One of the most cited series of experiments in social and evolutionary psychology are those conducted by Clark and Hatfield (1989, 2003) and Clark (1990). In these experiments, conducted to test the classic evolutionary theory notion that men are generally more receptive to sexual offers than are women (Symons, 1979; Trivers, 1972), college students approached fellow students and asked one of three questions: (a) “Would you go out with me tonight?” (b) “Would you come over to my apartment tonight?” or (c) “Would you go to bed with me tonight?” Men and women differed markedly in their receptivity to casual sexual offers (see Table 1).
Table 1
Type of Request
|
|||
---|---|---|---|
Gender of Respondent | Date | Apartment | Sex |
Men | 50.0% | 69.0% | 75.0% |
Women | 56.0% | 6.0% | 0.0% |
Note. The higher the number, the greater the percentage of people agreeing to the request.
Since its publication, this study has attracted a great deal of attention. Some examples: In 2003, Psychological Inquiry designated the study “a classic, an experiment that scholars will still be talking about 100 years from now.” For evolutionary psychologists, it has become part of the “canon” (see, for example, Buss, 2003), documenting the importance of gender in love and casual sexual relationships. Several TV science programs in America, Austria, Canada, and Germany, among others, have discussed the study and have attempted to conduct partial or full replications/variations of the original experiment (Hatfield, 2006; Molzer, 2003; Voracek, Hofhansl, & Fisher, 2005). In recent years, this study has sparked heated debates in Science, News in Brain and Behavioural Sciences, and Sexnet, the last two being scholarly Web-based discussion groups. In these postings, readers were both laudatory and annoyed by the study; some thought the results were obvious, others that they couldn’t possibly be valid; that the study was a landmark experiment, or that it was a trivial and frivolous experiment (see Clark & Hatfield, 2003 for a compendium of these comments).
In spite of the popularity (or notoriety) of this classic study, Hatfield (2006) acknowledged that, in and of itself, it had some severe limitations. Theorists have spent a great deal of time and effort speculating about the factors that shape men and women's willingness to engage in casual sex. This means that a simple test of "Who is most receptive to casual sex—men or women?" is a bit simplistic. Since Charles Darwin’s (1859) original conception of the theory of natural selection, evolutionary theory has become increasingly sophisticated. Many other mechanisms of selection have been identified or elaborated. Among these Neo-Darwinian understandings are: intrasexual and intersexual selection, good genes sexual selection (Møller & Alatalo, 1999), Fisher’s (1915) conceptualism of runaway sexual selection (see Pomiankowski, Iwasa, & Nee, 1991), Zahavi and Zahavi’s (1997) handicap principle, the immunocompetence handicap principle, sexual antagonistic co-evolution, and the life history theory (Kokko, Brooks, McNamara, & Houston, 2002), among others.
So too is the case with theories of mate selection. When discussing casual sex, social psychologists almost always cite Buss and Schmitt’s (1993) Sexual Strategy Theory (SST). In fact, however, the SST is only one of the many evolutionary theories of mate selection. Other prominent theories include Gangestad, Haselton, and Buss’ (2006) Theory of Evoked Culture, Gangestad and Simpson’s (2000) Strategic Pluralism Theory (see also Li & Kenrick's, 2006, discussion of "good genes" theories), Puts’ (2010) Intrasexual Selection Theory, and Apostolou’s (2012) Parental Selection Theory. Most agree that while people possess the same inherited and universal information processing mechanisms, preferences vary in predictable and adaptive ways across culture. For example, cultures differ in how critical attractiveness is in mate selection. The reason for this difference is that cultural preferences are mediated by parasite prevalence (how deadly various parasites are in a culture): People from cultures with high parasite prevalence are more likely to value attractiveness in a mate than are people from cultures with low parasite prevalence (Gangestad & Simpson, 2000). Men and women are also known to utilize different mate selection strategies in short term versus long-term encounters (Buss, 2003; Schmitt et al., 2012). Gangestad and Simpson (2000), for example point out that when looking for short-term mates, women generally look for characteristics such as masculine features and physical attractiveness; when searching for long-term mates, on the other hand, they search for men with parenting qualities (i.e., feminine features). Similar results have been found by a variety of other authors (Fisher & Cox, 2009). Given the importance of such social and ecological factors, Hatfield (2006) suggested that evolutionary scholars begin to ask the “Who, where, what, when, and why?” of this phenomenon—for example, to begin to investigate:
1. Who is doing the asking? Social psychologists might, for example, ask how old are the confederates? How good looking are they? (Fisher & Cox, 2009). What are their social class, religion, and race? Are they gay or straight? Are they drunk or sober? Could they have STIs or AIDS? How many people have they approached before the participant? Do they appear menacing?
2. Who is being asked? Scholars might consider the personality traits and social situation of the respondents. For example: Do they seem happy or depressed? Are they popular or social outcasts? Are they in a loving relationship or on the rebound? Do they have many/few opportunities for dates or sex?
3. Where does the “hit” take place? One might, for example, expect to secure quite different results if the invitation took place in a church or at a revival meeting, on a college campus, in an office, in a bar, or in a Greyhound station.
4. What is being asked? Would men and women be more receptive if the offer were preceded by flirtation and sexy conversation, rather than coming out of the blue?
5. Why do men and women accept/reject casual sexual offers? Which of the many reasons for engaging in sex (or refusing a sexual offer) motivate men’s and women’s acceptance or rejection of such offers? How do people expect to be treated if word gets out of their acceptance/rejection? Are they worried that the confederate may be peculiar, dangerous, or making fun of them?
6. Do cultural and historical factors shape men and willingness to engage in casual sex? Why haven’t social and evolutionary psychologists conducted more research of this kind? There are several reasons for the relative neglect of these obvious questions: First, this type of field research is very difficult and time consuming. Second, in this day and age, American social scientists are more sensitive to the possible pitfalls of such naturalistic experiments. They worry about public relations and the welfare of confederates and participants. Finally, even if scholars wish to stage replications, university IRBs generally refuse to approve such controversial field studies.
What to do? We decided that what is needed is a pencil and paper prototype that would prove a useful substitute for Clark and Hatfield’s (1989) procedure, without its drawbacks. We were hoping to replicate/extend using paper and pencil methods the main results of the classic study, thus demonstrating that we had crafted a prototype. Once we had developed such a prototype, we hoped that we (and other scholars) would be able to begin conducting the research that would help us to better understand the nature of men and women’s reactions to offers of casual sex. Several other social and evolutionary psychologists have attempted to develop such prototypes, but their attempts are not appropriate for use in a diversity of cultures and ethnic groups; did not ask participants about their willingness to date, go to apartments, or have sex (usually they only asked if people were interested in a sexual encounter); or failed in other ways to provide a close equivalent to the classic study. (See Discussion: Studies 1-3 for a list of these [generally] partial (although excellent) attempts at replication/development of prototypes.)
Studies 1-3 were designed to test the following hypotheses—in the hope of allowing us to see to what extent our paper and pencil measure would replicate the findings of the original Clark and Hatfield studies:
Hypothesis 1: Men will be more receptive to sexual offers than will women. Specifically, we expect to secure a main effect for participants’ Gender. This was, of course, the main (and most important) finding in the classic study.
Hypothesis 2: Participants will be differentially responsive to different Types of Requests. Specifically, we expect to secure a main effect for Type of Request, such that in general people (especially women) will be more likely to accept a date than an invitation to go to someone’s apartment or to engage in sex.
Hypothesis 3: Both Gender and Type of Request will interact in shaping men’s and women’s responsiveness to sexual offers. Specifically, we expect that men may be increasingly positive about sexual offers as they become more sexually explicit, while women will be increasingly negative as explicitness increases. This was the pattern of results secured in the original Clark and Hatfield (1989, 2003) studies.
Studies 1-3 must necessarily differ in several ways from the classic paradigm. First, in the 20 to 30 years since the original experiments were conducted, America has become a far more multi-ethnic society. We would be attempting to functionally replicate the classic study—not with the predominantly European-American population of the 1980s Tallahassee, Florida studies—but with the multicultural and predominantly Asian population of Honolulu, Hawai‘i. It is possible that the Asian influence may make Honolulu a more traditionally conservative place than Tallahassee when it comes to sexual matters (Forbes, 2012). Of course, Tallahassee is more conservative politically than is Honolulu. Second, in the past decades, casual sex has become far more common among college students than it once was. Some sex researchers claim that today the majority of college men and women engage in “one-night stands,” “friends-with benefits” relationships, “hook-ups”, and the like (see Hatfield, Hutchison, Bensman, Young, & Rapson, 2012, for a review of this research.) Third, in the classic studies, respondents could only accept or reject a proffer. In the following experiments, students were invited to give a one word answer (Yes or No) and to indicate how likely they would be to accept such an offer on a continuous Likert scale. Finally, they were asked to tell the researchers why they decided to accept or reject the sexual invitation.
Study 1
Participants
Participants were 90 women and 27 men (117 in all) from the University of Hawai‘i’s Mānoa campus. Respondents’ mean age was 21.7; they ranged from 18 to 49 in age. As is typical of Hawai‘i’s multicultural population, participants were ethnically varied: 35% were of mixed ethnicity, 35% were Caucasian, 7.7% were Japanese, 6% were Chinese, 5.1% were Hispanic, 3.4% were Korean, 2.6% were Filipino, and the remainder (5.2%) were African-American, American Indian, Hawaiian, Indian/Pakistani, Middle Eastern, Other Asian, Pacific Islander, or Portuguese.
Respondents also possessed an array of religious affiliations: Protestant (29.9%), Catholic (21.4%), Buddhist (6%), Jewish (2.6%), Atheist (0.9%) and none (39.2%). On average, they rated themselves as 3.9 in religiosity (on a scale ranging from 0 to 10). Students’ sexual orientations also varied, with 89.7% identifying themselves as heterosexual, 5.1% as bisexual, 3.4% as homosexual, and 1.7% being unsure of their sexuality. Fifty-six percent were in a committed relationship (54% of the female participants and 59% of the male participants). In this initial study, homosexual participants were dropped from the sample.
Procedure
When students reported for the session, they were given an IRB consent form and a two-page packet. (All three studies were, of course, approved by the UH IRB.) At the top of the first page of the packet appeared a picture of a fairly attractive college man or woman (see Figure 1).
Figure 1
The faces, which were designed to serve as a prototype of faces-in-general for college students of no particular ethnic group, were constructed using software originally developed by Bernard Tiddeman and David Perrett, in a program which is now available for scholars from Faceresearch.org (http:www.faceresearch.org). A sampling of 40 male and 40 female faces, taken from a sampling of the racial and ethnic groups typical of the UH campus, were selected. Using computer imaging, the faces were superimposed one upon one another. The result was two composite faces of equal attractiveness, as judged by a team of independent raters. Composite faces tend to be more attractive than are the individual faces comprising them (Langlois & Roggman, 1990). As a consequence, we might expect participants to be somewhat more receptive to the stimuli in Experiments 1-3 than they would have been to the students in the original study. Since in this study we are not exploring the impact of relative attractiveness on compliance with requests, the attractiveness of stimuli should not affect our hypotheses. All that is important is that the composites be equally attractive and representative of the population—which they were.
The images also controlled for factors such as differences in clothing worn, hairstyles, body type, and facial expression.
After glancing at this target face, participants were asked three questions:
Imagine this person comes up to you while you are walking to class and says: “I have noticed you around campus. I find you very attractive.” How would you respond to the following questions?
Will you go on a date with me?
Will you come back to my apartment?
Will you go to bed with me tonight?
Respondents indicated their answers on the following scale:
Respondents were asked to indicate the reasons for their responses to each of the three questions. Participants were then asked a series of demographic questions. Finally, they were debriefed and allowed to ask any questions they might have and make any comments they wished.
Results
Since participants were asked how receptive they would be to all three types of requests, in order to test our hypotheses, we conducted an ANOVA using gender and type of request as factors predicting receptivity to each offer. In Hypothesis 1, we predicted that men would be more receptive to offers of casual sex than would women. This hypothesis was confirmed: date, F1,115 = 9.70, p < .01; apartment, F1, 115 = 80.34, p < .01; sex, F1, 115 = 74.75, p < .01 (see Table 2 and Figure 2). We see, then, that our prototype study does indeed replicate the critically important findings of the classic studies.
Table 2
Type of Request
|
|||
---|---|---|---|
Gender of Respondent | Date | Apartment | Sex |
Men | 5.15 | 4.30 | 3.52 |
Women | 3.33 | 0.71 | 0.23 |
Note. The higher the number, the more certain people are that they would agree to the request.
Figure 2
Hypothesis 2 predicted that there would be a main effect for Type of Request. This hypothesis was also confirmed—although the means were not in the predicted pattern (i.e., the pattern secured in the original study). When examining within-subject contrasts, using a general linear model with repeated measures, we see that all comparisons are significant: Date versus Apartment: F1, 115 = 47.53, p < .001; Date + Apartment versus Offer of Sex: F1, 115 = 61.10, p < .001.
Hypothesis 3 predicted that we would secure a significant interaction between Gender x Type of Request. Here the results were ambiguous. As predicted, when we consider the Gender x Dating versus Apartment interaction, we find that, as predicted, it was significant: F1,115 = 12.34, p < .001. The Gender x Date + Apartment interaction did not reach significance, however: F1, 115 = 2.34, p = .13. More importantly, the pattern of results was entirely different than that secured in the classic study. In this simulation, both men and women declined in receptivity as the offer became more sexually explicit. Men’s interest declined less precipitously than did women’s (see Figure 2), but the pattern in this study was very different from that secured in the classic study.
Finally, participants who accepted or declined the offer of casual sex were asked why they had done so. Men who answered this question claimed to be most turned off by the fact that the women did not meet their standards of beauty, were too forward (and therefore too “easy”), and for various reasons were not their “type”. They rarely mentioned danger as a concern. Here is a sampling of the most common reasons men gave for refusing sexual offers:
“In a relationship”
“She’s not that attractive”
“Don’t know her”
“Not my type”
“Awkward”
Women were most put off by the fact that the man, who was a stranger, was too forward. They also worried that, given the oddness of his request, they might be in danger. The most common reasons women gave for declining sexual offers were:
“Have a boyfriend”
“Don’t know him”
“He’s creepy”
“Not sleeping with a stranger”
“Too forward”
“Stalkerish/dangerous”
Discussion: Study 1
The Study 1 paradigm possessed several advantages over the classic study: (1) In the classic study, the confederates naturally varied in attractiveness, general appearance, and presentation style. In this study, college students’ pictures and scripted offers were standardized. In that way, we controlled for attractiveness, general appearance, acting ability, and body language. (2) In the classic study, participants could only accept or reject offers. In this study, students were allowed to indicate whether or not they would accept or reject the various offers on an 11-point Likert scale—allowing us to assess small gradations in enthusiasm. (3) In the classic study, we had no idea why respondents reacted as they did. In this prototype, respondents were asked why they accepted or rejected the offer. This gave us some hints as to what to look for in subsequent research.
Yet, Study 1 was not a perfect functional replication of the classic study; also, we did not precisely replicate the findings of the original study. What could account for the differences? In retrospect, we realized that Study 1 differed from the classic study in one crucial respect: Originally, respondents were issued only one invitation and asked only one question—did they want to go on a date OR to an apartment OR to bed? In this study, respondents were asked all three questions. Thus, in the hope of developing a prototype as similar to classic study as possible, we decided to conduct a second study. In this study, participants would be asked only one question: did they or did they not wish to date, go to the apartment, or have sex with the target?
Study 2
Participants
Participants were 151 men and 176 women (327 in all) from the University of Hawai‘i’s Mānoa campus. Respondents’ mean age was 22.69; they ranged in age from 18-68. As before, respondents came from a variety of ethnic backgrounds: 29.8% were of mixed ethnicity, 23.9% were Caucasian, 16.9% were Japanese, 8.9% were Filipino, 7.7% were Chinese, 3.1% were Korean, 2.1% were Hawaiian or part-Hawaiian, and 7.6% came from a variety of other ethnic groups (African-American, American Indian, Hispanic, Middle Eastern, Other Asian, Indian/Pakistani, Pacific Islander, and Portuguese).
Participants also came from an array of religious groups. These were Protestant (35.9%), Catholic (20.1%), Buddhist (5.3%), agnostic (2.6%), Jewish (1%), Muslim (0.3%), other (3.6%), and none (31.3%). On the average, students rated themselves as 4.14 in religiosity (on a scale ranging from 0 to 10). Students were also asked about their sexual orientation: 91.0% were heterosexual, 4.3% were bisexual, 3.4% were homosexual, and 1.2% were unsure of their sexual orientation. (In Study # 2, participants who reported a homosexual orientation were shown a same sex-photograph, and rated their interest in that person.) Fifty-three percent of the respondents were in a committed relationship (61% of female participants and 45% of male participants).
Procedure
The procedure in this study was identical to that reported in Study #1, with one exception. When students reported for the experimental session, they were given a consent form and a two-page packet. After the usual procedure, respondents were asked:
Imagine this person comes up to you while you are walking to class and says “I have noticed you around campus. I find you very attractive.” How would you respond to the following question?
Then they were randomly assigned to answer just one of three questions: “Will you go on a date with me?”, “Will you come back to my apartment?”, or “Will you go to bed with me tonight?”
As before, respondents were asked to indicate their answers on the same 11-point scale used previously and to indicate their reason for their acceptance or rejection of the sexual invitation.
Results
We conducted a 2 x 3 ANOVA, predicting that we would secure significant main effects for Gender and Type of Request, and a significant interaction for Gender x Type of Request.
Once again, Hypothesis 1, which proposed that men would be more receptive to sexual offers than were women, received strong support (F1, 326 = 63.53, p < .01) (see Table 3).
Table 3
Type of Request
|
|||
---|---|---|---|
Gender of Respondent | Date | Apartment | Sex |
Men | 5.25 | 2.70 | 2.76 |
Women | 3.41 | 0.46 | 0.59 |
Note. The higher the number, the more certain people are that they would agree with the request.
Hypothesis 2, which predicted that there would be a main effect for Type of Request—specifically, that people would be more willing to accept the offer of a date than an invitation to an apartment or to have sex—was also supported (F2, 326 = 46.82, p < .01). However, again, contrary to the findings of the classic study, we did not secure a significant interaction between Gender and Type of Request (F2, 326 = .22, p = .81). An examination of the trends makes it clear that, once again, the pattern of acceptances is different than the one secured in the classic study. Both men and women were more likely to accept a date than a more sexually explicit offer (see Figure 3).
Figure 3
Once again, we asked men and women why they had accepted or rejected the various sexual offers. Responses were much the same as in Study 1. (We will discuss why people accept or reject sexual offers in more detail in our General Discussion).
Discussion: Study 2
With Study 2, it became clear that even with our improved procedure we did not secure the same significant 2-way interaction (or an identical pattern) that had appeared in the classic study. This could be the result of a number of different possibilities:
1. Perhaps Hawai‘i, which is a predominately an Asian-American, multicultural society, is simply more conservative than were the Florida students who participated in the classic study so many years ago (Forbes, 2012; Woo, Brotto, & Gorzalka, 2011). In the 1970s and 1980s, the vast majority of FSU students were Caucasian.
In order to test this hypothesis, we conducted a series of sub-analyses (comparing Asian and Caucasian students), but secured no significant main effects or interactions for any of the comparisons. In truth, we did not expect to secure any. We were really proposing that Honolulu (in general) might be a more conservative community than was Tallahassee (in general). In previous studies, researchers have never secured significant differences between the various Honolulu ethnic groups in the realm of love or sex (see Tang, Bensman, & Hatfield, 2012, for a further discussion of this point). Furthermore, Hawai‘i leads the nation in mixed marriages (“Hawaii still leads U. S. with highest rate of mixed marriages,” 2010). More than 50% of marriages are interracial (the figure rises to 70% if one includes Native Hawaiians, who are virtually all of mixed ancestry, in the statistics: Rapson, 1980), and thus many young people identify with multiple ethnic groups. In addition, once one begins to conduct sub-analyses of samples, the populations of Asian, Japanese, and Caucasian participants become too small to attain significant effects.
There were, of course, other possibilities:
2. In post-experimental interviews, some students mentioned that while most UH students date men/women of various ethnicities, they did not. It became clear that we should have provided stimulus photos appropriate to our diverse population.
3. In the classic study, the confederates presented more than just a pretty (or average) face. People form impressions of others very quickly—think of the speed dating research which suggests people know all they need to know after just a few seconds of interaction (Finkel, Eastwick, & Matthews, 2007). Perhaps in the classic study, participants could sense that the confederate was a “regular Joe or Jane”—not crazy, not dangerous, not a “loser” or a “creep”—during the brief period before he/she delivered his/her lines. Thus, perhaps we should have provided participants with a paragraph or two designed to provide a bit more information as to the confederates’ nature.
In Study 3 we set out to remedy these flaws.
Study 3
Participants
Participants were 178 men and 178 women (356 total) from the University of Hawai‘i’s Mānoa campus. Respondents’ mean age was 21.98; they ranged in age from 18 to 54. As in the first two studies, respondents came from a variety of ethnic backgrounds: 31.1% were Caucasian, 18.1% were Japanese, 15% were Filipino, 9% were Chinese, 5.9% were Hawaiian or part-Hawaiian, 4.8% were Hispanic, 4.2% were Pacific Islander, 3.7% were African American, 3.7% were Korean, and 4.5% came from a variety of other ethnic groups (American Indian, Middle Eastern, Other Asian, Indian/Pakistani, and Portuguese). In this study, unlike Studies 1 and 2, we asked participants to indicate the ethnic group with which they most identified.
The majority of participants, 57.1%, were raised in Hawai‘i; among those not raised on the islands, 34.7% were raised in the continental United States, 1.7% in Pacific Island nations, 1.4% in the Philippines, and 1.1% in Japan, and 4.1% elsewhere (in Africa, Central and South America, Korea, and the Middle East).
Participants also came from an array of religious groups: Protestant (21.7%), Catholic (14.1%), Buddhist (2.5%), Islam (0.8%), Jewish (0.8%), Atheist (1.1%), Agnostic (0.6%), and none (58.4%).
Students were also asked about their sexual orientation: 92% were heterosexual, 4.3% were bisexual, 2.8% were homosexual, and 0.9% were unsure of their sexuality. Once again, homosexual students were shown pictures of same sex targets. Forty-five percent of the respondents were in a committed relationship (53% of female participants and 36% of male participants).
Procedure
Once again, with few exceptions, the procedure was identical to that described in Studies 1 and 2. The surveys were distributed in lecture classes in a number of UH departments. Respondents were told that we were conducting a survey concerning sexual attitudes and behavior. Research assistants handed out the consent forms and instructed participants to read them before deciding on whether or not they wanted to participate. Respondents were instructed to select an appropriate questionnaire depicting the types of faces to which they were generally most attracted. Surveys were distributed in a randomized order, some asking about dates, others about apartment visits or sex.
The questionnaire itself was three pages long. The first page presented three faces, depicting men and women from different ethnic groups—Asian-American, Caucasian, and African-Polynesian. The faces were designed to mirror Hawai‘i’s ethnic composition. Once again, Faceresearch.org (http:www.faceresearch.org) was utilized to create appropriate faces (see Figure 4).
Upon receiving their questionnaire, participants were asked to select the person (from the three faces presented) that they found most attractive. Then they were told:
Imagine you are not currently in a relationship. This person approaches you while you are walking to meet some of your friends for lunch and says, “Hey, you were in my class last semester and I think you are really cute.” (You recognize him or her although you never caught his or her name. You have a short conversation about the class you took together.) He or she proceeds with, “You seem like you’re a pretty nice person. I wish I’d had a chance to talk with you before. I don’t mean to be forward and this is unusual for me, but I was wondering if….
„you want to go out sometime?” or
„you want to come over to my place tonight?” or
„you would want to have sex with me tonight?”
As before, which question the participant received was randomly determined.
Respondents were asked to indicate whether they would say “Yes” or “No” to such an offer and then to indicate how certain they were that they would do so using the following scale:
Respondents were also asked to indicate the reason for their acceptance or rejection of the sexual offer.
The questionnaires took roughly five minutes to complete. Finally, students were debriefed and allowed to ask any questions or make any comments they wished.
Figure 4
Results
In this study, we assessed people’s willingness to accept an offer in two ways: (1) as before, we asked participants to indicate their receptivity on Likert scale ranging from 0-10 and (2) we asked them to give a simple “Yes” or “No” answer to the questions.
When we consider our Likert scale data (see Table 4 and Figure 5), we find that once again, we secured a significant main effect for Gender (F1, 325 = 73.17, p < .01).
Table 4
Type of Request
|
|||
---|---|---|---|
Gender of Respondent | Date | Apartment | Sex |
Men | 6.68 | 5.74 | 3.96 |
Women | 5.71 | 2.90 | 1.29 |
Note: The higher the number, the more certain people are that they would agree to the request.
Figure 5
We secured much the same results when we consider participants’ Yes/No answers χ2 = 29.65, 1 and 352 d.f., p < .01 (see Table 5).
Table 5
Type of Request
|
|||
---|---|---|---|
Gender of Respondent | Date | Apartment | Sex |
Men | 80.0% | 61.0% | 25.0% |
Women | 63.0% | 13.0% | 5.0% |
Note. The higher the number, the greater the percentage of people agreeing to the request.
Hypothesis 2, which proposed that there would be a main effect for Type of Request, was also supported. People were more willing to go on a date than to accept invitations to go back to an apartment or to have sex (F2, 325 = 65.76, p <.01.) when we examine the Yes/No data we also see a significant main effect (χ2 = 64.15, 2 and 352 d.f., p < .001.)
Finally, in Hypothesis 3, we proposed that Gender and Type of Request would interact in determining acceptance/rejection. With the Likert data, this hypothesis was supported (F2, 325 = 5.70, p = .004)—albeit the form of the interaction differed from that in the classic study. However, when we consider the Yes/No data, however, this interaction is not significant, χ2 = 5.84, 2 and 352 d.f., P > .054, and the hypothesis was not supported.
Participants were asked to imagine themselves not currently in a relationship; however, this might have been difficult for some participants to do, so we re-ran the analyses using only the data from participants not currently in a relationship. We secured the same pattern of results, with almost identical means and no changes in statistical significance.
In Study 3, we posed an additional question. We asked: “If this person were interested in being your significant other, how would you respond?” Buss (2003), like other evolutionary psychologists, has argued that women often accept offers of casual sex because they are secretly hoping for a long-term relationship. Men are not. Would there be gender differences, we wondered, in how interested men and women were in a significant relationship (in each of the conditions)? If anything, we expected that men (in general) would be less interested in becoming the stimulus woman’s significant other, than women would be. In Tables 6 and 7 we see that there were significant Gender differences in respondents’ interest in a serious relationship.
Table 6
Type of Request
|
|||
---|---|---|---|
Gender of Respondent | Date | Apartment | Sex |
Men | 5.17 | 4.67 | 4.35 |
Women | 4.37 | 3.47 | 3.26 |
Note. The higher the number, the more certain people are that they would agree to the request.
Table 7
Type of Request
|
|||
---|---|---|---|
Gender of Respondent | Date | Apartment | Sex |
Men | 47.5% | 34.1% | 46.2% |
Women | 31.0% | 9.3% | 17.5% |
Note. The percentage listed is the percentage of participants who agreed to the request.
In any case, we did secure a significant Gender effect for both the continuous data (F1, 340 = 14.32, p < .01) and for the Yes/No data. Again, using a Generalized Linear Model, we found a significant main effect for Gender (χ2 = 15.97, 1 and 247 d.f., p < .01).
We also found a significant effect for Type of Request for continuous data (F2, 340 = 4.49, p < .01), and for the Yes/No data (χ2 = 7.04, 2 and 247 d. f., p < .05). We did not secure an interaction between Gender and Type of Request for the continuous data (F2, 340 = .19, p = .82) or the Yes/No data (χ2 = 1.75, 2 and 247 d. f., p = .42).
These results were a bit of a surprise. For both men and women, when the target was interested in a serious relationship, both men and women were less interested in accepting the offer of a date (Males: 80% vs. 47.5% [if the woman wanted a serious relationship]; Females: 63% vs. 31% [if the man wanted a serious relationship]) or of an apartment visit (Males: 64% vs. 34.1% [if the woman wanted a serious relationship]; Females: 13% vs. 9.3% [if the man wanted a serious relationship]), but more interested in accepting the offer of sex (Males: 25% vs. 46.2% [if the woman wanted a serious relationship; Females: 5% vs. 17.5% [if the man wanted a serious relationship]). Additionally, we had expected that men would be interested in casual affairs and that women would be more interested in serious relationships. This was not the case. Men were more eager than women to participate in both casual and serious relationships. To figure out exactly what this means, more research will have to be conducted.
As in Study 2, in Study 3 we asked men and women why they had accepted/refused offers of sex. Men’s reasons for rejecting invitations to go to the targets’ apartment or have sex with her included:
“I don’t like it when girls ask guys out.”
“I got to see the rest of the package.”
“Too forward, kind of weird, give me the sense that they have a screw loose. How many other people has she tried this with?”
“No idea if it’s safe, plus sex should be about intimacy, not blind lust.”
“It takes more than one conversation to get in my pants.”
Typical of women’s reactions were:
“I don’t know him that well. I’d feel awkward / unsafe”
“Don’t know them well. Maybe if it was in a more open place … coffee shop, dinner, or if it was a party and I could go with a few friends.”
“The guy is technically a complete stranger, and sex is something special shared with someone extremely special to you.”
“I’m not a slut.”
“I don’t like to give it up to a stranger. They could be a stalker and he didn’t even tell anything about himself. He seems creepy.”
Discussion: Study 3
Study 3 was the best crafted and the most informative of our attempts to construct a prototype that mirrors that utilized in the classic study. It provided a collection of prototypic photographs appropriate to a variety of cultures, added a bit of conversation that normalized the encounter, provided continuous measures of acceptance and rejection, and gave us a methodology for assessing the reasons lying behind men’s and women’s responses. This is the protocol we would use in exploring the questions with which we began this paper.
There was one finding that surprised us: Naively, we had assumed that men (eager for casual sex) would be more likely to seek out “no strings attached” sexual offers and less likely to be willing to get into a more serious relationship. We found we were wrong. Men were more eager to get into any kind of relationship—serious or not. We can think of several possible reasons for this unexpected finding. (1) Perhaps, although reluctant to get into relationships in general, men might be more eager to risk a relationship with a sexually available woman. (This finding reminds us of an earlier discovery that men most preferred women who were easy for them to get, hard for anyone else to get: Hatfield [Walster] et al., 1973). (2) Perhaps we see the operation of market conditions and sex ratio imbalances on men’s choices. In cross-cultural research, for example, we had expected Chinese men to be more individualistic in their sexual motives than were women. We discovered just the opposite. Chinese men, where women are in the minority, reported more concern with their partners’ pleasure than did American men or Chinese women: (Tang et al., 2012). Only future research can provide definitive information as to what accounts for this surprising result.
General Discussion: Studies 1-3
In the following discussion, we will define casual sex as:
A person mindfully engaging in sexual activities (such as mutual stimulation, oral sex, or sexual intercourse) outside of a “formal” relationship (dating, marriage, etc.), without a “traditional” reason (such as love, procreation, or commitment) for doing so (Hatfield, Luckhurst, & Rapson, 2012, p. 1).
There is compelling evidence that in America casual sex is on the rise. More than three-quarters of American college students have experienced at least one hookup (and typically more) with partners they did not consider to be romantic (see Garcia, Reiber, Massey, & Merriwether, 2012, and Hatfield, Luckhurst, et al., 2012 for a summary of this research). This change in the American sexual landscape means that it is of increasing interest to discover why men and women accept such casual sex offers. This study sheds some light on this question.
On this note, in Studies 1-3 we achieved a number of goals:
First, we developed a prototype that will allow scholars to investigate the “who, what, where, when, and why” of men’s and women’s willingness to accept (or reject) offers of casual sex. Most studies of casual sex have relied on survey or observational data. The existence of this experimental prototype will allow scholars to investigate the questions we posed in our Introduction or any others they might choose to investigate.
Second, we found some suggestive evidence that—even though men appear to be more eager to participate in all sorts of sexual activity than are women—culture may influence people’s receptivity to various types of sexual offers (i.e., the pattern of the Gender x Type of Request results might differ). Earlier, we reported that the classic study has been replicated and/or extended many times in America, Austria, Canada, and Germany (Clark & Hatfield, 2003). Spurred on by the findings of our Studies 1-3, we examined these “replications” and discovered that none of them were full replications. In fact, many scholars only attempted to replicate the last portion of the classic experiment—in which men and women are invited to participate in a sexual tryst. Other researchers asked respondents to indicate what they thought other men and women might do (see the list of partial replications later in this paper).
In light of the discovery, we searched the literature yet again to find out if any international (and complete) replications or prototypic replications were available. We found two—and they confirmed our suspicions. Hald and Høgh-Olesen (2009, 2010) conducted the first one-to-one replication of the Clark and Hatfield (1989) study. In Denmark, a “moderately attractive” confederate approached men and women, and issued one of the three sexual invitations. The authors report:
“For women, the results showed an inverted relation between explicitness of the sexual invitation and consenting rates, i.e. the more explicit the sexual invitation the lower the consenting rate. For men, the results showed a parabola-like relationship. That is, men did not respond favorably to sexual invitations of either too implicit ('date' condition) or too explicit ('bed' condition) a nature, but favorably to a sexual invitation fairly explicit in nature, yet still not 'in the box' ('home' condition)” (p. 366).
The authors attributed these results to the marked cultural differences between Denmark and the United States. These results are, however, more in tune with the Hawai‘i study than with the Florida study.
Recently Guéguen (2011) conducted a partial replication in France. He found that while almost all men were willing to agree to go back to a woman’s an apartment for a drink, they were a bit more hesitant to commit themselves to sex. Some women accepted the offer for drinks but all declined the sexual invitation. Interestingly, in our Study 1, one man nicely expressed the reason why he would prefer to accept an offer for the apartment to one for sex. He observed that if you accept an offer to visit a woman’s apartment you can always initiate sex if you wish to do so. If you accept an offer for sex, and discover your partner is mad, sad, or bad, you are stuck.
Our three studies seem to add credence to the argument that culture may affect the pattern of results—even though in all cultures thus far studied men appear to be more eager for sex than are women. In all three studies we see that in Hawaii (a sexually conservative, Asian influenced community), the pattern of results differed from the predominantly European American community in which the original study was run. In one of the later sections we will see that scholars secured pattern differences in some of the Northern-European and European countries as well.
Third, in Studies 1-3, we asked participants why they had accepted or rejected the sexual offers. Using the techniques of qualitative analysis (Rosch, 1978) to the extent possible, four research assistants developed a coding scheme for categorizing men’s and women’s reasons for refusing an invitation to the target’s apartment or to have sex into meaningful categories. They were able to classify responses into 10 categories:
-
No answer (Females: 30, Males: 54)
-
Not attractive enough (Females: 4, Males: 7)
-
Not my type (Females: 7, Males: 2)
-
Creepy/Stalkerish/Dangerous (Participants worried that the target might be a “psycho” or a “creep,” the invitation might be a trap, and acceptance might prove dangerous) (Females: 17, Males: 5)
-
Don’t know person/Need to know better (Females: 40, Males: 15)
-
Too forward (Participants considered the target to be rude, awkward, and in violation of accepted social norms) (Females: 15, Males: 13)
-
STI/Disease (Participants worried that the target might be unclean, might have “gotten around,” and might be carrying some sort of STI) (Females: 2, Males: 2)
-
Immoral (Participants who declined because such activities went against their religion’s prescriptions or own their personal moral code) (Females: 18, Males: 5)
-
In a committed relationship (Two participants failed to realize that they were supposed to “imagine themselves not in a committed relationship”) (Females: 1, Males: 1)
-
Other responses (Participants gave an answer that was unclassifiable or was so rare that it could not be included in any known category) (Females: 6, Males: 7)
We have indicated, next to a definition of each category, the number of men and women mentioning that as a reason for declining the sexual offer.
Fourth, we began this paper observing that once we had developed our newly-minted prototype, scholars might wish to investigate a few of questions we posed at the start of this paper. To assist in this enterprise, we will now list some of archival material that social psychologists might consider in formulating their hypotheses. (A note: These studies were inspired by a variety of theories and observations about life (ranging from “big picture” views, mini-theoretical observations, and methodological critiques). The authors interviewed men and women in a variety of countries, from a diversity of populations (casual acquaintances, people on the beach, and systematic surveys), asking totally different questions, and presenting their results in very different ways (ranging from casual journalistic observations to systematic analyses). Thus, while these studies may provide hints to the scholar seeking an understanding of the process underlying men’s and women’s receptivity to sexual offers, they can provide no definitive answers.
-
Who is most susceptible to offers of casual sex?
-
What sorts of targets are most likely to be accepted or rejected?
-
Greitemeyer (2005). [Germany]
-
Guéguen (2011). [France]
-
Hald and Høgh-Olesen (2010). [Denmark]
-
Schützwohl, Fuchs, McKibbin, and Shackelford (2009). [America, Germany, and Italy].
-
Why do men and women accept or reject offers of casual sex?
-
Hatfield [Walster], Walster, Piliavin, and Schmidt (1973). [General review].
-
Hatfield, Luckhurst, and Rapson (2010, 2011). [General review].
For a recent replication of the Clark and Hatfield study see:
(women asking men for sex) http://www.youtube.com/watch?v=5JJFBtHcBnM
(men asking women for sex) http://www.youtube.com/watch?v=gxyySRgrYsU
See also: http://www.huffingtonpost.com/2013/07/29/college-sex-social-experiment_n_3672320.html
If scholars wish to conduct research on the antecedents and consequences of agreeing to participate in casual sexual activities, the preceding paradigm and information ought to prove useful.
One additional point: Evolutionary psychologists have begun to compare people’s preferences in mates when measured by self-report, implicit measures of attitudes and preferences, and actual behavior (Hatfield, Forbes, & Rapson, 2012; Hatfield, Hutchison, et al., 2012; Thompson & O’Sullivan, 2012). This has proved profitable in increasing our understanding of choices in casual sexual encounters. An example: we have assumed that, in general, men care more about looks, women care more about power and status, in selecting mates in short term and long term relationships—although of course gender differences are less pronounced when considering long term relationships. In reality, those differences may be slightly exaggerated—especially when you are talking about casual encounters. When you ASK men and women how much they care about, say, looks, power, and status (when they sign up for, say, a matching service, or in a typical academic self-report study) they dutifully reflect the conventional wisdom. When you look at actual behavior, however, (at mixers, in speed dating, in bars, in hookups, etc.), however, you get a far different picture. When selecting real flesh-and-blood casual partners, men and women turn out to be surprisingly similar in their desire that he/she be good looking and “hot” (Hatfield [Walster], Aronson, Abrahams, & Rottmann, 1966). Good looks seem to be the sine qua non of a pick-up. This is true whether those “cruising the scene” are gay, straight, or lesbian; American or foreign born (see Finkel & Eastwick, 2009; Hatfield, Forbes, et al., 2012 and Hatfield, Hutchison, et al., 2012, for a summary of this research.)