Sexual double standards: a review and methodological critique of two decades of research.

Traditionally, men and women have been subjected to different "rules" guiding sexual behavior. Women were stigmatized for engaging in any sexual activity outside of heterosexual marriage, whereas for men such behavior was expected and rewarded. Boys had to "sow their wild oats," but girls were warned that a future husband "won't buy the cow if he can get the milk for free" (Crawford & Unger, 2000, p. 288). Women were faced with a Madonna-whore dichotomy: They were either pure and virginal or promiscuous and easy.

These rules have been the subject of considerable research. Reiss (1967) conducted the first large-scale and systematic study of sexual double standards. Although others had studied sexual behavior, Reiss focused on attitudes toward "various degrees of heterosexual permissiveness embodied in our premarital standards" (p. 6). His study included random samples of students from five selected high schools and colleges as well as a nationally representative sample of 1550 adults. His survey assessed attitudes toward "premarital sexual permissiveness" and their demographic and sociocultural correlates, such as age, race, social class, religion, and family characteristics.

Reiss (1960, 1964) classified attitudes toward premarital sexual activity into four general categories: abstinence (premarital intercourse considered wrong for both sexes), double standard (males considered to have greater right to premarital intercourse), permissiveness without affection (premarital intercourse considered right for both sexes regardless of emotional involvement), and permissiveness with affection (premarital intercourse considered right for both sexes if part of a committed relationship). Subtypes were delineated within each category. For example, within the double standard category, Reiss distinguished between an "orthodox" view that permitted premarital intercourse for males but not for females under any circumstances, and a "transitional" view that permitted premarital intercourse for females only if they were in love or engaged to be married.

Based on this attitude classification scheme, Reiss (1964) devised a pair of parallel 12-item scales. Participants responded to 12 items first with a male referent and then again with a female referent (e.g., "I believe that petting is acceptable for the male/female before marriage even if he/she does not feel particularly affectionate toward his/her partner"). As the sample item shows, the scales reflected Reiss's belief that degree of affection was an important factor in sexual attitudes.

Reiss classified participants into one of the four subtypes based on their responses to individual items and on the differences between their responses to the same items with a male versus a female referent. Overall, 42% of the student samples endorsed abstinence from premarital intercourse, with a sizeable minority endorsing permissiveness with (19%) and without (7%) affection. An orthodox double standard was endorsed by 11% and a transitional double standard by 14%. The adult sample was more conservative, with 77% endorsing abstinence, only 11% endorsing some form of permissiveness, and 9% endorsing some degree of double standard. These overall results were influenced by other variables. For example, in both student and adult samples, women were far more likely than men to endorse abstinence, and men more likely than women to endorse double standards.

Reiss maintained that although the orthodox double standard was a minority attitude, egalitarianism had not yet been achieved. Double standards were evident within the abstinence and permissiveness categories as well as the double standard category itself. For example, a respondent might endorse abstinence from intercourse for both sexes but believe that only men were entitled to kiss and pet without relational affection or commitment.

Throughout Reiss's account of his research, a progressive path toward greater sexual equality for males and females is assumed. Citing such factors as changing roles (more women in the workforce), liberalized norms (less condemnation of sexual activity in general), better health resources (leading to less risk of pregnancy and sexually transmitted diseases), and changing patterns of courtship (more autonomy and less familial control), Reiss (1967) suggested that "the realistic choice today is of the degree of equalitarianism the individual prefers rather than of full equalitarianism or full double standard" (p. 176). Although he recognized that factors such as the family structure impeded complete sexual equality, he predicted that the trend was toward increasing sex-role equality and decreasing double standards.

Is the progressive discourse tenable? Did the sexual revolution in behavior so visible in the 1960s lead to corresponding change in attitudes? If so, one might expect that, 35 years later, not only orthodox double standards but also transitional ones might be artifacts of the past.

By the 1970s, research seemed to indicate that individuals had come to hold virtually the same sexual standard for men and women (Peplau, Rubin, & Hill, 1977). Young people judged it equally acceptable for either sex to have premarital intercourse with affection, and a majority of young people thought that casual sex without affection was also acceptable (DeLamater & MacCorquodale, 1979). According to King, Balswick, and Robinson (1977), "taken as a whole, the evidence suggests that society is approaching a single premarital sexual standard as Reiss (1960) predicted" (p. 458). However, these conclusions did not end research on heterosexual double standards, which continued during the 1980s and 1990s, yielding mixed results (for reviews, see DeLamater, 1987; Oliver & Hyde, 1993).

The heterosexual double standard has been a "now you see it, now you don't" phenomenon. We hypothesize that despite the claims that double standards are progressively fading, they may exist in contemporary forms that influence the behavior of both women and men. Results of research on double standards may be influenced by methodological choices both general (e.g., experiment versus interview methods) and specific (e.g., linguistic and conceptual categories employed).

What follows is a review of evidence from a variety of research methods including experimental designs, ethnographies, interviews, focus groups, and linguistic analyses. We offer a critique of the strengths and weaknesses of each method, a presentation of the evidence for contemporary double standards that are nuanced and context dependent, a discussion of the implications of sexual double standards for women's well-being, and suggested directions for future research.

We generated a sample of quantitative studies by conducting computerized database searches of PsycInfo and Sociological Abstracts for the years 1981 to 2001 using the keywords sexual double standard, sex and double standard, sexual double stand, and sex stand. We supplemented this sample with ethnographies, focus group studies, interviews, and linguistic analyses. Sexual double standards have not often been the specific focus of research using such open-ended methods. However, double standards may be revealed in qualitative studies whose main purpose is a more general assessment of sexual behaviors and attitudes. We found qualitative studies that focused directly on double standards through our computerized database searches. We supplemented this sample by consulting the indexes of volumes reporting qualitative studies of more general sexual attitudes. The sample of qualitative studies used here is necessarily less systematic than the sample of quantitative studies; however, it includes every relevant study known to us.

In all, 30 published studies covering two decades and five different methods were included. Because differences in the frequency of specific sexual behaviors for men and woman may result from any number of sources (e.g., biological differences in sexual desire, gender socialization, etc.), studies were excluded that looked solely at the incidence of sexual behavior. Unpublished manuscripts, theses, and dissertations were also excluded.

Reiss (1960, 1964, 1967) first operationalized the double standard as differing standards of sexual permissiveness for women and men. We follow that definition, limiting its scope to attitudes about heterosexual activity, as did Reiss. Evidence for heterosexual double standards is demonstrated when study participants endorse different amounts or levels of sexual activity as acceptable for women and men in heterosexual relationships. Double standards may become evident in responses to questions about specific behaviors (e.g., participants may judge it less acceptable for a woman than for a man to have sex on a first encounter). They may also become evident in responses about the context, conditions, or motives for sexual activity (e.g., participants may judge it less acceptable for women than for men to have sex without affection, at an early age, or for selfish motives). Double standards also may be assessed by more global measures (e.g., measures asking whether women are entitled to as much sexual freedom as men are). Double standards may appear not just for heterosexual intercourse, but for other sexual behaviors as well (e.g., participants may judge sexual initiation less acceptable for a woman than for a man). Thus, the definition of heterosexual double standards used in this review is a broad one, encompassing Reiss's orthodox and transitional double standards, and reflecting the variety and breadth of ongoing research in this area.

An important aspect of any attitude research method is the language used. This is particularly obvious in experimental designs where participants are asked to respond to scale items or scenarios composed by the experimenter, with no easy way to contest the conceptual and linguistic categories they impose. More open-ended and qualitative research may impose the researchers' conceptual and linguistic schemas just as surely, although more subtly. The use of general, broad questions about freedom or equality may yield different indications of the strength of double standards than the use of specific behavioral comparisons. Another linguistic complication is that the language categories used in research vary considerably among studies and have changed over time. Reiss's use of the term premarital sex assumes that marriage is a universal and salient goal for participants, an assumption that may be less tenable today than when he conducted his research. Likewise, the word petting does not reflect contemporary usage. In line with our view that the choice of language is methodologically crucial, throughout this review we specify and sometimes critique the terms chosen by the original researchers as we report their results.

EXPERIMENTAL RESEARCH

Experimental research has been used to examine individuals' preferences for partners and their beliefs about acceptable levels of sexual activity for males and females at varying relationship stages. Experimental methods have been used to examine both individual double standards and perceptions of societal double standards. Individual double standards have been examined using within-subject designs, in which participants are asked to answer questions or make judgments about the sexual behavior of either male or female targets and are immediately asked the same questions about targets of the other sex. Perceptions of societal double standards have been examined directly by asking participants whether or not societal double standards exist. Sexual double standards also can be inferred from experiments using between-subjects designs, in which one group of participants is asked about male targets, and another group is asked the same questions about female targets.

Within-Subject Designs: Measures of Individual Double Standards

Within-subject designs provide the purest test of double standards because the same participants respond to the same set of questions for each target. We found only two within-subject experimental studies in our search of the literature. Sheeran, Spears, Abraham, and Abrams (1996) examined the individual double standards of 690 Scottish teenagers (277 males and 405 females). Participants were asked to estimate how many sexual partners most 20-year-old women or men have had and to rate a hypothetical young woman or man who changes sex partners a number of times during the year. Respondents predicted that women would have had significantly fewer sexual partners than men. Further, for both religious men and nonreligious women, sexually active women were considered less popular among both sexes than sexually active men. Finally, women who changed sex partners a number of times during the year were rated as more irresponsible and as having less self-respect than men who engaged in the same behaviors.

Sprecher and Hatfield (1996) found that men endorsed a double standard for women and men who were dating casually (defined as dating less than one month) but not for women and men who were dating seriously (defined as almost one year) or who were pre-engaged (had seriously discussed the possibility of getting married). In the U.S., 389 male and 654 female undergraduates (77% White) completed a premarital sexual permissiveness scale for themselves, "a male," and "a female." The permissiveness scale asked participants to rate the acceptability of sexual intercourse at each of five dating stages (first date, casually dating, seriously dating, pre-engaged, and engaged). Male participants held significantly more permissive attitudes for a male (the total score for items referring to a male) than for a female. Men's endorsement of a double standard was strongest at the first date but also existed for intermediate dating stages. Overall, U.S. men reported greater endorsement of double standards than did U.S. women.

This study was part of a cross-cultural comparison of double standards in the U.S., Japan, and Russia. Of note in the Japanese and Russian samples, which also consisted entirely of undergraduate college students, Russian students were more likely than U.S. students to show what the researchers termed a traditional double standard, especially at the first date, casual, and serious dating levels. Additionally, the greater endorsement of double standards by male participants than by female participants found in the U.S. sample was not found in the Russian and Japanese samples.

Between-Subjects Designs: Measures of Societal Double Standards

In between-subjects designs, each participant is presented with only one target; therefore, a direct measure of the double standard within the individual cannot be obtained. However, by comparing responses across participants by target sex, it is possible to determine whether, overall, women and men are judged differentially for engaging in the same behavior.

Person-perception tasks, which allow social scientists to measure stereotypes and impression formation unobtrusively and with few demand characteristics, have often been employed in research on double standards. Such tasks ask participants to make judgments or form impressions about a target after receiving information about him or her. Person-perception tasks allow researchers to examine the consequences of double standards on evaluations of males and females engaging in identical behavior. Evidence of double standards exists if women are evaluated differently than men for engaging in comparable levels of sexual activity under the same conditions.

Sprecher, McKinney, and Orbuch (1987) had 233 male and 320 female undergraduates (race and ethnicity unidentified) read a questionnaire ostensibly completed by another student. The questionnaire included manipulations for target sex, age at first coitus (16 or 21), and relationship stage at first coitus (a steady dating relationship that had lasted almost a year or a casual one that had lasted one week). Participants rated the fictional student on 23 bipolar scales that composed four factors: Sexual and Other Values (e.g., sexually experienced, sexually liberal, liberal in sex-roles), Maturity and Intelligence (e.g., responsible), Positive Personality (e.g., likable), and Dominance (e.g., dominant, active, masculine). There was a main effect for target sex such that female targets received higher scores on the Sexual and Other Values scale and lower scores on the Maturity and Intelligence, Positive Personality, and Dominance scales. There were also significant interactions between target sex and relationship stage at first coitus (Positive Personality and Dominance) and between target sex and age at first coitus (Maturity and Intelligence, Positive Personality, Dominance). Although having first coitus in a noncommitted relationship or at a young age had a negative effect on how both males and females were evaluated, the negative effect was greater for females. These results suggested that women were perceived more negatively than men for being more sexually experienced.

In a series of two studies, Oliver and Sedikides (1992) examined mate-selection preferences. It should be noted that, although date- and mate-selection paradigms can yield important information about double standards, they necessarily confound target sex and participant sex, because participants evaluate only opposite-sex targets.

In Study 1, 44 male and 44 female undergraduates (race and ethnicity unspecified) were asked to complete a sexual permissiveness scale as they would want either a blind date or spouse to complete it. Both male and female participants preferred low levels of sexual permissiveness in both blind dates and spouses. Male participants preferred lower levels of sexual permissiveness in a spouse than in a blind date; there were no significant differences for female participants. In Study 2, 80 male and 80 female undergraduates read a sexual permissiveness scale ostensibly filled out by an opposite sex student who was either sexually permissive or sexually nonpermissive. Participants then rated the target on 12 characteristics and likeability as a blind date and spouse. These items reflected two factors: Evaluation (e.g., morality) and Attraction (e.g., sexual attractiveness). Although the permissive target received lower evaluation ratings overall, female participants rated the male permissive target significantly more negatively than males rated the female permissive target. Male participants rated the female target as significantly more attractive when she was permissive than nonpermissive. Females rated the permissive male target as less desirable as both a spouse and a blind date. However, males rated the permissive female target as significantly less desirable for a spouse and significantly more desirable for a blind date than they rated the nonpermissive female target.

Milhausen and Herold's (1999) 165 female undergraduate participants (race and ethnicity unspecified) completed a questionnaire measuring both perceptions of a societal double standard and personal endorsement of double standards. Participants' beliefs in societal double standards were assessed with two items. Findings indicated that participants overwhelmingly (95%) definitely or probably believed that there is "a double standard for sexual behavior (a standard in which it is more acceptable for a man to have had more sexual partners than a woman)" (p. 363). Similarly, 93% of participants definitely or probably believed that "women who have had many sex partners are judged more harshly than men who have had many sex partners" (p. 363). When asked to explain their answers, 49% of respondents mentioned women's being penalized for their sexual behavior, and 48% mentioned men's being rewarded for theirs. Additionally, 42% of participants believed that it is women who "judge women who have had sex with many partners more harshly" (p. 363). Personal endorsement of double standards was measured by asking participants to indicate on a 5-point scale whether they would encourage or discourage a male or female friend from dating someone who has had intercourse with more than 10 different partners. Results showed that women were more likely to discourage a female friend from dating a man who had 10 previous sexual partners than to discourage a male friend from dating a woman who had 10 previous partners.

A smaller group of studies we reviewed showed little or no support for double standards. Jacoby and Williams (1985) presented 200 White undergraduates with five "sexual profiles" of opposite sex persons consisting of sets of answers to questions about sexual ideology and experience. Participants rated the dating and marriage desirability of the target in each profile. Findings indicated that there was no difference in the kinds of relationships in which sex was seen as acceptable for men and women and that no more than moderate sexual experience was desirable in either a male or female dating or marriage partner. This study had an unusual design that involved 287 different profiles for each sex. Little detail was provided on the actual questions participants responded to or the content of many of the profiles. Like other mate-selection paradigms, the design also confounded participant sex and target sex, making the results difficult to interpret.

Mark and Miller (1986) asked 467 undergraduates (210 males and 257 females whose race and ethnicity were unspecified) to read a supposed transcript of an interview with another student whose sex and sexual permissiveness (virgin, relationship sex, casual sex) were varied. After reading the transcript, participants rated the target on 54 bipolar items, which were shown to make up nine factors: Poor Adjustment, Unconventional, Likable, Agreeable, Caring, Assertive, Immoral, Conforming, and Trusting. An additional factor, Sexual (chaste-promiscuous, modest-immodest, nonseductive-seductive, moral-immoral), was created a priori and was not included in the factor analysis. Participants also completed 15 Likert-type items addressing general positive evaluations and mental health. Results indicated little support for sexual double standards. The only significant results were on the sexual and agreeable scales. Male participants judged female targets as more sexual than male targets if they engaged in casual sex. Liberal males and traditional females rated females who had had casual sex as marginally less agreeable than they rated males who had had casual sex.

Sprecher (1989) asked 666 undergraduates (race and ethnicity unspecified) to answer a premarital sexual permissiveness scale for 1 of 20 targets. Targets varied in gender, age, and relationship to the participant (i.e., a female or sister). The permissiveness scale asked participants to rate the acceptability of three sexual behaviors (heavy petting, sexual intercourse, and oral-genital contact) at four different relationship stages (first date, casually dating [dating less than 1 month], seriously dating [dating almost 1 year], and engaged). Although the findings did not support sexual double standards, they suggested that age (regardless of target sex) is important when evaluating sexual standards.

O'Sullivan (1995) had 110 male and 146 female undergraduate students (93.8% White) read a vignette about a target who varied in sex, number of past partners (low [2 for males and 1 for females] or high [13 for males and 7 for females]), and type of relationship (committed or noncommitted). (The methodological justification for using different numbers of past partners will be discussed later.) Participants rated targets on nine bipolar items: well adjusted, moral, feminine, masculine, likeable, aggressive, self-confident, desirable as a dating partner, and desirable as a marriage partner. There were main effects for all three independent variables. Male targets were rated somewhat less positively than female targets overall. Sexually experienced targets received the least positive evaluations, and targets in noncommitted sexual relationships were rated as less well-adjusted, less moral, less feminine, less likable, more aggressive, less self-confident, and less desirable as a dating partner and a spouse. However, little evidence of double standards was found. Women did not receive more negative evaluations than men did when described as having had high numbers of past sexual partners in noncommitted relationships. The methodology, however, makes the results difficult to interpret.

Sprecher, Regan, McKinney, Maxwell, and Wazienski (1997) asked 436 undergraduate college students (85% White) to complete one of three versions of a mate-selection preference list consisting of 18 partner traits. The target's sexual experience was manipulated using three versions of the sexuality items: sexual chastity (no previous sexual partners), some sexual experience (few sexual partners), or considerable sexual experience (several previous partners). Sprecher et al. found no evidence of a heterosexual double standard, with participants reporting no differences in preference for level of sexual experience based on target sex.

Gentry (1998) asked participants (254 undergraduates, 92% of whom were White) to read a portion of an interview with a fictional heterosexual college student focusing on relationships and sexuality. The target's sex, number of relationships (either an exclusive, monogamous sexual relationship or sexual relations with multiple partners), and level of sexual activity (above average, below average, or average) were manipulated. Participants then rated the target on 30 semantic differential scales measuring overall evaluation, sexual demeanor, subordination, traditionality, and physical and social appeal. Results showed that participants based their judgments about the targets on information about number of relationships and level of sexual activity rather than the target's sex.

Summary and Critique of Experimental Studies

Of the 11 studies reported in detail, 5 found evidence of sexual double standards. Only 2 of the 11 studies used within-subject designs; both of these found evidence of double standards.

Studies varied in the possible moderating variables they assessed. Only one study reported analyses based on target age; both target age and its interaction with target sex were significant (Sprecher et al., 1987). Seven reported analyses based on relationship stage; of these seven, three reported overall evidence of double standards. Additionally, all three found significant main effects of relationship stage and interactions between target sex and relationship stage. Two studies found that sexually active men were judged more harshly than were sexually active women: Oliver and Sedikides' (1992) study of mate-selection preferences, which confounded participant sex with target sex, and Milhausen and Herold's (1999) study, which involved female participants rating female and male targets.

Experimental designs have several advantages for studying sexual attitudes. They allow the researcher to manipulate or control variables believed to influence double standards, such as relationship stage. Unlike any other method, they permit causal inferences to be drawn. Because they typically present participants with hypothetical scenarios and use paper-and-pencil measures of dependent variables, they are relatively easy to use with large samples, which yield more statistical power. Their decontextualized scenarios also allow participants to reflect on their attitudes without personal emotional involvement.

Offsetting these theoretical and practical advantages are several limitations. An intrinsic problem with statistical comparisons is that the null hypothesis of similarity can never be taken as proven, whereas hypotheses of differences can be assessed probabilistically. The statistical emphasis on significant differences may foster the "file drawer problem" in which studies reporting similarities in attitudes toward a target's behavior regardless of target sex may remain unpublished, leading to an overestimate of double standards.

Other limitations of experimental methods are not intrinsic but occur frequently in practice. For example, virtually all the experimental studies described in this review relied on convenience samples of North American college students. This is a serious limitation, not just because these samples are unrepresentative of the general population, but because they are unrepresentative in particular ways that are rarely discussed or even acknowledged by researchers. Sears (1986) has discussed specific developmental issues within college student populations and has analyzed how uncritical over-reliance on student samples has affected psychology's view of human nature in general. His criticisms are particularly germane to the area of sexual attitudes, yet none of the research reports summarized here addressed sampling limitations in any depth.

In designing experimental studies of double standards, researchers have occasionally confounded target sex and participant sex (Jacoby & Williams, 1985; Oliver & Sedikides, 1992). Results of such studies are difficult to interpret. If women's evaluations of men differ from men's evaluations of women, the difference could be due to either participant sex or target sex. Participants might be endorsing different standards for women and men, or they might be endorsing a single standard, but one that is different for female and male participants. The former fits our definition of a double standard; the latter does not. However, even if individual women and men each hold a single standard, if the standard held by women is different from the standard held by men, heterosexual women and men could still be evaluated differently by potential partners. These complexities can be teased out by systematically varying both participant sex and target sex.

Another theoretical and practical issue in experimental designs is that the conceptual and linguistic categories used to measure double standards are preformed by the experimenter. This factor, intrinsic in experimental designs, may have two unwanted and undesirable consequences. First, the response categories available to participants may be inadequate to represent the variable under consideration. Second, even when the categories seem adequate, the method assumes that they have the same meaning to all participants and that the meaning ascribed to the category by participants is the same as that ascribed by the experimenter. We discuss examples of these two problems in turn.

In designing experimental studies, there is a tension between the need to represent the complexity of a concept and the need to keep the number of experimental cells manageable. For example, Oliver and Sedikides (1992) had only two relationship levels: "blind date" and "spouse." This reductionist approach necessarily omits all of the various relationship stages in between. At the other extreme, Jacoby and Williams (1985) represented level of sexual activity with four variables--target's sexual standards, number of sexual partners, number of partners for whom affection was felt, and number of partners for whom love was felt--resulting in 574 different targets. Not only was this a very complex design, it also, as Oliver and Sedikides (1992) pointed out, required participants to perform calculations to determine the target's behavior with respect to relationship affection and commitment (number of partners with whom the target was emotionally connected / number of total partners). It is not clear that participants recognized the required calculations or were invested enough in the task to perform them.

Mark and Miller (1986) did not find evidence of a double standard. However, similar to Jacoby and Williams (1985), the emotional context of the targets' sexual activity was confounded with degree of sexual activity. Targets were described as either a virgin, sexually experienced in committed relationships, sexually experienced in casual relationships, or no information. This design confounds sexual experience and level of commitment and does not allow separate tests for their main effects or interactions between them. Such tests would require crossing at least two levels of sexual experience (such as inexperienced and experienced) with at least two levels of commitment (such as casual and committed).

O'Sullivan (1995) asked participants to make judgments about a target after reading a vignette depicting either a male or a female target as having a low or high number of past sexual partners within committed or noncommitted relationships. Instead of using general terms like a few partners or many partners, O'Sullivan used specific numbers of prior partners. However, males were described as having had either 2 or 13 past partners whereas females were described as having had either 1 or 7 past partners. O'Sullivan argued that using the same number of partners (i.e., 13) for both males and females would indicate a high number of partners for males and an extremely high number of partners for females. Although pilot research did suggest gender-specific norms--the average number of partners among participants in the same pool as those in this study was higher for males--the research did not explore how these pilot data themselves suggest a double standard, which was then encoded in the methodology. Using different numbers of prior sexual partners for males and females leads to a particular kind of comparison, one that does not correspond to our definition of double standards.

Similar to Jacoby and Williams (1985) and Mark and Miller (1986), Gentry's method allowed for only a limited test of the effect of relationship stage on the double standard. Level of commitment was manipulated by indicating that the target had either one exclusive partner or multiple partners at the same time. This operationalization of relationship commitment assumes that participants will associate one partner with a high level of commitment and multiple partners with no commitment. However, there are many different relationship possibilities not represented, such as having one partner with a low level of commitment or having multiple partners but a high level of commitment. Again, this study does not rule out the possibility that contemporary double standards are moderated by relationship affection or commitment.

A more general problem with experimenter-specified categories is that it is impossible to know if the category has the same meaning to all participants or to know whether their interpretation is congruent with the experimenter's. Although this is potentially problematic in any research relying on verbal stimuli (Landrine, Klonoff, & Brown-Collins, 1992), it may be particularly so in sexuality research. Sexuality remains a relatively taboo topic of conversation, with frequent use of euphemism, avoidance, and indirect speech in talk about sex (Mulkay, 1988). Because much sexual activity takes place privately, participants may know very little about the actual sexual activities of their peers. Therefore, it is unclear that there is social consensus about the meaning of sexuality-laden terms. Researchers may only believe they know what students mean when they respond to targets described as "sexually permissive/restrictive" or relationships described as "steady/casual" (Sprecher et al., 1987).

The meanings of terms such as date, dating partner, and virgin have changed over time. Anecdotal reports from our own students suggest that dating is not current usage to describe heterosexual relationships. (One student told us it brought to mind a 1950s sitcom.) Instead, students use terms such as hooking up and seeing someone. Scripted activities related to hooking up may be quite different from those related to dating (Paul, McManus, & Hayes, 2000; Rose & Frieze, 1989). Moreover, the meaning of a term may vary by geographic locale and may even change depending on its immediate context. The possibility that words such as dating, permissive, virgin, and casual may have quite different connotations to participants at different times and in different contexts presents an empirical question that has not been investigated. O'Sullivan's (1995) use of pilot research to explore participants' understanding of "many past sexual partners" in terms of local norms is a step in the right direction.

Finally, there is a notable gap in research on sexual double standards in the area of attention to racial and ethnic diversity. Only four of the studies we reviewed mentioned the racial or ethnic composition of their sample, and none reported analyses by race or ethnicity of participant. Not one manipulated the race or ethnicity of targets. Although participants' race, ethnicity, and racial attitudes may affect perceptions of others' sexual behavior just as surely as do participants' sex and gender-role attitudes, experimental studies of double standards have thus far failed to recognize or address these variables.

QUALITATIVE AND INTERPRETIVE DESIGNS

In contrast to experimental methods, qualitative studies more readily lend themselves to contextually sensitive phenomena. The questions they can address are more open-ended and diffuse, less abstract and hypothetical. Here we review studies using ethnographic, interview, focus group, and linguistic analyses.

Ethnographic Studies

Ethnographic researchers, by definition, attempt to capture the belief systems of a community through close and sustained observation. Therefore, individuals' accounts of their communities' norms and their own beliefs about sexuality may emerge as part of broader ethnographic studies not specifically focused on double standards.

Heterosexual double standards are salient in ethnographies of adolescent culture in U.S. middle schools. In a 3-year study, Eder, Evans, and Parker (1995) used a variety of data sources: They observed lunchtime activity, attended extracurricular activities, conducted individual and small group interviews, and recorded audio and video tapes of students' spontaneous conversations with peers. Within the middle school peer culture, girls (but never boys) sometimes were negatively labeled simply because they showed interest or assertiveness with respect to sexuality. Girls who initiated any kind of sexual overture (e.g., "making a pass") were labeled "bitches," "sluts," and "whores." The sanction against female sexual agency extended to wearing attractive clothing or makeup, which also could earn the label of whore. The routine use of sexual insults aimed at girls by boys, and to a lesser extent by girls, suggested that middle school students "do not believe that girls should be sexually active or have a variety of boyfriends, while such behaviors are viewed as normal and acceptable for boys" (p. 131). Eder et al. (1995) concluded, "labeling young girls in this manner becomes part of a continual attempt to limit their sense of sexual autonomy and identity" (p. 153).

Focusing on girls, Orenstein (1994) conducted a yearlong ethnographic study of two middle schools, one suburban, White, and affluent, the other urban, largely ethnic minority, and poor. Girls' accounts described their fear of the slut label and their shamed silence around sexual desire. Orenstein vividly captured how double standards can be communicated even in a sex education class. The teacher, Ms. Webster, is trying to illustrate the risk of sexually transmitted diseases:

   "We'll use a woman," she says, drawing the Greek symbol for
   woman on the blackboard. "Let's say she is infected, but she hasn't
   really noticed yet, so she has sex with three men."

   (As she draws symbols for men on the board) a heavyset boy in a
   Chicago Bulls cap stage whispers, "What a slut," and the class
   titters.

   "Okay," says Ms. Webster, who doesn't hear the comment. "Now
   the first guy has three sexual encounters in six months." She turns
   to draw three more women's signs, her back to the class, and several
   of the boys point at themselves proudly, striking exaggerated
   macho poses.

   "The second guy was very active, he had intercourse with five
   women." As she turns to the diagram again, two boys stand and
   take bows..

   During the entire diagramming process, the girls in the class
   remain silent. (p. 61)

Double standards have emerged in ethnographies of college students as well. Moffat (1989) studied peer culture on a university campus, focusing on sexual beliefs and attitudes and using a variety of methods. He conducted participant observation while "passing" as a student living in a men's dormitory, and he analyzed sexual autobiographies written by students in sexuality classes. He reported that the majority of students believed in heterosexual double standards and classified women into dichotomous categories of "good" women or sluts. Moffat characterized the attitudes of the male students he studied as follows:

   Men have the right to experiment sexually for a few years. There
   are a lot of female sluts out there with whom to so experiment.
   And once I have gotten this out of my system, I will then look for
   a good woman for a long-term relationship (or for a wife). (1989,
   p. 204)

Interview and Focus Group Studies

Individual interviews and focus group discussions are open-ended methods of self-report. In these settings, participants can express beliefs and attitudes in their own terms and provide contextual information to justify or explain their positions. One such study relied on interviews conducted with 55 adolescents, both boys and girls, with a mean age of 16 years (Martin, 1996). Participants were students at three high schools, one working-class public and two upper-middle-class private institutions. They were interviewed privately in the school setting for about an hour each. Questions assessed identity, self-esteem, and body image, as well as experiences with the physical changes of puberty, intimate relationships, and sexual intercourse. The researcher coded responses by theme using a grounded theory approach (Glaser & Strauss, 1967).

Both girls and boys distinguished between girls they called sluts, "hos", or "Sally off the street" and girls they called "regular" or "normal" girls, those who could become someone's girlfriend. Distinctions such as these were not made for boys. However, having sex was not enough to earn the label of slut. Instead, peers' distinctions were made based on whether the gift was "too young" or had "too many" partners. Martin concluded that "the double standard in sex is still firmly rooted in teenage culture" (pp. 85-86).

In addition to the double standards of peers, many girls perceived parental double standards. For example, one respondent who expressed confusion and guilt about her sexual activity reported that her boyfriend's mother categorized any girl who "slept with" a boyfriend or even used a tampon during menstruation as a slut. Martin's female participants appeared to "take the distinction of slut to heart and fear it." She concluded, "Regardless of its particular contextual meaning, the word slut holds a lot of power. Being called a slut or a ho--or feeling like one--is to feel degraded and dirty" (pp. 86-87).

In a study of attitudes toward condom use among 105 young (median age 18) drug users in Western Australia, both male and female participants were asked "What do young men think of young women who carry condoms?" and "What do young women think of young men who carry condoms?" (Loxley, 1996). Results showed discrepancies between participants' reports of the attitudes of people of their own sex and the other sex. Participants of each sex believed that those of the other sex held more negative attitudes than were actually reported (e.g., women thought men had more negative attitudes toward women than men actually reported having). This mismatch suggests that perceptions of societal double standards may influence behavior. However, the transparent within-subject design, with each participant being asked about both sexes' attitudes, may have induced unknown biases in responding. Loxley's brief report provides little information about how the purpose of the study was justified or explained to participants or about the context in which the two key questions were presented.

Other studies, using single-sex samples, also show that adolescents perceive and are affected by sexual double standards. Thompson (1995), using a snowball sampling technique and open-ended interviews, gathered stories from 400 girls representing a variety of geographical locations, class backgrounds, and ethnicities. Like Martin (1996), Thompson concluded that the heterosexual double standard remains "virulently alive and well" (p. 31) but that it is less absolute and more contextually negotiated than in the past. Girls described innumerable fine lines as to what constituted good and bad sexual behavior, lines that mapped local and individual constructions of sex and gender. They appeared to use these maps to define and orient themselves as "good girls," as opposed to "easy" girls or sluts.

A smaller study of 17 male college students (ages 1822) used a grounded theory approach to analyze transcripts from interviews (Fromme & Emihovich, 1998). These respondents endorsed a double standard that divided women into two groups. "Good women" were those who say no to casual sex or to intercourse early in a relationship; these women were seen as acceptable for longer term relationships. "Bad women" were those who had sex on a first date or sex with many partners. The authors noted that their respondents did not seem to recognize that they derogated women for behaviors they accepted for themselves, as in this comment:

   If I met a woman in a bar and had sex with her chances are I
   wouldn't call her because I wouldn't have any respect for her.
   Because if she did something like that ... would I want someone
   like that for the rest of my life? No, of course not. (p. 174)

One other source of first-person accounts deserves mention. Espin (1997), drawing on her knowledge of Hispanic/Latin culture and on experiences as a therapist with Latina clients, notes that Hispanics in the U.S. continue to attribute a great deal of importance to female sexual purity. Particularly among the upper social classes, women's virginity before marriage is a cultural imperative, and women's sexual behavior is an important marker of a family's honor. Married women are expected to remain completely monogamous while accepting their husbands' extramarital sexual affairs. Indeed, experiencing sexual pleasure and gratification, even in marriage, may be interpreted as evidence of a lack of purity and virtue in women.

Using a focus group technique, Fullilove, Fullilove, Haynes, and Gross (1990) recruited low-income urban African American women to participate in small-group discussions of sexuality. The researchers reported data from six homogeneous groups, some comprised of adolescents and some of adult women. Participants spontaneously described a dichotomy of sexual roles for women: the "good girl/madonna" role and the "bad girl/whore" role. A comparable dichotomy did not exist for men. For the participants in this study, sexual intercourse outside marriage did not necessarily relegate a woman to the bad girl/whore category. Rather, bad women were those who engaged in casual sex or offered sex in return for money or drugs.

Ward and Taylor (1994), in a study of sexuality education for minority students, conducted homogeneous focus groups with urban adolescent boys and girls of six ethnic or cultural groups: Vietnamese, Portuguese, African American, White, Haitian, and Hispanic. Each group was facilitated by an adult from the relevant cultural background. Without exception, groups from all ethnic backgrounds described heterosexual double standards that were "limiting and oppressive to females" (p. 63). There were some variations across ethnic groups. For example, some groups stressed premarital chastity more than others did. However, contrasting expectations for females and males, with more restrictions on females' behavior, were clearly expressed in all groups.

The importance of local community norms is echoed in a study of 512 high school students (ages 15-19) from rural Australian small towns (Hillier, Harrison, & Warr, 1998). Using a combination of survey and focus group methods, participants' attitudes about safe sex were explored. Fourteen single-sex focus groups were conducted in seven towns, each led by a same-sex facilitator. Analyses of content and themes were conducted and "cross-verified by project members" (p. 18).

Participants judged many of the risks of having sex to be greater for young women than for young men. Chief among these was the "sullied reputation" risk, a pervasive concern expressed in every female focus group. Loss of one's reputation was perceived to lead to sexual harassment, loss of both female and male friends, shame, and alienation. Respondents made clear their belief that losing one's good reputation could happen quickly and easily in a small town where everyone's activities were known and talked about. The risk of a bad reputation was perceived by both sexes to apply only to girls. Such double standards were clearly captured in these remarks from two focus groups:

   Girl: They (boys) have a one-night stand and nothing happens.
   We're more in fear of getting labeled like a tart or a slut or
   something. Whereas the boys if they have it they don't get
   labeled ... and we're more ashamed of it if we do.

   Boy: You do it for the feeling and to brag about it afterwards.
   (Hillier et al., 1998, p. 26)

The researchers concluded that when young women in these small towns had sexual intercourse, or even when others wrongly believed that they did, they risked losing friends, family, and the opportunity for future relationships. These risks were exacerbated by the fact that females were more socially defined through their relationships than males were. Therefore, the young women were unequally positioned with their sexual partners with respect to their power to negotiate.

Attitudes toward sexual risk taking also provided the impetus for a study of U.S. inner city youth (Stanton, Black, Kaljee, & Ricardo, 1993). The primary method was focus groups, conducted with young people in two age groups (9-10 and 13-14), supplemented by interviews with parents and ethnographic background study. The focus group sample was drawn from patients at an urban pediatric clinic serving a largely low-income African American population.

Attitudes about girls' and boys' sexual activity seemed to be more similar than different in this population. High levels of sexual activity were perceived as normative for both sexes. One difference that did emerge was that sexual activity for even very young boys was socially accepted by peers and adults, whereas attitudes toward girls' sexual activity were more variable. Although some girls chose to abstain from sexual intercourse and reported that peers respected their choice, others reported a great deal of pressure to become sexually active. Moreover, participants agreed that girls who became pregnant received considerable positive attention and acceptance. Thus, these results suggest that peer pressure on boys strongly discouraged abstinence from intercourse. Although there was similar peer pressure on girls, at least some girls perceived abstinence as a viable option.

Taken together, these analyses of diverse ethnic and cultural groups suggest that contemporary sexual double standards are local and subcultural constructions rather than a universal mandate.

Language and Discourse Analyses

Language both reflects and reinforces social reality (Crawford & Unger, 2000). Studies of terms for sexually active women and men show that such terms differ in frequency and connotation. The English language has many more terms describing women than men in specifically sexual ways, and the great majority of these are negative (Adams & Ware, 1989). In a study of slang from a U.S. university, almost 90% of the words for women and only 46% of the words for men had negative connotations. Many negative words for women described them as being very sexually active ("turboslut," "roadwhore," "skag," "wench"). There was only one term for very sexually active men ("Mr. Groin"), and it was not seen as unambiguously negative (Munro, 1989). In a language corpus from another university, Sutton (1995) found that students generated many negative words for women when asked simply to collect 10 slang terms that they and their friends used often. By far the largest category was terms denoting sexually active women: "bait," "beddy," "ho," "hooker," "hootchie," "scud," "skank ho," "slag," "slut," "tramp," and "whore" are examples. Sexually active women were also denoted by their genitals: "pelt," "slam hole," "stimey hole," and "tuna." The only comparable male term offered by students was "hoebuck" (a parallel to "hobag" but less negative in connotation). Thus, the linguistic resources available to assess the sexual behavior of women and men differ considerably and define women's (more than men's) morality and desirability in terms of their sexuality.

The only experimental study to include an open-ended question about language and sexuality was conducted by Milhausen and Herold (1999). Participants were asked what words they would use to describe a man or woman who has had many sexual partners. Two raters independently coded responses into categories. Virtually all the words listed for highly experienced men and women were deemed negative. Terms for men most often connoted sexual predation ("player") and promiscuity ("sleazy"). Terms for women most often connoted promiscuity ("slut") and psychological dysfunction ("insecure"). Milhausen and Herold tallied the number of negative terms applied to female and male targets. Finding no difference, they concluded that their linguistic analysis provided no support for double standards. All the participants in this study were women. It would be interesting to see comparable data from men. As Milhausen and Herold pointed out, supplementing quantitative measures with open-ended items that tap participants' own language use provides "a richer context for understanding ... perceptions of the double standard" (p. 367).

Another recent study of language use showed that men are more likely to be portrayed positively and agentically in everyday talk about sex. Weatherall and Walton (1999) studied college students' metaphors for sexual activity by asking students to keep daily dimes of their sex talk for a week. Common types of metaphors were food and eating, animals, sport and games, and war and violence. Although most of these metaphors would seem to be potentially gender neutral, males were two and a half times more likely to be the actor in mundane talk about sexual activity. These linguistic data may reflect the double standard in agency and initiation that is a frequently recurring theme in qualitative and quantitative studies that assess attitudes about sexuality.

Research on the lexicon, usage patterns, and metaphors of sexual activity is a promising and as yet underdeveloped area. In addition to the measures described here, other techniques such as semantic differential scales and discourse analysis could provide insights into the connotative meaning and social function of sexuality-laden language.

Summary and Critique of Qualitative Studies

The majority of the ethnographies, interview and focus group studies, and linguistic analyses reviewed here found evidence for double standards, and sometimes indicated the influence of some of the same contextual factors shown in quantitative studies. As a group, the qualitative studies used a much more varied spectrum of participants with respect to ethnicity, race, social class, age, and social position than did the quantitative studies. Although there is no intrinsic reason that quantitative studies could not draw from diverse populations, this has not been the case in practice.

Qualitative researchers' use of diverse samples yields some benefits. It adds to the limited database of quantitative studies using college students and potentially can correct for some of the developmental and social biases introduced by over-relying on college samples (Sears, 1986). By drawing on participants from varying social positions, this research also helps portray the variety of socially constructed double standards. Open-ended methods have the potential to show the dynamic aspects of double standards: how they are conveyed in interaction and actively employed as a means of social control for both sexes. In the middle school sex education class described by Orenstein (1994) and discussed earlier in this paper, both boys and girls were provided with models that might encourage them to incorporate sexual double standards into their value systems. Some boys engaged in macho posturing about high levels of male sexual activity, and some voiced negative labels for females who behaved similarly. These behaviors occurred without sanction from an authority figure and with approval from other students.

Of course, interpretive methods have their drawbacks, too. They are costly in terms of time. They require researchers who are members of the group being studied and/or highly trained in developing rapport and trust. Finally, unlike laboratory studies, they cannot manipulate variables to establish causality, and they cannot systematically examine the interaction of several factors.

There is potential for selection bias in choosing interpretive studies for a review, and this review is no exception. Because interpretive methods such as ethnographies may be very broad in scope, with the goal of capturing the ethos of a community, a narrow topic such as sexual double standards may be embedded in the (often book-length) report but not accessible by keyword or index searches. Another potential selection bias is that interpretive studies that look for but do not find evidence of double standards may not mention the topic in their reports, so that the published literature overrepresents the incidence of double standards. This is analogous to the file-drawer problem in quantitative studies. In this review, our criteria for including a qualitative study were necessarily less precise than our criteria for including a quantitative study.

CONTEMPORARY DOUBLE STANDARDS

The women's movement, along with the sexual revolution, made sweeping changes in the way women acted and were perceived. One aspect of these changes was the weakening of the orthodox heterosexual double standard (Reiss, 1967). However, the belief that double standards would fade away entirely may have been overly optimistic. Although sexual attitudes have moved toward more egalitarian standards, the conditions under which particular behaviors are acceptable remain somewhat different for men and women.

Within contemporary U.S. culture, some social and ethnic groups still endorse the absolute or orthodox double standard that prohibits sexual intercourse outside marriage for women (Espin, 1997; Ward & Taylor, 1994). In other groups, the declining importance of the orthodox double standard has made way for double standards that are subtler but perhaps equally effective as a means of social control. Using a variety of quantitative and qualitative methods, the studies reviewed here have shown that contemporary sexual double standards are multidimensional. Women may be judged more negatively than men for having had sex with many partners in the past, having sex outside of a committed relationship, or for having sexual experience at an early age (Sprecher et al., 1987). Martin (1996) found that merely having sexual intercourse was not enough to elicit a negative evaluation for the adolescent girls in her sample; however, a girl was labeled a slut if she was believed to be sexually active and was judged "too young" by her peers.

Men in particular may judge women's sexual behavior in more restrictive ways when they are asked to view the target as a potential marriage partner versus a dating or casual sex partner. Sexually permissive women may be preferred as dating partners; however, those who say no to casual sex may be more acceptable for long-term relationships (Fromme & Emihovich, 1998). When it comes to marriage partners, men may prefer women who have less sexual experience (Oliver & Sedikides, 1992). Moffat's (1989) findings suggested that the college men he studied felt entitled to have casual sex with "sluts" until they were ready to settle down with a "good woman."

An important aspect of sexual double standards in many social groups is that women are judged more negatively than men for initiating heterosexual interactions (Green & Sandos, 1983). Eder et al. (1995) found that middle school girls who initiated any type of sexual activity were labeled bitches, sluts, and whores. The double standards of sexual experience and agency lead to differential judgments of culpability when sexual activity results in negative consequences. Women who initiate physical contact on a date or who are more sexually experienced are judged as more culpable than passive or inexperienced women if they are the victims of rape (L'Armand & Pepitone, 1982; Muehlenhard & MacNaughton, 1988). Contemporary double standards allow some women greater sexual freedom than in the past. However, these standards still represent a covert means of controlling women's sexuality by judging its expression more harshly than men's sexual expression is judged.

With the exception of some religious and ethnic immigrant groups, an absolute standard of female abstinence outside marriage is rare. Instead, heterosexual double standards are local social constructions, made up of innumerable "fine lines" (Thompson, 1995) that may be inadvertently crossed. A middle school student may be labeled a slut for wearing too much makeup, initiating a conversation with a boy, or using a tampon during menstruation. A White female college student who has had N sexual partners may still be marriageable, but a student with N + 1 partners may cross the fine line and be judged promiscuous and undesirable as a mate. A Black working-class woman who has many sexual partners may be viewed as acceptable unless she trades sex for money (Fullilove et al., 1990). A Hispanic woman may be constrained against expressing sexual pleasure even within marriage (Espin, 1997). Thus, contemporary sexual double standards are matters of ongoing negotiation and meaning making within particular social groups.

The multidimensional perspective provides a way to understand why some studies find evidence of double standards whereas others do not. Of course, failure to demonstrate evidence for double standards could mean that participants do not hold a double standard on any dimension. It could also occur because the study did not include dimensions that mattered to the participants. Or perhaps the study did include relevant dimensions, but their levels were insufficient to show the effect. For example, college students might not differentiate between male and female targets that had two past partners, but might evaluate the targets differently if they were said to have had 20 partners. Dimensions also may interact in complex ways. Number of prior partners may matter little when evaluating a target as a friend or a blind date but may matter a great deal when evaluating him or her as a potential spouse.

The "disappearing act" of heterosexual double standards may be related to the methods used to study them. Overall, our review found that experimental studies were somewhat less likely to show double standard effects than were studies done with more qualitative methods, such as interviews, focus groups, ethnographies, and linguistic analyses. Experimental methods often abstract behavior from its social context, asking questions about hypothetical males and females. If, as we have argued, contemporary double standards are multidimensional social constructions that vary widely across communities, it would be expected that interpretive and context-dependent ways of asking questions would reveal their presence more than abstract and general ones. To varying degrees, interpretive and qualitative methods incorporate participants' identities as members of a community. They encourage participants to reflect on culturally relevant lived experiences as references for their responses and to answer questions in their own words. The immediacy and force of the many negative terms for sexually active women revealed by these methods adds a new dimension to understanding the repressive power of normative double standards.

IMPLICATIONS FOR WOMEN'S WELL-BEING

Differential evaluation of women's and men's sexual activity has important effects on the lives of women. Here we briefly discuss two areas that have received considerable research attention: high-risk sexual behavior and sexual identity. Although a complete review of these topics is beyond the scope of this paper, we cite representative studies.

Influence of Heterosexual Double Standards on High-Risk Sexual Behavior

Unprotected sexual intercourse has always posed a risk for women with respect to unwanted pregnancy. With the advent of the HIV/AIDS epidemic, both men and women are increasingly concerned about the risks of sexually transmitted diseases. The incidence of HIV infection and AIDS is growing rapidly among heterosexual women. As a result of these increased risks, public health campaigns specifically targeting women have focused on the importance of condom use (Hynie & Lydon, 1995).

Despite these efforts, the acceptance of heterosexual double standards may have detrimental effects on women's condom use. For example, in Hynie and Lydon's (1995) study, women who read a narrative evaluated a female character more negatively if she provided a condom before sexual intercourse than if she had unprotected sex. In addition, women rated a female condom provider in a casual encounter as less nice, less socially acceptable, and less appropriate than when her male partner provided the condom. These findings suggest that social norms discourage women's contraceptive preparedness in initial sexual encounters.

Interestingly, Campbell, Peplau, and Debro (1992) found that women's attitudes toward condom use are not only positive, but more positive than men's attitudes toward condom use. It is surprising that even given increased public health campaigns and a greater risk of HIV/AIDS, some women still rely on men to provide condoms in sexual situations (Hynie & Lydon, 1995). This could result from a heterosexual double standard that forces women to choose between being contraceptively prepared and socially desirable.

Hynie & Lydon (1995) also found that a woman's belief about her partner's endorsement of double standards is more important than her own beliefs in predicting condom use. If a woman believes her partner endorses a double standard, she is less likely to provide a condom during intercourse (Hynie & Lydon, 1995). Women who are contraceptively prepared outside of a long-term committed relationship are perceived as indicating a willingness and desire to engage in casual sexual intercourse (Williams et al., 1992). This preparedness is less acceptable for women than for men because of differing standards for judging the appropriateness of casual sex for men and women. Women are left to choose between risky, unprotected sex or negative judgments about their character.

Impact of Double Standards on Women's Sexual Identity

Double standards may be internalized and used by women to evaluate themselves. Women often believe that they should conform to societal-based standards of morality by resisting their own sexual urges (Katz & Farrow, 2000). Sexuality and the desire to be attractive as a dating partner present conflicting ideals for women. Girls learn to look sexy but say no, to be feminine but not sexual, and to attract boys' desire but not to satisfy their own. They experience dissonance when they are depicted as sexual objects yet defined in terms of sluts and whores when they express their sexuality (Durham, 1998). These conflicting messages may cause many women to develop a negative sexual identity (Katz & Farrow, 2000).

Research on the likelihood of sexual initiation has shown that both men and women report that males are more likely to initiate sexual behaviors including kissing and intercourse (DeLamater & MacCorquodale, 1979; O'Sullivan & Byers, 1992). Traditional sex role norms give men greater freedom to initiate sexual intercourse and to act as sexual experts, whereas women are expected to refuse sex, acting as sexual gatekeepers and limit-setters (Lips, 1981; Safilios-Rothschild, 1977). The role of gatekeeper can have many implications for women's sexuality. It is a less powerful image than the man's role of expert or teacher, further objectifying women and inhibiting them from taking responsibility for their own sexuality (Lips, 1981). This encourages women to think of themselves as objects or victims of the sexual encounter as opposed to willing, active participants.

Experimental person-perception research has shown that violating a gender role expectation can result in social rejection, even when the norm violations are seemingly small. Green and Sandos (1983) showed that male targets who took the initiative for either a conversation or a dinner invitation were rated more positively and more socially acceptable than female targets who did the same. The effects of norm violation extend to perceived culpability for sexual violence. When women initiate dates, they are perceived as more willing to engage in sexual intercourse and their rapes are rated as more justified than when men initiate the dates (Muehlenhard, 1988a). Women who initiate physical contact on a date are likely to be viewed as more responsible for being raped (Muehlenhard & MacNaughton, 1988). Further, if the victim of a rape is described as sexually experienced, the rape is judged as less serious, the perpetrator receives less blame, and the victim receives more blame (L'Armand & Pepitone, 1982).

Women who want to engage in sexual intercourse with their partners but believe that their partners endorse a sexual double standard are forced to make a choice between openly acknowledging their desire for sex (which can result in being labeled easy or a slut) or refusing and saving their reputations. Such a refusal acts as a way to legitimize women's sexual desires within the confines of a sexual double standard. A woman's initial resistance to sexual intercourse and eventual concession to her partner's wishes is known as token resistance: refusing or resisting sexual activity while intending to engage in that activity (Muehlenhard & Rodgers, 1998). Muehlenhard and McCoy (1991) found that women who used token resistance were more likely to believe their partners accepted sexual double standards than were women who had openly acknowledged their sexual desires. Interestingly, women's own endorsement of sexual double standards was unrelated to their use of token resistance.

Sexual double standards may lead females to sacrifice their sexual autonomy in exchange for social desirability. A woman who only engages in sexual activity when her partner wants to may find it difficult to become aroused when she desires sexual activity (Muehlenhard, 1988b). Refusing intercourse when they are willing can cause women to become disconnected from their own desires.

NEW DIRECTIONS FOR RESEARCH

This review of 2 decades of research on heterosexual double standards suggests several directions for future research: a focus on the double standard from the target's perspective; investigation of racial, ethnic, and cultural differences in double standards; and an investment in methodological plurality.

Heterosexual Double Standards From the Target's Perspective

Traditionally, research on heterosexual double standards has focused on the perspective of the perceiver. Little if any empirical research has examined heterosexual double standards from the target's perspective. Perhaps the more interesting question is not whether individuals hold different standards for male and female sexual behavior but what the effects are for individuals who believe that others in their social community endorse double standards and whose linguistic community is replete with discourse representing double standards.

Goffman (1963) defined characteristics as stigmatizing when negative stereotypes about the characteristic exist and the characteristic is devalued by society. Heterosexual double standards and the negative consequences of sexual agency and activity for women suggest that women who are perceived as sexually experienced are stigmatized. Social psychology and sociology have long been interested in the negative affective responses to social stigma. More research is needed on how sexual double standards impact women affectively and behaviorally and on what consequences result from being stigmatized for sexual behaviors (for an anecdotal account, see White, 2001).

Racial-Ethnic and Cultural Diversity

There are several important reasons for studying diverse groups. Heterosexual double standards are locally constructed within communities, yet virtually none of the experimental studies we reviewed looked at how they are constructed and conveyed within diverse communities in the U.S. Interview studies and ethnographies have demonstrated that participants from Hispanic, African American, working-class, or middle-class communities and from different age cohorts may have very different beliefs about sexual double standards. Future research should examine how double standards are culturally constructed and reinforced within other diverse communities.

Methodological Plurality

Gender researchers have long called for the recognition of gender as a set of social processes, not merely an attribute of individuals (Crawford & Marecek, 1989; Kimmel & Crawford, 1999). When gendered attributions are being analyzed, as in the case of research on heterosexual double standards, it is advisable to consider the advantages of many different methods, each of which may be uniquely useful in capturing some aspects of the phenomena in question. Ideally, multiple methods can be employed within the same study, as in Tolman and Szalacha's (1999) comparison of quantitative and qualitative methods in a study of adolescent female sexuality. Methods can also be triangulated, or compared across studies, with an eye to how their results complement and contradict each other.

Sexuality and its expression take place in a larger context of structural gender inequality that characterizes most societies (Crawford & Unger, 2000; Lips, 2003). The social control of sexuality and gender relations are complex phenomena, deeply enmeshed in culturally specific ways of making meaning, and they require sophisticated and varied methods of study.

REFERENCES

Adams, K. L., & Ware, N. C. (1989). Sexism and the English language: The linguistic implications of being a woman. In J. Freeman (Ed.), Women: A feminist perspective (4th ed., pp. 470-484). Mountain View, CA: Mayfield.

Campbell, S. M., Peplau, L. A., & DeBro, S. C. (1992). Women, men, and condoms: Attitudes and experiences of heterosexual college students. Psychology of Women Quarterly, 16, 273-288.

Crawford, M., & Marecek, J. (1989). Psychology reconstructs the female. Psychology of Women Quarterly, 13, 147-166.

Crawford, M., & Unger, R. (2000). Women and gender: A feminist psychology (3rd ed.). New York: McGraw-Hill.

DeLamater, J. (1987). Gender difference in sexual scenarios. In K. Kelley (Ed.), Females, males and sexuality: Theories and research. Albany: State University of New York Press.

DeLamater, J., & MacCorquodale, P. (1979). Premarital sexuality. Madison: University of Wisconsin Press.

Durham, M. G. (1998). Dilemmas of desire: Representations of adolescent sexuality. Youth and Society, 29, 369-400.

Eder, D., Evans, C. C., & Parker, S. (1995). Gender and adolescent culture. New Brunswick, NJ: Rutgers University Press.

Espin, O. M. (1997). Latina realities: Essays on healing, migration and sexuality. Boulder, CO: Westview Press.

Fromme, R. E., & Emihovich, C. (1998). Boys will be boys: Young males' perceptions of women, sexuality and prevention. Education and Urban Society, 30, 172-188.

Fullilove, M. T., Fullilove, R. E., Haynes, K., & Gross, S. (1990). Black women and AIDS prevention: A view towards understanding the gender rules. The Journal of Sex Research, 27, 47-64.

Gentry, M. (1998). The sexual double standard: The influence of number of relationships and level of sexual activity on judgments of women and men. Psychology of Women Quarterly, 22, 505-511.

Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded theory: Strategies for qualitative research. Chicago: Aldine.

Goffman, E. (1963). Stigma: Notes on the management of spoiled identity. Englewood Cliffs, NJ: Prentice-Hall.

Green, S. K., & Sandos, P. (1983). Perceptions of male and female initiators of relationships. Sex Roles, 9, 849-851.

Hillier, L., Harrison, L., & Warr, D. (1998). "When you carry condoms all the boys think you want it": Negotiating competing discourses about safe sex. Journal of Adolescence, 21, 15-29.

Hynie, M., & Lydon, J. E. (1995). Women's perceptions of female contraceptive behavior: Experimental evidence of the sexual double standard. Psychology of Women Quarterly, 19, 563-581.

Jacoby, A. P., & Williams, J. D. (1985). Effects of premarital sexual standards and behavior on dating and marriage desirability. Journal of Marriage and the Family, 47, 1059-1065.

Katz, J., & Farrow, S. (2000). Discrepant self views and young women's sexual and emotional adjustment. Sex Roles, 781-796.

Kimmel, E. B., & Crawford, M. (Eds.). (1999). Innovations in feminist psychological research. Cambridge: Cambridge University Press.

King, K., Balswick, J. O., & Robinson, I. E. (1977). The continuing premarital sexual revolution among college females. Journal of Marriage and the Family, 39, 455-459.

Landrine, H., Klonoff, E. A., & Brown-Collins, A. (1992). Cultural diversity and methodology in feminist psychology: Critique, proposal, empirical example. Psychology of Women Quarterly, 16, 145-163.

L'Armand, K., & Pepitone, A. (1982). Judgments of rape: A study of victim-rapist relationship and victim sexual history. Personality and Social Psychology Bulletin, 8, 134-139.

Lips, H. M. (1981). Women, men, and the psychology of power. Englewood Cliffs, NJ: Prentice-Hall.

Lips, H. M. (2003). A new psychology of women: Gender, culture, and ethnicity (2nd ed.). Mountain View, CA: Mayfield.

Loxley, W. (1996). "Sluts" or "sleazy little animals"?: Young people's difficulties with carrying and using condoms. Journal of Community & Applied Social Psychology, 6, 293-298.

Mark, M. M., & Miller, M. L. (1986). The effects of sexual permissiveness, target gender, subject gender, and attitude toward women on social perception: In search of the double standard. Sex Roles, 15, 311-322.

Martin, K. A. (1996). Puberty, sexuality, and the self: Girls and boys at adolescence. London: Routledge.

Milhausen, R. R., & Herold, E. S. (1999). Does the sexual double standard still exist? Perceptions of university women. The Journal of Sex Research, 36, 361-368.

Moffat, M. (1989). Coming of Age in New Jersey. New Brunswick, NJ: Rutgers University Press.

Muehlenhard, C. L. (1988a). Misinterpreted dating behaviors and the risk of date rape. Journal of Social and Clinical Psychology, 6, 20-37.

Muehlenhard, C. L. (1988b). "Nice women" don't say yes and "real men" don't say no: How miscommunication and the double standard can cause sexual problems. Women and Therapy, 7, 95-108.

Muehlenhard, C. L., & MacNaughton, J. S. (1988). Women's beliefs about women who "lead men on". Journal of Social and Clinical Psychology, 7, 65-79.

Muehlenhard, C. L., & McCoy, M. L. (1991). Double standard/double bind: The sexual double standard and women's communication about sex. Psychology of Women Quarterly, 15, 447-461.

Muehlenhard, C. L., & Rodgers, C. S. (1998). Token resistance to sex: New perspectives on an old stereotype. Psychology of Women Quarterly, 22, 443-463.

Mulkay, M. (1988). On humor. New York: Basil Blackwell.

Munro, P. (1989). Slang U. New York: Harmony Books.

Oliver, M. B., & Hyde, J. S. (1993). Gender differences in sexuality: A meta-analysis. Psychological Bulletin, 114, 29-51.

Oliver, M. B., & Sedikides, C. (1992). Effects of sexual permissiveness on desirability of partner as a function of low and high commitment to relationship. Social Psychology Quarterly, 55, 321-333.

Orenstein, P. (1994). Schoolgirls: Young women, self-esteem, and the confidence gap. New York: Doubleday.

O'Sullivan, L. F. (1995). Less is more: The effects of sexual experience on judgments of men's and women's personality characteristics and relationship desirability. Sex Roles, 33, 159-181.

O'Sullivan, L. F., & Byers, E. S. (1992). College students' incorporation of initiator and restrictor roles in sexual dating relationships. The Journal of Sex Research, 30, 270-282.

Paul, E. L., McManus, B., & Hayes, A. (2000). "Hookups": Characteristics and correlates of college students' spontaneous and anonymous sexual experiences. The Journal of Sex Research, 37, 76-88.

Peplau, L. A., Rubin, Z., & Hill, C. T. (1977). Sexual intimacy in dating relationships. Journal of Social Issues, 33, 86-109.

Reiss, I. L. (1960). Premarital sexual standards in America. New York: The Free Press.

Reiss, I. L. (1964). The scaling of premarital sexual permissiveness. Journal of Marriage and the Family, 26, 188-198.

Reiss, I. L. (1967). The social context of premarital sexual permissiveness. New York: Holt, Rinehart, and Winston.

Rose, S., & Frieze, I. H. (1989). Young singles' scripts for a first date. Gender & Society, 3, 258-268.

Safilios-Rothschild, C. (1977). Love, sex, and sex roles. Englewood Cliffs, NJ: Prentice-Hall.

Sears, D. O. (1986). College sophomores in the laboratory: Influences of a narrow data base on social psychology's view of human nature. Journal of Personality and Social Psychology, 51, 515-530.

Sheeran, P., Spears, R., Abraham, S. C. S., & Abrams, D. (1996). Religiosity, gender, and the double standard. Journal of Psychology, 130, 23-33.

Sprecher, S. (1989). Premarital sexual standards for different categories of individuals. The Journal of Sex Research, 26, 232-248.

Sprecher, S., & Hatfield, E. (1996). Premarital sexual standards among U.S. college students: Comparison with Russian and Japanese students. Archives of Sexual Behavior, 25, 261-288.

Sprecher, S., McKinney, K., & Orbuch, T. L. (1987). Has the double standard disappeared? An experimental test. Social Psychology Quarterly, 50, 24-31.

Sprecher, S., Regan, P. C., McKinney, K., Maxwell, K., & Wazienski, H. (1997). Preferred level of sexual experience in a date or mate: The merger of two methodologies. The Journal of Sex Research, 34, 327-337.

Stanton, B. F., Black, M., Kaljee, L., & Ricardo, I. (1993). Perceptions of sexual behavior among urban early adolescents: Translating theory through focus groups. Journal of Early Adolescence, 13, 44-66.

Sutton, L. A. (1995). Bitches and skankly hobags: The place of women in contemporary slang. In K. Hall & M. Bucholtz (Eds.), Gender articulated: Language and the socially constructed self (pp. 279-296). New York: Routledge.

Thompson, S. (1995). Going all the way: Teenage girls' tales of sex, romance, and pregnancy. New York: Hill and Wang.

Tolman, D. T., & Szalacha, L. A. (1999). Dimensions of desire: Bridging qualitative and quantitative methods in a study of female adolescent sexuality. In E. B. Kimmel & M. Crawford (Eds.), Innovations in feminist psychological research (pp. 7-40). Cambridge, England: Cambridge University Press.

Ward, J. V., & Taylor, J. (1994). Sexuality education for immigrant and minority students: Developing a culturally appropriate curriculum. In J. M. Irvine (Ed.), Sexual cultures and the construction of adolescent identities (pp. 51-68). Philadelphia: Temple University Press. Weatherall, A., & Walton, M. (1999). The metaphorical construction of sexual experience in a speech community of New Zealand university students. British Journal of Social Psychology, 38, 479-498.

White, E. (2001). Fast girls: Teenage tribes and the myth of the slut. New York: Scribner.

Williams, S. S., Kimble, D. L., Covell, N. H., Weiss, L. H., Newton, K. J., Fisher, W. A., et al., (1992). College students use implicit personality theory instead of safer sex. Journal of Applied Social Psychology, 22, 921-933.

Manuscript accepted December 12, 2002

Address correspondence and reprint requests to Mary Crawford, Department of Psychology, 406 Babbidge Road, U-1020, University of Connecticut, Storrs, CT 06269-1020; email: [e-mail].