|Home | About | Journals | Submit | Contact Us | Français|
This study compared single (n = 129) and partnered gay men (n = 114) to determine if they differed in their concerns over traditional masculine roles and interest in casual sex, and to measure the relationship between concerns over masculine roles and interest in casual sex. Additionally, a regression model to predict relationship satisfaction was tested. Participants were recruited at two Southern California Gay Pride festivals. Group comparisons showed single men were more restrictive in their affectionate behavior with other men (effect-size r = .14) and were more interested in casual sex than partnered men (effect-size r = .13); and partnered men were more concerned with being successful, powerful, and competitive than single men (effect-size r = .20). Different masculine roles were predictive of interest in casual sex among the two groups of men. Finally, a hierarchical regression analysis found that interest in casual sex and the length of one’s current relationship served as unique predictors of relationship satisfaction among the partnered gay men (Cohen’s f2 = .52).
Many men are interested in casual and emotionally uncommitted sex (Clark & Hatfield, 1989; Schmitt, 2005). This interest has repeatedly been shown to differentiate the sexes whereby men are far more interested in casual sex than women are (Buss, 2003; Simpson & Gangestad, 1991). Furthermore, gay men and heterosexual men are nearly identical in their degree of interest (Bailey, Gaulin, Agyei, & Gladue, 1994; Schmitt, 2006).
Opinions differ on why men exhibit this pattern of interest in casual sex (ICS). Some believed that biological and evolutionary processes influence ICS (Buss, 1995; Gangestad & Simpson, 2000). Others argued that men’s ICS is merely the result of socialized gender roles (Eagly & Wood, 1999; Levant & Brooks, 1997). While several studies support the former position (e.g., Eisenberg et al., 2007; Lim et al., 2004; van Anders & Watson, 2006; Walum et al., 2008), only three peer-reviewed studies have related ICS with masculine gender roles (Simpson, Wilson, & Winterheld, 2004).
The first was by Walker, Tokar, and Fischer (2000) who found that concerns with expressing emotions and with being affectionate with other men were predictive of ICS. Mahalik and his colleagues (2003) reported that scores on a measure assessing ICS was significantly related to concerns with being successful, with restricting emotions, and with restricting affection with other men. Finally, Cunningham and Russell (2004) reported that the endorsement of masculine typical traits (e.g., being dominant and competitive) was positively correlated with ICS. Overall, limited evidence supports the position that traditional masculine roles are associated with men’s ICS.
In terms of gay men, many practitioners have hypothesized that ICS impedes gay men’s ability to find and maintain long-term relationships (Baron, 1996; Brown, 1995; Nichols, 1989). Inherent in this assumption is that single gay men express higher ICS than partnered gay men. Some empirical findings seem to indirectly support this belief. For instance, men with higher ICS desire greater sexual novelty, are more prone to sexual sensation seeking, and report more sexual partners compared to men with lower ICS (Gaither & Sellbom, 2003; Schmitt, 2006). However, no study has compared single and partnered gay men.
It has further been hypothesized that because gay men exhibit high ICS, their relationships suffer—especially over time (Shelton, 2008; Shernoff, 1999). Again, some empirical findings seem to support this belief. For instance, men with higher ICS report less relationship satisfaction, and they feel less commitment and love for a current partner compared to men with lower ICS (Klusmann, 2001; Seal, Agostinelli, & Hannett, 1994). However, these findings are based on samples of heterosexual men.
Thus, we conducted a field survey with three main goals. First, we sought to compare single and partnered gay men to determine whether the two groups differed in their ICS while replicating and expanding previous findings that examined gender role conflict (GRC) among gay men (Wester, Pionke, & Vogel, 2005). Second, we sought to examine the relationship between GRC factors and ICS for both groups of men. Lastly, we sought to test the degree to which GRC factors, ICS, and relationship length could predict relationship satisfaction among the partnered gay men.
Two main hypotheses guided this investigation. First, because masculine norms encourage men to be sexually promiscuous, it was believed that concerns with fulfilling one’s gender role would be positively associated with ICS. Second, given that a high degree of ICS appears detrimental to a committed romantic relationship, we hypothesized that relationship length would contribute significant variance in predicting relationship satisfaction beyond the variance explained by GRC and ICS.
The men (N=243) were recruited at two California Gay Pride Festivals (Long Beach and San Diego). The average age was 34.08 (SD=10.94; range=18–67) and on average they had openly identified as gay for 14.61 years (SD=10.28; range=0–49 years). We used the Kinsey Scale (Kinsey, Pomeroy, & Martin, 1948/1975) to verify sexual orientation: Scores range from 0–6 (0=exclusively heterosexual; 6=exclusively gay) and those with a score of 5 or 6 were retained—a method commonly used by sex researchers (e.g., Tiggermann, Martins, & Kirkbride, 2007). Racial/ethnic composition was as follows: 57.6% White (Non-Latino), 21.8% Hispanic/Latino, 12.8% Asian American, 3.7% African American, 0.8% Native American, and 3.3% Other. Most (55.1%) had at least a bachelor’s degree and the median annual individual income bracket was $40,000–$49,999. Almost half (46.9%) were in a same-sex relationship (mean length=4.48 years; SD=5.99) with 58.2% cohabiting with their partner, and 82.9% in a closed/monogamous relationship. The partnered and single men did not differ significantly on any key demographic characteristic.
The Gender Role Conflict Scale (O’Neil, Helms, Gable, David, & Wrightman, 1986) was used to measure concerns with fulfilling four factors associated with traditional masculine roles: The 13-item Success, Power, and Competition subscale assesses one’s emphasis on personal achievement and control/authority over other people (e.g., “I strive to be more successful than others”). The 10-item Restrictive Emotionality subscale assesses discomfort with disclosing one’s feelings (e.g., “I have difficulty telling others I care about them”). The 8-item Restrictive Affectionate Behavior Between Men subscale assesses discomfort with affection between men (e.g., “Hugging other men is difficult for me”). The 6-item Conflict Between Work and Family Relations subscale assesses difficulty in balancing work/school demands with family/leisure life (e.g., “My career, job, or school affect the quality of my leisure or family life”). A 6-point scale is used to respond to each item (1=strongly disagree; 6=strongly agree). The subscale reliabilities ranged from .81–.90.
We used the Interest in Uncommitted Sex Scale (Bailey et al., 1994) to assess interest in casual sex. Respondents use a 7-point scale (1=strongly disagree; 7=strongly agree) to rate their level of agreement with 10 different statements (e.g., “I could easily imagine myself enjoying one night of sex with someone I would never see again”). The scale reliability was .81.
The demographic form contained questions for partnered gay men, which asked how long they had been with their partner and how satisfied they were. Relationship satisfaction questions were from the 7-item Relationship Assessment Scale (Hendrick, Dicke, & Hendrick, 1998). Participants use a 5-point scale with varying anchor labels (e.g., 1=poorly, 5=extremely well) to respond to the items (e.g., “How well does your partner meet your needs?”). The scale reliability was .88.
Given the personal questions, we used a shortened version of the Marlowe-Crowne Social Desirability Scale (SDS; Reynolds, 1982) to control for social desirability. This 13-item true-false scale assesses the tendency to distort answers in order to present oneself in a favorable light (e.g., “I’m always willing to admit when I make a mistake”). The scale reliability was rKR-20=.55.
The results will focus on the two groups of interest: single and partnered gay men. However, ten supplemental tables are available online, which include statistics for the combined group of gay men. One partnered participant did not report relationship length; however, he was included for the correlational analysis.
We assessed the distributions for normalcy (Tabachnik & Fiddell, 2006). Among the partnered men, distributions for the Restrictive Affectionate Behavior Between Men subscale and relationship length were positively skewed (p <.05). Square-root transformations were used to reduce the skew to non-significant levels. The single men’s scores were also transformed for the group comparison.
First, we compared the groups to determine whether they differed in GRC factors and ICS. To control for the effects of social desirability, we conducted separate ANCOVAs where relationship status served as the fixed factor and SDS scores served as the covariate (see Table 1). Effect sizes are reported as Pearson’s r.
Three significant differences were found. On average, the single men (M=20.61, SE=0.72) reported slightly more concern about expressing affection with other men compared to partnered men (M=17.99, SE=0.68): F(1,240)=4.78, p<.05, r=0.14. The effect size was similar to the finding by Wester et al. (2005; η2=0.05), suggesting that there may be a real—albeit slight—difference between single and partnered gay men. Interestingly, partnered men (M=50. 48, SE=1.23) experienced slightly more concern about being successful, powerful, and competitive compared to single men (M=46.49, SE=1.05): F(1,240)=10.15, p <.01, r= 0.20. Finally, the single men (M=42.35, SE=1.06) reported slightly greater ICS compared to partnered men (M=38.45, SE=1.22): F(1,240)=4.02, p <.05, r=.13. Overall, while single and partnered gay men varied statistically in GRC factors and ICS, these differences will need to be replicated to see if there is a true effect.
Second, we calculated the partial correlations (controlling for social desirability) between GRC factors and ICS for each group (see Table 2). The two restrictive GRC factors were positively related with ICS for both groups. To further explore this relationship, we conducted a regression analysis to determine which GRC factor best predicted ICS. Given the correlation between the Restrictive Emotionality and Restricted Affectionate Behavior Between Men subscales, we combined them into a composite score (Tabachnik & Fiddell, 2006), which we term Restricted Intimacy. For both groups, SDS scores were entered in Step 1 to control for the effects of social desirability. In Step 2, the three GRC factors were entered as predictors with ICS serving as the criterion. The GRC factors accounted for 6% of the variance in ICS scores among single men, and 13% among partnered men (regression tables available online). However, for single men, no GRC factor predicted ICS whereas for partnered men, success concern was a significant unique predictor (β=.33). Overall, while specific GRC factors were correlated to ICS, a large percentage of the variance was unaccounted for.
Lastly, a hierarchical multiple regression analysis was conducted using relationship satisfaction as the criterion variable (see Table 3). The predictor variables were entered in four steps: 1) SDS scores were entered to control for social desirability, 2) GRC factors, 3) ICS, and 4) relationship length. The GRC factors contributed a non-significant variance to relationship satisfaction. In Step 3, ICS accounted for a significant 23% of the variance in relationship satisfaction, F(1,107) change=34.36, p<.001. In Step 4, relationship length accounted for a significant 7% of the variance in relationship satisfaction above and beyond the GRC factors and ICS, F(1,106) change=10.95, p<.01. The entire regression model accounted for 34% of the variance in relationship satisfaction, after controlling for social desirability (Cohen’s f2=.52).
This study investigated whether single and partnered gay men differed in GRC factors and ICS. We also analyzed the degree to which GRC factors and ICS were related. Finally, a regression model was tested to determine the degree to which relationship satisfaction could be predicted based on GRC factors, ICS, and relationship length.
First, we found that single men reported slightly greater ICS and more concern over being affectionate with other men compared to partnered men. This offers very tentative support for the proposal that a fear of intimacy accounts for many gay men who are single (Baron, 1996; Brown, 1995). We further found that partnered men were more concerned with success, power, and competition when compared to the single men. A possible explanation for this is that people associate being in a romantic relationship with feelings of success and achievement (Maass, 2006; Pillemer, Ivcevic, Gooze, & Collins, 2007). Thus, some men who are concerned about feeling superior may actively seek out this “status symbol.”
Second, we found that the restrictive GRC factors (i.e., men should be unemotional and unaffectionate with other men) were positively correlated with ICS. Thus, as in Walker et al.’s (2000) analysis consisting of predominately heterosexual men, gay men who were less comfortable with emotionality and affection were more interested in ICS. While this seems to offer support for the idea that men’s ICS is due to normative male socialization, the regression analysis revealed that this was only true for partnered gay men. Thus, GRC factors may operate differently for single and partnered gay men and further research is needed to understand this relationship.
Lastly, ICS and relationship length uniquely predicted relationship satisfaction, whereas as GRC factors did not. While numerous factors can affect relationship satisfaction—such as internalized homophobia (Frost & Meyer, 2009)—these findings are consistent with previous findings among heterosexual men. For instance, Burn and Ward (2005) found that heterosexual men’s interest in having multiple sexual partners was negatively related to relationship satisfaction. Thus, gay men who express high ICS may be less satisfied being in a relationship because it limits their ability to express that interest. Furthermore, the longer a person has been in a relationship, the more satisfied they are likely to be regardless of their ICS.
Several limitations should be noted. First, the majority of the sample identified as White (non-Latino). Furthermore, collecting data at Pride festivals introduced the confound of self-selection bias. Consequently, the results may not generalize beyond those who engage with the gay community (Meyer & Wilson, 2009). Second, collecting data at a festival imposed time limits. Thus, additional scales that may have enhanced the analysis were not administered. Third, the low SDS reliability may have attenuated or inflated the resulting coefficients (Cohen, Cohen, West, & Aiken, 2003). Different measures that assess this construct should be considered in future research (Loo & Loewen, 2004). Finally, the cross-sectional and correlational nature of the data prohibits determining causation.
Notwithstanding these limitations, the results offer some implications for practice and research. In regards to practice, it may benefit gay clients to discuss what role casual sex plays in their lives and how it affects their relationships. For instance, may clients who have difficulty being emotional and affectionate with others substitute casual sex for intimacy (Shernoff, 2005)? For those engaged in couples counseling, it may be worthwhile exploring if ICS is affecting relationship satisfaction and what such interest may be rooted in (e.g., sexual dissatisfaction). Finally, given that some gay partners mutually agree to sexual non-exclusivity, practitioners must remain mindful of their personal views when conceptualizing such relationships (LaSala, 2004).
In regards to research, these findings raise questions regarding gay men’s romantic relationships. First, some have suggested that gay men who are single struggle with fears of intimacy or sexual compulsions (e.g., Driggs & Finn, 1990). Thus, research can continue to compare these two groups of men to determine how they differ and to what degree. Second, there is a need to examine what other factors influence relationship satisfaction. For instance, how do sexual interests (e.g., interest in pornography and preference for younger partners) affect relationship satisfaction? Further, quantitative research among sexually non-exclusive gay couples is needed. Specifically, how do men in these relationships negotiate the terms of their non-exclusivity and how do they maintain their primary relationships?
Altogether, gay men’s interest in casual sex plays a significant role in their dating and romantic relationships. Yet, it is inaccurate to assume that single men are more interested in casual sex than partnered men. Rather, both groups of men likely have similar interest in casual sex that partly stems from traditional masculine roles. Still, for partnered gay men, a high degree of interest in casual sex may be a source of dissatisfaction with one’s relationship.
This study was supported by the first author’s NIH training grant 5 T32 HD07228: 25 (Neural Regulation of Reproduction/Laboratory of Neuroendocrinology). We thank Maria Alaniz, Dave Bryant, Lawrence Chiu, Emmanuel Delot, Brandy Hamill, Eileen Luders, Michael Nanaszko, and Luong Thai for their assistance in the collection of the data. We also thank William Ming Liu, James O’Neal, and Letitia Anne Peplau for their feedback on the project and report. Some of the results in this article were presented at the 115th Annual Convention of the American Psychological Association, San Francisco, CA, August 2007.
Publisher's Disclaimer: The following manuscript is the final accepted manuscript. It has not been subjected to the final copyediting, fact-checking, and proofreading required for formal publication. It is not the definitive, publisher-authenticated version. The American Psychological Association and its Council of Editors disclaim any responsibility or liabilities for errors or omissions of this manuscript version, any version derived from this manuscript by NIH, or other third parties. The published version is available at www.apa.org/pubs/journals/men.