Introduction

In contemporary society, it is widely believed that men and women are evaluated differently for engaging in similar levels of sexual activity (Milhausen & Herold 2001) and identical sexual acts (Sprecher 1989). This is called the sexual double standard (SDS: Reiss 1960, 1967): men are evaluated more positively or less negatively than women who have similar sexual histories. We survey southwestern U.S. college students to examine the possibility that the SDS still exists when evaluating those who engage in uncommon sexual acts (e.g., threesomes) as compared to common ones (e.g., monogamous sex) in contemporary Western society. The purpose of this study is to test for the continued existence of the SDS, albeit under highly specific and context-dependent conditions (Crawford & Popp 2003). We aim to demonstrate that changing social norms regarding sex acts have obscured the reality of a continued SDS. Such tests would provide evidence for the continued usefulness of the person-perception paradigm in SDS research. Our contention is that the erratic evidence for the SDS and the questionable usefulness of the person-perception paradigm within this topic area are driven, in part, by the commonness of the sexual acts targets engage in.

The SDS is a special case of general double standards (e.g., Wilcox 1997). A better understanding of the SDS may lead to a better understanding of double standards overall. Moreover, it is important to study the SDS because the impact of individuals’ sexual histories affects the perception and treatment of men and women (Crawford & Popp 2003; Marks & Fraley 2005, 2007), and the considerable impact this information has on people’s social and professional lives. For instance, when women in television shows engage in sexual acts, they are more often punished than men in those same shows (Aubrey 2004).

The SDS has been a popular topic for researchers over the last two decades (e.g., Gentry 1998; Jacoby & Williams 1985; Marks & Fraley 2005; Sprecher 1989), but evidence for its existence has been elusive. For instance, Gentry (1998) employed a person-perception task and found that raters judged men and women who had an above average amount of sexual activity similarly. In another person-perception study, Marks and Fraley (2005) found that even in the absence of valenced sexual information regarding sexual behaviors (e.g., insinuating that having many partners is a negative characteristic), individuals tended to equally derogate men and women who had similar sexual histories.

One proposed reason for this inconsistency is that the person-perception paradigm may be flawed (Crawford & Popp 2003; Marks & Fraley 2007). Although we do not discount this critique, we argue that what might be responsible for the erratic nature of the SDS is changing norms regarding sex across the genders on American college campuses (where most sexuality research is done; Oliver & Hyde 1993). Traditional research on the SDS assesses evaluations of targets who have engaged sexual acts that are routine by today’s standards (Oliver & Hyde 1993), such as erotic touching, sexual intercourse, oral sex, sex on the first date, casual dating sex, and serious dating sex (Sprecher et al. 1991). Prior authors have asserted that the SDS may have disappeared for common sexual acts (Sprecher & Hatfield 1996), but may still be present for unusual sexual acts (Browning, Hatfield et al. 2000). If this is the case, the person-perception paradigm should still be a useful tool to study the SDS, albeit at a greater level of specificity of the conditions under which the SDS should emerge (Crawford and Popp 2003).

There is some evidence consistent with our contention that the SDS will be found for uncommon sexual acts. For instance, when using ‘number of sex partners’ as the criterion to assess the SDS, researchers tend to use numbers that are only slightly discrepant from the median reported by most sex researchers (e.g., Brown & Sinclair 1999). For instance, the inability of Marks and Fraley (2005) to find evidence of the SDS for their 19 sexual partner condition may be the result of a weak manipulation, in that 19 sexual partners is not discrepant enough from the median number of sex partners. In contrast, Jonason (2007) found that when the number of sex partners was either 1 or 50, evidence of the SDS emerged. A man with 50 sexual partners was rated as having more status than a woman with 50 sexual partners and a man with 1 sex partner. A man and woman with 1 lifetime sexual partner were evaluated similarly. Thus, 50 sex partners, is sufficiently discrepant from the median number of sex partners to be considered unusual (in a statistical sense), hence the findings of Jonason (2007).

If deviation from the norm in heterosexual sexual encounters is a necessary condition to elicit the SDS within the person-perception paradigm, we first must choose a sexual act that is uncommon. One uncommon sexual act is a mixed-gender threesome, or a sexual act involving three people where at least one member of each gender is present. Such sex acts are rare on college campuses, with less than 8% of heterosexuals reporting one in their lifetime (Browning et al. 2000; Browning et al. 1999; Gallup et al. 2006; Hughes et al. 2004; Hughes & Gallup 2003; Laumann et al. 1994). Because of the rarity of the threesome, it is an ideal candidate for studying the SDS.

If evaluations are based on commonness of acts, then those who engage in threesomes should be derogated more than those who engage in monogamous sex. If this is the case, prior evidence for the SDS may have emerged because the acts studied were considered uncommon at that time. Specifically, we predict that targets who engaged in a threesome will be evaluated more favorably and less negatively than targets who engaged in monogamous sex (H1).

For a given sexual act, comparisons by gender should highlight the presence or absence of the SDS today (Mark & Miller 1986; Marks & Fraley 2005). Universally speaking, women are evaluated in a more negative or derogatory fashion for the same acts or the same level of sexual activity when compared to men (Baumiester & Twenge 2002). We expect that men will be evaluated more favorably than women, and women will be derogated more strongly than men who engaged in the same sexual acts (H2).

One potential individual difference of importance when individuals who engage in sexual activities are being evaluated is the gender of the participant (Crawford & Popp 2003; Jonason & Fisher 2008; Marks & Miller 1986). Women tend to evaluate others’ sexual activity in a derogatory fashion, whereas men tend to look more favorably on others who are sexually active (Jonason 2007; Jonason & Fisher 2008). We predict that derogation of targets will be strongest in women and favorable evaluations of targets will be strongest in men (H3).

We expect that threesome type will provide both main effects and interactions to predict ratings of targets. We suspect that main effects are likely because those who are sexually successful are evaluated more positively than other targets (Jonason 2007; Jonason & Fisher 2008). Specifically, we predict that a target who had a threesome with two members of the opposite gender will be evaluated more positively than the target who had a threesome with a member of each gender (H4).

We also expect two interactions concerning threesome type. When there are two opposite gender others present in a target’s threesome, individuals will have engaged in sexual acts that “count.” Sexual acts between two women may not “count” as sexual acts (Sanders & Reinisch 1999). If a man has a threesome with two women, he is able to add two people to his list of lifetime sex partners. When a woman has sex with two men she does the same, but when she has sex with a man and a woman, the woman may only count the man she had sex with, and not count the woman. When men evaluate those who have had high amounts of sexual activity they provide more favorable evaluations whereas women tend to derogate the same targets (Jonason 2007; Jonason & Fisher 2008). Therefore, we predict an interaction of gender of the participant and type of threesome on evaluations of targets, such that men will rate targets who have engaged in threesome with two opposite gender others more favorably than women (H5).

Additionally, a man who has had a threesome with one man and one woman may be considered bisexual or to have homosexual tendencies and may be evaluated in a more derogatory fashion than a man who has a threesome with two members of the opposite gender. In contrast, a woman who had a threesome with a man and a woman may be evaluated more favorably than a woman who engaged in a threesome with two opposite gender others because she is perceived to be bisexual or to have lesbian tendencies. Such an evaluation may be based on the eroticization of lesbians (Louderback & Whitley 1997; Whitley et al. 1999). Therefore, we predict an interaction of gender of the target and type of threesome on evaluations of targets, such that a woman who has a threesome with one man and one woman will be rated more favorably than a man who had a threesome with one man and one woman (H6).

Because threesomes have not yet been evaluated in reference to the SDS we wanted to avoid imposing any experimenter bias. Hence, we did not use author-generated scales as has often been done in research on the SDS. We employ the act-frequency paradigm (Buss & Craik 1983) in part 1 of Study 1 where participants were asked to report their perceptions of a target who engaged in a mixed-gender threesome in a qualitative fashion. Part 2 of Study 1 and Study 2 were based on these participant-generated responses.

Although we use a different methodology to generate evaluative terms, we expect them to reflect dimensions that have been assessed in the past. Adjectives from prior work appear to fall into two dimensions: positive or favorable evaluations and negative or derogatory evaluations. For instance, when participants have been asked how ‘popular’ they felt the target was (Marks & Fraley 2005) this item may have been reflective of this positive or favorable dimension of the evaluations individuals make regarding others who engage in sexual behaviors. We expect the terms generated in part 1 of Study 1 to conform to favorable and derogatory dimensions that prior authors have implicitly used in their single-item evaluations (RQ1). Additionally, we expect that men and women will be evaluated along similar lines (RQ2).

The use of these two measures can better highlight the nature of the SDS. Traditional research on the SDS rarely addresses both favorable and derogatory evaluations simultaneously and tends to assume that both are equivalent measures of the SDS, simply in different directions. This assumption warrants testing. By measuring the two dimensions in these studies we can test this assumption. If they are measuring the same thing, we would expect them to be correlated. Thus, we predict that the two dimensions will be correlated (H7).

Overview of the Current Studies

The current studies were designed to further assess the existence of the SDS, a phenomenon that has been described as elusive (Crawford & Popp 2003). We assess how evaluations differ across different types of sexual acts (monogamous sex vs. threesome) and how gender of participant and target drive favorable and derogatory evaluations of other’s sexual histories. Both studies utilize the person-perception paradigm and the first study utilizes an exploratory study to create measurement instruments for both.

Study 1

Because people can simultaneously make positive and negative evaluations of sexually active people, the SDS may be reflected in one or both types of evaluations, as opposed to some monolithic derogation of sexually active women. That is, men may be evaluated more favorably and derogated less for the same sexual acts as women. Women, in contrast, may be derogated more and evaluated less favorably than men for engaging in the same sexual acts (H2). The evaluations across male and female targets are also expected to vary in a similar fashion (H3). When considering threesome type, we predict that those who engage in threesomes with two members of the opposite gender will be evaluated more favorably (H4). Both target and participant gender should also interact with threesome type in predicting both positive and negative evaluations of targets (H5 & H6). Last, we test the implicit assumption in most SDS research that both the negative and positive evaluations of targets are equivalent, opposite-direction measures of the SDS (H7).

Method

Participants

One hundred twenty participants (54 men, 66 women; M Age = 20, SDAge = 3.28, Range = 18–37) from a large public university in southwestern U.S. participated in exchange for extra credit in their psychology class. Our sample size fits within a range set by Hynie and Lyndon (1995) and Marks and Fraley (2005) for studying the SDS. Based on the effect size reported by these authors (R 2 H&L = .13; R 2 M&F = .15) and an alpha of .05, our current sample size should be powerful enough to detect differences among our three predictor variables (.95 H&L;.98 M&F), as per Cohen et al. (2003).

Procedure

Participants were assigned to one of two between-subjects conditions that reflected the possible combinations of mixed-gender threesomes for a man (named John) and women (named Jane). Men evaluated a woman (n = 14) and a man (n = 13) who had a threesome with two opposite gender others and a woman (n = 14) and a man (n = 13) who had a threesome with a member of each gender. Women evaluated a woman (n = 17) and a man (n = 16) who had a threesome with two opposite gender others and a woman (n = 17) and a man (n = 16) who had a threesome with a member of each gender.

A 2 (participant gender) × 2 (threesome type) × 2 (target gender) design was used, where participants were given a vignette; one of which read: “John has had a threesome with 2 women.” Participants were asked how much do the following terms apply to your perceptions of him (1 = not at all; 5 = very much). Participants completed the evaluations alone in a room with a closed door. First, participants evaluated their assigned target. Second, participants responded to demographic questions. At the end of this session, participants were thanked and debriefed.

Measure

Evaluative terms were derived in an exploratory, act-nomination (Buss & Craik 1983) study of 36 (13 men, 23 women; M Age = 19, SDAge = 2.22, Range = 18–25) undergraduate psychology students from the southwestern U.S. who were not allowed to participate in any subsequent portion of this study. They were asked to list all the adjectives they could in response to two questions in this order: (a) What are your perceptions of a man who has had a threesome (where three people engage in simultaneous sexual acts) and (b) What are your perceptions of a woman who has had a threesome (where three people engage in simultaneous sexual acts).

A second group (the group corresponding the results below) was asked how much (1 = not at all; 5 = very much) these terms applied to targets who engaged in some type of threesome. These items were subjected to a Principle Components Analysis, revealing two dimensions, on their face, reflecting favorable (10% of the variance) and derogatory (38% of the variance) evaluations of targets. The sample size was adequate (Kaiser–Meyer–Olkin Measure of Sampling Adequacy = .89; Bartlett’s Test of Sphericity χ 2 (861) = 3,557.76, p < .01).

Measuring Derogations of Targets

Thirty-five items that reflected derogatory evaluations of targets (sinful, arrogant, being used, cocky, crazy, desperate, dirty, drunk, emotionally lacking, freak, greedy, gross, immoral, lost, low self-esteem, slut, whore, no self-respect, odd, scandalous, scary, selfish, sexual abuse victim, sick, stinky, has too many sexual desires, unpopular, undependable, unfaithful, unintelligent, unnatural, unstable, vain, wants attention, weird). These items were averaged together to create a single index for derogatory evaluations of the targets (Cronbach’s α = .96; M = 3.50, SD = .83).

Measuring Favorable Evaluations of Targets

Six items comprised favorable evaluations of the targets (confident, happy, horny, lucky, outgoing, wild). These items were averaged together to create a single index for favorable evaluations of the targets (α = .85; M = 2.69, SD = .95).

The two measures conform to the positive–negative distinction that other authors (e.g., Marks & Fraley 2005, 2007) have implicitly used in their research, confirming RQ1. When we examined the factor structures by gender of target we found that they were approximately equal, confirming RQ2. The factor structure, along with factor loadings, for the complete dataset and ones by participant gender has been omitted here for space, but can be obtained by contacting the first author.

Results and Discussion

In contrast to H7, both the dimensions were uncorrelated. Next, we report the results from a 2 (gender of participant) × 2 (gender of target) × 2 (type of threesome) MANOVA on two DVs (favorable evaluations and derogatory evaluations). We report only the Pillai’s Trace, as other tests (Wilks’ Lambda, Hotellin’s Trace, and Roy’s Largest Root) had identical results. Then we report the univariate tests (see Table 1). Results largely confirmed our predictions.

Table 1 Summary of means and standard deviations for Study 1.
Full size table

To test H2, we examined the role of target’s gender on evaluations. Significant multivariate main effects were found for gender of the target (F(2, 108) = 5.13, p < .01, η p 2 = .09) on the DVs. Gender of target had a significant main effect on derogatory (F(1, 117) = 9.51, p < .01, η p 2 = .08) and not favorable evaluations of targets. Jane was evaluated more negatively than John. Results partially confirmed our prediction.

To test H3, we examined the role of participant’s gender on evaluations. Significant multivariate main effects were found for gender of the participant (F(2, 108) = 4.14, p < .05, η p 2 = .07) on the DVs. Gender of the participant had significant main effects on both derogatory (F(2, 117) = 5.47, p < .05, η p 2 = .05) and favorable (F(1, 117) = 4.46, p < .01, η p 2 = .04) evaluations of targets. Women provided more derogatory evaluations of targets than men. Men provided more favorable evaluations of targets than women. Results confirmed our prediction.

To test H4, we examined the role of threesome type on evaluations of targets. Significant multivariate main effects were found for threesome type (F(1, 108) = 6.22, p < .01, η p 2 = .10) on the DVs. Threesome type had a significant main effect on favorable (F(1, 117) = 10.97, p < .01, η p 2 = .09) and not derogatory evaluations of targets. Target’s who had sex with two opposite-gender others were evaluated more positively than those who had a threesome with a member of each gender. Results partially confirmed our prediction.

To test H5, we examined the interaction of the gender of the participant and threesome type on evaluations of targets. Threesome type and gender of the participant interacted (F(2, 108) = 4.05, p < .05, η p 2 = .07) on the DVs. There was a significant interaction between gender of the participant and threesome type on positive evaluations (F(1, 117) = 7.72, p < .01, η p 2 = .07) and not derogatory evaluations. Men evaluated having a threesome with two same-gender others as more favorable than women. Men also evaluated those with same-gender threesomes more favorably than both men and women for those with one man and woman in their threesome. No other significant effects were found. Results partially confirmed our prediction (see Table 2).

Table 2 Interaction of threesome type by participant gender.
Full size table

To test H6, we examined the interaction of gender of the target and threesome type on evaluations of targets. There was no significant multivariate interaction of gender of target and threesome type. There was a significant interaction between the gender of the target and the threesome type on derogatory (F(1, 117) = 4.97, p < .05, η p 2 = .04) and favorable (F(1, 117) = 6.66, p < .05, η p 2 = .06) evaluations of targets. Jane, whether engaged in a threesome with two men or one man and one woman, was evaluated in a more derogatory fashion than John who had a threesome with two women, but not a man who had a threesome with a man and woman. When John engaged in a threesome with two women he was evaluated more favorably than when John had a threesome with a man and a woman, a woman who had a threesome with two men, and a woman who had threesome with a man and woman. The woman who had a threesome with a man and a woman was also rated less favorably than a women who had a threesome with two men or a man who had a threesome with a man and a woman. Results partially confirmed our prediction (see Table 3).

Table 3 Interaction of threesome type by target gender.
Full size table

To summarize Study 1, we assessed effects of participant gender, target gender, and threesome type to further our understanding of the SDS and to make the case that the SDS still exists, albeit under highly specific conditions. The target Jane was derogated more for engaging in the same sexual acts as John. On the other hand, it was for the target John that differences in the favorable evaluations emerged: Jane was evaluated neutrally across threesome type in favorable evaluations. Power analysis suggests 15 people per cell was required to detect the differences in this study, thus we were slightly underpowered in some tests. Despite this, we had no trouble detecting differences.

Study 2

In Study 2, we re-examine the presence or absence of the SDS, but ask participants to evaluate targets who engage in monogamous sexual activity. We again assess cross-gender evaluations for target (H3) and participant (H2). Such evidence will highlight how cross-gender evaluations will differ among those engaged in threesomes but not in monogamous sex. If the SDS is still present for monogamous sex, we would predict that men who engage in monogamous sex will be evaluated more favorably and derogated less than women who also engage in monogamous sex. We also expect that women will evaluate the targets in a more derogatory and less favorable fashion than men. We assess H1, which contends that common sexual acts are going to be evaluated in a more favorable fashion than uncommon sexual acts by assessing data from Study 1 and 2. We also re-examine the correlation between the two measurement dimensions, predicting that they will be correlated (H7).

Method

Participants

One hundred and five undergraduates (36 men, 69 women; M Age = 21, SDAge = 5.55, Range = 18–45) from a large public university in southwestern U.S. participated in exchange for extra credit in their psychology class. Participants from Study 1 were restricted from participation. A power analysis, based on the average effect size from Study 1 (Average R 2 = .13), an alpha of .05, and three predictors, revealed that this sample had adequate power (.92), as per Cohen et al. (2003).

Procedures and Measures

A 2 (participant gender) × 2 (target gender) design was used, where participants read a profile about a man (John) and woman (Jane) who was in a monogamous relationship where he or she only had sex with each other. Men evaluated a man (n = 18) and woman (n = 18) who engage in monogamous sex. Women evaluated a man (n = 34) and woman (n = 35) who engage in monogamous sex. Participants made between-subjects ratings of either John or Jane using measures from above on favorable (α = .79, M = 3.23, SD = .73) and derogatory (α = .97, M = 2.97, SD = .99) dimensions. The procedures were repeated from the act-frequency portion of Study 1.

Results and Discussion

In contrast to H7, but consistent with evidence from Study 1, both the dimensions were uncorrelated. A 2 (participant gender) × 2 (target gender) MANOVA was run on two DVs (derogatory and favorable evaluations), but only one main effect was found when targets were evaluated for involvement in a monogamous sexual relationship, men were evaluated more favorably for engaging in monogamous sex than women (see Table 4). Our inability to find evidence for derogatory evaluation of Jane when she engaged in monogamous sex, further supports H1, that the SDS is effectively absent for monogamous sex but not for threesomes. Results confirm our predictions that the SDS was relatively absent for sex in monogamous relationships.

Table 4 Summary of means and standard deviations for Study 2.
Full size table

To test H1, we examined data from both Study 1 and Study 2. Threesomes (M = 3.50) were derogated more (t(223) = 4.37, p < .01, d = .58) than monogamous sex (M = 2.97). Threesomes (M = 2.69) were also evaluated less favorably (t(223) = 4.37, p < .01, d = .64) than monogamous sex (M = 3.23). Results confirmed our prediction. The SDS appears to not be present for those who engage in monogamous sex but appears to still be present in those engaging in mixed-gender threesomes. It is also clear, that those who engage in monogamous sex are evaluated more favorably and derogated less than those who engage in mixed-gender threesomes. In contrast, the Study 1, we had sufficient power in this study (15 people per cell). Despite adequate power, we detected few differences when we considered evaluations of a man and a woman who engaged in monogamous sexual activity.

General Discussion

Most research on the SDS has focused on how the number, frequency, or type of sexual acts affects perceptions of men and women. However, the commonality of a sexual act may also result in differential evaluations of men and women who engage in that act. Specifically, because of changing attitudes about sex and the roles associated with it (Oliver & Hyde 1993), studying more novel sexual acts may have a higher probability of revealing the SDS. We found that the SDS tends to be present in uncommon sexual acts but not in common ones. It also yields evidence of the SDS using the person-perception paradigm. Last, the SDS appears to be focused on derogatory evaluations while concurrent favorable assessments may be present, perhaps, reflective of an approach-avoidance facet to the evaluations of other’s who are sexually active.

Our main thesis was that evaluations of sexual acts are sensitive to social norms regarding sex and, thus, acts that were derogated yesterday will no longer generate the same evaluations tomorrow. We found that the SDS can be found in uncommon sexual acts but cannot be found in common sexual acts. Men and women were derogated similarly when engaged in monogamous sexual acts, but differently when they were engaged in threesomes. Although ratings of targets in monogamous sexual acts and threesomes were done by different participants, these participants can reasonably be treated as equal (20–21 year old psychology students from southwestern U.S.).

Women derogated targets more than men. Men, on the other hand, rated the targets more favorably than women. This was even the case in Study 2, where no significant differences were found. Additionally, this is consistent with Jonason (2007) and Jonason and Fisher (2008), in that women derogated targets (regardless of targets’ level of sexual activity) more than men. It may be that college-aged women look down upon sexual acts more so than college-aged men and, thus, derogate those who engage in them more than men. It may also be the case that college-aged men want to have short-term sexual relationships with women who may be sexually available (Jonason 2007; Sprecher et al. 1997). Knowledge that she engaged in a threesome with two men suggests such availability, which possibly drove the favorable evaluations of Jane.

The SDS appears to be driven by the derogatory evaluation of women in less common sexual practices. A woman who had a threesome with a member of each gender was derogated more than the man who engaged in identical activity. These results suggest that the SDS appears in evaluations of individuals who engage in threesomes. These derogatory evaluations are likely the result of conservative sexual attitudes; where sexual acts are viewed in a negative fashion. There is some evidence that suggests that conservatism in the evaluations of targets who are sexually active is localized in women (Jonason & Marks, under review).

In contrast to the results concerning derogatory evaluations, a man who had a threesome with two women was rated more favorably than a man who had a member of each gender in his threesome. These results may reflect an effect of what has been called the eroticization of lesbianism (Louderback & Whitley 1997; Whitley et al. 1999). This is the idea that individuals, especially heterosexual men, tend to perceive sex among two women to be erotic. Although the evaluations were of the targets, it is possible that the difference could be driven by a halo effect of favorable evaluations of the two women in the threesome onto the male target. Alternatively, it may be that a man who has a threesome with a man and a woman is perceived to be bisexual or homosexual. There might be an undercurrent of homophobia for a man who is involved in the evaluations of others who have threesomes. Because the manipulation did not specify if the target had sex with all others in the threesome, some participants may have inferred that the man being evaluated was bisexual or homosexual and, thus, rated him less favorably. However, in theory there should have also been a corresponding horn/devil effect (Asch 1946) for women in the derogatory evaluations, but this was not observed. These results suggest that men are evaluated along the favorable dimension based on the number of women that are in the threesome. This is similar to what has been found previously: men with more sex partners are perceived as having more status (Jonason 2007). This final possibility may reflect Western conceptualizations of masculinity in that being sexually active is part of being a powerful man and may be reflective of the concept of hypermasculinty and how that is associated with reports of more sexual behaviors (Fisher 2007).

Crawford and Popp (2003) called for new methodologies to study the SDS. Our act-frequency study partially answered this call. The use of the act-frequency technique allowed for us to assess both derogatory and favorable evaluations simultaneously, unlike most work on the SDS, which tends to focus on one dimension at time. For instance, Marks and Fraley (2005) assessed favorable evaluations in terms such as ‘popularity.’ Our study demonstrates that concurrent assessments in both positive and negative directions and that the SDS may really more be about derogation than favorable evaluations. We showed in Study 2 that men who engaged in monogamous sex were evaluated more favorable than women. We may be tapping into approval of men for conforming to traditional sexual behavior. Additionally, we demonstrated that favorable and derogatory evaluations were uncorrelated in both studies, suggesting that those interested in the SDS should not assume that favorable and derogatory evaluations are simply opposite-direction measures of the same evaluative phenomenon.

Limitations and Future Directions

Although the act-frequency technique is superior at exploring novel topics via the removal of experimenter bias and increased ecological validity, it can only partially answer the call of Crawford and Popp (2003) for new methodology to study the SDS. The most notable problem is that the act-frequency technique is still rather decontextualized; a problem discussed in Crawford and Popp (2003) and partially addressed in Marks and Fraley (2005, 2007). On its face, our evidence suggests that the SDS is stronger for unusual sexual acts than more common ones. However, the decontextualized person-perception paradigm could have driven this effect (Crawford & Popp 2003). Perhaps ethnographic observations and discourse analysis might yield evidence of the SDS for a larger variety of sexual acts. Additionally, future work might include a social desirability scale because it is likely that ratings on scales in decontextualized experimental manipulations might be affected by egalitarian norms.

There are a few other limitations worthy of note. The most notable is the relatively small effect sizes in the interactions, and, thus, caution is necessary in interpretation. Both had small effect sizes, and while significant, the small effect sizes may denote spurious relationships or chance effects. We feel the main effects across and within the studies are more informative as to the presence and absence of the SDS today. Another limitation is the relatively small sample sizes. In Study 1, where we were slightly underpowered, we had no trouble detecting differences. In Study 2, we had adequate power and found few difference. This further suggests to us that the SDS is effectively absent for common sexual acts but remains for uncommon one.

The sample was constrained to U.S. psychology undergraduates and to one uncommon sexual act (i.e., threesome). Although in some ways results from college-student samples might be overestimates of sexual activities and attitudes when comparing college students to non-student samples (Wiederman & Whitely 2002), other evidence suggests the student and non-student samples can be comparable in sexuality research (Anthill & Russell 1982; Farky & Muellerms 1978). By assessing only one uncommon sexual act we are (1) unable to make generalization across other forms of uncommon sexual acts (e.g., partner swapping; Jenks 1998) and (2) to generalize beyond the SDS to general double standards (e.g., Wilcox 1997). However, there is evidence to support the notion that things that are different are judged in a more derogatory fashion, for instance, in prejudice research (e.g., Allport 1979) and, thus, we feel that our results should generalize to other sexual acts.

In conclusion, the common belief (Milhausen & Herold 2001) and recent empirical work (Marks & Fraley 2007) that argues for the continued existence of a SDS was confirmed. However, in contrast to prior work, this study took the approach that the erratic support for the SDS has been the result of changing social norms regarding sex and, thus, sexual acts that occur infrequently offer a more viable avenue to find evidence for it. It is possible that with more social liberalization, double standards like the SDS will become meaningless.