Sex differences in attitudes toward gay men and lesbians: a multidimensional perspective.

In their recent meta-analysis, Kite and Whitley (1996) affirmed the often-cited finding that men hold more negative attitudes toward homosexuality than do women. They also determined that men's attitudes toward homosexuality are particularly negative when the person being rated is a gay man rather than a lesbian. Their review of the literature also highlighted the complex nature of attitudes toward homosexuality noted by others. The construct of attitudes toward homosexuality is not unidimensional but is composed of several factors including beliefs that gay people are threatening or dangerous, that gays deserve condemnation, and that contact with gay people should be avoided (e.g., Millham, San Miguel, & Kellogg, 1976; Plasek & Allard, 1984). These complexities, and how they influence sex differences in attitudes toward homosexuality, remain largely unexplored. The present research examines two issues: (1) whether sex differences in attitudes toward homosexuality vary by attitude component and (2) whether, within each component, the sex of the person being rated influences these attitudes.

Gender Belief Systems and Attitudes Toward Homosexuality

Our analysis was guided by Kite and Whitley's (1996) proposition that heterosexuals' evaluations of gay men and lesbians are influenced by a generalized gender belief system. According to this model, people's expectations about gender reflect the belief that gender-associated attributes are bipolar: What is masculine is not feminine and vice versa (Deaux & Kite, 1987; see also Bem, 1993). That is, people expect someone who is described by stereotypically masculine traits also to possess stereotypically masculine physical characteristics and to adopt stereotypically masculine roles. Similarly, the knowledge that a person is stereotypically feminine on one dimension leads to the inference that the person is stereotypically feminine on other dimensions (e.g., Deaux & Lewis, 1984; Foushee, Helmreich, & Spence, 1979).

Beliefs about homosexuality are influenced by this generalized gender belief system. Men who are described in stereotypically feminine terms are more likely to be judged homosexual than are men described in stereotypically masculine terms. To a lesser extent, women described in stereotypically masculine terms are more likely to be judged lesbian than women described in stereotypically feminine terms (Deaux & Lewis, 1984). Similarly, gender-based judgments of gay people reflect the belief that male homosexuals are similar to female heterosexuals and that female homosexuals are similar to male heterosexuals (e.g., Kite & Deaux, 1987; Storms, Stivers, Lambers, & Hill, 1981). As Kite (1994) has argued, separating gender-role beliefs from attitudes toward homosexuality is extremely difficult and perhaps impossible.

If gender-role beliefs influence people's perceptions of homosexuality, why would the sexes differ in their evaluations of gay people? One explanation comes from considering the consequences of men's versus women's gender-role violations. Although all sex-role deviants are disliked relative to gender-role conformists (e.g., Laner & Laner, 1979, 1980), gender-role violation is apparently taken more seriously when the violator is a man (e.g., Feinman, 1981; Herek, 1986a; Stockard & Johnson, 1979). Moreover, the linkages between gender-associated characteristics and perceived sexual orientation appear to be stronger for men than for women (e.g., Kite & Deaux, 1987; Page & Yee, 1985). As Herek (1986a) and others (e.g., Oliver & Hyde, 1993; Thomson, Grisanti, & Pleck, 1985) have argued, men's greater gender-role rigidity creates pressure for men to avoid expressing feminine traits or engaging in feminine activities and to reject others (e.g., gay men) who are perceived as gender-role deviants. In contrast, women are allowed more fluid gender roles and are less likely to perceive pressure to reject gender-role nonconformists. That society responds differently to men's and women's gender-role nonconformity can be explained by the generally higher status associated with the American male gender role compared with the American female gender role (Bem, 1993; Lewin & Tragos, 1987, but see McCreary, 1994 for another viewpoint). Simply put, men may have more to lose if they overstep their gender-role boundaries by accepting homosexual behavior. Women, in contrast, may be allowed greater gender-role flexibility and, hence, may be allowed to hold more tolerant attitudes toward gender-role violators. This should be true regardless of the sex of the gender-role violator.

This gender role analysis leads to the prediction that men will be especially condemning of gay men, who violate the male gender role. And men do respond particularly negatively toward gay men (see Kite & Whitley, 1996). In contrast, heterosexual men may see lesbianism in erotic terms, and the positive value associated with this eroticism may positively influence their attitudes toward lesbians (Louderback & Whitley, 1997). More generally, if women's roles are viewed as lower in status than are men's, prejudice toward lesbians should not be as strongly culturally sanctioned as is prejudice toward gay men. Therefore, both sexes may be relatively accepting of lesbians. Unfortunately, attitudes, toward lesbians have rarely been considered in the literature on attitudes toward homosexuality, rendering this prediction tentative. One important aspect of the present research is to consider sex differences in attitudes toward lesbians and gay men and to examine systematically whether the pattern of the interaction between the sex of rater and the sex of target varies by attitude component.

Sex Differences in Attitude Components

Kite and Whitley (1996) argued that gender-associated beliefs should be more likely to influence some components of attitudes toward homosexuality than other components. The results of their meta-analysis showed that ratings of homosexual persons appeared to be based on the gender belief system. In contrast, they argued that the failure to find sex differences in attitudes toward homosexual civil rights occurred because people's ratings on those dimensions were influenced by more global beliefs about civil rights than about gender roles per se. Because attitudes toward homosexuality may serve different functions for different people (e.g., Herek, 1986b), and because different types of judgments about an attitude object can be based on different sources of information (e.g., Esses, Haddock, & Zanna, 1993), a gender role analysis might not be the best theoretical framework for all components of attitudes toward homosexuality. For example, heterosexuals' attitudes toward contact with gay people may be explained by a psychodynamic perspective, which posits that intolerance for homosexuality stems from repression or denial of one's own sexual or homosexual impulses.

To examine these possibilities, we began by reviewing the available literature, including the major scales measuring attitudes toward homosexuality summarized by Beere (1990). We expected that when the attitude was based on gender-role beliefs, the sex difference would be most pronounced. Gender-role beliefs might be reflected in two components of attitudes toward homosexuality: the morality of homosexuality and a generalized condemnation of homosexual persons. These items tap beliefs that homosexuals threaten the social order, such as the institution of the family and American youth, and that gay people are deviant and deserve rejection. Although negative attitudes on these components might also be linked to religious beliefs (Batson & Burris, 1994; Herek, 1988) or to more conservative attitudes in general (Whitley, 1987), these perspectives do not explain why gay men are especially derogated compared with lesbians. We viewed these items as similar to those represented in Kite and Whitley's (1996) analysis of attitudes toward homosexual persons; we therefore expected to replicate their finding that men's ratings are more negative than women's on these items, particularly when the target is a gay man. We also included an assessment of homosexuals' civil rights, expecting (again, following Kite and Whitley) that attitudes toward this dimension would be based on broader beliefs and, therefore, that women and men would be equally accepting of those rights for both gay men and lesbians.

Relatively little attention has been paid to how the sexes view contact with homosexuals, but theoretically this variable should be an important one. Typically, fear of contact is considered from a psychodynamic perspective. That is, heterosexuals' negative attitudes toward lesbians and gay men are thought to stem from a denial of sexual impulses in general or from denial of attraction to same-sex others more specifically. The failure to recognize latent homosexual impulses purportedly produces irrational, negative responses toward gay people (e.g., Adams, Wright, & Lohr, 1996; Morin & Garfinkle, 1978). There is scant evidence for sex differences in this tendency; most research focuses on heterosexual men's defensive reactions to gay men (see also Franklin, 1998). However, contact may be less threatening for women who generally have more experience rejecting sexual advances (e.g., McCormick, 1979). Whitley (1988) found that although heterosexual men and women were equally averse to the possibility of homosexual advances, only men's reactions to such contact were correlated with fear of homosexual advances. Other theoretical perspectives might also be relevant here. A gender-role analysis, for example, could posit that men's rejection of contact with gay men stems from the generalized rejection of all things feminine. Given the lack of research on sex differences in attitudes toward contact with lesbians and gay men, we do not speculate as to which analysis might better account for our data.

Finally, we wanted to explore stereotypic beliefs about lesbians and gay men. Research on stereotypes of gay people suggests that these beliefs are gender-associated: People believe gay men are feminine and lesbians are masculine (e.g., Kite & Deaux, 1987). A gender-role analysis might apply here. If so, men would hold more stereotypic beliefs overall, and particularly about gay men. Although their hypothesis could not be tested with the available data, Kite and Whitley (1996) noticed that when measures of attitudes toward homosexuality included many items assessing homosexual stereotypes, women appeared to be more negative toward lesbians than toward gay men. Hence, this may be one area where women are more negative toward lesbians than toward gay men. Because the few studies that have directly compared the sexes' stereotypes about gay men and lesbians have not found a sex difference (e.g., Kite & Deaux, 1987; Page & Yee, 1985; Taylor, 1983), these possibilities are speculative.

In summary, we hypothesized that men would hold more negative attitudes than would women on the condemnation/tolerance factor and on the morality factor and that men's responses on these items would be particularly negative when they evaluated gay men. In contrast, women were expected to rate gay men and lesbians similarly on these components. We did not expect a sex difference in attitudes toward gay people's civil rights. The available literature on contact toward homosexuals suggested that men would report more negative contact toward gay men than would women. It was less clear how women would respond to the lesbian contact items, so we did not offer a specific hypothesis about this outcome. Finally, we did not make a firm prediction about the women's and men's ratings on the stereotype items.

METHOD

Participants

Two hundred seventy students (133 were women) from Ball State University participated in the study in return for experimental credit for their introductory psychology course. Breakdown of the 96% who reported ethnicity was as follows: 89.8% Caucasian, 1.9% Black, 1.5% Asian American, and 2.6% Hispanic American. Ninety-four percent of participants were of traditional college age (17-22 years). Data from five participants (three women, two men) who reported that they were gay, lesbian, or bisexual were excluded from the analysis.

Questionnaire

A 174-item questionnaire was compiled using questions from previous measures of attitudes toward homosexuality, summarized by Beere (1990), as well as new items. We selected items a priori to represent components of attitudes toward homosexuality that were of conceptual interest: condemnation/tolerance, morality, civil rights, contact, and stereotypes. To accomplish this, we compiled a list of non-overlapping items from the available measures and wrote new items for those categories not well represented by the available measures.(1) We then independently categorized each item as to which component it represented (some were categorized as not of interest); we agreed on 73% of those categorizations. Of the 32 items on which we disagreed, 20 were dropped from the study and 12 were categorized on the basis of discussion. Our final measure included 92 items assessing attitudes toward gay men and lesbians. Of these, 42 items addressed attitudes toward gay men, and 42 parallel items addressed attitudes toward lesbian women. An additional eight items referred to homosexuality or same-sex interaction; these items did not vary by sex of target. Participants responded to these items using a five point Likert scale ranging from I strongly agree to 5 strongly disagree. Participants also reported their gender, race, date of birth and sexual orientation. One female experimenter administered the questionnaire to all participants in groups of 5 to 15 students. Respondents were made aware both verbally and in the written instructions that all responses were anonymous.

RESULTS

Factor Analysis

We first reverse-scored relevant items so that higher numbers always indicated more negative attitudes toward homosexuality. We then submitted all 91 items assessing attitudes toward homosexuality to a varimax factor analysis. Results revealed five factors that accounted for 56.8% of the variance. Factor 1, Condemnation/Tolerance, contained 22 items and accounted for 41.7% of the variance. Factor 2, Morality, contained 23 items and accounted for 5.9% of the variance. Factor 3, Lesbian Contact, contained 14 items and accounted for 3.7% of the variance. A parallel factor, Gay Male Contact, also contained 14 items and accounted for 2.8% of the variance.(2) Factor 5, Stereotypes, contained 14 items and accounted for 2.6% of the variance. These factors represented four of the dimensions we had determined a priori to be important. However, civil rights did not emerge as a separate factor; these items loaded on the Condemnation/Tolerance factor.

Based on these analyses, we formed subscales representing the four identified domains: Condemnation/ Tolerance, Morality, Contact, and Stereotypes. Items were placed on the subscale for which they had the highest loading. Only one item had a factor loading below .35 ("If I were propositioned by a member of my sex, I would know how to respond This item was excluded from the subscales. Seventy-five percent of the factor loadings exceeded .50. With the exception of the Lesbian and Gay Male Contact factors, most parallel items for the gay male and lesbian targets loaded on the same factor. For each of the five parallel items for which this was not the case, we placed the item on the subscale with the highest factor loading. If the loadings were similar, we placed the item on the subscale designated by our conceptual item placement from the item-selection stage.

Because we were interested in comparing attitudes toward gay male and lesbian targets, we constructed a subscale for gay males and a subscale for lesbians for each factor. Hence, for the purposes of analyses, the Gay Male Contact and Lesbian Contact factors were combined. The Morality and Contact subscales also contained three and four neutral items, respectively. These items were combined to form separate subscales. For each subscale, mean scores were computed and used as dependent variables in subsequent analyses. Sample items, number of items, and Cronbach's alpha for each subscale are provided in Table 1. Correlations among the subscales are reported in Table 2.

Table 1. Sample Items, Number of Items, and Cronbach's Alpha for Subscales

Gay Male Condemnation/Tolerance (n = 11 items, Cronbach's alpha = .92)

Gay men should be required to register with the police department where they live.

Gay men should not be discriminated against.

Gay men should not hold responsible positions.

Lesbian Condemnation/Tolerance (n = 11 items, Cronbach's alpha = .89)

Job discrimination against lesbians is wrong.

Lesbians should be allowed to serve in the military.

Apartment complexes should not accept lesbian renters.

Gay Male Morality (n = 10 items, Cronbach's alpha = .92)

The increasing acceptance of gay men in our society is aiding in the deterioration of morals.

Gay men just can't fit into society.

Gay men are a viable part of our society.

Lesbian Morality (n = 10 items, Cronbach's alpha = .93)

Homosexual behavior between two women is just plain wrong.

The idea of lesbian marriages seems ridiculous to me.

Lesbians endanger the institution of the family.

Neutral Morality (n = 3, Cronbach's alpha = .80)

Homosexuality is not sinful.

Homosexuality is a perversion.

I find the thought of homosexual acts disgusting.

Gay Male Contact (n = 14 items, Cronbach's alpha = .96)

It would be upsetting to find out I was alone with a gay man.

I would feel comfortable working closely with a gay man.

I would feel nervous in a group of gay men.

Lesbian Contact (n = 14 items, Cronbach's alpha = .95)

If I knew someone were lesbian, I would go ahead and form a friendship with that individual.

If a lesbian approached me in a public restroom, I would be disgusted.

I enjoy the company of lesbians.

Neutral Contact (n = 4, Cronbach's alpha = .75)

I would feel comfortable knowing I was attractive to members of my sex.

I would feel uncomfortable if a member of my sex made an advance toward me.

I would feel comfortable if I found myself attracted to a member of my sex.

Gay Male Stereotypes (n = 7 items, Cronbach's alpha = .78)

A gay man's mother is probably very domineering.

Most gay men have a life of one night stands.

Most gay men like to dress in opposite-sex clothing.

Lesbian Stereotypes (n = 7 items, Cronbach's alpha = .75)

Most lesbians have some identifiable masculine characteristics.

The love between two lesbians is quite different from the love between two persons of the opposite sex.

Lesbians have weaker sex drives than heterosexuals.

Table 2. Correlations Among Subscales of Attitudes Toward
Homosexuality

                         Gay Condemnation/    Lesbian Condemnation/
                             Tolerance               Tolerance

Gay Condemnation/
Tolerance                        --                     .90

Lesbian Condemnation/
Tolerance                                                --

Gay
Morality

Lesbian
Morality

Gay
Contact

Lesbian
Contact

Neutral
Contact

Gay
Stereotypes
N                               263                     264

                              Gay      Lesbian      Gay     Lesbian
                            Morality   Morality   Contact   Contact

Gay Condemnation/
Tolerance                     .83        .75        .83       .55

Lesbian Condemnation/
Tolerance                     .79        .79        .78       .67

Gay
Morality                       --        .91        .85       .64

Lesbian
Morality                                  --        .76       .75

Gay
Contact                                              --       .61

Lesbian
Contact                                                        --

Neutral
Contact

Gay
Stereotypes
N                             264        264        263       263

                         Neutral         Gay           Lesbian
                         Contact     Stereotypes     Stereotypes

Gay Condemnation/
Tolerance                  .46           .62             .61

Lesbian Condemnation/
Tolerance                  .50           .62             .63

Gay
Morality                   .58           .63             .61

Lesbian
Morality                   .59           .62             .63

Gay
Contact                    .67           .60             .56

Lesbian
Contact                    .63           .54             .54

Neutral
Contact                     --           .42             .42

Gay
Stereotypes                               --             .79
N                          264           264             264

Note. All correlations significant at p < .001.

Fisher's z' transformation revealed significant sex differences in the correlations between the Stereotypes factor for gay men and Condemnation/Tolerance for both gay men (z = 2.76, p [is less than] .01) and lesbians (z = 2.61, p [is less than] .01). In both cases, correlations were stronger for male raters (r = .71 and .70, respectively) than for female raters (r = .50 for both). The correlation between Neutral Contact and Lesbian Contact also differed by participant sex (z = -2.92, p [is less than] .01) with women's correlation (r = .78) exceeding men's (r = .59).

Relation of Participant's Sex and Target's Sex to Attitude Component

Mean subscale scores were subjected to 2 (Participant's Sex) x 2 (Target's Sex) mixed-factor ANOVA, with Target's Sex as the within-subjects factor.' Men's attitudes were more negative than women's on the Condemnation/Tolerance factor, F(1, 261) = 16.41, p [is less than] .0001; the Morality factor, F(1, 261) = 7.09, p [is less than] .0 1; and the Contact factor, F(1, 261) = 4.65, p [is less than] q - 05; but not on the Stereotypes factor, F(1, 258) = 1.41, ns. On all factors, attitudes toward gay men were more negative than attitudes toward lesbians: Condemnation)Tolerance F(1, 261) = 27.45, p [is less than] .0001; Morality F(1, 261) = 26.08, p [is less than] .000 1; Contact F(1, 261) = 32.05, p [is less than] .0001; and Stereotypes F(1, 258) = 8.83, p [is less than] .01.

For three of the four factors, these main effects were qualified by significant Participant's Sex x Target's Sex interactions (see Table 3). For the Condemnation/Tolerance subscale, simple main effects revealed that men evaluated gay men more negatively than lesbians, F(1, 261) = 72.20, p [is less than] .0001, but women evaluated gay men and lesbians similarly, F(1, 261) = 1. 80, ns. The same pattern emerged for the Morality subscale: Men rated gay men more negatively than they did lesbians, F(1, 261) = 77.50, p [is less than] .0001, but women did not differ in their ratings of gay men and lesbians, F(1, 261) = 2.83, ns. For the Contact subscale, a cross-over pattern emerged: men rated contact with gay males more negatively than contact with lesbians, F(1, 261) = 291.86, p [is less than] .0001, whereas women rated contact with lesbians more negatively than contact with gay males, F(1, 261) = 81.57, p [is less than] .0001.

Table 3. Mean Scores on Subscales of Attitudes Toward Homosexuality, by Target's Sex and Participant's Sex

                                                        Women's
                Men's Attitudes   Men's Attitudes       Attitudes
Subscale        Toward Gay Men    Toward Lesbians    Toward Gay Men

Condemnation/        2.60               2.36               2.10
Tolerance            (.87)              (.76)              (.67)
                     3.34               3.07               2.89
Morality             (.93)              (.92)              (.81)
                     3.59               2.80               2.77
Contact              (.93)              (.84)              (.80)
                     3.02               2.91               2.90
Stereotypes          (.60)              (.54)              (.56)

                        Women's
                       Attitudes             F           df
   Subscale         Toward Lesbians

Condemnation/             2.14            46.71(*)      1,261
Tolerance                 (.64)
                          2.94            51.98(*)      1,261
Morality                  (.88)
                          3.19           333.53(*)      1,261
Contact                   (.85)
                          2.87             3.27         1,258
Stereotypes               (.58)

Note. n = 132 women and 131 men for all subscales except Stereotypes (n = 129 men and 129 women). Numbers in parentheses are standard deviations. Higher numbers indicate more negative attitudes toward homosexuality.

(*) p < .0001

Comparisons of the gender-neutral subscales across participant's sex indicated that men held more negative attitudes than did women on both the gender-neutral morality subscale, F(1, 260) 4.96, p [is less than] .05 (men's M = 3.50, SD = 1.06; women's M 3.21, SD = 1.06), and on the gender-neutral contact subscale, F(1, 260) = 6.10, p [is less than] .05 (men's M = 4.18, SD = .80; women's M = 3.92, SD = .88). Finally, because the item "I would know how to respond if a member of my sex made sexual advances toward me" did not load on any factor, it was analyzed separately. Women (M = 2.77, SD = 1.27) reported higher agreement with that item than did men (M = 2.35, SD = 1.16), F(1, 261) = 7.92, p [is less than] .01.

DISCUSSION

We hypothesized that attitude toward homosexuals was a multidimensional construct and that the appearance of sex differences might depend on attitude component. We further hypothesized that, within each component, attitudes might differ toward gay men and lesbians. Our factor analysis revealed four components of attitudes toward homosexuality: Condemnation/Tolerance, Morality, Contact, and Stereotypes. For all factors except Stereotypes, men were less tolerant of homosexuality than were women. Moreover, on all factors except Stereotypes, heterosexual men were less accepting of gay men than of lesbians; heterosexual women reported more negative attitudes only toward contact with lesbians compared with contact with gay men. To understand fully why the sexes differ in their attitudes toward homosexuality, we must consider responses to both gay men and lesbians on these separate components.

Condemnation/Tolerance toward gay people is an important component of most measures of attitudes toward homosexuality (e.g., Herek, 1988; Kite & Deaux, 1987; Larsen, Reed, & Hoffman, 1980). In our factor analysis, items reflecting that dimension accounted for more than 40% of the variance in attitudes toward homosexuality. The items represented by these factors overlap significantly with the types of items that Kite and Whitley (1996) labeled attitudes toward homosexual persons. Following Kite and Whitley (1996), we proposed that generalized gender-role beliefs would account for sex differences on this factor. A gender-role analysis predicts men's rejection of sex-role violators, including gay men. Accordingly, as we expected, we found that men were less tolerant than were women, particularly when they evaluated gay men. In contrast, we hypothesized that women would be allowed greater latitude in their attitudes toward homosexuality and, indeed, heterosexual women were less condemning of both gay men and lesbians. Similarly, as expected, the belief that homosexuality is immoral and violates society's norms produced a parallel set of results. Because the pattern of results between the Condemnation/Tolerance subscale and the Morality factor were similar, and because these factors were highly intercorrelated, it is likely unnecessary to distinguish these two components of attitudes toward homosexuality--at least when sex differences are of interest.

Attitude toward contact with gay men and lesbians emerged as a distinct component of attitudes toward homosexuality, with results that further highlight the need to consider attitudes toward gay men and lesbians separately. Both sexes reported more negative attitudes toward contact with a same-sex gay person than with an other-sex gay person. For men, this result mirrored the results of the Condemnation/Tolerance and morality factors, but for women this outcome represents a departure from those assessments: Only on the Contact component were heterosexual women more negative toward lesbians than toward gay men. It is also noteworthy that this was the only component that produced separate factors for attitudes toward gay men and lesbians.

Attitude toward contact with gay people is typically viewed from a psychodynamic perspective, which posits that heterosexual men in particular are repressing a same-sex attraction and, therefore, display a defensive reaction to gay men by expressing condemnation of homosexuals (e.g., Adams et al. 1996; Morin & Garfinkle, 1978). Heterosexual women's attitudes toward contact with lesbians are rarely considered. And, although Steffensmeier and Steffensmeier (1974) also found that both heterosexual men and women reject contact with a same-sex other, no work we know of explains why heterosexual women, specifically, would fear contact with lesbians. Although Whitley (1988) found only men's rejection of contact with gays to be related to fear of homosexual advances, it is possible that heterosexual women also reject contact with lesbians more than contact with gay men for psychodynamic reasons (e.g., a reaction formation). Gender-role attitudes likely play some role in heterosexuals' attitudes toward contact with gay people, but were this the only explanation, we would expect results more similar to those on the Condemnation/Tolerance factor. Our data cannot test these theoretical possibilities, but they do highlight contact as a distinct component of attitudes toward homosexuality and suggest directions for further study.

To understand fully how the sexes perceive contact with gay people, researchers must consider the complexities of that construct. The contact items specific to gay men and lesbians assessed reactions to potential interaction with a gay person whom the rater may or may not have known previously (e.g., working with a gay person or having a gay person for a neighbor). These interactions were not sexual in nature. The neutral contact items differed in that they are directly linked to potential sexual interaction with a same-sex other--a prospect that is likely more threatening to heterosexuals than other forms of contact. Because of item wording, responses to those items reflect attitudes toward such interactions with members of the raters' own sex; hence, the observed sex difference indicates men's greater rejection of sexual interactions with gay males compared with women's rejection of sexual interactions with lesbians. Men reacted more negatively toward this type of contact, perhaps because such contact raises the possibility that they would find such interactions arousing (Adams et. al., 1996), or because, compared with women, they are less certain about how to reject sexual advances (McCormick, 1979). In our data, women were more likely than men to report knowing how to respond to a sexual advance from a same-sex other.

Finally, consideration of moderators of the predicted relationships was beyond the scope of this article. Consequently, we did not assess whether our respondents actually had contacts with homosexuals. Yet, because heterosexuals who report actually knowing more than one gay person are typically more accepting of gay people than are those who do not report this knowledge (Herek & Glunt, 1993), this factor might mediate the obtained sex differences in attitudes toward contact with homosexuals. This question is of interest for future research.

We expected that attitudes toward gay people's civil rights would emerge as a separate factor and, following earlier meta-analyses (Kite & Whitley, 1996; Oliver and Hyde, 1993), predicted no sex difference on that dimension. Our factor analysis showed that civil rights items loaded on the Condemnation/Tolerance factor, suggesting that raters do not view these constructs as distinct (see Herek, 1988, for similar results). As a further check of our hypothesis, we examined separately the items we had selected a priori to represent attitudes toward homosexual civil rights. Not surprisingly, results followed the same pattern as those reported for the Condemnation/Tolerance factor. One possibility is that men and women view only attitudes toward homosexuals' free speech similarly, but that men are less tolerant than women of other civil rights for gays. Kite and Whitley (1996) provide evidence that the sexes do not differ in their beliefs about homosexuals' free speech, but that men may be less willing than women to grant gays other civil rights. Another possibility is that for the work summarized in the meta-analyses, attitudes toward civil rights was confounded with population; in those studies, civil rights were typically assessed in general survey samples, and other components were examined only in college student samples. Perhaps sex differences in attitudes toward homosexuality are more prominent among college students than among the general population (but see Moore, 1993, for an exception). Finally, sex differences may be less likely to emerge on the single-item measures commonly used to assess attitudes toward gay people's civil rights than on multi-item measures commonly used for other attitude components (see Herek & Capitanio, 1995). Our data cannot disentangle these varied possibilities, but they suggest that male and female college students view homosexuals' civil rights differently, particularly when they evaluate gay men's civil rights.

We found no sex differences on those items reflecting stereotypic beliefs about gay men and lesbians. This is consistent with other research examining these stereotypes (e.g., Kite & Deaux, 1987; Page & Yee, 1985; Taylor, 1983). The relationship between stereotypes of and attitudes toward gay people has received little attention (see Simon, 1998, for an exception), but our findings suggest that this relationship would be similar for women and men. The stereotype items we included reflect gender-associated beliefs (e.g., gay people have characteristics of the other sex), psychodynamic beliefs (e.g., gay people have domineering mothers), and beliefs about gay people's sexuality (e.g., sexually weak, deviant). A more fine-grained analysis of these beliefs might produce sex differences on some items but not others, but we have no direct evidence to support this possibility. Our results did not support Kite and Whitley's (1996) suggestion that women are particularly negative toward lesbians on stereotype items.

Our results support the often-cited contention that it is essential to distinguish between attitudes toward gay men and attitudes toward lesbians (e.g., Herek, 1994). Also, although condemnation/tolerance of gay people is a central aspect of people's attitudes toward homosexuality, other attitude components are also important, with women's and men's responses varying by the attitude component being assessed. Our gender-role analysis accounted for responses to the condemnation/tolerance items, but other theoretical explanations may better account for sex differences in attitudes toward contact with gay people. In addition, factors such as gender-role beliefs and the erotic value attributed to lesbianism (e.g., Louderback & Whitley, 1997) have been found to mediate sex differences in attitudes toward homosexuality. These and other correlates of attitudes toward gay people should be examined as potential mediators of our results. Until more detailed analyses consider responses to these components separately, and until potential confounding factors (e.g., sample characteristics) are ruled out, theoretical explanations for all aspects of attitudes toward homosexuality will be incomplete.

(1) A complete list of items is available upon request.

(2) Four neutral contact items also loaded on the contact factors.

(3) Changes in degrees of freedom are due to missing values that vary by attitude component.

REFERENCES

Adams, H. E., Wright, L. W., & Lohr, B. A. (1996). Is homophobia associated with homosexual arousal? Journal of Abnormal Psychology, 105, 440-445.

Batson, C. D., & Burris, C. T. (1994). Personal religion: Depressant or stimulant of prejudice and discrimination. In M. P. Zanna, & J. M. Olson, (Eds.), The Psychology of Prejudice: The Ontario Symposium, Volume 7 (pp. 149-169). Hillsdale, NJ: Lawrence Erlbaum.

Beere, C. A. (1990). Sex and gender issues: A handbook of tests and measures. New York: Greenwood Press.

Bem, S. L. (1993). The lenses o gender: Transforming the debate on sexual inequality. New Haven, CT: Yale University Press.

Deaux, K., & Kite, M. E. (1987). Thinking about gender. In B. B. Hess and M. M. Ferree (Eds.), Analyzing gender: A handbook of social science research (pp. 92-117). Newbury Park, CA: Sage.

Deaux, K., & Lewis, L. L. (1984). Structure of gender stereotypes: Interrelationships among components and gender label. Journal of Personality and Social Psychology, 46, 991-1004.

Esses, V. M., Haddock, G., & Zanna, M. P. (1993). Values, stereotypes, and emotions as determinants of intergroup attitudes. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and stereotyping: Interactive processes in group perception (pp. 137-166). San Diego, CA: Academic Press.

Feinman, S. (1981). Why is cross-sex-role behavior more approved for girls than for boys? A status characteristic approach. Sex Roles, 7, 289-300.

Foushee, H. C., Helmreich, R. L., & Spence, J. T. (1979). Implicit theories of masculinity and femininity: Dualistic or bipolar? Psychology of Women Quarterly, 3, 259-269.

Franklin, K. (1998). Unassuming motivations: Contextualizing the narratives of anti-gay assailants. In G. M. Herek (Ed.), Psychological perspectives on lesbian and gay issues: Vol. 4. Stigma & sexual orientation: Understanding prejudice against lesbians, gay men, and bisexuals (pp. 1-23). Thousand Oaks, CA: Sage.

Herek, G. M. (1986a). On heterosexual masculinity: Some psychical consequences of the social construction of gender and sexuality. American Behavioral Scientist, 29, 563-577.

Herek, G. M. (1986b). The instrumentality of attitudes: Toward a neofunctional theory. Journal of Social Issues, 42, 99-114.

Herek, G. M. (1988). Heterosexuals' attitudes toward lesbians and gay men: Correlates and gender differences. The Journal of Sex Research, 25, 451-477.

Herek, G. M. (1994). Assessing heterosexuals' attitudes toward lesbians and gay men: A review of empirical research with the ATLG scale. In B. Green and G. M. Herek (Eds.), Psychological perspectives on lesbian and gay issues Vol. 1: Lesbian and gay psychology: Theory, research, and clinical applications (pp. 206-228). Thousand Oaks, CA: Sage.

Herek, G. M., & Capitanio, J. P. (1995). Black heterosexuals' attitudes toward lesbians and gay men in the United States. The Journal of Sex Research, 32, 95-105.

Herek, G. M., & Glunt, E. K. (1993). Interpersonal contact and heterosexuals' attitudes toward gay men: Results from a national survey. The Journal of Sex Research, 30, 239-244.

Kite, M. E. (1994). When perceptions meet reality: Individual differences in reactions to lesbians and gay men. In B. Green & G. M. Herek (Eds.), Psychological perspectives on lesbian and gay issues Vol. 1: Lesbian and gay psychology: Theory, research, and clinical applications (pp. 25-53). Thousand Oaks, CA: Sage.

Kite, M. E., & Deaux, K. (1987). Gender belief systems: Homosexuality and the implicit inversion theory. Psychology of Women Quarterly, 11, 83-96.

Kite, M. E., & Whitley, B. E., Jr. (1996). Sex differences in attitudes toward homosexual persons, behavior, and civil rights: A meta-analysis. Personality and Social Psychology Bulletin, 22, 336-353.

Laner, M. R., & Laner, R. H. (1979). Personal style or sexual preference: Why gay men are disliked. International Review of Modern Sociology, 9, 215-228.

Laner, M. R., & Laner, R. H. (1980). Sexual preference or personal style: Why lesbians are disliked. Journal of Homosexuality, 5, 339-356.

Larsen, K. S., Reed, M., & Hoffman, S. (1980). Attitudes of heterosexuals toward homosexuality: A Likert-type scale and construct validity. The Journal of Sex Research, 16, 245- 257.

Lewin, M., & Tragos, L. M. (1987). Has the feminist movement influenced adolescent sex-role attitudes? A reassessment after a quarter century. Sex Roles, 16, 125-135.

Louderback, L. A., & Whitley, B. E., Jr. (1997). Perceived erotic value of homosexuality and sex-role attitudes as mediators of sex differences in heterosexual college students' attitudes toward lesbians and gay men. The Journal of Sex Research, 34, 175-182.

McCormick, N. B. (1979). Come-ons and put-offs: Unmarried students' strategies for having and avoiding sexual intercourse. Psychology of Women Quarterly, 4, 194-211.

McCreary, D. R. (1994). The male role and avoiding femininity. Sex Roles, 31, 517-531.

Moore, D. W. (1993, April). Public polarized on gay issue. Gallup Poll Monthly, pp. 30-34.

Millham, J., San Miguel, C. L., & Kellogg, R. (1976). A factor-analytical conceptualization of attitudes toward male and female homosexuals. Journal of Homosexuality, 2, 3-10.

Morin, S. F., & Garfinkle, E. M. (1978). Male homophobia. Journal of Social Issues, 34, 29-47.

Newman, B. S. (1989). The relative importance of gender role attitudes to male and female attitudes toward lesbians. Sex Roles, 21, 451-465.

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

Page, S., & Yee, M. (1985). Conception of male and female homosexual stereotypes among university undergraduates. Journal of Homosexuality, 12, 109-118.

Plasek, J. W., & Allard, J. (1984). Misconceptions of homophobia. Journal of Homosexuality, 10, 23-37.

Simon, A. (1998). The relationship between stereotypes of and attitudes toward lesbians and gays. In G. M. Herek (Ed.), Psychological perspectives on lesbian and gay issues: Vol. 4. Stigma and sexual orientation: Understanding prejudice against lesbians, gay men, and bisexuals (pp. 62-81). Thousand Oaks, CA: Sage.

Steffensmeier, D., & Steffensmeier, R. (1974). Sex differences in reactions to homosexuals: Research continuities and further developments. The Journal of Sex Research, 10, 52-67.

Stockard, J., & Johnson, M. M. (1979). The social origins of male dominance. Sex Roles, 5, 199-218.

Storm, M. D., Stivers, M. L., Lambers, S. M., & Hill, C. A. (1981). Sexual scripts for women. Sex Roles, 7, 699-707.

Taylor, A. (1983). Conceptions of masculinity and femininity as a basis for stereotypes of male and female homosexuals. Journal of Homosexuality, 9, 37-53.

Thompson, E. H., Jr., Grisanti, C., & Pleck, J. H. (1985). Attitudes toward the male role and their correlates. Sex Roles, 13, 413-427.

Whitley, B. E., Jr. (1987). The relationship of sex-role orientation to heterosexuals' attitudes toward homosexuals. Sex Roles, 17, 103-113.

Whitley, B. E., Jr. (1988). Sex differences in heterosexuals' attitudes toward homosexuals: It depends on what you ask. The Journal of Sex Research, 24, 287-291.

Manuscript accepted on 7/8/97.

We thank Bernard E. Whitley, Jr. for comments on an earlier version of the manuscript. Portions of this paper were presented at the meeting of the American Psychological Association, Toronto, Canada, August, 1996.

This research supported by an undergraduate fellowship to the first author from Ball State University.

Address correspondence to Mary Kite, Department of Psychological Science, Ball State University, Muncie, IN 47306; e-mail: [e-mail].