Social Norms: Are Sexually Aggressive Men More Likely to Misperceive Other Men’s Sexual Desires and Behavior?

Abstract

Separate lines of research show that men overestimate the extent of male peers’ sexual activity, and independently, that sexually aggressive men believe that other men approve of coercive behavior. This study examined the intersection of these lines of inquiry, testing whether the degree of male participants’ misperception of other men’s sexual behavior differs as a function of perpetrator status. In a national sample, we presented heterosexually active men (n=497) with sexual scenarios varying in sexual acts, partner types, and circumstances. Results showed that participants significantly overestimated the typicality of all types of sexual situations for other men. Participants also misjudged the desirability of scenarios consistent with a traditional masculinity sexual script to other men; these scenarios reflected an adversarial perspective on relationships and an impersonal approach to sexuality – a known risk factor for sexual aggression. Further, sexually aggressive men overestimated the desirability of these traditional masculinity scenarios to a greater extent than non-aggressive peers. Findings suggest that interventions that provide accurate knowledge about social norms, or “typical” sexual desires and behaviors among other men, may reduce pressure to live up to perceived but perhaps inaccurate masculine ideals, as well as reduce social norm-related risks for sexually aggressive behavior.

The degree to which men believe their peers approve of or perpetrate sexual assault has emerged as a consistent and powerful predictor of sexually aggressive behavior toward women. This has been documented among college-enrolled and community-based samples of men (Dardis, Murphy, Bill, & Gidycz, 2016; Jacques-Tiura et al., 2015) and is associated with persistence in sexually aggressive behavior (Thompson, Kingree, Zinzow, & Swartout, 2015). Given consistent findings that between 1/3 and ½ of men report engaging in some form of sexually aggressive behavior over time (Abbey, Wegner, Pierce, & Jacques-Tiura, 2012; Parkill & Abbey, 2008; White & Smith, 2004), with around 5-12% committing completed or attempted rape (Abbey et al., 2012; Dardis, Murphy, Bill & Gidycz, 2016), better understanding the role of men’s beliefs about their peers’ sexual conduct is an important element of untangling the etiology of sexual aggression.

In a simultaneous but separate vein of sexual behavior research, studies show that, in general, men tend to perceive peers’ sexual behavior inaccurately. More specifically, men tend to over-estimate the number of partners other men have, and the frequency with which other men engage in a range of specific sexual behaviors (e.g. Lewis, Litt, Cronce, Blayney, & Gillmore, 2014; Martens et al., 2006). Less well explored, however, are the intersections of these two avenues of research. For example, very little research has examined whether sexually aggressive men’s beliefs about their peers’ behavior is accurate. Second, to our knowledge, no studies have examined whether misperceiving the desirability and frequency of various non-aggressive sexual situations among male peers, and particularly those associated with an “impersonal” approach to sex (which is further explained below), is related to risk for sexual aggression in the same way that believing in peers’ endorsement of sexual coercion is. This study attempts to bring these two lines of inquiry together and examines men’s misperceptions of male peers’ sexual desires and behavior in a national sample of heterosexually active young men. More specifically, we explore whether the magnitude of misperceptions differs between sexually aggressive and non-aggressive men.

Men’s Sexual Behavior and Perceived Social Norms

Men’s beliefs about the sexual behavior and desires of other men can be characterized as perceived social norms, which are a widely documented influence on sexual behaviors. Cialdini Kallgren, & Reno (1991) identified two types of influential norms: injunctive and descriptive. Injunctive norms are perceptions of the degree to which others (peers, family members, partners) approve of specific behaviors or find them desirable. In contrast, descriptive norms are beliefs about the frequency with which others actually engage in particular behaviors. Both categories of norms are implicated in sexual behavior, and peers are particularly relevant referents for both injunctive and descriptive normative perceptions. The influence of perceived peer approval (injunctive norms) has been documented regarding behaviors including, but not limited to, college student condom use (Asare, 2015), and “hooking-up” and casual sex among heavy-drinking college-enrolled men (Napper, Kenney, & LaBrie, 2015). Descriptive norms also shape behavior; perceptions of the frequency with which male peers engage in casual, non-committed sex (Lewis et al., 2014), sex while drinking (Lewis, Lee, and Patrick, 2007), and specific sexual acts (Martens et al., 2006; Scholly, Katz, Gascoigne, & Holck, 2005) are related to men’s own frequency of engaging in these behaviors. Men are also more likely to report valuing consent in their own sexual and romantic relationships when they perceive that their peers behave in this way (Fabiano, Perkins, Berkowitz, Linkenbach, & Stark, 2003).

At the same time, research suggests that young people are not particularly skilled at accurately assessing the frequency of their peers’ sexual behaviors. Martens et al. (2006) found that college-enrolled men and women both over-estimated the frequency of oral, anal, and vaginal intercourse among peers, as well as peers’ overall number of sexual partners. Similarly, Barriger and Vélez-Blasini (2013) asked college students about “hook-ups,” or casual sexual encounters, and found that both men and women overestimated the frequency with which their friends engaged in hook-ups, as well the prevalence of each of the specific sexual behaviors involved in those encounters (including kissing, non-genital touching, and oral/vaginal intercourse). College students also inaccurately gauge the number of casual sexual partners and sexual encounters involving alcohol among their peers (Lewis et al., 2007; Lewis et al., 2014) while underestimating their peers’ condom use (Scholly et al., 2005). Much less research has examined accuracy in assessing injunctive norms, i.e., the degree to which important others approve of specific behaviors. One exception is Barriger and Vélez-Blasini (2013), who found that undergraduate men underestimate the extent to which peers are “comfortable” with sexual behaviors such as kissing and non-genital touching but are more accurate in their assessments of more intimate behaviors such as oral and vaginal sex.

Social norm theory suggests that people tend to overestimate the prevalence of health risk behaviors among others, and, importantly, this over-estimation creates pressure to engage in these behaviors in order to be more “normative” (Perkins & Berkowitz, 1986). This is borne out in the research described above, showing that perceptions of peer sexual behavior are related to one’s own sexual behavior. Overestimating the frequency of others’ sexual behaviors (and the degree to which they approve of sexual risk behaviors) could exacerbate this, by creating misplaced pressure to engage in more, and potentially more risky, sexual behavior.

Sexual behavior perceptions and sexual aggression.

This misplaced pressure is particularly problematic in the context of sexual aggression. As noted above, perpetrators are more likely than non-perpetrators to believe that their peers both engage in and approve of sexual aggression. Only two studies have examined the accuracy of sexually aggressive men’s perceptions of peer sexual behaviors or attitudes, and whether perpetrators are more inaccurate than other men, however. First, in a study of college-enrolled men, perpetrators of sexual assault were more likely to mischaracterize friends as sexually aggressive and to over-estimate their peers’ belief in rape myths than were non-perpetrators (Dardis et al., 2016). Similarly, Neighbors et al. (2010) showed that men who had perpetrated intimate partner violence were more likely than non-perpetrators to overestimate rates of forcing sex on a partner. While additional research is needed to assess the magnitude and mechanisms of these misperceptions, these studies suggest that perpetrators may justify their behavior, in part, through the mistaken assumption that their peer referents condone it.

The importance of sexually aggressive men’s perceptions of peer sexual behavior may extend to non-aggressive situations as well. In the widely replicated Confluence Model of sexual assault perpetration, an “impersonal” approach to sex has been consistently associated with risk for sexual aggression (Malamuth, Linz, Heavey, Barnes, & Acker, 1995). Impersonal sex is operationalized as a non-intimacy-based approach to sexual activity, centered around “scoring” or accumulating numerous sexual partners (Malamuth et al., 1995). While a non-intimacy-oriented or casual approach to sexuality can certainly include consensual, non-aggressive sexual encounters, evidence suggests that when impersonal sexuality is paired with hostile attitudes toward women and adherence to traditional gender attitudes (“hostile masculinity”), risk for sexual aggression increases (Abbey, Jacques-Tiura & LeBreton, 2011). It is well documented that men’s own enactment of an impersonal approach to sex can be related to risk for aggression (see for review, Tharp et al., 2012). To our knowledge, however, studies have not examined whether men’s perceptions of non-aggressive but impersonal peer sexual behavior also function as a risk for or a reinforcer of men’s own sexually aggressive conduct in the same way that perceptions of peer sexual coercion do. Because of the gender expectation-related links between impersonal sex and traditional notions of masculinity (e.g. that men are expected to both want and get a lot of sex; Seal & Ehrhardt, 2003), perceived descriptive and injunctive norms related to these non-aggressive forms of sexual behavior could be doubly powerful influences on men’s sense of what their sexual behavior should look like. This may be particularly true for perpetrators of sexual assault, whose behavior is already related, in part, to perceptions of peer aggression. Yet, the degree to which perpetrators of sexual violence over-perceive the prevalence of non-aggressive but “impersonal” sexual behavior among their peers is under-explored in current research.

Summary and Study Aims

In brief, the influence of both descriptive and injunctive norms on individual sexual behavior is well documented, as is young people’s tendency to overestimate how much sex and how many sexual partners their peers have. Separately, research also shows that men who engage in sexual aggression are more likely than non-perpetrators to believe their peers approve of or engage in sexual assault, and some evidence suggests that these perceptions are inaccurate. This study builds upon these findings. Participants were presented with non-aggressive sexual scenarios varying in sexual acts, partner types, and circumstances and were asked to rate the commonness and desirability of those scenarios both for themselves, and for other men their age. We first calculated the sample mean of men’s ratings of the commonness and desirability of each scenario for themselves. We then tested two hypotheses related to how individuals estimated the commonness and desirability of the scenarios for other men. The hypotheses were 1) that, consistent with previous literature, men in the sample would, on average, overestimate both the desirability and prevalence of sexual situations presented in the scenarios, and 2) that men who had engaged in sexually aggressive behavior would overestimate the desirability and prevalence of the sexual scenarios to a greater degree than those who had not, especially for scenarios presenting situations consistent with an impersonal approach to sexual encounters.

Methods

Participants and Recruitment Procedures

Data used for these analyses were collected as part of a parent study with a primary goal of examining heterosexually active men’s relationship and sexual behaviors, adherence to sexual scripts, and use of sexual and intimate partner violence. Findings presented here represent unique secondary data analysis; previous publications present the process used to derive the sexual script scores used in this paper’s analysis (Morrison et al., 2015), or use different variables from the larger data set and examine relationships between intimate partner violence perpetration and sexual risk behaviors (Casey et al., 2016), sexual safety strategies (Masters et al., 2015), or participants’ endorsement of various indicators of masculine identities (Casey et al., 2016). All procedures were approved by the University of Washington’s internal review board.

Recruitment, eligibility screening, consent procedures, and survey administration occurred exclusively online. Eligibility criteria included identifying as male, being between the ages of 18-25, having past sexual contact with a woman (defined as touching below the waist or having oral, vaginal, or anal sex), intending future sexual relationships with women, and being a US resident during adolescence. Participants were recruited via ads on Facebook targeted to young men, and through advertisements on Craigslist in 14 different regions of the US selected to increase the racial diversity of potential viewers. Ads invited potential respondents to “share [their] views” in a “web survey on relationships with women.” Stock photos accompanying these ads displayed male/female pairs representing several racial/ethnic identities.

Respondents who clicked on the advertisement were then presented with a series of screening questions. In addition to the above criteria, we imposed quotas so as to recruit relatively equal numbers of eligible African American, Asian American, European American/white, Latino, and Multiracial or “other” racially-identified men into the survey; once approximately 100 men of a particular racial identity group had responded, the survey screened out further potential respondents from that group. Steps to enhance data quality and integrity included tracking respondents’ internet protocol (IP) addresses to prevent duplicate responses, and programming random, intermittent questions as “required” to prevent respondents from simply clicking through the survey. Respondents who attempted to repeat the survey were not automatically excluded, but their response patterns were carefully assessed using the criteria described below.

Participants

A total of 662 individuals began the survey. Of these, 14 were excluded due to nonsensical responding (contradictory or uniform responding), and 151 because they stopped the survey before responding to the sexual scenario items that were the core of these analyses (and which appeared later in an admittedly long survey). This left 497 participants available for analysis. We did not detect that this analysis sample was biased from the full sample; comparisons of the excluded and included respondents showed that they did not differ significantly in age, race/ethnicity, or education. Sample sizes for each of the bivariate analyses described below range from 475-497 due to occasional item skipping. Of the 497 respondents, 19.5% were African American, 19.3% Asian American, 21.3% European American/white, 21.5% Latino, and 18.3% Multiracial or of other racial/ethnic backgrounds. Participants came from all regions of the U.S. and ranged in age from 18-25 (M=20.6, SD=2.1). Approximately 6.2% had less than a high school education; 26.2% had completed high school or obtained their general equivalency diplomas, 47.2% had some post-secondary education but no degree, 6% had obtained an Associate’s degree, and 14.4% had a Bachelor’s degree or higher. Most participants had an annual income under $12,000 (63%), and 56% of the sample was enrolled in higher education either full or part-time at the time of the survey. This exceeds the 39% of 18-24-year-old men estimated by the U.S. Department of Education (2014) to be enrolled in college at the time, suggesting that this sample may be slightly better educated and/or higher income than the general population. Finally, although all participants were considered “heterosexually active” because of their reported desire for future sexual contact with women, 88% reported that they are primarily sexually attracted to women, 9.8% that they are sexually attracted to both men and women, and .4% (n=2) that they are primarily sexually attracted to men (1.8% of the sample reported they were unsure of their primary sexual attraction, or were attracted to “no one.”).

Measures

Sexual scenario items.

We measured perceptions of sexual situations through the vehicle of both specific sexual scenarios, and through larger sexual scripts. Sexual scripts are normative cognitive schemas that guide people in sexual situations. They reflect the social norms that shape individuals’ sexual beliefs and behavior (Gagnon, 1990; Gagnon & Parker, 1995). Sexual scenarios, and the larger sexual scripts they reflect, are used here to extend findings from previous research which has focused primarily on perceptions of peers’ engagement in specific sexual behaviors (e.g. kissing, intercourse), or numbers of sexual partners. The survey drew from previous work by the research team in which we elicited and distilled common sexual scenarios for young heterosexual men in the U.S. The work to identify and test these scenarios is described in detail in previously published work by the team (Morrison et al., 2015). In brief, however, we followed established procedures for identifying sexual scripts as outlined by Maticka-Tindale et al. (2005). This involved qualitative interviews with a similar sample of heterosexually active young men who were asked to describe stories of actual sexual and romantic relationships. We coded these stories for elements including who was involved, when and where the encounters occurred, how they proceeded, and what norms and gender-related expectations were embedded in the way the scenes unfolded. Using the totality of these themes, we constructed short sexual scenarios representing common situations across participants.

Table 1 shows synopses of the situations reflected in each of the 11 scenarios. Men in the sample read each scenario, then responded to four questions regarding the situation depicted. Two items assessed the relevance of the scenario to individual respondents’ own behavior and desires: “How common is this situation for you?” (behavior) and “How desirable is this situation for you?” (desire). Next, to assess respondents’ perceptions of the behaviors and desires of men similar to them, they were asked, “How common is this situation for guys your age?” and “How desirable is this situation for guys your age?” Response options for the “common” questions were 0 (never happens), 1 (rarely happens), 2 (happens sometimes), 3 (happens fairly often), and 4 (happens very often); options for the “desirable” questions were 0 (very undesirable), 1 (undesirable), 2 (neither undesirable nor desirable), 3 (desirable), and 4 (very desirable).

Table 1.

Means (SD) and Paired Samples T-tests for “Common/desirable for me” Versus “Common/desirable for guys your age”

	Common for … .			Desirable for . . .
	You	Guys your age	t (df)	You	Guys your age	t (df)
Traditional masculinity sexual script
Mean across scenarios	1.11 (.89)	2.34 (.65)	14.28*** (484)	1.38 (.80)	2.75 (.91)	12.48*** (488)
Scenarios
1. One-night stand with new acquaintance after drinking. Loses interest because partner appears less attractive the next day and agreed to have sex too easily.	.78 (1.00)	2.54 (.78)	32.98*** (496)	1.16 (1.16)	2.29 (1.13)	20.23*** (494)
2. One-time casual hookup with recent acquaintance, loses interest once they’ve had sex	1.19 (1.27)	2.24 (1.04)	17.88*** (470)	1.49 (1.14)	1.92 (1.04)	9.30*** (473)
3. One-time sex with casual friend as revenge for girlfriend cheating	.55 (.95)	2.37 (.97)	28.79*** (490)	.81 (1.12)	1.69 (1.19)	17.51*** (490)
4. One-time sex with new casual dating partner, but loses interest once they’ve had sex.	1.24 (1.24)	2.52 (.97)	21.47*** (488)	1.76 (1.09)	2.27 (.92)	11.23*** (491)
5. Ongoing casual sex with recent acquaintance, no interest in long-term relationship - man doesn’t want to be tied down.	1.29 (1.31)	2.54 (1.03)	20.27*** (478)	1.63 (1.22)	2.51 (1.03)	15.49*** (479)
6. On-going significant casual sex with acquaintance who is dating someone else	1.00 (1.08)	2.08 (1.04)	19.41*** (485)	1.13 (1.17)	1.89 (1.10)	15.69*** (485)
7. Boring, unsatisfactory sex with monogamous partner; man has to coax woman into sex.	1.43 (1.25)	2.35 (.95)	16.09*** (479)	1.26 (1.10)	1.62 (1.13)	8.26*** (479)
8. On-going, intermittent sexual encounters with ex-romantic partner, but no chance of getting back together.	1.36 (1.25)	2.29 (1.01)	30.13*** (481)	1.78 (1.12)	2.45 (1.03)	21.79*** (480)
Positive woman sexual script
Mean across scenarios	1.62 (1.00)	2.38 (.75)	21.79*** (487)	2.75 (.91)	2.87 (.74)	3.67*** (489)
Scenarios
9. On-going, mutually fun sex between good friends	1.45 (1.28)	2.38 (1.03)	16.14*** (483)	2.52 (1.21)	2.90 (.97)	7.78*** (487)
10. On-going, mutually good sex in committed monogamous relationship	1.76 (1.38)	2.39 (.99)	10.84*** (479)	2.85 (1.08)	2.93 (.93)	2.08 (480)
11. Mutually good sex in new monogamous relationship	1.64 (1.24)	2.37 (.97)	12.30*** (489)	2.88 (1.20)	2.80 (1.00)	1.71 (493)

^***p <.001

Previous published exploratory factor analyses of these data detected two underlying factors in the set of scenarios (Morrison et al., 2015). Each factor characterizes a larger sexual “script,” or conceptualization of what is desirable and normative in sexual situations (Gagnon, 1990; Gagnon & Parker, 1995). Eight of the scenarios (first set in Table 1) comprised a sexual script we previously named “traditional masculinity;” these vignettes described situations consistent with a more casual, male-centered, non-intimacy-based approach to sexual activity that is also consistent with an “impersonal” approach to sex. The remaining three scenarios formed a factor we termed the “sex positive woman” script, as these situations collectively described an approach to sexual situations involving mutuality (though not always monogamy), and concern for both parties’ sexual satisfaction (second set in Table 1). Based on the aforementioned factor analysis demonstrating scenarios’ adherence to the respective scripts, we computed mean commonness and desirability scores across scenarios within the traditional masculinity and sexual positive woman scripts, respectively, for analyses.

Sexual assault perpetration.

We measured past use of sexual aggression with seven items from the Sexual Experiences Survey (Koss, Gidycz, &Wisniewski, 1987). These items used behaviorally specific anchors to assess whether participants had ever (yes or no) forced sexual contact by using continual force or arguments; attempted to force sex by using force or alcohol/drugs; or forced sexual intercourse by using physical force, continual arguments and pressure, or by using alcohol/drugs. Identically to Dardis et al. (2016), who also examined the accuracy of perceptions of peer norms, we created a sexual assault perpetration variable (coded as yes or no) in which a “yes” response to any question resulted the participant being identified as having used sexual aggression at least once.

Overestimation scores.

To test our second hypothesis, we compared men’s perceptions of the desires and behavior of “guys their age” to the actual mean scores of the desires and behaviors of men in the sample. Because the men in the sample were similar to one another along many dimensions (age, adolescence in the U.S. as a cultural reference point, identifying as heterosexually active, etc.), we view the sample as a valuable source of information about injunctive and descriptive norms for this cohort of young men. Specifically, to examine men’s relative degree of overestimation of what is common and desirable for other men, we calculated variables reflecting the difference between men’s scenario ratings about commonness/ desirability among other men versus the actual average self-rating of all of the men in the sample. We also created these scores for the mean ratings across scenarios within a particular script (see Table 1). For example, for commonness of the traditional masculinity script, we subtracted the sample mean of the “how common for you” ratings (in this example, 1.11), from each participant’s mean “how common for other men” score across scenarios in that script. Positive scores represent an over-estimation of how common that script was for the men in the sample, with higher scores indicating greater levels of over-estimation. In the same way, we computed an overestimation variable for desirability ratings for each of the two types of scripts.

Analysis Approach

We conducted all analyses in IBM SPSS V19.0. To examine hypothesis 1, we conducted paired sample t-tests examining differences between the sample mean of men’s ratings of the commonness and desirability of each scenario for themselves versus participants’ perceptions about other men their age. We considered participants’ perceptions to be inaccurate if they were statistically significantly different than the sample mean. We also conducted these analyses using mean scores (averaged across the scenarios). Because we computed multiples t-tests (one for each of the 11 items), we used a Bonferroni correction within the 11 commonness comparisons, and the 11 desirability comparisons. As a result, the significance cut-off for each t-test was .004.

To examine hypothesis 2, or whether inaccuracy (i.e., overestimation of other men’s behavior and desires) was related to sexual assault perpetration status, we conducted independent samples t-tests testing whether men’s overestimation scores (mean ratings across the scenarios) for the commonness and desirability of the two sexual scripts differed significantly between men who had perpetrated sexual aggression compared to those who had not. A Bonferroni correction was applied; across the four t-tests, the significance cut-off was therefore p < .012.

Results

Accuracy of Men’s Perceptions of Other Men’s Sexual Behavior and Desire

Table 1 presents results for hypothesis 1, that men would overestimate both the prevalence and the desirability of all sexual situations described in the vignettes. Men overestimated the commonness of both the overall traditional masculinity script and the sex positive woman script, as represented by mean commonness scores. Effect sizes were large (d = 1.37 and .80 for the traditional masculinity and the sex positive woman scripts, respectively). Concomitantly, men significantly overestimated the frequency of every scenario within both scripts. Specifically, the mean “commonness” rating for the men themselves fell somewhere between “rarely happens,” and “happens sometimes” for most vignettes (with two “traditional masculinity” scenarios falling between “never” and “rarely happens”). However, the perception of the norm for the commonness of the scenarios fell between “happens sometimes” and “happens fairly often” for all vignettes. The largest inaccuracies emerged for vignettes within the traditional masculinity sexual script, and included scenarios describing an alcohol-fueled one-night stand, and a one-time “revenge” hookup.

Findings regarding the accuracy of men’s perceptions of the average desirability of scenarios to other men were similar to the commonness findings, with a notable exception. As with perceptions of commonness, participants significantly over-estimated the mean desirability both of scenarios within the traditional masculinity sexual script (with a very large effect size; d = .98), and of all the individual scenarios in that script. Men’s average perception of the norm for the desirability of the sex positive woman script was statistically significantly higher than the sample mean, although small in effect size (d = .17). In terms of individual scenarios, men only significantly overestimated the desirability of one sex positive woman scenario. This scenario described a mutually enjoyable on-going sexual relationship between friends. In contrast, men accurately rated the desirability of sex in the context of a new monogamous romantic relationship, and of sex in a long-term monogamous relationship.

Accuracy of the Perceptions of Perpetrators vs. Non-perpetrators

Of the men in the sample, 161 (32.4%) reported using at least one form of sexually aggressive behavior. For 24 men (4.9% of sample), the most severe type of sexual aggression was unwanted sexual touching, 90 (18.3%) committed sexual coercion (using pressure or continual arguments to gain sexual access), and 47 (9.6%) committed attempted or completed rape. These rates are largely comparable, though slightly higher, than those found in college-specific samples; a recent systematic review of studies of college-enrolled men found an overall sexual aggression rate of 29%, with 6.5% of the men across studies reporting rape (Anderson et al., 2019). Whether men in our sample had ever used any form of sexual aggression was not significantly related to racial identity (χ² = 5.88, df = 4, p = .21), age (t= 1.85, df=487, p=.06), or participants’ mothers’ education (a proxy for socioeconomic status; χ² = 8.21, df=7, p = .32). Because our independent variable – sexual aggression – did not vary as a function of these demographic variables, demographics were not included as covariates in subsequent analyses.

Table 2 presents mean commonness and desirability difference scores within each sexual script for men who have vs. have not used any type of sexual aggression. Results for mean scores across scenarios are presented for parsimony and power, because of the consistent within-script findings from hypothesis 1. Hypothesis 2, that men who have used sexual aggression would overestimate the prevalence and desirability of the scenarios to a greater extent than non-perpetrators, was only partially supported. There was no statistically significant difference between perpetrators and non-perpetrators in their over-estimation of the commonness of the situations associated with either script. Men with a history of sexual aggression did, however, overestimate the desirability of the traditional masculinity script – but not the sex positive woman script -- to a greater extent than did men without such a history (with a modest size difference; d = .27). Because of this statistically significant finding for the traditional masculinity script, we examined the individual scenarios. Men with a history of sexual aggression were more likely to overestimate the desirability of five specific scenarios. These were scenario 2 (a one-time hook-up with an acquaintance; t = 2.46, p = .014), scenario 3 (revenge sex; t = 2.32, p = .02), scenario 4 (one-time sex with a casual dating partner; t = 2.58, p = .01), scenario 6 (sex with someone who is in a relationship with someone else; t = 2.43, p = .016), and scenario 7 (boring sex with a partner who needs to be coaxed; t = 3.18, p = .002) to a greater extent than did men without a sexual aggression history.

Table 2.

Means (Standard Deviations) and Paired Samples T-tests: Overestimation of Commonness/desirability by Sexual Assault (SA) Perpetration Status

	Common			Desirable
Script	No SA	SA	t (df)	No SA	SA	t (df)
Player	1.21 (.67)	1.28 (.62)	1.08 (480)	.65 (.61)	.81 (.59)	2.65** (480)
Sex positive	.74 (.73)	.82 (.78)	1.20 (489)	.14 (.72)	.09 (.79)	.66 (489)

^**p< .01

Discussion

This analysis examined the degree to which men in this racially diverse sample were inaccurate in their assessments of other men’s sexual desires and behaviors, and whether sexually aggressive men were particularly prone to overestimating peers’ sexual activity. Consistent with our first hypothesis, men in the sample significantly over-estimated the prevalence of all types of sexual situations presented, inclusive of both casual, “impersonal” sexual situations, as well as mutually satisfying sex within on-going relationships. In short, men in the sample believe that other men their age are encountering more of these sexual situations, and more often, than they actually are. Additionally, men over-estimated how desirable most of the sexual scenarios are to other men, although this was primarily true for the more impersonal scenarios consistent with the traditional masculinity sexual script. In contrast, men as a whole were more accurate in their perceptions of the desirability of sexual scenarios characterized by mutuality and intimacy – they also rated these scenarios, on average, as most desirable and common for themselves.

These findings are consistent with previous research showing that men overestimate peers’ number of sex partners (Lewis et al., 2014) and the frequency of discrete sexual behaviors such as kissing (Martens et al., 2006). To this previous literature, our findings add evidence that men also overestimate the typicality of different kinds of sexual situations and relationships. The reason for and impact of this overestimation is perhaps less clear. On the one hand, this evidence suggests that many men may see “typical” male sexuality as being different from their own. They perceive other men as both having (commonness) and wanting (desirability) more sex than themselves, particularly more adversarial or impersonal sex. Given the well-established impact of norms on behavior, this means that young men generally may feel pressure to live up to the fictional lives of their peers and particularly to the inaccurate notion that other guys are frequently experiencing “impersonal” hook-ups and casual sexual encounters. Further, their own, fully normative choices to pursue mutually satisfying sexual relationships may feel atypical.

Collectively, these findings reinforce the importance of attention to social norms in sexual health promotion. Accurate knowledge about social norms, and about the sex that other men have and want, could constitute an additional support for men’s own sexual choices and safety. Social norms approaches which use surveys to establish accurate normative information within social networks, and then media and social marketing to publicize this information, can be one approach to challenging men’s misperceptions about “typical” peer sexual behavior and to promoting sexual health (see, for example, Perkins & Perkins, 2018). Additionally, techniques from motivational enhancement approaches – successfully used for other behaviors such as substance abuse and risky sex -- may show promise (e.g., Picciano, Roffman, Kalichman, Rutledge, & Berghuis, 2001). Such approaches guide participants through comparing their own estimates of what is typical for their peers vs. what data shows is truly the case.

On the other hand, it is important to consider other possible explanations for the gap between men’s self-ratings, and their beliefs about guys their age. As noted above, men universally overestimated peers’ sexual behavior, which meant that ratings of their own desires and behaviors were consistently lower than what they believed to be true of their peers. This does not exclude the possibility that men’s own behavior is influenced by perceptions of peers but does reinforce that normative perceptions are but one correlate of sexual behavior. It may be that men in the study were aware of larger cultural scripts that link masculinity with frequency of sexual activity and an impersonal approach to sex and assume those scripts to be relevant to other men. Some participants in this study may not have access to enacting these scripts or do not themselves ascribe to or care about those cultural expectations. In short, for some men, impersonal sex may exist as something that “other” men commonly desire or do. This is consistent with some emerging research showing that men can have a variety of relationships to prevailing cultural sexual scripts, and may variously strive to enact or comply with those scripts (Dworkin & O’Sullivan, 2005), feign them (Murray, 2018), or define themselves against them (Masters et al., 2013).

Better understanding variations in men’s sense making and internalization of sexual scripts may assist with tailoring social norms-based interventions to men whose behavior may be negatively influenced by misperceptions of their peers’ desires and behavior. Additional research is needed that examines the directions and mechanisms of influence between men’s normative perceptions of peer sexual behavior and their own. For example, previous research has identified endorsing a traditional masculinity ideology, gender role stress, and membership in all-male peer social networks as risk factors for sexual risk taking (Casey et al., 2017; Reidy et al., 2016; Santana et al., 2006) as well as sexual aggression (Tharp et al., 2012). Thus, these potential mediators’ role in the interplay between injunctive and descriptive social norms and sexual behavior could be better elucidated in research. The qualitative meaning that men ascribe to the sexual scripts to which they are exposed is also an important on-going line of inquiry, as is understanding the role of cultural context and indicators of social positionality in both the relative power of social norms, and the specific ways of enacting sexuality that men feel pressured to emulate.

Our second hypothesis, that men with a history of sexual aggression would overestimate other men’s sexual activity to a greater extent than non-perpetrators, had mixed support. Perpetrators and non-perpetrators were both inaccurate in their perceptions of the frequency with which other men encountered the range of sexual situations and relationships depicted in the scenarios. It may be that these inaccurate messages about male sexuality are so pervasive that they are internalized across the board. It may also be that aggregating all types of sexual assault perpetration obscured differences in perceptions between men who have used different “levels” of sexual aggression (e.g. unwanted sexual contact vs. rape); an analysis prohibited by power limitations here.

Sexually aggressive men did overestimate the desirability of scenarios consistent with the traditional masculinity sexual script to a greater extent than their non-aggressive peers. While depicting consensual situations, these scenarios are consistent with an adversarial perspective on relationships between men and women and therefore with an “impersonal sex” approach to sexuality. As noted in the introduction, an impersonal approach to sex is a well-established correlate of sexual aggression (e.g., Tharp et al., 2012), and the content of the scenarios in the traditional masculinity script incorporated elements of both this non-intimacy-based approach to sex and masculinity-related expectations to have a lot of sexual encounters. Previous research has already demonstrated that perpetrators believe that their peers approve of coercive tactics to gain sexual access to women (Dardis et al., 2016; Thompson et al., 2015). We extend this with the finding that sexually aggressive men may also be inaccurate in their view that “guys their age” desire and approve of non-intimacy-based but non-aggressive sexual situations.

It may be important to sexual violence prevention endeavors to add explicit attention to misperceptions of the degree to which peers endorse “impersonal” approaches to sexual relationships as widely desirable. It is important to note that an impersonal or casual approach to sexual activity is not fundamentally incompatible with respectful, consent-based sexual encounters. At the same time, our findings suggest that it is important to talk about all kinds of sexual situations in the context of sexual violence prevention – not just situations involving sexual aggression. This may provide space for men to talk about the role of intimacy and consent in their own lives and to see those behaviors and desires reflected among their peers. Additionally, as bystander-based approaches to sexual violence prevention continue to be refined, it is important to add skill development related to challenging or interrupting talk in some male spaces that perpetuates misperceptions of men’s approval of non-intimacy based sexual encounters – even ones that are non-aggressive.

Limitations

Limitations to the study include the fact that participants were recruited and surveyed online and may not therefore constitute a representative sample of emerging adult men in the U.S. Although evidence suggests that Facebook users’ demographic profile is similar to the larger U.S. population (Chang, Rosenn, Backstrom, & Marlow, 2010), Craigslist users are anonymous, and the overall generalizability of findings presented here cannot be assured. Generalizability may be further compromised by our attempt to recruit a highly racially diverse sample - while a strength of the sample, the participants in this study do not precisely reflect the breakdown of the U.S. population by race/ethnicity. Additionally, this sample may be slightly higher educated than the larger U.S. population (DOE, 2014), although a strength of the sample is that it draws from both college and non-college enrolled young people. Finally, it is not possible to assess whether the desirability and commonness scores of men in the sample, which were used as an indicator of peer desires and behavior, would have been similar to the actual peer referents of men in the sample, although many demographic similarities are present in the group.

Conclusion

This study reinforces previous research showing that men misperceive and overestimate the amount of sexual activity in which their male peers are engaging and adds evidence that these inaccurate perceptions extend to a variety of specific sexual situations and relationship types. Given the power of social norms on behavior, these findings add to extant literature that implies that heterosexually active men feel unwarranted pressure to conform to mistaken ideas about masculine sexuality – thus underscoring the importance of including social normative components in sexual health promotion efforts. Perhaps more importantly, we found that men with a history of sexual aggression were more inaccurate in their perceptions of other men’s desire to engage in impersonal, non-intimate sexual scenarios consistent with a stereotypically “traditional masculinity” approach to sexuality than were their non-aggressive peers. Thus, sexual violence prevention efforts may need to extend to include attention not only to men’s beliefs about other men’s sexually aggressive behavior, but to also address men’s beliefs about the typicality of impersonal and non-intimate ways of engaging in sexual activity. Refining these efforts will be supported by additional research that examines attitudinal and behavioral mediators of the relationships between inaccurate assessments of peer sexual behavior, and sexual aggression.