M. Sanahan (Eds.), Handbook of the life course (pp. 3-19). New York: Springer.      119  Emery, C., Meeuwisse, W., & McAllister, J. (2006). Survey of sport participation and sport injury in Calgary and area high schools. Clinical Journal of Sport Medicine, 16, 20-26. Erikson, E. (1963). Childhood and society. New York: Newton. Fagot, B. (1977). Consequences of moderate cross-gender behaviour in preschool children. Child Development, 48, 902-907. Fergusson, D., Horwood, L., & Beautrais, A. (1999). Is sexual orientation related to mental health problems and suicidality in young people? Archives of General Psychiatry, 56, 876-880. Fetner, T., Elafros, A., Bortolin S., & Drechsler, C. (2012). Safe Spaces: Gay-Straight Alliances in High School. Canadian Review of Sociology , 49, 188-207. Fiedler, K., Armruster, T., Nickel, S., Walther, E., & Asbeck, J. (1996). Constructive biases in social judgment: Experiments on the self-verification of question contents. Journal of Personality and Social Psychology, 71, 861-873. Friedman, M., Koeske, G., Silvestre, A., Korr, W., & Sites, E. (2000). The impact of gender-role nonconforming behaviour, bullying and social support no suicidality among gay male youth. Journal of Adolescent Health, 38, 621-623. Frosh, S., Phoenix, A., & Pattman, R. (2002). Young masculinities. New York: Palgrave.  Fumia, D. ?I Do? Belong in Canada: Same Sex Relationship and Marriage. In D. Cheal (Ed.) New Canadian Families, Toronto: Oxford University Press, Second Edition, pp. 167- 182. Furman, W. (1999). Friends and lovers: The role of peer relationships in adolescent romantic relationships. In W. A. Collins & B. Laursen (Eds.), The Minnesota Symposia in Child     120  Psychology: Vol. 30. Relationships as developmental contexts (pp. 133-154). Mahwah, NJ: Erlbaum. Gibson, P. (1989). Gay male and lesbian youth suicide. Report on the Secretary's Task Force on Youth Suicides, 3, 110-142. Washing, DC: US Department of Health and Human Services.   Goffman, E. (1959). The Presentation of Self in Everyday Life. New York: Double Day and Company. Goffman, E. (1963). Stigma: Notes on the management of spoiled identity. New York, NY: Simon and Schuster.   Gottschalk, L. (2003). Same-sex sexuality and childhood gender non-conformity: A spurious connection. Journal of Gender studies, 12, 35-50. Graham, J. (2009). Missing data analysis: Making it work in the real world. Annual Review of Psychology, 60, 549-576. Green, A. (2008). Health and sexual status in an urban gay enclave: An application of the stress process model. Journal of Health and Social Behaviour, 49, 436-451. Green, A. (2011). Playing the (sexual) field: The interactional basis of systems of sexual stratification. Social Psychology Quarterly, 74, 244-266. Green, R. (1976). One-hundred ten feminine and masculine boys: Behavioural contrasts and demographic similarities. Archives of Sexual Behaviour, 5, 425-446. Green, A., & Halkitis, P. (2006). Crystal methamphetamine and sexual sociality in an urban gay subculture: An elective affinity. Culture, Health, and Sexuality, 8, 317-333.     121  Grossman, T. (2002). Prahomosexuelle Kindheiten.ine empriische Untersuchung uber Geschlechtsrollenkonformitat und ?non- konformitat bei homosxuellen manner. Zeitschrift fur Sexualforschung, 15, 98-119. Hammelman, T. (1993). Gay and lesbian youth: Contributing factors to serious attempts or considerations of suicide. Journal of Gay and lesbian Psychotherapy, 2(1), 77-89.  Harris, F., & Harper, S. (2010). Masculinities go to community college: Understanding male identity socialization and gender role conflict. In S. Harper & F. Harris, College men and masculinities: Theory, research and implications for practice (pp. 66-76). San Francisco, CA: John Wiley & Sons, Inc. Harry, J. (1983). Parasuicide, gender, and gender deviance. Journal of Health and Social Behaviour, 24, 350?361. Harry, J. (1989). Parental physical abused and sexual orientation in males. Archives of Sexual Behaviours, 18, 251-261. Hartup, W. W. (1993). Adolescents and their friends. In B. Laursen (Ed.), Close friends in adolescence (pp. 3-22). San Francisco: Jossey-Bass.  Hatzenbuehler, M., McLaughlin, K., & Nolen-Hoeksema, S. (2008). Emotion regulation and internalizing symptoms in a longitudinal study of sexual minority and heterosexual adolescents. Journal of Child Psychology and Psychiatry, 49, 1270-1278.  Hatzenbuehler, M., McLaughlin, K., & Xuan, Z. (2012). Social networks and risk for depressive symptoms in a national sample of sexual minority youth. Social Science and Medicine, 75, 1184-1191. Heath, M. (2003). Soft-boiled masculinity: Renegotiating gender and racial ideologies in the Promise Keepers movement. Gender and Society, 17(3), 423-444.     122  Heath, M. (2009). State of our unions: Marriage promotion and the contested power of heterosexuality. Gender and Society, 23(1), 27-48. Heck, N., Flentje, A., & Cochran, B. (2011). Offsetting risks: High school gay-straight alliances and lesbian, gay, bisexual, and transgendered (LGBT) youth. School Psychology Quarterly, 26(2), 161-174. Hennen, P. (2005). Bear bodies, bear masculinity: Recuperation, resistance, or retreat? Gender and Society, 19, 25-43. Henning-Stout, M., James, S., & Macintosh, S. (2000). Reducing harassment of lesbian, gay, bisexual, transgender, and questioning youth in schools. School Psychology Review, 29(2), 180-192. Honaker, J., & King, G. (2010). What to do about missing values in time series cross-sectional data. American Journal of Political Science, 54(2), 561-581. Honaker, J., King, G., & Blackwell, M. (2011). Amelia II: A program for missing data. Journal of Statistical Software, 45(7), 1-47. Jeffreys, S. (2005). Beauty and misogyny. New York: Routledge.  Johnson, A. (2005). The gender knot: Unraveling our patriarchal legacy. Philadelphia: Temple University Press.  Kaufman, M. (1993). Cracking the armour: Power and pain in the lives of men. Toronto: Viking Canada.  Kimmel, M. (2004). Masculinity as homophobia. In E. Disch (Ed.), Reconstructing gender: A multicultural anthology (pp. 103-109).  Kimmel, M. (2008). Guyland: The perilous world where boys become men. New York, NY: HarperCollins Publishers.     123  Korobov, N. (2004). Inoculating against prejudice: A discursive approach to homophobia and sexism in adolescent male talk. Psychology of Men and masculinity, 5, 178-189. Kosciw, J., Greytalk, E., & Diaz, E. (2009). Who, what, where, when, and why: Demographic and ecological factors  contributing to hostile school climates for lesbian, gay, bisexual, and transgendered youth. Journal of youth and Adolescence, 38, 976-988. Kraus, S. (1995). Attitudes and the prediction of behaviour: A meta-analysis of the empirical literature. Personality and Social Psychology Bulletin, 21, 58-75. Kujala, U., Taimela, S., Antii-Poika, I., Prava, S., Tuminen, R., & Myllellen, P. (1995). Acute injuries in soccer, ice hockey, volleyball, basketball, judo, and karate: An analysis of national registry data. British Medical Journal, 311, 1465-1468.  Kupersmidt, J. B., Coie, J. D., & Dodge, K. A. (1990). The role of poor peer relationships in the development of disorder. In S. R Asher & J. D. Coie (Eds.), Peer rejection in childhood, (pp. 274-305). Cambridge England: Cambridge University press. La Piere, R. (1934). Attitudes vs. actions. Social Forces, 13, 230-237. Landolt, M., Bartholomew, K., Saffrey, C., Oram, D., & Perlman, D. (2004). Gender nonconformity, childhood rejection, and adult attachment: A study of gay men. Archives of Sexual Behaviour, 33, 117-128. Leary, M. R., Koch, E. J., & Hechenbleiker, N. R. (2001). In M. R. Leary (Ed.), Interpersonal rejection (pp. 145-166). New York, NY: Oxford University Press. Lee, C. (2002). The impact of belonging to a high school gay/straight alliance. The High School Journal, 85(3), 13-26.     124  Lindley, L., Walsemann, K., & Carter, J. (2011). The association of sexual orientation measures with young adults' health-related outcomes. American Journal of Public Health, published ahead of print.  Link, B., & Phelan, J. (2001). Conceptualizing stigma. Annual Review of Sociology, 27, 363-385. Link, B., Cullen, F., Struening, E., Shrout, P., & Dohrenwend, B. (1989). A modified labeling theory approach to mental disorders: An empirical assessment. American Sociological Review, 54(3), 400-423. Link, B., Strueing, E., Rahav, M., Phelan, J., & Nuttbrock, L. (1997). On stigma and its consequences: Evidence from a longitudinal study of men with dual diagnoses of mental illness and substance abuse. Journal of Health and Social Behaviour, 38(2), 177-190. Mac an Ghaill, M. (1994). The making of men: Masculinities, sexualities, and schooling. Buckingham: Open University Press. MacDonald, C. (2012). ?That?s just what people think of a hockey player, right??: Manifestations of masculinity among major junior ice hockey players. MA Thesis, Concordia University.  McCarry, M. (2007). Masculinity studies and male violence: Critique or collusion? Women's Studies International Forum, 30, 404-415. Mead, G. (1934). Mind, self, and society. Chicago: University of Chicago Press.  Messner, M. (1990). Boyhood, organized sports, and the construction of masculinities. Journal of Contemporary Ethnography, 18(4), 416-444. Mihalik, G. J. (1991). Homosexuality, stigma, and biocultural evolution. Journal of Gay and Lesbian Psychotherapy, 1, 15-29. Morrell, R. (1998). Gender and education: The place of masculinity in South African schools. South African Journal of Education, 18, 218-224.     125  Morrow, D. (1993). Social work with gay and lesbian adolescents. Social Work, 38(6), 655-660. Mosher, C., Levitt, H., & Manley, E. (2006). Layer of leather: The identity formation of leathermen as a process of transforming meanings of masculinity. Journal of Homosexuality, 51, 93-123. Munren, S. K., & Smolak, L. (1997). Femininity, masculinity, and disordered eating: A meta-analytic review. International Journal of Eating Disorders, 22, 231?242. Nebenzahl, D. (January 30, 2012). Beyond macho: Defining a man?s world. Montreal Gazette. Ollendick, T., Weist, M., Borden, M., & Greene, R. (1992). Sociometric status and academic, behavioural, and psychological adjustment: A five-year longitudinal study. Journal of Consulting and Clinical Psychology, 60(1), 80-87. Olweus, D. (1993). Bullying in School: What We Know and What We Can Do. Oxford: Blackwell. Parker, J. G., & Asher, S. R. (1993). Friendship and friendship quality in middle childhood: Links with peer group acceptance and feelings of loneliness and social dissatisfaction. Developmental Psychology, 29, 611-621.  Parker, J. G., Rubin, K. H., Price, J. M., & DeRoiser, M. E. (1995). Peer relationships, child development and adjustment: A developmental psychopathological perspective. In D. Cicchetti & D. Cohen (Eds.), Developmental Psychopathology: Vol 2. Risk disorder, and adaptation (pp. 96-161). New York: Wiley. Parker, J., Rubin, K., Erath, S., Wojslawowicz, J., & Buskirk, A. (2006). Peer relationships, child development, and adjustment: A developmental psychopathology perspective. In D. Cicchetti & D. Cohen (Eds.), Developmental psychopathology (2nd ed.), Vol. 1 (pp. 419-493). Hoboken, New Jersey: John Wiley & Sons, Inc.       126  Pascoe, C. (2007). Dude, you're a fag: Masculinity and sexuality in high school. Los Angeles: University of California Press.  Pleck, J., Sonenstein, F., & Ku, L. (1993). Masculinity ideology: Its impact on adolescent males' heterosexual relationships. Journal of Social Issues, 49, 11-29. Plodel, M., & Fartacek, R. (2009). Childhood gender nonconformity and harassment as predictors of suicidality among gay, lesbian, bisexual, and heterosexual Australians. Archives of Sexual Behaviour, 38, 400-410. Radloff, L. (1977). The DES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement 1, 385-401. Ragin, C. (1994). Constructing social research. Thousand Oaks, CA: Pine Forge Press.  Reeser, T. (2010). Masculinities in theory: An introduction. Malden, MA: Blackwell Publishing. Reinherz, H., Giaconia, R., Silverman, A., Friedman, A., Pakiz, B., Frost, A., et al. (1995). Early psychosocial risk for adolescent suicidal ideation and attempts. Journal of the American Academy of Child and Adolescent Psychiatry, 34, 599-611. Rekers, G. A., Bentler, P. M., Rosen, A. C., & Lovaas, I. (1977). Child gender disturbances: A clinical rationale for intervention. Psychotherapy: Theory, Research, and Practice, 14, 2?11. Remafedi, G., Farrow, J., & Deisher, R. (1991). Risk factors for attempted suicide in gay and bisexual youth. Pediatrics, 87(6), 869-875. Remafedi, G., French, S., Story, M., Resnick, M., & Plum, R. (1998). The relationship between suicide risk and sexual orientation: Results of a population-based study. American Journal of Public Health, 88, 57-60.     127  Richer, M., Leppin, A., & Gabhainn, S. (2006). The relationship between parental socio-economic status and episodes of drunkenness among adolescents: Findings from a cross-national survey. BMC Public Health, 6, 289-297. Roffman, D. (2000). A model for helping schools address policy options regarding gay and lesbian youth. Journal of Sex Education and Therapy, 25(2/3), 130-137.  Ross, M. (1980). Retrospective distortion in homosexual research. Archives of Sexual Behaviour, 9, 523-531. Rubin, D. (1994). Multiple imputation for nonresponse in surveys. New York, NY: John Wiley & Sons.  Russell, S., Seif, H., & Trung, N. (2001). School outcomes of sexual minority youth in the United States: Evidence from a national sample. Journal of Adolescence, 24, 111-127. Safren, S., & Heimberg,R. (1999). Depression, homelessness, suicidality and related factors in sexual minority and heterosexual adolescents. Journal of Counseling and Clinical Psychology, 67, 859-866. Sanchez, S., Greenburg, F., Liu, W., & Vilain, E. (2009). Reported effects of masculinity on gay men. Psychology of men and masculinity, 10(1), 73-87. Savin-Williams, R. (2001). A critique of research on sexual minority youths. Journal of Adolescence, 24, 5-13. Schafer, J. (1997). Analysis of incomplete multivariate data. London: Chapman & Hall. Schilt, K., & Connell, C. (2007). Do workplace gender transition make gender trouble? Gender, Work and Organization, 14(6), 596-618.  Schrock, D., & Schwalbe, M. (2009). Men, masculinity, and manhood acts. Annual Review of Sociology, 35, 277-295.     128  Schwalbe, M. (2005). Identity stakes, manhood acts, and the dynamics of accountability. Studies in Symbolic Interaction, 28, 65-81. Schwartz, D., Dodge, K. A., & Coie, J. D. (1993). The emergence of chromic peer victimization in boys? play groups. Child Development, 64, 1755-1772. Seidler, V. (2006). Young men and masculinities: Global cultures and intimate lives. London: Zed Books. Shoveller, J., Knight, R., Johnson, J., Oliffe, J., & Goldenberg, S. (2009). ?Not the swab!? Young men?s experiences with STI testing. Sociology of Health and Illness, 32(1), 57-73. Snively, C., Kreuger, L., Stetch, J., Watt, J., & Chadha, J. (2004). Understanding homophobia: Preparing for practice realities in urban and rural settings. Journal of gay and Lesbian Social Services, 17(1), 59-81. Snyder , M., & Gangestad, S. (1981). Hypothesis-testing processes. IN J. H. Harvey, W. Ickes, & R. R. Kidd (Eds.), New direction in attribution research (Vol. 3, pp. 171-198). Hillsdale, NJ: Erlbaum. Sport Canada (1998). Sport Participation in Canada, 1998 report. Ottawa: Minster of Public works and government services Canada.  Springer, K., & Mouzon, D. (2011). "Macho men" and preventive health care: Implications for older men in different social classes.  Journal of Health and Social Behaviour, 52(2), 212-227. Steer, S. (2001). Reconsidering the concept of hegemonic masculinity: Discursive psychology, conversation analysis, and participants? orientation. Feminisms and Psychology, 11, 107-135.     129  Sutton, R. (1994). Hearing Us Out: Voices From The Gay and Lesbian Community. Canada: Little, Brown and Company. Synnott, A. (2001). Men and masculinities. American Anthropologist, 103(1), 212-217. Synnott, A. (2009). Re-thinking men: Heroes, villains and victims. Aldershot: Ashgate.  Tavakol, M., & Dennick, R. (2011). Making sense of Cronbach's alpha. International Journal of Medical Education, 2, 53-55. Thompson, S., & Johnson, L. (2003). Risk factors of gay, lesbian, and bisexual adolescents: Review of empirical literature and practical implications. Journal of Human Behaviour in the Social Environment, 8(2/3), 111-128. Thurlow, C. (2001). Naming the 'outsider within': Homophobic pejoratives and verbal abuse of lesbian, gay, and bisexual high-school pupils. Journal of Adolescence, 24, 25-38. Udry, J. R., & Chantala, K. (2002). Risk assessment of adolescents with same-sex relationships. Journal of Adolescent Health, 31, 84-32. Udry, J. R., & Chantala, K. (2004). Masculinity-femininity guides sexual union formation in adolescents. Personality and Social Psychology Bulletin, 30(1), 44-55. von Hippel, P. (2009). How to impute interactions, squares, and other transformed variables. Sociological Methodology, 39(1), 265-291. Walls, E., Kane, S., & Wisneski, H. (2010). Gay-straight alliances and school experiences of sexual minority youth. Youth Society, 41(3), 307-332. Washburn-Ormachea, J., Hillman, S., & Sawilowsky, S. (2004). Gender and gender-role orientation differences on adolescents? coping with peer stressors. Journal of youth and Adolescence, 33, 31-40.     130  Wechsler, H., Davenport, A., Dowdall, G., Moeykens, B., & Castillo, S. (1994). Health and behavioural consequences of binge drinking in college: A nation survey of students at 140 campuses. Journal of the American Medical Association, 272, 1672-1677. West, C., & Zimmerman, D. (1987). Doing Gender. Gender and Society, 1(2), 125-151. Weston, K. (1995). Get thee to a big city: Sexual imaginary and the great gay migration. GLQ, 2, 253-277. Whitam, F., & Zent, M. (1984). A cross-cultural assessment of early cross-gender behaviour and familial factors in male homosexuals. Archives of Sexual Behaviour, 13(5), 427-439. Wicker, A. (1969). Attitudes versus action: The relationship of verbal and overt behavioural responses to attitude objects. Journal of Social Issues, 25, 41-78. Wilkinson, W. (2004). Authoritarian hegemony, dimensions of masculinity, and male antigay attitudes. Psychology of Men and Masculinity, 5, 121-131. Yunger, J., Carver, P., & Perry, D. (2004). Does gender identity influence children?s psychological well-being? Developmental Psychology, 40, 572-582. Zucker, K. (2005). Commentary on Gottschalk?s (2003) ?Same sex sexuality and childhood gender non-conformity: A spurious connection.? Journal of Gender Studies, 40, 55-60. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0165695/manifest

Comment

Related Items