Evolutionary gender differences in sexual preferences
Kodi B. Arfer and Nicholas R. Eaton
Created 27 Oct 2014 • Last modified 27 Jan 2016
We investigated a variety of hypothesized gender differences, particularly hypotheses grounded in evolutionary theory, in what people like sexually. Over 900 Internet users living in the United States rated how appealing they found several kinds of sexual partners, activities, and themes. There were 41 items grouped into five categories of hypotheses: social distance, interpersonal power, reproductive health, degree of sexual experience, and pornography. We assessed item-level gender differences as well as overall support for each theory-derived category. Our findings were generally supportive of evolutionary theory. Men reported more interest than women in impersonal sex, with little romantic commitment or knowledge of partners. Men reported more interest than women in dominating or active roles, and women reported more interest in submissive or passive roles. We also found that men were more sensitive to cues of female reproductive value such as youth and virginity and were more interested in visual pornography, whereas women found being sexually desired by their partner more appealing. Overall, our results support the value of evolutionary theory (particularly concerning minimal parental investment and paternal uncertainty) for understanding human sexual interests. Our results also suggest a close relationship between power and sexuality, consistent with longstanding feminist arguments.
Introduction
Gender differences in sexuality are perennial research topics (e.g., Baumeister, Catanese, & Vohs, 2001; Chivers, Seto, Lalumière, Laan, & Grimbos, 2010; Petersen & Hyde, 2010). Several theoretical frameworks imply certain gender differences in sexual interests. Relevant theories include those grounded in evolution such as differential parental investment (Buss & Schmitt, 1993) and paternal uncertainty (Buss, 1988), as well as feminist theory (e.g., Brownmiller, 1984). In this study, we tested predictions of these theories about men's and women's sexual preferences using a single dataset, and we assessed overall support for each of several categories of hypotheses.
Our concern is with a large and inclusive notion of sexual preferences. This term, "sexual preferences", often refers specifically to preferred gender of sexual partner (and thus is to some degree synonymous with "sexual orientation"). However, in sexuality as in all domains of human life, such as food, sports, and research, there are widespread individual differences in what people like; sexual partner gender is not the only relevant dimension. This diversity is illustrated by the diversity of Internet pornography (e.g., Ogas & Gaddam, 2011) and Internet communities devoted to particular sexual interests (e.g., Scorolli, Ghirlanda, Enquist, Zattoni, & Jannini, 2007) or even the lack of sexual interests (Chasin, 2011). This is not to say that sexual preferences are entirely idiosyncratic. A substantial body of theory can be used to make detailed predictions about sexual preferences on the basis of gender. Below we review this theory and our own hypotheses, with an emphasis on gender differences that can be examined with the dataset used in this study. Each section describes a related category of hypotheses.
Theory and hypotheses
Social distance from partners
Evolutionary theory contends that the challenges men and women face to successfully reproducing differ importantly in terms of minimal parental investment. For men, the minimum investment is only an act of coitus and some semen. Women, on the other hand, must support the developing fetus for the duration of pregnancy and then, usually, breastfeed the child for some time after birth. Therefore, from the perspective of evolutionary fitness, men have the most to gain by impregnating large numbers of women, whereas women have the most to gain by conceiving only in the circumstances most favorable to the success of the resulting child. Sexual strategies theory (Buss & Schmitt, 1993) reasons that natural selection should have ensured that men, on average, are more interested in shorter-term, less committed sexual relationships, and are less selective of partners.
We hypothesize that men will be more tolerant of social distance from sex partners—that is, lack of commitment and intimacy—than women will. Specifically, men should be more interested than women in sexual activity with people to whom they are less committed, in situations implying less knowledge of the partner and lower probability of a lasting partnership. Given this evolutionary reasoning, men should be more interested in sex with friends, acquaintances, and strangers, and more interested in group sex. This is suggested by men's greater receptivity to offers of casual sex (Clark & Hatfield, 1989; Conley, Ziegler, & Moors, 2013), their greater concern for attractiveness than agreeable personality in opposite-sex friends (whereas women show the opposite pattern; Lewis et al., 2011), and their higher rate of fantasizing about group sex (Briere, Smiljanich, & Henschel, 1994). Women, conversely, should be more interested in sex in a committed relationship, and more interested in romantic gestures that suggest commitment, such as hugging and kissing. There is evidence that intimacy and romance are particularly important for women's sexuality (Meana, 2010). We hypothesize also that solo masturbation, as a sexual act with no partner at all, should appeal more to men than women. More men than women report masturbating regularly (Laumann, Gagnon, Michael, & Michaels, 1994). Evolutionary theory can still be applied to non-reproductive sexual behavior, such as masturbation, using arguments that such behavior can ultimately increase fitness (Arfer, 2014a).
Power dynamics with partners
Evolutionary theory also emphasizes the asymmetry between the sexes concerning knowledge of a child's parents. Women, through the experience of giving birth, can readily tell whether a child is theirs or not. Men, on the other hand, can never be sure (without modern technology) that what they believe to be their child was not actually fathered by another man. Thus arises the danger of a man being cuckolded, spending resources on a woman and her child without spreading his own genes. The threat of cuckoldry is one way in which men are incentivized to mate-guard, that is, to restrict the sexual freedom of women they hope to mother their children, and to restrict other men's sexual access to these women (Buss, 1988; Marlowe, 2000). Because, in general, power over women is helpful to men's reproductive interests (by also, e.g., allowing them to bypass women's desire for commitment), Smuts (1995) (see also Vandermassen, 2011) has argued that these evolutionary forces have driven men to desire power over women generally. This biosocial line of thought converges with the feminist argument (e.g., Brownmiller, 1975; Brownmiller, 1984) that men seek to dominate women, especially in sexual domains.
Conversely, we might expect women to be interested in cooperating with men's power structures over them, or at least appearing to cooperate. Women seek, per the theory discussed in the previous section, the best men available, who in turn may demand submission to certain regulations as a price for investing their resources. For example, Dickemann (1979) discussed how in cultures practicing purdah (the separation of women from men through veiling and exclusion from public spaces), higher-status men enforce purdah more consistently. Further, by making a show of submitting to men's authority, women may influence their mates to be less vigilant, allowing them to have sex with other men when it is advantageous for them to do so.
We hypothesize that men have a general sexual interest in domination, and women in submission. Accordingly, for any sexual act in which one partner is in some way controlling or humiliating the other, or simply where one is the actor and the other is the target, men should be more interested than women in the dominant or active role, and women should be more interested than men in the submissive or passive role. These hypotheses are supported by the finding of Joyal, Cossette, and Lapierre (2014) that more women than men report fantasizing about being sexually dominated, and more men than women about sexually dominating someone else (see also Leitenberg & Henning, 1995; Critelli & Bivona, 2008). Similarly, men should be more interested in opportunities for exploitative sex, such as sex with an unconscious person or other forms of rape. This would be consistent with the studies of sexual fantasy just cited as well as with men's higher rates of committing sexual abuse (e.g., Brousseau, Bergeron, Hébert, & McDuff, 2011).
Reproductive health of partners
Reproductive value depends on age more for women than men, because women take a long time to produce each child and eventually become infertile due to menopause. Hence, according to evolutionary logic, men should be more interested in younger sex partners (Buss & Schmitt, 1993). This would be consistent with Hayes's (1995) study of personal advertisements, which found that men seeking women preferred partners younger than themselves. Men should be more interested in large buttocks because the fat deposits of a mother are associated with healthier children (Caro & Sellen, 1990). Men should also be more interested in small feet, because smaller feet indicate youth and nulliparity (Fessler et al., 2005). Since women should have no special foot-size preferences except perhaps for typically sized feet (Fessler et al., 2005), we weakly hypothesize that women will find large feet more attractive than men will.
Sexual experience of partners
Because of the threat of cuckoldry, men should find virginity particularly appealing, on the assumption that a less sexually active woman will be less likely to cheat (Buss, 1989). Women, on the other hand, should find sexual experience attractive. According to the sexy-son hypothesis, women stand to gain from reproducing with more sexually successful men because their sons will then be more likely to be sexually successful (Weatherhead & Robertson, 1979).
Pornography
Ogas and Gaddam (2011) observed that visual forms (pictures and videos) of pornography or erotica are more popular among men than women, whereas written forms (such as romance novels) are more popular among women than men. Ellis and Symons (1990) theorized this difference arises from differences in minimal parental investment: Women care more about emotions and context, whereas men care more about raw fertility cues.
Additional items
As part of the concerns with mate quality discussed above, women should be particularly interested in wealthy, high-status partners. In their study of men's pornography, Ogas and Gaddam (2011) (p. 57) noticed a tendency for emphasis on women's sexual pleasure, and claimed men are particularly concerned that a woman's expressed sexual pleasure is genuine in order to help ensure her faithfulness. Conversely, we may expect women to find it appealing to be desired or attractive, in order to attract better mates or to better match passive sexual roles as described above.
The present study
In this study, we tested hypotheses of gender differences in sexual preferences by comparing the responses of men and women on 41 self-report items. We used a large and diverse sample, and instead of examining only whether observations significantly deviated from null hypotheses, we predicted population distributions of responses and compared these distributions between genders. This approach allows us to, for example, distinguish between the case of both genders rating one item very highly but one slightly higher than the other, and the case of one gender rating an item highly and the other rating it low. Finally, we quantified and assessed overall support for each category of theoretical hypotheses.
Method
See http://arfer.net/projects/galaxy for task code (including the complete instrument) and data-analysis code.
Subjects
We recruited and ran subjects on Amazon Mechanical Turk (MTurk), a data-collection website. MTurk "workers" are ordinary Internet users aged 18 and above who volunteer to complete short unskilled online tasks for nominal fees. Increasingly, MTurk is being used by researchers to study human behavior. For example, Suri and Watts (2011) studied social networks, and Eriksson and Simpson (2010) examined gender differences in risk preferences. Evidence that MTurk can provide data comparable in quality to that produced in conventional laboratory settings is provided by studies such as Buhrmester, Kwang, and Gosling (2011), which replicated previous estimates of the retest reliability of several personality measures. While samples of Americans on MTurk are not random samples of the US population, they are likely to be more diverse (Ipeirotis, 2010), and therefore more representative of the US population, than the undergraduate subject pools used in most studies with human subjects. Moreover, we expected that subjects using an anonymous online platform would be more honest about their sexual preferences and less influenced by social desirability.
A total of 1,001 MTurk users completed the study. The study was described as "a survey about sexual feelings and sexual behavior". Subjects were required to live in the United States. The median completion time was 13 minutes. Subjects were compensated with $2.
Instrument
The survey subjects completed had eight major parts. In this study, we are concerned chiefly with part 2 (concerning sexual preferences) and use data only from part 2, part 1 (concerning demographics), part 7 (concerning sexual orientation), and part 8 (concerning data integrity).
Part 1 asked questions about gender, age, etc. The gender item had four fixed options—"Male", "Female", "Transgender (male-to-female)", and "Transgender (female-to-male)"—and a free-response option labeled "Other".
In part 2, subjects were told "We would like you to consider how sexually appealing you find each of these activities." They were presented with a number of short descriptions and rated each on a 7-point scale with anchors "Not at all appealing", "Somewhat appealing", and "Very appealing". There were 99 items in total, but for many of these, we had no theoretically meaningful hypothesis of which gender should rate the item as more appealing. In these cases, either the direction was trivially predictable from the principle that people are on average more sexually interested in the opposite gender than their own gender, or we had no hypothesis at all. We were able to formulate hypotheses for 40 items, as well as for solo masturbation, for which we constructed a 41st by simply combining the male- and female-specific solo-masturbation items. See the appendix for complete instructions and the text of these 41 items.
In part 7, we asked "What is your sexual orientation?" The options were "Heterosexual or straight", "Gay or lesbian", "Bisexual", "Asexual", "Not sure", and a free-response option.
Part 8 repeated two questions from part 2 as validity items: one about fellating someone else and one about sex with a friend. A third validity item asked "How honest were you able to be while answering this survey? (We won't withhold payment based on your answer to this question.)" Subjects responded on a 5-point scale with anchors "Not at all honest", "Somewhat honest", and "Entirely honest".
Analytic sample
To reduce the effect of random or invalid responses on our results, we excluded subjects from analysis on the basis of response time and their responses to the three validity items in part 8. Subjects were excluded if they completed the task in less than 7 minutes (n = 46), they rated their honesty as "Somewhat honest" or less (n = 20), or their responses to both repeated items differed from their original responses by 3 units or more (n = 5). Finally, since this study concerns sex differences and gender differences (without trying to distinguish the effects of anatomy, endocrinology, social role, gender identity, etc.), we excluded all subjects who chose an option for the gender item other than "Male" or "Female" (n = 15). The resulting analytic sample comprised 924 subjects.
Note that subjects were not excluded on the basis of sexual orientation identity or gender preferences. Because these constructs are not as concrete, stable, or readily observed as gender, it is most useful to be able to predict sexual preferences on the basis of gender alone. Besides, although evolutionary theory of mating preferences generally concerns heterosexual relations, people with preferences or orientations other than strict heterosexuality have the same species and gender as heterosexuals, so perhaps their sexual emotions operate similarly.
Results
Sample
Of the analytic sample, 460 subjects (50%) were female. Ages ranged from 18 to 82 (median 31). Subjects identified as 80% White, 10% Black, 8% Asian, 6% Hispanic, and 2% Native American (they were allowed to select more than one of these categories). With regard to education, 12% had a graduate degree, 39% had an undergraduate degree, 37% had had some college education, 11% had graduated from high school, and the remaining 1% had not completed high school. The median annual-income category was $20,000 to $29,999. Most subjects (97%) identified as native speakers of English, and the remainder as fluent non-native speakers of English. With regard to sexual orientation identity, 79% were heterosexual, 13% were bisexual, 4% were gay or lesbian, 1% did not know, 1% made a free response (e.g., pansexual), and less than 1% (3 subjects) were asexual.
Preferences
We examined endorsements of the preferences items as a contingency table, with the 7 possible ratings as columns and the 82 item–gender combinations as rows (82 being the product of the 41 preferences items and the 2 genders). We tested the null hypothesis that ratings were independent of item–gender combination. The results were significant, χ2(486) = 29,115, p < .001. Separate χ2 tests for each item, testing the independence of gender and rating, yielded p < .05 for all but 6 of the 41 items.
To interpret gender differences on each item, we examined the cell-level data of the contingency table (Table 1 and Figure 1), which provides richer information than the simple presence of absence of significant gender differences.1 For each theoretically relevant category of items (e.g., social distance), we also performed a two-tailed binomial test of the null hypothesis that for 50% of the possible items in that category, the gender difference (as judged from the contingency table) was in the hypothesized direction.
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | |
|---|---|---|---|---|---|---|---|
| Social distance from partners | |||||||
| Love M ✔ | 0 | 1 | 1 | 5 | 8 | 12 | 73 |
| Love F ✔ | 1 | 0 | 1 | 2 | 3 | 9 | 85 |
| Intimacy M ✔ | 1 | 1 | 1 | 7 | 7 | 17 | 66 |
| Intimacy F ✔ | 1 | 0 | 0 | 2 | 4 | 9 | 83 |
| Kissing M ✔ | 1 | 1 | 3 | 9 | 14 | 21 | 52 |
| Kissing F ✔ | 1 | 1 | 1 | 7 | 6 | 13 | 71 |
| Hugging M ✔ | 2 | 6 | 10 | 22 | 12 | 14 | 34 |
| Hugging F ✔ | 2 | 3 | 4 | 15 | 9 | 11 | 56 |
| With relationship partner M ✔ | 1 | 0 | 0 | 3 | 6 | 16 | 74 |
| With relationship partner F ✔ | 1 | 0 | 1 | 2 | 3 | 10 | 83 |
| With friend M ✔ | 5 | 6 | 5 | 21 | 15 | 19 | 29 |
| With friend F ✔ | 27 | 11 | 14 | 22 | 10 | 7 | 9 |
| With acquaintance M ✔ | 9 | 8 | 11 | 19 | 17 | 13 | 24 |
| With acquaintance F ✔ | 40 | 15 | 10 | 16 | 9 | 6 | 5 |
| With stranger M ✔ | 13 | 7 | 9 | 21 | 15 | 14 | 21 |
| With stranger F ✔ | 43 | 12 | 9 | 17 | 6 | 6 | 7 |
| Group sex M ✔ | 10 | 6 | 7 | 20 | 13 | 14 | 31 |
| Group sex F ✔ | 33 | 10 | 11 | 13 | 8 | 9 | 17 |
| Solo masturbation M ✖ | 4 | 4 | 9 | 25 | 20 | 18 | 19 |
| Solo masturbation F ✖ | 6 | 4 | 5 | 16 | 11 | 16 | 41 |
| Power dynamics with partners | |||||||
| With rebel M ✖ | 6 | 3 | 5 | 17 | 19 | 19 | 32 |
| With rebel F ✖ | 10 | 4 | 5 | 18 | 16 | 14 | 32 |
| Being raped M ✔ | 81 | 7 | 4 | 3 | 1 | 1 | 2 |
| Being raped F ✔ | 75 | 6 | 5 | 7 | 3 | 2 | 2 |
| Raping M ✔ | 80 | 7 | 3 | 3 | 2 | 2 | 2 |
| Raping F ✔ | 90 | 5 | 1 | 1 | 1 | 1 | 1 |
| Being controlled M ✔ | 20 | 14 | 12 | 20 | 11 | 11 | 13 |
| Being controlled F ✔ | 13 | 7 | 9 | 15 | 10 | 16 | 29 |
| Controlling M ✔ | 10 | 9 | 14 | 20 | 14 | 13 | 20 |
| Controlling F ✔ | 20 | 13 | 12 | 22 | 10 | 9 | 13 |
| Being tied up M ✔ | 28 | 12 | 12 | 19 | 11 | 9 | 10 |
| Being tied up F ✔ | 18 | 7 | 9 | 14 | 11 | 12 | 29 |
| Tying up M ✖ | 20 | 11 | 9 | 23 | 10 | 11 | 16 |
| Tying up F ✖ | 21 | 10 | 10 | 19 | 12 | 9 | 19 |
| Exhibition M ✔ | 45 | 9 | 11 | 13 | 7 | 7 | 9 |
| Exhibition F ✔ | 33 | 5 | 9 | 17 | 11 | 7 | 16 |
| Voyeurism, known M ✔ | 15 | 9 | 9 | 18 | 15 | 12 | 23 |
| Voyeurism, known F ✔ | 42 | 8 | 7 | 12 | 7 | 8 | 15 |
| Voyeurism, secret M ✔ | 19 | 12 | 9 | 17 | 13 | 12 | 19 |
| Voyeurism, secret F ✔ | 33 | 10 | 10 | 14 | 7 | 9 | 16 |
| Insulting partner M ✔ | 33 | 15 | 11 | 14 | 8 | 9 | 10 |
| Insulting partner F ✔ | 57 | 12 | 9 | 8 | 6 | 4 | 4 |
| Being insulted M ✖ | 40 | 15 | 13 | 14 | 7 | 4 | 7 |
| Being insulted F ✖ | 52 | 9 | 8 | 9 | 8 | 6 | 7 |
| Urinating on partner M ✔ | 71 | 8 | 5 | 5 | 3 | 4 | 3 |
| Urinating on partner F ✔ | 84 | 5 | 3 | 3 | 1 | 2 | 2 |
| Being urinated on M ✖ | 76 | 6 | 4 | 5 | 4 | 2 | 3 |
| Being urinated on F ✖ | 87 | 3 | 3 | 4 | 1 | 1 | 1 |
| With unconscious partner M ✔ | 65 | 12 | 7 | 7 | 4 | 2 | 3 |
| With unconscious partner F ✔ | 85 | 6 | 1 | 4 | 1 | 1 | 1 |
| With 70-year-old M ✔ | 71 | 14 | 8 | 3 | 2 | 1 | 1 |
| With 70-year-old F ✔ | 79 | 8 | 5 | 5 | 1 | 1 | 0 |
| With 8-year-old M ✔ | 94 | 2 | 0 | 1 | 0 | 1 | 1 |
| With 8-year-old F ✔ | 97 | 1 | 0 | 0 | 0 | 0 | 0 |
| Incest M ✔ | 80 | 5 | 5 | 6 | 2 | 1 | 1 |
| Incest F ✔ | 94 | 3 | 1 | 1 | 0 | 0 | 0 |
| Pain M ✔ | 51 | 21 | 8 | 9 | 5 | 3 | 3 |
| Pain F ✔ | 45 | 12 | 11 | 12 | 9 | 6 | 5 |
| Causing pain M ✔ | 51 | 16 | 10 | 12 | 4 | 2 | 5 |
| Causing pain F ✔ | 58 | 13 | 8 | 10 | 5 | 2 | 3 |
| Reproductive health of partners | |||||||
| With 14-year-old M ✔ | 80 | 7 | 5 | 4 | 1 | 1 | 3 |
| With 14-year-old F ✔ | 96 | 2 | 1 | 0 | 1 | 0 | 0 |
| Small feet M ✔ | 17 | 9 | 15 | 33 | 12 | 7 | 7 |
| Small feet F ✔ | 38 | 15 | 12 | 28 | 4 | 2 | 1 |
| Big feet M ✖ | 32 | 20 | 17 | 17 | 5 | 4 | 6 |
| Big feet F ✖ | 38 | 15 | 14 | 20 | 5 | 5 | 3 |
| Big butt M ✔ | 9 | 8 | 9 | 17 | 14 | 14 | 29 |
| Big butt F ✔ | 28 | 16 | 15 | 21 | 9 | 3 | 8 |
| Sexual experience of partners | |||||||
| With experienced partner M ✖ | 5 | 4 | 10 | 18 | 21 | 18 | 25 |
| With experienced partner F ✖ | 10 | 7 | 12 | 24 | 13 | 14 | 21 |
| With virgin M ✔ | 3 | 7 | 10 | 21 | 13 | 16 | 30 |
| With virgin F ✔ | 34 | 11 | 10 | 22 | 4 | 7 | 12 |
| Pornography | |||||||
| Porn, written M ✔ | 5 | 9 | 8 | 23 | 18 | 14 | 23 |
| Porn, written F ✔ | 10 | 4 | 6 | 17 | 14 | 13 | 36 |
| Porn, visual M ✔ | 2 | 2 | 3 | 17 | 16 | 21 | 38 |
| Porn, visual F ✔ | 10 | 6 | 7 | 14 | 16 | 18 | 31 |
| Additional items | |||||||
| With rich person M ✖ | 5 | 5 | 6 | 22 | 18 | 15 | 30 |
| With rich person F ✖ | 11 | 6 | 9 | 25 | 12 | 13 | 24 |
| Being desired M ✔ | 0 | 0 | 0 | 2 | 3 | 13 | 82 |
| Being desired F ✔ | 1 | 1 | 1 | 1 | 2 | 6 | 88 |
| Giving pleasure M ✔ | 0 | 0 | 0 | 2 | 4 | 8 | 84 |
| Giving pleasure F ✔ | 1 | 0 | 0 | 3 | 5 | 11 | 79 |
Social distance from partners
As expected, men were more interested than women in sex with people who are not relationship partners (namely friends, acquaintances, and strangers) and in group sex. Women's modes for these items were 1, whereas men's were 4 or 7, depending on the item. Conversely, women gave somewhat higher ratings to items implying commitment: Although love, intimacy, hugging, kissing, and sex with a relationship partner were all near ceiling, women were 9 to 23 percentile points more likely to give a rating of 7. Contrary to our hypothesis, however, solo masturbation was rated more highly by women (mode 7) than men (mode 4). In summary, 9 of 10 item-level gender differences were consistent with our hypotheses regarding social distance, which was significantly greater than 50%, p = .02.
Power dynamics with partners
Generally, we saw the expected pattern of men preferring items related to domination and women preferring items related to submission. Women were less likely to choose 1 for being raped, experiencing pain, or exhibiting themselves, and more likely to choose 7 for being controlled and being tied up. Men were less likely to choose 1 for controlling their partner, raping their partner, insulting their partner, urinating on their partner, being a voyeur, or causing pain. Men were also less likely to choose 1 for other possibly exploitative items, namely sex with unconscious people, 70-year-olds, 8-year-olds, and relatives. Four items, however, showed a gender difference in the opposite of the hypothesized direction. Men were less likely than women to choose 1 for sex with a rebel, being insulted, and being urinated on, and women were slightly more likely to choose 7 for tying up their partner. In summary, 16 of 20 items were consistent with theory, which was significantly greater than 50%, p = .01.
Reproductive health of partners
Men were less likely than women to choose 1 for sex with a 14-year-old and reported substantially more attraction to people with large buttocks. They were more interested than women in small feet. But they were also slightly more interested in big feet, contradicting our weak expectation for this item. In summary, 2 of 3 items (66%) were consistent with theory. (We do not conduct binomial tests for this and the following categories because a two-tailed binomial test needs at least 6 trials to yield a p-value less than .05.)
Sexual experience of partners
The gender difference for sex with a virgin was large, with a mode of 7 for men and 1 for women. For sexually experienced partners, contrary to hypothesis, women were more likely to choose 1. In summary, 1 of 2 items (50%) were consistent with theory.
Pornography
As hypothesized, men were more interested in visual forms of pornography; the gender difference is clearest in that women were much more likely to choose 1. For written forms of pornography, women were more likely to choose 7. (Women were also more likely to choose 1, but equally likely as men to choose either of 1 or 2, making interpretation of the low end of the scale unclear.) In summary, both items were consistent with theory.
Additional items
Contrary to our hypothesis, men were more interested in sex with a rich, high-status person. The items for being sexually desired and giving sexual pleasure were near ceiling, with over 75% of subjects choosing 7 for both, but the expected gender differences were still slightly evidenced, with women more likely to choose 7 for being desired and men more likely to choose 7 for giving pleasure.
Discussion
Over 900 Internet users living in the United States rated how appealing they found several kinds of sexual partners, activities, and themes. As hypothesized, we found generally that men were more interested than women in sexual situations implying lower romantic commitment and less knowledge of the partner. In situations implying power differentials, we found that men were interested in dominating or active roles more than women, and women were interested in submissive or passive roles more than men. We also found that men were more sensitive to cues of female reproductive value such as youth and virginity and were more interested in visual forms of pornography, whereas women found being sexually desired by their partner more appealing. Some results contrary to our hypotheses include women's greater interest in solo masturbation and men's greater interest in sex with a rich person. Overall, however, our results support the value of evolutionary theory (particularly concerning minimal parental investment and paternal uncertainty) for understanding human sexual interests.
Many previous studies, cited earlier, have examined similar variables to those included in our study, such as interest in casual sex. Our study adds to the existing literature with its large, diverse, and thoroughly anonymous sample; the measurement of many different aspects of sexual preferences (from foot size to incest) in the same sample with parallel items; the focus on distributions rather than central tendency (so that trends such as bimodality and between-gender differences in variability can be seen); and its occasional unanticipated results, such as men's greater interest in sex with a rich person and women's particularly low interest in virgins (although in that case the gender difference is in the expected direction).
Perhaps the most important findings of our study, along with Joyal et al. (2014), are those concerning interpersonal power. As discussed in the introduction, power, sexuality, and the interplay thereof play key roles in both evolutionary and feminist theory. These findings of generalized sexual interest in domination by men and submission by women raise the possibility that sexual motivation underlies gender differences in power generally. This is not to say that men's oppression of women is good or inevitable, any more than our finding of some subjects who report sexual interest in being raped implies that rape is good or inevitable. Rather, these findings add to a growing literature suggesting influential roles for sexual motivation in domains of human behavior that are apparently unrelated to sex (for a review, see Arfer, 2014b). Changing society and ourselves to better match our ideals is likely to require understanding, and struggling with, human sexual motivation.
Limitations
There are two chief ways in which our method limits the conclusions that can be drawn from our study. First, the sample was a convenience sample of Americans. Representative samples would be necessary to more accurately measure base rates of item endorsement (particularly by people who do not tend to participate in sex surveys), and other cultures must be examined to check whether any of the observed effects are peculiar to the United States or the industrialized West. This said, our Internet sample was large, it was more diverse than the typical undergraduate sample used in psychological research, and it provided more anonymity than in-person assessments, potentially facilitating honesty and reducing the effects of social desirability.
Second, we used exclusively self-report items. Subjects may not be completely truthful about their sexual interests, and they may not have perfect insight into their own preferences (Ariely, Loewenstein, & Prelec, 2006). On the other hand, self-report has the virtue of being the most specific way to measure subjectively perceived interest, and our validity items helped to identify inconsistent, random, hurried, or dishonest responding.
References
Arfer, K. B. (2014a). The adaptive value of sexuality is an open question. In Empirical sexual attitudes. Retrieved from http://arfer.net/w/esa/evolution
Arfer, K. B. (2014b). Sexuality has many odd and possibly pernicious effects on cognition and decision-making. In Empirical sexual attitudes. Retrieved from http://arfer.net/w/esa/cognition
Ariely, D., Loewenstein, G., & Prelec, D. (2006). Tom Sawyer and the construction of value. Journal of Economic Behavior and Organization, 60, 1–10. doi:10.1016/j.jebo.2004.10.003
Baumeister, R. F., Catanese, K. R., & Vohs, K. D. (2001). Is there a gender difference in strength of sex drive? Theoretical views, conceptual distinctions, and a review of relevant evidence. Personality and Social Psychology Review, 5(3), 242–273. doi:10.1207/S15327957PSPR0503_5
Briere, J., Smiljanich, K., & Henschel, D. (1994). Sexual fantasies, gender, and molestation history. Child Abuse and Neglect, 18(2), 131–137. doi:10.1016/0145-2134(94)90115-5
Brousseau, M. M., Bergeron, S., Hébert, M., & McDuff, P. (2011). Sexual coercion victimization and perpetration in heterosexual couples: A dyadic investigation. Archives of Sexual Behavior, 40(2), 363–372. doi:10.1007/s10508-010-9617-0
Brownmiller, S. (1975). Against our will: Men, women, and rape. New York, NY: Simon and Schuster. ISBN 0-671-22062-4.
Brownmiller, S. (1984). Femininity. New York, NY: Linden Press. ISBN 0-671-24692-5.
Buhrmester, M., Kwang, T., & Gosling, S. D. (2011). Amazon's Mechanical Turk: A new source of inexpensive, yet high-quality, data? Perspectives on Psychological Science, 6, 3–5. doi:10.1177/1745691610393980
Buss, D. M. (1988). From vigilance to violence: Tactics of mate retention in American undergraduates. Ethology and Sociobiology, 9(5), 291–317. doi:10.1016/0162-3095(88)90010-6
Buss, D. M. (1989). Sex differences in human mate preferences: Evolutionary hypotheses tested in 37 cultures. Behavioral and Brain Sciences, 12(1), 1–14. doi:10.1017/S0140525X00023992. Retrieved from http://www.psy.cmu.edu/~rakison/BussSexDifferencesinHuman.pdf
Buss, D. M., & Schmitt, D. P. (1993). Sexual Strategies Theory: An evolutionary perspective on human mating. Psychological Review, 100(2), 204–232. doi:10.1037/0033-295X.100.2.204
Caro, T. M., & Sellen, D. W. (1990). The reproductive advantages of fat in women. Ethology and Sociobiology, 11(1), 51–66. doi:10.1016/0162-3095(90)90005-Q
Chasin, C. D. (2011). Theoretical issues in the study of asexuality. Archives of Sexual Behavior, 40(4), 713–723. doi:10.1007/s10508-011-9757-x
Chivers, M. L., Seto, M. C., Lalumière, M. L., Laan, E., & Grimbos, T. (2010). Agreement of self-reported and genital measures of sexual arousal in men and women: A meta-analysis. Archives of Sexual Behavior, 39(1), 5–56. doi:10.1007/s10508-009-9556-9
Clark, R. D., & Hatfield, E. (1989). Gender differences in receptivity to sexual offers. Journal of Psychology and Human Sexuality, 2(1), 39–55. doi:10.1300/J056v02n01_04
Conley, T. D., Ziegler, A., & Moors, A. C. (2013). Backlash from the bedroom: Stigma mediates gender differences in acceptance of casual sex offers. Psychology of Women Quarterly, 37(3), 392–407. doi:10.1177/0361684312467169
Critelli, J. W., & Bivona, J. M. (2008). Women's erotic rape fantasies: An evaluation of theory and research. Journal of Sex Research, 45(1), 57–70. doi:10.1080/00224490701808191
Dickemann, M. (1979). The ecology of mating systems in hypergynous dowry societies. Social Science Information, 18(2), 163–195. doi:10.1177/053901847901800201
Ellis, B. J., & Symons, D. (1990). Sex differences in sexual fantasy: An evolutionary psychological approach. Journal of Sex Research, 27(4), 527–555. doi:10.1080/00224499009551579
Eriksson, K., & Simpson, B. (2010). Emotional reactions to losing explain gender differences in entering a risky lottery. Judgment and Decision Making, 5, 159–163.
Fessler, D. M., Nettle, D., Afshar, Y., Pinheiro Ide, A., Bolyanatz, A., Mulder, M. B., … Zbarauskaite, A. (2005). A cross-cultural investigation of the role of foot size in physical attractiveness. Archives of Sexual Behavior, 34(3), 267–276. doi:10.1007/s10508-005-3115-9
Gelman, A., Carlin, J. B., Stern, H. S., & Rubin, D. B. (2004). Bayesian data analysis (2nd ed.). Boca Raton, FL: Chapman & Hall/CRC. ISBN 1-58488-388-X.
Gelman, A., & Stern, H. (2006). The difference between "significant" and "not significant" is not itself statistically significant. The American Statistician, 60(4), 328–331. doi:10.1198/000313006X152649
Hayes, A. F. (1995). Age preferences for same- and opposite-sex partners. Journal of Social Psychology, 135(2), 125–133. doi:10.1080/00224545.1995.9711415
Ipeirotis, P. G. (2010). Demographics of Mechanical Turk (Technical Report No. CeDER-10-01). New York, NY: New York University. Retrieved from http://hdl.handle.net/2451/29585
Joyal, C. C., Cossette, A., & Lapierre, V. (2014). What exactly is an unusual sexual fantasy? Journal of Sexual Medicine. Advance online publication. doi:10.1111/jsm.12734
Kruschke, J. K. (2010). What to believe: Bayesian methods for data analysis. Trends in Cognitive Sciences, 14(7), 293–300. doi:10.1016/j.tics.2010.05.001
Kruschke, J. K. (2011). Doing Bayesian data analysis: A tutorial with R and BUGS. San Diego, CA: Elsevier Academic Press. ISBN 978-0-12-381485-2.
Laumann, E. O., Gagnon, J., Michael, R. T., & Michaels, S. (1994). The social organization of sexuality: Sexual practices in the United States. Chicago, IL: University of Chicago Press.
Leitenberg, H., & Henning, K. (1995). Sexual fantasy. Psychological Bulletin, 117(3), 469–496. doi:10.1037/0033-2909.117.3.469
Lewis, D. M. G., Conroy-Beam, D., Al-Shawaf, L., Raja, A., DeKay, T., & Buss, D. M. (2011). Friends with benefits: The evolved psychology of same- and opposite-sex friendship. Evolutionary Psychology, 9(4), 543–563. Retrieved from http://www.epjournal.net/articles/friends-with-benefits-the-evolved-psychology-of-same-and-opposite-sex-friendship
Marlowe, F. (2000). Paternal investment and the human mating system. Behavioural Processes, 51(1, 3), 45–61. doi:10.1016/S0376-6357(00)00118-2
Meana, M. (2010). Elucidating women's (hetero)sexual desire: Definitional challenges and content expansion. Journal of Sex Research, 47(2, 3), 104–122. doi:10.1080/00224490903402546
Ogas, O., & Gaddam, S. (2011). A billion wicked thoughts: What the world's largest experiment reveals about human desire. New York, NY: Dutton. ISBN 978-0-525-95209-1.
Petersen, J. L., & Hyde, J. S. (2010). A meta-analytic review of research on gender differences in sexuality, 1993–2007. Psychological Bulletin, 136(1), 21–38. doi:10.1037/a0017504
Scorolli, C., Ghirlanda, S., Enquist, M., Zattoni, S., & Jannini, E. A. (2007). Relative prevalence of different fetishes. International Journal of Impotence Research, 19(4), 432–437. doi:10.1038/sj.ijir.3901547
Smuts, B. (1995). The evolutionary origins of patriarchy. Human Nature, 6(1), 1–32. doi:10.1007/BF02734133
Suri, S., & Watts, D. J. (2011). Cooperation and contagion in web-based, networked public goods experiments. PLOS ONE. doi:10.1371/journal.pone.0016836
Vandermassen, G. (2011). Evolution and rape: A feminist Darwinian perspective. Sex Roles, 64(9, 10), 732–747. doi:10.1007/s11199-010-9895-y
Weatherhead, P. J., & Robertson, R. J. (1979). Offspring quality and the polygyny threshold: "The sexy son hypothesis". American Naturalist, 113(2), 201–208. Retrieved from http://www.jstor.org/stable/2460199
Notes
Actually, rather than using the raw contingency table for Table 1, Figure 1, and the proportions quoted in this section, we added 1 to each cell first. This is equivalent to modeling each item and gender in a Bayesian fashion as an independent categorical distribution (i.e., a 1-trial multinomial distribution) with a flat prior (i.e., the Dirichlet distribution with all parameters set to 1). The cell proportions are then posterior probabilities of a future subject rating the given item at the given level, so we can predict population behavior straightforwardly. Additionally, this analysis renders the χ2 tests superfluous, which is why in our interpretation of the results we do not distinguish items with significant χ2 tests from those without (moreover, see Gelman & Stern, 2006). For references on Bayesian data analysis, see Kruschke (2010), Kruschke (2011), and Gelman, Carlin, Stern, and Rubin (2004).