A comparison of African American and White college students' affective and attitudinal reactions to lesbian, gay, and bisexual individuals: an exploratory study.

Attitudes and affective reactions to lesbian, gay, and bisexual (LGB) individuals are generally negative throughout the United States. Herek and Glunt (1993), for example, found from a national telephone survey that roughly half the respondents indicated that homosexuality was "perverse" and that gay and lesbian individuals were "disgusting." Various authors have noted that anti-gay attitudes and sentiments may be even more pronounced among African Americans. For example, Fullilove and Fullilove (1999) have commented that "homophobia is very common in the African American community" (p. 1,123). That sentiment was echoed by Kennamer, Honnold, Bradford, and Hendricks (2000), who reported that homophobia appears to be "a major part of the African American culture, driven by both religious forces and political forces" (p. 522).

Knowing more about African Americans' views toward LGBs is important. Such information may aid clinicians and counselors who provide services to African American clients for whom homosexuality is or forms part of their presenting problem(s). Also, knowing more about African Americans' attitudes toward LGBs and homosexuality may provide some insight into the cultural experiences of African American LGBs, who typically must contend and interface with their heterosexual counterparts in the Black community (Herek & Capitanio, 1995). Further, homophobia among African Americans has been criticized for being an impediment of African American LGBs' ability to come to terms with their sexual orientation (Lewis, 2003).

Finally, various critics contend that homophobia among African Americans is partly responsible for slowing African American mobilization against the AIDS epidemic in their communities (e.g., Brandt, 1999; Fullilove & Fullilove, 1999; Morales & Fullilove, 1992; Peterson & Marin, 1988).

A modest number of studies have investigated African Americans' views of LGBs and have yielded contradictory findings. Some of those studies have failed to find significant differences in homphobia between African Americans and Whites (e.g., Glenn & Weaver, 1979; Herek & Capitanio, 1995; Irwin & Thompson, 1977; Marsiglio, 1993). Other studies have found African Americans, on average, to be more homophobic than Whites (e.g., Hudson & Ricketts, 1980; Lewis, 2003; Schneider & Lewis, 1984; Tiemeyer, 1993; Waldner, Sikka, & Baig, 1999).

The more noteworthy of those studies is the one by Lewis (2003), who attempted to compare the opinions of approximately 7,000 African Americans and 43,000 Whites on homosexual relationships, civil liberties for gays and lesbians, and employment rights of homosexuals. Lewis compiled data from 31 national surveys conducted between 1973 and 2000, mostly by news or popular survey organizations, such as the Times Mirror and Gallup polls. His goal was to identify demographic variables, including education and commitment to religion, that may account for racial differences in opinions in these three areas.

The findings were somewhat paradoxical. Even after controlling for frequency of church attendance, education, age, and gender, he found that African Americans were more homophobic than Whites. More specifically, Lewis found that African Americans were 11 percentage points more likely than Whites to condemn homosexual relations as "always wrong" and 14 percentage points more likely than Whites to see LGBs as deserving of "God's punishment" in the form of AIDS. Moreover, African Americans indicated that they would support removing pro-gay books from their public library by 6 percentage points more than Whites and would be less willing to allow an openly gay person make a speech in their community by 4 percentage points more than Whites. Ironically, however, African Americans were more supportive than Whites of gay civil liberties and significantly more opposed to antigay employment discrimination than Whites. Lewis commented that "Blacks appear to be more likely than Whites to both see homosexuality as wrong and to favor gay rights laws" (p. 66), and he interpreted those findings in light of African Americans' historically strong opposition to discrimination in political and economic spheres. Given that religiosity, education, age, and gender did not meaningfully eliminate African Americans' relatively high levels of homophobia, Lewis concluded that additional research is needed to understand the variables at the heart of African Americans' homophobia, particularly as a means for developing more effective, culture-specific campaigns against homophobia.

In light of the inconsistent results obtained from empirical studies examining African American and White homophobia, this study represents an attempt to explore this question further with an eye toward exploring other variables that may influence homophobia among African Americans. We pursued this endeavor by examining African American university students' affective and attitudinal reactions to LGBs vis-a-vis the reactions of a fairly comparable group of Whites and by exploring aspects of African Americans' culture, religious beliefs, and sociodemographics in relation to those reactions.

Culture is believed to influence the way individuals construe their worlds, including both approach and reaction to others and to external events (Angel & Williams, 2000; Hall & Barongan, 2002). Betancourt and Lopez (1993) have suggested that culture be included in any endeavor to understand human behavior in the context of race, ethnicity, and social class. They also indicated that, when ethnic groups are found to differ on study variables, attributing those differences to culture cannot be assumed without having measured aspects of culture thought to be involved in the behavior and controlling for them statistically. We attempted to do this by including an acculturation scale in the set of questionnaires for African Americans. The acculturation scale provides some information about the degree to which African Americans have acculturated toward mainstream White culture, or conversely, have experienced enculturation into traditional African American culture. This allowed us to examine the correlation between African Americans' enculturation/ acculturation and negative reactions to LGBs. It also facilitated the possible disentangling of religiosity from other aspects African American culture regarding their influence of homophobia and homonegativity among African Americans.

Homophobia refers to an affective response, including fear, anxiety, anger, and discomfort, a person experiences when interacting with LGB individuals, which may or may not involve a cognitive component (Adams, Wright, & Lohr, 1996). The cognitive component--sometimes referred to as homonegativity--refers to negative attitudes, beliefs, or actions toward homosexuality (Hudson & Ricketts, 1980). In theory, homophobia and homonegativity are distinct constructs. In reality, however, they likely overlap so much that it is difficult to separate them. In this study, we elected to include a separate measure for each type of negative reaction (affective and cognitive) to LGBs to ensure that we adequately assessed both constructs, and because it is conceivable that individuals may be homophobic but not homonegative, or vice versa.

Because questions designed to measure negative affect and attitudes toward LGBs may elicit socially desirable responses, a measure of social desirability was included in this study so that the response set of social desirability could be taken into consideration in the comparative data analyses. Also, although enculturation and acculturation refer to opposite ends of the same continuum, we elected to use the term enculturation in this paper to reflect our interest in African Americans' immersion and socialization into African American culture. Age, gender, class standing, and the educational attainment of their parents were included in the analyses in order to provide additional descriptive information about the participants and as a means for knowing if the two ethnic groups differed on these variables. Any of these variables on which the two groups differed was treated as a covariate in subsequent analyses. Lastly, two measures of religiosity were included in the study--frequency of church attendance and commitment to religion--because religiosity commonly has been found to predict negative reactions to LGBs (Fisher, Derison, Polley, Cadman, & Johnston, 1994; Larson, Reed, & Hoffman, 1980).

Because of the contradictory empirical literature on African Americans' and Whites' reactions to LGBs, and because of the exploratory nature of this study on how enculturation and religiosity may conjointly influence negative reactions to LGBs among African Americans, we did not make any formal hypotheses.

METHOD

Participants

Seventy African Americans (22 males, 48 females) and 143 non-Hispanic Whites (38 males, 105 females) attending a public university in the southeastern region of the United States participated in the study. Their mean ages were 18.65 years for African Americans (SD = 1.05) and 18.36 years for Whites (SD = .81). Almost all of the participants were freshmen or sophomores, and the vast majority of African Americans (90%) and Whites (94%) reported that their sexual orientation was either exclusively or mostly heterosexual. Also, 89% of African American participants indicated that their religious affiliation was Christianity (48% Baptist, 30% Other Christian, and 11% Catholic), and 76% of White respondents indicated their religion to be Christianity (35% Catholic, 27% Other Christian, and 14% Baptist). All of the participants were from general psychology classes and had volunteered to participate for extra credit.

Instruments

Participants completed the following questionnaires.

Demographic sheet. On a single page, participants indicated age, gender, self-designated ethnicity, and level of education attained by each parent. The number of years of education by whichever parent had the highest education was used in the analyses, serving as an index of socioeconomic status (SES) for each participant. They also indicated their religious affiliation and their class standing based on the number of hours of college courses already completed at the time of the study, using a 4-point scale. Finally, the participants indicated their sexual orientation on a 7-point Likert scale, with responses ranging from 1 (exclusively heterosexual) to 7 (exclusively homosexual).

Religiosity. We used two measures to assess participants' commitment to religion. One measure required respondents to indicate the frequency with which they attended church; this index of religiosity is commonly used by other researchers (e.g., Davidson, Darling, & Norton, 1995). The other measure consisted of 7 items assessing respondents' intrinsic reasons for believing in a religion and was created by Batson, Schoenrade, and Ventis (1993). Among our sample of African American and White participants, this scale obtained Cronbach's alphas of .78 and .92, respectively.

Balanced Inventory of Desirable Responding. The Balanced Inventory of Desirable Responding (BIDR; Paulhus, 1988, 1991) was included in the set of questionnaires to assess the response set of social desirability. The BIDR contains 40 items to which respondents indicate their agreement using a 7-point Likert scale. The BIDR measures two constructs. One construct, self-deceptive enhancement, assesses the tendency to respond honestly to items, but in a positively biased manner. The other construct, impression management, assesses a deliberate self-presentation and can be viewed as a measure of defensiveness. The BIDR was scored in accordance to instructions by Paulhus. Only extreme responses to the items comprised the person's total score, with higher scores reflecting higher levels of socially desirable responses. Reliabilities for the BIDR range from .67 (test-retest; 5-week interval) to .83 (Cronbach's alpha). Further, the BIDR has been found to correlate .71 with the Marlowe-Crowne scale and .80 with the Multidimensional Social Desirability Inventory of Jacobson, Kellogg, Cause, and Slavin (1977). Among our sample of African American and White participants, this scale obtained Cronbach's alphas of .81 and .79, respectively.

Index of Homophobia. All partcipants completed the Index of Homophobia (IHP; Hudson & Ricketts, 1980). This is a 25-item scale assessing one's affective response to gay and lesbian people. Participants respond to items using a 5-point Likert scale. Scores are derived by dividing the total scores on the items by 25, therefore ranging from 1 to 5, with higher scores reflecting higher levels of homophobia. The IHP is the most widely used measure of homophobia (O'Donahue & Caselles, 1993). Based on our sample of African American and White participants, the IHP obtained Cronbach's alphas of .95 and .93, respectively.

Heterosexual Attitudes Toward Homosexuality. All participants completed the Heterosexual Attitudes Toward Homosexuality questionnaire (HATH; Larsen, Reed, & Hoffman, 1980). This is a 20-item scale assessing attitudes and beliefs in response to gay and lesbian people. Participants respond to items using a 5-point Likert scale. Scores are derived by dividing the total scores on the items by 20, therefore ranging from 1 to 4, with higher scores reflecting higher levels of homonegativity. Based on our sample of African American and White participants, the HATH obtained Cronbach's alphas of .94 and .95, respectively.

African American Acculturation Scale--Revised. In addition to the other questionnaires, African American participants completed the African American Acculturation Scale--Revised (AAAS-R; Kloniff & Landrine, 2000). The AAAS-R is a 47-item questionnaire designed to assess the degree to which African American respondents have been socialized or immersed into a traditional African American culture. The AAAS-R is scored by averaging subscale scores, and total scores can range from 1 to 7, with higher scores reflecting more enculturation toward African American culture.

The original AAAS contained 74 items and measured 8 theoretically-derived subscales. The AAAS-R was developed with the goal of reducing the number of items, eliminating items perceived to be offensive to some respondents, and making the instrument empirically derived (based on factor analysis) rather than theoretically derived. Based on factor analysis, the AAAS-R appears to measure eight distinct constructs, thus constituting 8 subscales. They are (1) Religious Beliefs and Practices; (2) Preference for Things African American; (3) Interracial Attitudes; (4) Family Practices; (5) Health Beliefs and Practices; (6) Cultural Superstitions; (7) Segregation; and (8) Family Values. The AAAS-R total score has been found to correlate .97 with the original AAAS and to have good reliability (Cronbach's alpha = .93); the 8 AAAS-R subscales have been found to have Cronbach's alphas ranging from .67 to .89. Based on our sample of African American participants, the AAAS-R (total score) obtained a Cronbach alpha of .82. Cronbach alphas for individual subscales were marginal to good, ranging from .52 (Family Values) to .84 (Interracial Attitudes), with an average alpha of .64.

RESULTS

Table 1 shows the means and standard deviations of the study variables by ethnicity and gender. To determine if African Americans and Whites differed on demographic variables, we performed a MANOVA with ethnicity serving as the independent variable, and age, class standing, SES, frequency of church attendance, religious commitment, and socially desirable responding (BIDR) as dependent variables. Ethnicity was associated significantly with an effect (using Wilks' Lambda, F [6, 210] = 6.50, p = .000; [eta.sup.2] = .157). Subsequent univariate analyses clarified the MANOVA results and indicated that, relative to Whites, African Americans attended church more frequently (F [1, 215] = 15.68, p = .000; [eta.sup.2] = .068), were more committed to their religion (F [1, 215] = 27.13, p = .000; [eta.sup.2] = .112), and had significantly lower SES (F [1, 215] = 4.26, p < .05; [eta.sup.2] = .019).

To determine if African Americans would obtain higher scores than Whites on homophobia (IHP) and homonegativity (HATH), and if men and women differed on these measures, we performed a MANOVA with ethnicity and gender serving as independent variables and scores on IHP and HATH as the dependent variables (it is noted here that IHP and HATH significantly correlated with each other for both African Americans [r = .83, p < .01] and for Whites [r = .84, p < .01]). Ethnicity was not significantly associated with an effect on IHP and HATH (F [2, 232] = 2.35, p = .09). Despite that ethnicity, overall, had only approached significance, univariate tests for each dependent measure separately indicated that African Americans had higher IHP scores (F [1, 233] = 4.56, p < .05; [eta.sup.2] = .019) and HATH scores (F [1, 233] = 3.93, p < .05; [eta.sup.2] = .017) than Whites. Gender was significantly associated with an effect on IHP and HATH (F [2, 232] = 7.26, p = .001; [eta.sup.2] = .059). Men had significantly higher IHP scores (F [1, 233] = 13.60, p = .000; [eta.sup.2] = .055) and HATH scores (F [1, 233] = 12.91, p = .000; [eta.sup.2] = .017) than women. There was no significant ethnicity by gender interaction (F [1, 232] = .37, ns).

Because of the significant univariate tests on IHP and HATH between African Americans and Whites and the fact that the two ethnic groups differed significantly on frequency of church attendance, commitment to religion, and SES, we performed a MANCOVA with ethnicity and gender as independent variables and IHP and HATH serving as dependent variables. Frequency of church attendance, commitment to religion, and SES were entered as covariates. Ethnicity was not associated significantly with an effect on IHP and HATH (F [2, 209] = .87, ns). With the covariates entered, neither of the univariate tests was statistically significant. Gender continued to be associated significantly with an effect on IHP and HATH (F [2, 209] = 14.78, p = .000; [eta.sup.2] = .124), with men having higher IHP scores (F [1, 210] = 26.60, p = .000; [eta.sup.2] = .112) and HATH scores (F [1, 210] = 25.83, p = .000; [eta.sup.2] =. 110) than women. No significant ethnicity by gender interaction effect emerged (F [2, 209] = .49, ns).

Consistent with one of the main purposes of this study, which was to explore the confluence of religiosity and cultural variables on negative reactions to LGBs, we performed an additional series of regressions. Specifically, homophobia (IHP) and homonegativity (HATH) were regressed on the study variables (age, gender, class standing, SES, socially desirable response set [BIDR], frequency of church attendance, and religious commitment) separately for each ethnic group (see Tables 2 and 3). For both African Americans and Whites, gender and religious commitment significantly predicted IHP (Multiple Rs = .52 and .62, respectively [ps < .01]). For African Americans, gender and frequency of church attendance significantly predicted HATH (Multiple R = .52, p < .01), whereas for Whites, gender and religious commitment significantly predicted HATH (Multiple R = .63, p < 01). That is, in addition to gender, African Americans' religious commitment significantly predicted negative affective responses to LGBs, but frequency of church attendance significantly predicted negative attitudes toward LGBs. For Whites, in addition to gender, religious commitment significantly predicted both negative affective responses and negative attitudes toward LGBs.

To explore further the relative influence of religiosity and the potential influence of cultural variables on African Americans' negative reactions to LGBs, we conducted a series of regression analyses with the African Americans' data. First, we performed two separate standard multiple regressions predicting IHP, then HATH, from the 8 enculturation subscales of the AAAS-R (see Tables 4 and 5). The AAAS-R subscales significantly predicated both IHP (Multiple R = .46, p < .05) and HATH (Multiple R = .46, p < .05). Only one AAAS-R subscale, Family Practices, significantly contributed to the prediction of IHP (standardized B coefficient = .326, p < .05). Two AAAS-R subscales, Preference for Things of African American and Interracial Attitudes, significantly predicted HATH (standardized B coefficients = -.300 and .264, ps < .05, respectively).

For the purpose of specifically distilling religiosity from cultural variables, we then performed two stepwise multiple regressions, predicting IHR then HATH, from 7 of the AAAS-R subscales (see Tables 6 and 7). Frequency of church attendance, religious commitment, and the AAASR subscale Religious Beliefs and Practices were controlled for statistically by force-entering them into the analysis at Step zero. For the prediction of IHP, after Step one, with Family Practices in the equation, [R.sup.2] = .27, F (4, 68) = 6.20, p = .000. After Step one, adding the remaining AAAS-R subscales did not result in a significant increase in [R.sup.2]. For the prediction of HATH, none of the 7 AAASR subscales resulted in a significant increase in R2 beyond the three religiosity variables that were force-entered into the equation. Stated differently, none of the cultural variables measured by AAAS-R predicted homonegativity after controlling for religiosity; one cultural variable, Family Practices (which is explained more in the Discussion), significantly predicted homophobia after religiosity was controlled for statistically. Finally, to confirm the predictive significance of Family Practices to homophobia, we conducted a Pearson product-moment correlation analysis between IHP and Family Practices while partialing out the three religiosity variables (frequency of church attendance, religious commitment, and Religious Beliefs and Practices). The correlation was statistically significant (r = .31, p < .01).

DISCUSSION

The initial results of our analyses suggested that African American college students may have slightly more negative affective and attitudinal reactions to LGBs than White students. However, additional analyses clarified those results by statistically controlling for three variables on which the ethnic groups had differed significantly. Specifically, after frequency of church attendance, religious commitment, and SES were taken into consideration, any previously suggested ethnic differences in homophobia and homonegativity vanished. These results are consistent with previous studies (e.g., Glen & Weaver, 1979; Herek & Capitanio, 1995; Irwin & Thompson, 1977; Marsiglio, 1993) in which no appreciable difference in negative reactions to LGBs had been found between African Americans and Whites.

That notwithstanding, other empirical and anecdotal evidence suggests that anti-gay sentiments may be more prevalent among African Americans than among Whites. One of the primary reasons for conducting the current study was to explore how religiosity and enculturation toward the African American culture may underlie some, if not much, of African Americans' negative reactions to LGBs. With that goal in mind, we explored if sociodemographic and religiosity variables would significantly predict homophobia and homonegativity. For both African Americans and Whites, gender and religious commitment predicted homophobia. For African Americans, gender and frequency of church attendance predicted homonegativity (for Whites, gender and religious commitment predicted homonegativity). In essence, it was African Americans' commitment to their religious faith that correlated most with how they respond affectively toward LGBs: the more they attended church, the more negative their attitudes were toward LGBs. Knowing the fallacy of assuming causality with correlational data, we can only speculate that, for African Americans, repeated church attendance possibly reinforces negative attitudes about LGBs.

We also found that the more African Americans were immersed and socialized into the African American community or culture, the more they expressed homophobia and homonegativity. As a means to disentangle the influence of religiosity from enculturation on reactions to LGBs, we conducted stepwise regressions in which the three religiosity variables (frequency of church attendance, religious commitment, and the one AAAS-R subscale that assessed religious beliefs and practices) were partialed out of the regression of the enculturation subscales on homophobia and homonegativity, respectively. The results indicated that, after controlling for religiosity, none of the enculturation subscales predicted homonegativity. However, after controlling for religiosity, one enculturation subscale, Family Practices, persistently predicted homophobia.

The items constituting Family Practices warrant closer examination in order to possibly glean some understanding of what aspects of African American culture might be involved with homophobic sentiments. The Family Practices subscale consists of 4 items assessing participation in traditional African American family practices. Specifically, the items ask if respondents shared a bed with siblings or other relatives when they were children, if they stayed with relatives for days or weeks and then returned home to their parents, if they had extended family members living in their home while growing up, and if, as children, they bathed with siblings, parents, or other relatives. At the surface, it seems difficult to understand the linkage between such practices and subsequent homophobia. A psychodynamic explanation for this finding is that early childhood experiences that include sharing a bed and bathing with others, particularly of the same gender, may arouse unwanted sexual impulses of a homoerotic nature. Such experiences might evoke negative emotions in response to LGBs later in life as a defense against earlier impulses (Christopher, 2004; Garnets & Kimmel, 2003). Another possible explanation, more cognitive-behaviorally based, is that such intrafamilial experiences might reflect a closeness within the family in which children are highly socialized by parental and family values, including their parents' biases. Stated more directly, parents who maintain close relationships with their children likely are more successful at transmitting their personal values and prejudices to their children (Baker, 2002; Holden, 1997; Sheinberg, 2004). Additional studies are needed to explicate the correlation between Family Practices and homophobia.

CONCLUSION

Lewis (2003) has indicated that, in light of the relative pervasiveness of homophobia within African American communities, it is time to achieve a greater "understanding of the roots of Black attitudes toward homosexuality" (p. 75). The results of this study suggest that religiosity may be the most critical component of African Americans' negative reactions to LGBs. These results also suggest that aspects of early childhood experiences may be involved in those reactions.

On one hand, we do not contend in this paper that negative affective and attitudinal responses to LGBs are the sole province of African Americans. Indeed, the social scientific literature has consistently documented that most ethnic groups in the United States are not very accepting of homosexuality. On the other hand, it may be unrealistic to believe that all ethnic groups are equally unaccepting of homosexuality. Cultures in different parts of the world vary in their acceptance levels of LGBs (see Duberman, Vicinus, & Chauncey [1989] for an extensive review). It seems reasonable to speculate that subcultural variations in homophobia and homonegativity may exist among diverse ethnic groups within the United States as well.

The African American community, like most ethnic communities, has little to lose and much to gain by reexamining its intolerance toward homosexuality. As indicated earlier, in addition to causing needless suffering among LGB African Americans who already must contend with racism and discrimination for being Black, African Americans' general lack of acceptance of homosexuality has been implicated in the country's struggle to curb HIV infection among African Americans.

With current movements to permit gays and lesbians to marry in various regions of the country, it seems that the United States may be, albeit slowly, becoming more accepting of homosexuality. It is hoped that this gradual tolerance for LGBs will inevitably penetrate subcultures within the United States and as a result, ethnic groups that traditionally may have not accepted homosexuality will move toward more acceptance of diversity with respect to LGB people.

Manuscript accepted March 8, 2005

REFERENCES

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

Angel, R. J., & Williams, K. (2000). Cultural models of health and illness. In I. Cuellar & F. A. Paniagua (Eds.), Handbook of multicultural mental health: Assessment and treatment of diverse populations. San Diego, CA: Academic Press.

Baker, J. M. (2002). How homophobia hurts children: Nurturing diversity at home, at school, and in the community. New York: Harrington Park Press.

Batson, C. D., Schoenrade, P., & Ventis, W. L. (1993). Religion and the individual: A social-psychological perspective. New York: Oxford University Press.

Betancourt, H., & Lopez, S. R. (1993). The study of culture, ethnicity, and race in American Psychology. American Psychologist, 48, 629-637.

Brandt, E. (Ed.) (1999). Dangerous Liaisons: Blacks, gays, and the struggle for equity. New York: New Press.

Christopher, E. (2004). Sexuality: Psychoanalytic perspectives. Psychodynamic Practice: Individuals, Groups, & Organisations, 10, 544-549.

Davidson, J. K., Darling, C. A., & Norton, L. (1995). Religiosity and the sexuality of women: Sexual behavior and sexual satisfaction revisited. The Journal of Sex Research, 32, 235-243.

Duberman, M. B., Vicinus, M., & Chauncey, Jr., G. (Eds.). 1989). Hidden from History: Reclaiming the gay & lesbian past. Ontario, Canada: NAL Books.

Fisher, R. D., Derison, D., Polley, C., F. III, Cadman, J., & Johnston, D. (1994). Religiousness, religious orientation, and attitudes towards gays and lesbians. Journal of Applied Social Psychology, 24, 614-630.

Fullilove, M. T., & Fullilove, R. E. III. (1999). Stigma as an obstacle to AIDS action. American Behavioral Scientist, 42, 1,117-1,129.

Garnets, L. D., & Kimmel, D. C. (Eds.). (2003). Psychological perspectives on lesbian, gay, and bisexual experiences (2nd ed.). New York: Columbia University Press.

Glenn, H. D., & Weaver, C. N. (1979). Attitudes toward premarital, extramarital and homosexual relations in the U.S. in the 1970s. The Journal of Sex Research, 15, 108-118.

Hall, G. C. N., & Barongan, C. (2002). Multicultural psychology. Upper Saddle River, NJ: Prentice Hall.

Herek, G. M., & Capitanio, J. P. (1995). Black heterosexuals' attitutudes 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.

Holden, G. W. (1997). Parents and the dynamics of child rearing. Boulder, CO: Westview Press.

Hudson, W. W., & Ricketts, W. A. (1980). A strategy for the measurement of homophobia. Journal of Homosexuality, 5, 357-372.

Irwin, P., & Thompson, N. L. (1977). Acceptance of the rights of homosexuals: A social profile. Journal of Homosexuality, 3, 107-121.

Jacobson, L. L., Kellogg, R. W., Cauce, A. M., & Slavin, R. (1977). A multidimensional social desirability inventory. Bulletin of the Psychonomic Society, 9, 109-110.

Kennamer, J. D., Honnold, J., Bradford, J., & Hendricks, M. (2000). Differences in disclosure of sexuality among African American and White gay/bisexual men: Implications for HIV/AIDS prevention. AIDS Education and Prevention, 12, 519-531.

Klonoff, E. A., & Landrine, H. (2000). Revising and improving the African American acculturation scale. Journal of Black Psychology, 26, 235-261.

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.

Lewis, G. B. (2003). Black-White differences in attitudes toward homosexuality and gay rights. Public Opinion Quarterly, 67, 59-78.

Marsiglio, W. (1993). Attitudes toward homosexual activity and gays as friends: A national survey of heterosexual 15- to 19-year old males. The Journal of Sex Research, 30, 12-17.

Morales, E. S., & Fullilove, M. T. (1992). "Many are called": Participation by minority leaders in an AIDS intervention in San Francisco. Ethnicity and Disease, 2, 389-401.

O'Donahue, W., & Caselles, C. E. (1993). Homophobia: Conceptual, definitional, and value issues. Journal of Psychopathology and Behavioral Assessment, 15, 177-195.

Paulhus, D. L. (1988). Assessing self-deception and impression management in self-reports: The Balanced Inventory of Desirable Responding. Unpublished manual, University of British Columbia, Vancouver, Canada.

Paulhus, D. L. (1991). Measurement and control of response bias. In J. P. Robinson, P. R. Shaver, and L. S. Wrightsman (Eds.), Measures of personality and social psychological attitudes (Volume 1). San Diego: Academic Press, Inc.

Peterson, J. L., & Marin, G. (1988). Issues in the prevention of AIDS among Black and Hispanic men. American Psychologist, 43, 871-877.

Schneider, W., & Lewis, I. A. (1984, February/March). The straight story on homosexuality and gay rights. Public Opinion, 7, 16-20, 59-60.

Sheinberg, N. (2004). How should I raise my child? Assessing the parental beliefs and practices of Latino and African American mothers in the context of a parenting intervention program. Unpublished doctoral dissertation, Harvard University, Cambridge, MA.

Tiemeyer, P. E. (1993). Relevant public opinion. In Sexual Orientation and U.S. Military Personnel Policy: Options and Assessments. National Defense Research Institute. Santa Monica, CA: RAND.

Waldner, L, K., Sikka, A., & Baig, S. (1999). Ethnicity and sex differences in university students' knowledge of AIDS, fear of AIDS, and homophobia. Journal of Homosexuality, 37, 117-133.

Charles Negy

University of Central Florida

Russell Eisenman

University of Texas--Pan American

Address correspondence to Charles Negy, Department of Psychology, University of Central Florida, Orlando, FL, 32816; e-mail: cnegy@pegasus. cc.ucf.edu.

Table 1. Means and Standard Deviations of Study
Variables by Ethnicity and Gender

                       Age                           Class
Ethnicity            (years)        SES (a)        Standing

African American      M (SD)         M (SD)         M (SD)

  Males            18.67 (1.02)   15.62 (2.13)    1.24 (.70)
   (n = 22)
  Females          18.64 (1.07)   14.62 (2.45)    1.26 (.53)
   (n = 48)
  Combined         18.65 (1.05)   14.93 (2.39)    1.25 (.58)
   (n = 70)

White
  Males            18.59 (.90)    15.59 (2.19)    1.24 (.68)
   (n= 38)
  Females          18.28 (.77)    15.74 (2.42)    1.17 (.47)
   (n = 105)
  Combined         18.36 (.81)    15.70 (2.36)    1.19 (.53)
   (n = 143)

Effects
  Ethnicity
  F values (b)         3.76          4.26 *           .57
  Gender
  F values (c)         n/a            n/a             n/a

                    Frequency
                    of Church      Religious
Ethnicity           Attendance     Commitment        BIDR

African American      M (SD)         M (SD)         M (SD)

  Males            3.33 (1.62)     3.80 (.57)    10.29 (5.33)
   (n = 22)
  Females          3.70 (1.40)     4.07 (.68)    11.57 (6.09)
   (n = 48)
  Combined         3.59 (1.47)     3.99 (.65)    11.18 (5.86)
   (n = 70)

White
  Males            2.78 (1.53)     3.32 (1.07)   11.68 (5.56)
   (n= 38)
  Females          2.67 (1.52)     3.37 (.91)    10.65 (5.35)
   (n = 105)
  Combined         2.70 (1.52)     3.36 (.96)    10.92 (5.41)
   (n = 143)

Effects
  Ethnicity
  F values (b)      15.68 ***      27.13 ***         0.00
  Gender
  F values (c)         n/a            n/a             n/a

Ethnicity              IHP            HATH

African American      M (SD)         M (SD)

  Males             3.31 (.90)     2.80 (.94)
   (n = 22)
  Females           3.08 (.78)     2.48 (.71)
   (n = 48)
  Combined          3.15 (.82)     2.57 (.80)
   (n = 70)

White
  Males             3.34 (.73)     2.84 (.92)
   (n= 38)
  Females           2.75 (.65)     2.17 (.74)
   (n = 105)
  Combined          2.90 (.72)     2.35 (.84)
   (n = 143)

Effects
  Ethnicity
  F values (b)        .53 (d)       1.62 (d)
  Gender
  F values (c)      26.60 ***      25.83 ***

Note. BIDR = Balanced Inventory of Desirable Responding;

IHP = Index of Homophobia; HATH = Heterosexuals' Attitudes
Toward Homosexuality.

(a) SES was based on whichever parent had the highest years of education.

(b) df for MANOVA with demographic variables as DVs were 6, 210.
df for MANOVA with IHP and HATH as DVs were 2, 232.

(c) df for MANOVA with IHP and HATH as DVs were 2, 232.

(d) F values after controlling for demographic variables.

* p <.05 *** p <.001

Table 2. Standard Multiple Regression of Demographic
Variables on IHP (AA = African Americans; WA = White Americans)

Predictor                            B
Variables

                             AA            WA

Gender                     -.439         -.645
Class Standing             -.355          .000
SES                         .000          .000
Church Frequency            .126          .000
Religious Commitment        .397          .477
BIDR                       -.000         -.000
Age                         .149          .000

Predictor                            Beta
Variables                      (standardized)

                             AA            WA

Gender                     -.246          -.395
Class Standing             -.245          -.052
SES                        -.078          -.026
Church Frequency            .224          -.199
Religious Commitment        .317           .633
BIDR                       -.103          -.039
Age                         .188           .004

Predictor                               t-test values
Variables

                             AA            WA

Gender                     -2.17 *       -5.72 ***
Class Standing             -1.38          -.53
SES                         -.69          -.37
Church Frequency            1.54         -1.80
Religious Commitment        2.24 *        5.68 ***
BIDR                        -.903         -.550
Age                         1.06           .03

R value =                   .52 **         .62 **
[R.sub.2] Adjusted =        .27            .39
[R.sub.2] =                 .19            .36

* p <.05 ** p < .01 *** p < .001

Table 3. Standard Multiple Regression of Demographic
Americans)

Predictor Variables                   B
                             AA            WA

Gender                     -.425         -.676
Class Standing             -.130          .000
SES                         .000          .000
Church Frequency            .176          .000
Religious Commitment        .253          .434
BIDR                        .000          .000
Age                         .000          .000

Predictor Variables                  Beta
                                     (standardized)
                             AA            WA

Gender                     -.248          -.356
Class Standing             -.093          -.026
SES                        -.006          -.055
Church Frequency            .326           .024
Religious Commitment        .210           .496
BIDR                       -.051          -.093
Age                         .125           .015

Predictor Variables        t-test        values
                             AA            WA

Gender                    -2.19 *        -5.19 ***
Class Standing             -.52           -.27
SES                        -.05           -.80
Church Frequency           2.24*           .22
Religious Commitment       1.49           4.48 ***
BIDR                       -.45          -1.34
Age                         .70            .15
R value =                   .52 **         .63 **
[R.sup.2] Adjusted =        .27            .39
[R.sup.2] =                 .19            .36

* p < .05 ** p < .01 *** p < .001

Table 4. Standard Multiple Regression of African American
Acculturation Scale-Revised (HAAS-R) on IHP

Predictor Variables                         Beta        t-test
(AAAS-R Subscales)               B     (standardized)   values

Religious Beliefs/Practices     .000        .159        1.35
Preference for African Things   .000        .026         .92
Interracial Attitudes           .000        .071         .58
Family Practices                .148        .326        2.63 *
Health Beliefs/Practices        .000        .083         .51
Cultural Superstitions          .000        .045         .31
Segregation                     .000       -.148       -1.15
Family Values                   .000        .037         .29
                                   R =                   .46 *
                                   [R.sub.2] =           .21
                                   [R.sub.2] Adjusted =  .13
* p < .05

Table 5. Standard Multiple Regression of African American
Acculturation Scale-Revised (AAAS-R) on HATH

Predictor
Variables
(AAAS-R                                Beta           t-test
Subscales)               B        (standardized)      values

Religious
  Beliefs/
  Practices            .114            .210            1.78
Preference for
  African Things      -.159           -.300           -2.20 *
Interracial
  Attitudes            .167            .264            2.16 *
Family
  Practices            .000            .203            1.64
Health Beliefs/
  Practices            .000            .125             .77
Cultural
  Superstitions        -.000          -.062            -.42
Segregation            -.000          -.019            -.15
Family Values           .000           .066             .51
                                R =                     .46 *
                                [R.sub.2] =             .21
                                [R.sub.2] Adjusted =    .11
* p < .05

Table 6. Stepwise Regression Analysis predicting IIIP from
African American Acculturation Scale-Revised (AAAS-R)

                              Cumulative   [R.sup.2]        Beta
Step   Variable Entered       [R.sup.2]     Change     (standardized)

1 (a)  Church Frequency                                    0.232
       Religious Commitment                                0.225
       Religious Beliefs/
         Practices (b)          .19 **       0.19          0.039
2      Family Practices         .27 ***      0.08          0.293

(a) Church Frequency, Religious Cornrnitment, and Religious Beliefs/
Practices were forced entered into the equation at Step One.

(b) Religious Beliefs/Practices is an AAAS-R subscale.

** P < .01

*** P = .000

Table 7. Stepwise Regression Analysis
predicting HATH from African American
Acculturation Scale-Revised
(AAAS-R)
                              Cumulative   [R.sup.2]        Beta
Step   Variable Entered       [R.sup.2]     Change     (standardized)

1 (a)  Church Frequency                                    0.360
       Religious Commitment                                0.154
       Religious Beliefs/
         Practices (b)          .21 *         .21         -0.023

(a) Church Frequency, Religious Commitment, and Religious Beliefs/
Practices were forced entered into the equation at Step One.

(b) Religious Beliefs/Practices is an AAAS-R subscale.

* p < .05