Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2014 Sep 4.
Published in final edited form as: J Homosex. 2007;53(4):135–161. doi: 10.1080/00918360802103399

Sexual Prejudice among Puerto Rican young adults

José A Bauermeister 1, Mercedes M Morales 2, Gretchen Seda 3, Milagritos González-Rivera 4
PMCID: PMC4154190  NIHMSID: NIHMS621495  PMID: 18689195

Abstract

Sexual prejudice is linked to hate crimes, mental health, risk behaviors, and stigma. Few studies have examined sexual prejudice among Latinos. We surveyed 382 college students in Puerto Rico. A structural model tested whether contact and positive experiences with homosexuals, perceived similarities with peers' attitudes toward homosexuality, and religiosity were predictive of sexual prejudice among Puerto Rican young adults. Sex differences in the structural model were explored. With the exception of peers' attitudes toward homosexuality, all study variables predict sexual prejudice. No sex differences were found. Implications for decreasing sexual prejudice among Puerto Rican youth in a college setting are discussed.

Keywords: Sexual prejudice, discrimination, homophobia, homosexuality, gender, gay & lesbian, college, Latino, attitudes, religion

Introduction

Perceptions of homosexuality play a crucial role in shaping the social and political climate in society and affect how policies at the local and national level are created to protect or marginalize homosexual men and women. Current public debates and legislation on the inclusion of statutes against discrimination due to sexual orientation, hate crimes, and gay marriage and civil unions highlight the controversial range of beliefs and attitudes regarding the civil rights granted to homosexual populations in society. National surveys in the United States have found high levels of sexual prejudice against homosexuals (Herek & Capitanio, 1999; Herek & Capitanio, 1996). Herek (2000a) defined sexual prejudice as negative attitudes toward an individual because of his/her sexual orientation. Researchers studying sexual minorities have found prejudice increases the inequalities between homosexual and heterosexual individuals' educational achievement outcomes (Russell, Seif, Truong, 2001) and wage earnings (Badgett, 1995). Furthermore, prejudice toward homosexual men and women has been linked to increased incidence of hate crimes (Garofalo, 1997; Herek, Gills, & Cogan, 1999; Herek, 1989), mental health problems (Mays & Cochran, 2001; Rosario, Hunter, Maguen, Gwadz, & Smith, 2001), risk behaviors like drug use and unprotected sexual intercourse (Meyer & Dean, 1998), and stigma (Díaz & Ayala, 2001; Peterson & Carballo-Diéguez, 2000; Grossman, 2001; Cole, Kemeny, Taylor, & Visscher, 1996).

Researchers have found consistent associations between sexual prejudice and heterosexuals' frequency and appraisal of interactions with homosexual men and women, perceived similarity with peer attitudes toward homosexuality, and religiosity (Herek, 2004; Sherrod & Nardi, 1998; Sakalli & Ugurlu, 2001; Altemeyer, 2001). Sexual prejudice decreases when individuals know more lesbians and gay men and believe their interactions to be positive. On the other hand, individuals more likely to comply with peers' negative beliefs about homosexuality tend to report higher levels of sexual prejudice. Aside from these normative beliefs, heterosexuals' attitudes toward homosexuality have been found to be strongly associated with respondents' religiosity. The condemnation of homosexuality in most religious traditions promotes increased negative appraisals of gay men and lesbians. Consequently, researchers have consistently found increased sexual prejudice among individuals who have stronger religious beliefs (Herek, 2004). Nonetheless, the strength of these associations have varied by respondents' sociodemographic characteristics.

Researchers have documented sex differences in the associations between sexual prejudice and its correlates (Herek, 2002a; Herek, 2000b). Herek (2000a Herek (2002b) found females reported lower negative attitudes toward gay men and lesbians than their male counterparts. Among males, negative attitudes toward gay men were stronger than those held toward lesbians. These same findings have been replicated in other studies (Sakalli, 2002; Sherrod & Nardi, 1998). Herek (2002a, p. 60) suggests that, at the individual level, sexual prejudice sex differences may illustrate differing psychological functions:

“To the extent that heterosexual men's attitudes toward gay men reflect their fear of being perceived as homosexual and their concerns about being the target of attraction of a gay man (Herek 1986; Kimmel 1997; Kite and Whitley 1998; Loudberback and Whitley 1997), they are likely to function as psychological defenses against feelings of anxiety and threat. Such defenses might not play a role in heterosexual men's attitudes toward lesbians. Instead, the latter may be shaped mainly by men's attitudes toward women generally and their interest in lesbian sexuality (Kite and Whitley 1998; Louderback and Whitley 1997).”

Ethnographic research on Latinos' understanding of homosexuality suggests Latino gay men and lesbians are stigmatized continuously because of their sexual orientation (Díaz & Ayala, 2001; Díaz, 1998; Carrier, 1995). Latinos' attitudes toward gay men and lesbians have been compared against other racial and ethnic groups in the U.S. (Sherrod & Nardi, 1998). Nonetheless, limited studies have examined the psychological components of attitudes toward homosexuality among Latinos in the U.S. and even fewer studies have acknowledged and explored potential cultural variations across different Latino ethnic groups (Herek & González-Rivera, in press; Carballo-Diéguez, 1998). Moreover, the absence of culturally appropriate psychometric instruments for different Latino populations in the U.S. has handicapped efforts to measure sexual prejudice effectively (Herek & González-Rivera, in press; Carballo-Diéguez, 1989).

This study focuses on a representative sample of college students attending a public undergraduate institution in Puerto Rico. The motives in selecting college students as a population are threefold: (a) college students tend to have open and active sexual lives and are nested within the age group of highest risk for HIV infection; (b) college students have been studied previously, allowing for the comparison of the present work to prior research; and, (c) students' attitudes toward homosexual men and women will influence their future work environments, fostering either tolerance or discrimination. Moreover, to the authors' knowledge, Puerto Rican's attitudes toward homosexuality have yet to be explored. Following the Mexican population, Puerto Ricans comprise the second largest Latino group in the U.S. (U.S. Census, 2000). Moreover, Puerto Rico is within the top 10 states with highest AIDS prevalence in the U.S. (CDC, 2002). Because stigmatizing attitudes toward homosexuality have been identified as a barrier to HIV/AIDS prevention efforts among Puerto Ricans (Carballo-Diéguez, 1995; Carballo-Diéguez & Dolezal, 1996), a deeper understanding of the psychological construction of Puerto Rican heterosexual men and women's attitudes toward gay individuals would facilitate the development of effective psychosocial interventions to reduce HIV and sexual prejudice.

The study's objectives were (a) to begin the adaptation and validation of the Attitudes toward Lesbians and Gays Scale (ATLG) in its Spanish version for Puerto Ricans; (b) to examine the direction and intensity of Puerto Rican young adults' attitudes toward homosexuality and their interactions and positive experiences with homosexuals, perceived similarity to peers' attitudes toward homosexuals, and religiosity; and (c) propose and test sex differences in a model with theoretically relevant constructs using a structural equation modeling approach. The ATLG scale was translated and used. Based on the absence of data on Puerto Rican men and women's attitudes toward homosexuality, a model to address the remaining two goals was developed from the body of literature on sexual prejudice among non-Hispanic U.S. samples (Herek, 1984; Herek & Glunt, 1993; Herek, 1996). Sex differences between sexual prejudice and its predictors were then explored in the model.

We proposed the associations among constructs based on previous findings in the scientific literature. Using this information, we tested the following five hypotheses in our exploratory structural model. Figure 1 below displays the interrelationship between our hypotheses.

Figure 1.

Figure 1

Open in a new tab

Exploratory Structural Equation Model of the effects of interaction and positive experiences with homosexuals, perceived similarity with peers' attitudes toward homosexuality, and religiosity on Puerto Rican college students' sexual prejudice. All solid lines represent statistically significant paths at the .05 level. Magnitudes of association are presented with the standard errors in parentheses; X2 (61, N = 360) = 209.48. Normed (NFI), non-normed (NNFI), and comparative (CFI) goodness-of-fit are .92, .92, .94, respectively; RMSEA misfit is .08.

  1. Favorable attitudes toward gay men and lesbians will increase as participants' report more interactions with homosexual men and women. Akin to Allport's (1954) contact hypothesis on intergroup contact, increased interpersonal contact with homosexual people decreases heterosexuals' sexual prejudice (Herek & Capitanio, 1996; Herek & Glunt, 1993; Sherrod & Nardi, 1998). The applicability of this “contact hypothesis” to attitudes toward gay and lesbians among Puerto Rican individuals has not been tested.

  2. Participants will report fewer negative attitudes toward lesbians and gay men if they perceive their friends hold similar attitudes toward homosexuals. Peer influence on attitudes and behaviors has been documented consistently by social psychologist (Michener, DeLamater, & Myers, 2004). This pattern has been observed among samples of non-Hispanic college students (Sakalli & Ugurlu, 2001).

  3. Increased positive interactions with gay and lesbian persons will decrease participants' self-reported negative attitudes toward homosexuality. Moreover, their recollections of positive interactions will be moderately associated with the number of homosexual men and women with whom they have interacted and their perceptions of peers' acceptance of homosexuality. Participants' exposure and recollection to positive interactions with gay people have been found to decrease negative esteem toward this population and to increase their desire to interact with homosexual men and women in the future (Nelson & Krieger, 1997).

  4. Participants will have stronger negative attitudes toward homosexual men and women to the extent that they bear stronger religious beliefs. Religiosity has been a reliable negative predictor of non-Hispanic White heterosexuals' attitudes toward homosexual men and women (Herek, 1994; Sherrod & Nardi, 1998). Because religiosity has been found to be a predictor of sexual prejudice among other racial/ethnic groups, we expect this negative relationship to hold in this Puerto Rican sample.

  5. When compared to female participants, male participants will report stronger negative associations in their attitudes toward gay men and lesbians across all study variables. Sex differences in attitudes toward gay men and lesbians have been found in samples of non-Hispanic Whites (Herek, 2002).

Method

Design and respondents

Study data were collected in the second largest public undergraduate campus in Puerto Rico during the months of January and February, 2001. The campus is located outside of the San Juan metro area and has approximately 13,000 students, mainly undergraduates, stratified under four Colleges: Arts and Sciences (60%), Engineering (25%), Business Administration (10%), and Agricultural Sciences (5%). The absence of a complete list of registered students on campus and the limited financial resources for the study led to a stratified random sample of class sections within each college rather than a random sample of students. Selection of class sections ensured the equal probability of each class section from the total sections offered by each college during the second semester. Twelve course sections from Arts and Sciences, five from Engineering, two from Business Administration, and one from Agricultural Sciences were chosen. We obtained a sample of 382 undergraduate students enrolled in these 20 course sections.

Participants

Twenty-two participants were excluded from the analyses because they self-reported as bisexual (n=9; 2.4%), homosexual (n=7; 1.8%), or in search of their sexual identity (n=6; 1.6%). This study included 360 heterosexual undergraduate students: 153 men (42.5%) and 207 women (57.5%). Participants' average age was 20.8 years of age (SD = 1.74). When asked about their marital status, most students (93%) reported being single.

Students reported attending college an average of 3.37 years (SD=1.42). The distribution of the sample by years of college attended was 11% freshmen, 17% sophomores, 23% juniors, 28% seniors, and 21% fifth-years or longer. Distribution of students by college differed from the distribution of class sections because requirements for all degrees can cross colleges. The sample, divided by College, consisted of 72 students from Business Administration (20.0%), 130 students from Engineering (36.1%), 126 students from Arts and Sciences (35.0%), and 32 students from Agricultural Sciences (8.9%). To correct for the over sampling of students from other colleges in the Arts and Sciences courses, we computed statistical weights for the expected proportion of students by college and applied them to our analyses.

Data collection procedures

Research staff contacted professors whose courses were selected for participation via campus telephone or after class if they were unreachable over the phone. We gave professors a brief description of the study and asked for their authorization to use their class time to administer a survey to students. All contacted professors supported the study. Study staff and professors scheduled a class time where researchers could ask students to participate in our study. At each classroom, the researchers introduced themselves, explained the voluntary nature of the participation, as well as the students' rights to refuse. Participants were informed of the survey's topic and assured of the data's confidentiality through a consent form. While there was no incentive for participation, all students agreed to participate (100% response rate). Nonetheless, we were unable to determine if all eligible students within a classroom attended class the day the questionnaire was administered.

Students filled a self-administered questionnaire in the classroom. As expected from our pilot testing of the questionnaire, the self-administered questionnaire's mean response time was 15 minutes. To avoid participants from filling out two questionnaires, researchers asked students to abstain from answering a survey if they had already filled out the questionnaire. Three students stated already filling out the survey in another class.

Measures

The self-administered questionnaire consisted of six sections (all written in Spanish). With the exception of participants' quality of interactions with gay individuals, all constructs used multi-item indices. All constructs attained a Cronbach's α over .70 (Nunally & Bernstein, 1994).

Demographic and background measures

Demographic information was collected using standard questions for age, sex, marital status, and sexual orientation. Participants were also asked to report the number of completed years of college, their college, and their program of study.

Interactions with homosexual individuals

Experiences with homosexual individuals were measured using two items: “Please check the number that best approaches the number of gay men and women with whom you have interacted (that you know)” and “Thinking about your best friend, how many gay and women do you think your best friend knows?” Participants reported on both questions from six categories: None, One, Two, Three, Four, or Five or more. Both items have been used in previous studies and found to have good reliability (Herek & González-Rivera, in press). In our study, both items had a Cronbach's α of .81.

Quality of interaction with homosexual individuals

Quality of interactions with homosexual men and women was collected with an item asking participants to recall previous interactions with gay men and women and report their perceived interaction in these instances. The item read, “Generally thinking of the homosexual individuals with whom you remember having interacted, select the number from 1 (Very bad) to 5 (Very good) that best indicates how you perceived this interaction.” Because we only used one indicator for the construct, random measurement error was corrected by fixing the indicator's error term to equal the variance times one minus a reliability of .70 (Vinokur, Pierce, & Buck, 1999).

Perceptions of peer's acceptance of homosexual individuals

Participants' perception of peers' similarity of attitudes toward homosexuals was assessed using two identical items, except on the peers' sex. The questions read: “Circle the number that best describes how likely your attitude toward homosexuals resembles the attitude most of your [male/female] friends have toward homosexuals.” Both questions were answered using a 5-point scale ranging from 0 (Nothing at all) to 4 (A lot). Previous studies have found these items to mirror respondents' normative attitudes toward homosexual individuals (Herek & Capitanio, 1996). These items showed a Cronbach's α of .70.

Religiosity

We measured religiosity using four items: (a) importance of religious or spiritual beliefs; (b) frequency of church or temple attendance in the past year; (c) frequency to events sponsored by their religious group in the past year; and (d) prayer frequency in the past month. Importance of spiritual beliefs was measured using a 4-point scale ranging from 1 (Not important at all) to 4 (Very important). The item measuring participants' attendance to a church or temple in the past year read, “How often have you attended a church or temple in the past year?” Participants answered this by choosing one of the following six categories: Never, Once or twice a year, A few times a year, Once or twice a month, Once a week, and More than once a week. The six answer categories for church attendance were also used for attendance to sponsored spiritual events. Prayer frequency was assessed with six categories: Never, Once or twice a month, A few times a month, One or two times a week, Once a day, or More than once a day. Other studies have used similar measures of religiosity in their questionnaires and found them to be consistently reliable (Herek & González-Rivera, in press). Religiosity items had a Cronbach's α of .80 in our study.

Attitudes toward Lesbians and Gays

The final section of the questionnaire included the Spanish-translation of the original version of the Attitudes Toward Lesbians and Gay (ATLG) Scale (Herek, 1984). This 20-item scale is answered using a 5-point scale ranging from 1 (Strongly disagree) to 5 (Strongly agree). Respondents' score is computed by taking the mean across items such that higher scores reflect more negative attitudes toward gay men and women. In its original version, the scale can be divided into two 10-item subscales: the Attitudes toward Lesbians (ATL) and the Attitudes toward Gays (ATG) subscales. We translated the ATLG scale into the Spanish-language and then back translated from Spanish into English to ensure equivalency between the two versions. Spanish-speaking Puerto Rican men and women, who volunteered in a pre-test pilot study, considered the use of language appropriate. Spanish and English versions of the items' content can be found in Appendix A.

Exploratory principle axis factor analyses with oblique rotation of the Spanish version of the ATLG Scale had three factors: a Condemnation-Tolerance Factor (17 items; 47.67%), a Self-Expressive Factor (one item; 6.75%), and a Beliefs Factor (two items; 6.51%). These three components explained 60.9% of the total variance. The explained variance of these factors was very similar to those obtained in Herek's (1984) original study. Following Herek's (1984) recommendation, we selected the items loading high only on the component matrix of the Condemnation-Tolerance factor. After removing 3 items (items 2, 7, 12), we decided to also remove an additional item (item 17) because it decreased the scale's reliability greatly. The ATLG scale ended with 16-items and explained 55% of the total variance. The item loadings for three factors are presented in Table 1.

Table 1.

Three-Factor Loadings for the ATLG Scale.

Item Number Loadings
Condemnation Tolerance Self-expressive Belief
Item 1 .632 -.077 .215
Item 2 .174 .085 .492
Item 3 .786 -.151 .123
Item 4 .564 -.280 .090
Item 5 .765 -.184 -.142
Item 6 .725 -.189 -.083
Item 7 .332 .109 .452
Item 8 .745 -.318 -.004
Item 9 .710 -.282 -.084
Item 10 .677 -.198 .013
Item 11 .621 .129 -.053
Item 12 .550 .460 -.048
Item 13 .573 .126 .160
Item 14 .831 .083 .015
Item 15 .717 .215 -.021
Item 16 .686 .020 -.114
Item 17 .576 .303 -.136
Item 18 .819 .165 -.206
Item 19 .809 .128 -.114
Item 20 .815 .082 .111
Open in a new tab

Note. Bold typeface indicates items loading within each factor.

Herek's ATLG Scale has been administered in various racial groups throughout the United States with consistent Cronbach's alpha coefficients higher than .87 (Herek, 2000). While the scale had never been used for a Puerto Rican sample, the ATLG-Spanish version showed reliability scores (α=.93) similar to those obtained by Herek in an undergraduate sample (Herek, 1996). To decrease the inflation of the Cronbach's alpha reliability coefficient due to the number of items, we parceled the 16 items into four composite scores. Each score represented the mean value for each four-item ATL and ATG indicator, respectively. Alternate placement of items by their rank ordered factor loadings were used to include the indicators for the ATG and ATL indicators, respectively. All four-item indicators had a .90 Cronbach's alpha coefficient.

Results

Analytic strategy

We used EQS version 6.1 (Bentler, 2004) for our exploratory latent-variable structural modeling. Contrary to regression or other methods, structural equation modeling permitted the testing of various endogenous factors simultaneously (Klem, 2000). Furthermore, this analytic strategy allowed us to examine the multiple structural path models while adjusting for the reliability of the observed measures (Bedeian, Day, Kelloway, 1997). Our estimation of model coefficients was improved by using an estimated covariance matrix generated from the variables' observed covariance matrix. The comparisons between the observed and estimated covariance matrices provided overall goodness-of-fit measures, allowed for model modifications, and provided a straightforward approach for group comparisons. We present our findings following Raykov, Tomer, and Nessel's (1991) proposed guidelines for adequate reporting of structural equation modeling and provide three goodness-of-fit indices: Bentler-Bonnet's Normed Fit Index (NFI), Bentler-Bonnet's Non-Normed Fit Index (NNFI), and the Comparative Fit Index (CFI). We also provide information on the root mean-square error of approximation (RMSEA) as an index of misfit (Boomsma, 2000). Hu & Bentler (1999) suggest values of .90 or higher among fit indices and values of .06 or lower for RMSEA as acceptable indications of well-fitting models.

A small number of cases (n=45; 12%) had missing data. Because structural equation modeling theory assumes no missing data, we proceeded to run maximum likelihood estimation using the EM algorithm to impute the missing cases using EQS (Bentler, 2004). Missing data diagnostics found 15 patterns of missing data. Study variables were assumed to be missing at random and followed a multivariate normal distribution (Ragunathan, 2004). Computed statistical weights were then used for all analyses to ensure the samples' representativeness of the student population.

Testing the measurement model of study variables

Prior to testing our structural models, we proceeded to test the measurement models' adequacy. The measurement model allowed the ascertainment of the chosen indicators' appropriateness for the various latent factors in our analyses. The measurement model used the ML solution for its estimation. Table 2 provides a summary of all study variables' correlations, and Table 3 details study variables' means and standard deviations by sex.

Table 2.

Measured variables' (indicators) correlations for all cases.

Measured Variable 1 2 3 4 5 6 7 8 9 10 11 12 13
1. Interaction (a) 1.00
2. Interaction (b) .687 1.00
3. Quality of Interaction .214 .151 1.00
4. Peer Similarity (a) .079 .094 .026 1.00
5. Peer Similarity (b) .196 .212 .204 .538 1.00
6. Religiosity (a) -.072 -.079 .023 .013 -.030 1.00
7. Religiosity (b) .015 -.012 -.004 .113 -.011 .544 1.00
8. Religiosity (c) -.033 -.022 -.011 .110 -.012 .409 .712 1.00
9. Religiosity (d) -.061 -.097 .036 -.001 -.015 .574 .521 .509 1.00
10. ATL (a) -.227 -.121 -.330 -.034 -.213 .300 .347 .309 .281 1.00
11. ATL (b) -.241 -.113 -.265 -.040 -.183 .376 .438 .454 .388 .815 1.00
12. ATG (a) -.301 -.194 -.404 .002 -.131 .233 .318 .358 .251 .713 .769 1.00
13. ATG (b) -.289 -.181 -.181 .012 -137 .263 .356 .383 .241 .717 .750 .811 1.00
Open in a new tab

Note. Letters in parentheses indicate the presence of multiple indicators for each construct.

Table 3.

Mean (standard deviation) of study variables by sex.

Factor and Subscale Males Females
Interaction with homosexuals
 Personal interaction 3.09(1.73) 3.18(1.59)
 Best friend's interaction 3.06(1.59) 3.12(1.65)
Positive experiences with homosexuals
 Quality of Interaction 3.60(.91) 4.17(.68)*
Similarities with Peer Attitudes
 Male Peers 1.35(1.06) 1.15(1.02)
 Female Peers 1.45(1.04) 1.66(1.13)
Religiosity
 Importance of Beliefs 3.38(.79) 3.58(.60)
 Church Attendance 2.43(1.60) 2.93(1.54)
 Attendance to church events 1.23(1.35) 1.81(1.62)*
 Prayer Frequency 2.64(1.67) 3.54(1.31)*
Sexual Prejudice
 Attitudes toward lesbians (a) 2.53(.98) 2.52(.92)
 Attitudes toward lesbians (b) 2.78(1.10) 2.83(.97)
 Attitudes toward gay men (a) 3.31(1.04) 2.94(1.04)*
 Attitudes toward gay men (b) 3.05(.87) 2.83(.90)
Open in a new tab
*

p < .004 (Bonferroni Correction for 13 tests at p < .05)

The error variances for e1 (item on participants' personal interaction with gay men and lesbians) and e5 (item on perceived female peer similarities toward homosexuals) were close to zero in the measurement model. Consequently, we fixed the residual variance for both terms to .005. The measurement model converged in 8 iterations. The measurement model showed the data had an acceptable fit, [X2(58, N = 360) = 208.93, p < .001; NFI = .92, NNFI = .92, CFI = .94, RMSEA = .09]. In light of the model's fit, we used this measurement model in our subsequent structural model tests. Table 4 and Table 5 present the factor loadings for all indicators, and the estimated correlations among latent-variables, respectively.

Table 4.

Factor Loadings of Indicators for participants' Interaction with homosexuals, Quality of interactions with homosexuals, Perceived Peer Acceptance, Religiosity, and Sexual prejudice.

Indicator Loading Error
1. Interaction (a) .99 .04
2. Interaction (b) .68 .74
3. Quality of Interaction .83 .56
4. Peer Similarity (a) .58 .81
5. Peer Similarity (b) 1.00 .06
6. Religiosity (a) .63 .78
7. Religiosity (b) .86 .52
8. Religiosity (c) .82 .57
9. Religiosity (d) .75 .67
10. ATL (a) .89 .54
11. ATL (b) .90 .46
12. ATG (a) .90 .43
13. ATG (b) .88 .47
Open in a new tab

Table 5.

Estimated correlations among latent factors including interaction with homosexuals, positive experiences with homosexuals, peer similarity, religiosity, and sexual prejudice.

Latent factor 1 2 3 4 5
1. Interactions with homosexual individuals -
2. Positive experiences with homosexuals .28 -
3. Peer Similarity on Homosexuality .19 .18 -
4. Religiosity -.01 .01 .01 -
5. Sexual Prejudice -.30 -.17 -.28 .17 -
Open in a new tab

Testing the structural model

The structural model represented in Figure 1 proposed a good fit, [χ2 (61, N = 360) = 209.48; NFI = .92, NNFI = .92, CFI = .94, RMSEA = .08]. Consideration was given to potential model modifications suggested by the Lagrange Multiplier Test for adding parameters and the Wald Test for dropping parameters, but we opted to avoid making any changes because they were unsubstantiated by the literature and reduced the likelihood of capitalizing on chance (Kline, 1998).

We found evidence for our first hypothesis. Participants interacting with a greater number of gay and lesbian individuals tended to report lower negative attitudes toward homosexual men and women than those experiencing fewer interactions with gays and lesbians (β=-.20, se = .03, p < .05). Participants' sexual prejudice score decreased by a fifth of a scale point (out of a total of 5) for every homosexual individual the participant recalled interacting with.

While we found a moderate association between participants' interaction with gay men and lesbians and their perception of their peers' similarity in their attitudes toward homosexuals (r = .24, se = .05, p < .05), we found no evidence to support our second hypothesis (β=-.05, se = .03, n.s.). Participants perceiving their friends held similar attitudes toward gay and lesbians did not report lower negative attitudes toward gay men and lesbians.

We believed positive experiences with homosexual men and women would decrease participants' negative attitudes toward gay men and lesbians. We found a moderately strong negative association (β=-.45, se = .07, p < .05) between quality of participants' interactions with gay and lesbian individuals and negative attitudes toward homosexual; thus, confirming our third hypothesis. A one unit increase in participants perceived positive experiences during their interactions with homosexual men and women decreased their sexual prejudice score by half a point. Moreover, we found significant correlations between positive experiences with gay men and lesbians and previous interactions with homosexual men and women (r = .26, se = .05, p < .05), as well as with participants' perceived similarities in their friends' attitudes toward gay men and lesbians (r = .24, se = .07, p < .05). While moderately low, the association between these three latent factors point to the multifaceted nature of participants' attitudes toward gay and lesbian people.

Our fourth hypothesis stated participants with stronger religious convictions would hold stronger negative attitudes toward gay men and lesbians. We found religiosity to be the strongest predictor of participants' negative attitudes toward gay men and lesbians (β=.50, se = .11, p < .05). For every unit increase in participants' assessment of the importance of their religious beliefs in their lives, their sexual prejudice score increased by half a scale point.

Given the non-significant prediction of peers' similarities in their attitudes toward homosexuals, we attempted removing this path but the model was unable to converge adequately after 500 iterations. Therefore, we kept this factor in our model to ensure successful model stability. The final model showed an R2 of 56% for sexual prejudice's variance.

Testing for sex effects

In order to test whether the exploratory structural model provided an equally good fit for males and females, we re-ran the structural model estimation procedures running each group's covariance matrix simultaneously. All factor loadings, paths, and variances were constrained to be equal in the initial model. The sex differences model indicated a relatively acceptable fit for both sexes, [χ 2 (141, N-males = 153, N-females = 207) = 292.12; NFI = .88, NNFI = .93, CFI = .94, RMSEA = .055]. We then freed each path consecutively to test whether sex differences existed between the significant latent-factors and sexual prejudice. Our findings suggest no differences in the model's path vary due to participants' sex. After freeing the path for participants' interaction with homosexuals and sexual prejudice, we found no difference across male and female participants (Δ χ 2(1) = 1.27, n.s.). Subsequently, we freed the path between positive experiences with homosexuals and sexual prejudice but we found no difference by participants' sex (Δ χ 2(1) = .05, n.s.). Finally, we tested whether sex differences existed between religiosity and sexual prejudice but no difference was found (Δ χ 2(1)= 0.27, n.s.).

An Alternative Model: What if sexual prejudice predicts the study variables?

Although our analyses find a good fit for the data, we tested whether another model could fit the data just as well or better (MacCallum, Wegener, Uchino, & Fabrigar, 1993). Theoretically, it is just as plausible that individuals with greater negative attitudes toward homosexuality would abstain from interacting with gay men and lesbians, rating their interactions as negative, perceiving their friends as having different attitudes toward homosexual individuals, or find reassurance about their beliefs in their religiosity. Figure 2 presents this inversed causation alternate model below.

Figure 2.

Figure 2

Open in a new tab

An alternative exploratory structural model: What if sexual prejudice predicts interaction and positive experiences with homosexuals, perceived similarity with peers' attitudes toward homosexuality, and religiosity. All solid lines represent statistically significant paths at the .05 level. Magnitudes of association are presented with the standard errors in parentheses; X2 (61, N = 360) = 242.31. Normed (NFI), non-normed (NNFI), and comparative (CFI) goodness-of-fit are .91, .91, .93, respectively; RMSEA is .09.

The alternate model had a good fit, [χ 2(61, N = 360) = 242.31, p < .001; NFI = .91, NNFI = .91, CFI = .93, RMSEA = .09]. Sexual prejudice was predictive of all latent factors: interaction with homosexuals (β=-.34, se = .08, p < .05; R2=11%); positive experiences with gay men and lesbians (β=-.50, se = .06, p < .05; R2=25%); perceived peer similarities in their attitudes toward homosexuals (β=-.19, se = .08, p < .05; R2=4%); and, religiosity (β=.48, se = .04, p < .05; R2=23%). Significant correlations were found for the disturbances of perceived similarities with peers' attitudes toward homosexuals and their previous interactions with gay men and lesbians (r=.18, se = .06, p < .05), as well as their positive experiences with homosexuals (r=.17, se = .05, p < .05). The correlation between the disturbances of positive experiences with homosexuals and participants' interactions with homosexuals was not significant (r=.11, se = .04, n.s.). Sexual prejudice accounted for an 11% of participants' interaction with homosexual individuals (R2=.11); 25% of participants' positive experiences with homosexual individuals (R2=.25); 4% of peer similarities with homosexual men and women (R2=.04); and 23% of religiosity (R2=.23).

While these results are noteworthy, we compared which model fitted the data better using the Model AIC and Model CAIC coefficients for non-nested model comparisons and each models' fit indices. The original model had a better fit (Model AIC=87.48; Model CAIC=-210.58; NFI = .92; NNFI = .92; CFI = .94; RMSEA = .08) than the alternative model (Model AIC = 120.31 & CAIC=-177.74; NFI = .91; NNFI = .91; CFI = .93; RMSEA = .09). The difference in fit across the two models suggests our original model's theoretical rationale fits the data better (Byrne, 1994).

Discussion

We developed a structural model to test whether interactions and positive experiences with homosexuals, perceived similarities with peers' attitudes toward homosexuality, and religiosity were predictive of sexual prejudice among a population of Puerto Rican young adults. Our findings suggest all latent variables, with the exception of peers' attitudes toward homosexuality, predicted participants' sexual prejudice toward homosexual men and women. In addition, although sex differences were expected, we found no differences between males and females. While previous researchers have found sexual prejudice tends to be higher for men than women (Kite, 1992; Herek, 2000b), we were unable to replicate these results.

The number of homosexual persons known by the participants reduced their reported sexual prejudice toward homosexual men and women. This finding suggests that Puerto Rican youth's sexual prejudice may be influenced by the lack of contact with gays and lesbians. The stigmatization of homosexual men and women in the Puerto Rican community may limit some students' contact with gay men and lesbians. Within this population, providing in-class panels with gay men and lesbians, for example, may facilitate students' interactions with gay men and lesbians. Other researchers have found these classroom activities decrease some heterosexual's negative appraisals toward gay men and lesbians (Sakalli & Ugurlu, 2001).

Our findings also suggest that participants' positive experiences with gay men and lesbians exert an influence in their sexual prejudice toward homosexual people. Participants' recollection of positive experiences with gay men and lesbians was strongly predictive of their sexual prejudice. Moreover, the presence of a positive association between participants' interactions with homosexuals and their recollection of positive experiences with this group suggests an iterative process by which Puerto Rican young adults could increase or decrease their intention to interact with homosexual men and women at later times. These findings suggest that Puerto Rican young adults' appraisals of their contact with homosexual men and women are particularly important during the design and implementation of activities to decrease sexual prejudice. Activities and settings promoting an environment where acceptance and open dialogue is normalized may facilitate positive interchanges. Research on what constitutes positive experiences between Puerto Rican heterosexual and homosexual men and women is needed.

While participants' perceived similarities with peers' attitudes toward homosexuality proved to be non-predictive of sexual prejudice, moderate associations between this predictor and participants' interactions and positive experiences with homosexual men and women were found. Perceived peer similarities have been found to reflect respondents' normative beliefs toward homosexuality and to be predictive of sexual prejudice (Herek & González-Rivera, in press). Nonetheless, we propose two arguments for why sexual prejudice was not predicted by participants' peer similarities in attitudes toward homosexuality. First, participants in this sample may disregard friends' attitudes toward homosexuality and construct their own beliefs and values about homosexual men and women. Second, participants may socialize in networks where peers hold similar attitudes toward homosexual men and women. Additional research on potential relationship between sexual prejudice and social network characteristics is warranted to assess these effects.

Religiosity was found to be the strongest predictor of sexual prejudice in our analyses. Most Puerto Ricans follow some interpretation of the Christian tradition. Given Christianity's stance on homosexuality as a sinful lifestyle, it is not surprising to find a strong negative association between religiosity and sexual prejudice (Díaz, 1998; Carrier, 1995). Moreover, we questioned whether religiosity could increase sexual prejudice by increasing participants' moral opposition to interact with homosexual people and/or perceive any interactions with gay men and women as positive. Nonetheless, a structural model with religiosity predicting sexual prejudice alongside the other study variables was found to be unreliable and unstable. Therefore, we do not believe religiosity is the driving factor for all other study variables. While an intervention to decrease individuals' religiosity is unethical, we do suggest open discussions of Christianity's stance on sexual orientation versus homosexual behaviors within religious settings (Wager, Serafini, Rabkin, Remien, & Williams, 1994). These activities may reduce Puerto Rican's negative attitudes toward homosexual men and women by removing the focus of stigma from the homosexual individual to a particular sexual behavior.

Finally, we found no sex differences in the associations between sexual prejudice and the study variables. Nonetheless, the absence of sex difference may be due to our measurement of sexual prejudice. The differences in item content within our ATL and ATG subscales make accurate comparisons between the two subscales unfeasible. Therefore, we limited our analyses to participants' attitudes toward gay people, instead of specific attitude comparisons between gay men and women. Previous studies finding sex differences in attitudes toward homosexuality adapted the ATLG item wording to be identical with the exception of the reference group (i.e., “[Male/Female] homosexuality is a perversion.”) (Herek, 2002a). Conversely, perhaps the differences in sexual prejudice for our study population are not due to a sex dichotomy but to differences across gender roles (Sakalli, 2002). Men with greater salience in machismo and in their perceptions of masculinity should promote greater sexual prejudice among men (Cunningham, 1998). Future research should explore how item wording, cultural values, and gender differences may influence sexual prejudice among Puerto Ricans.

Our study has several limitations. First, the study's cross-sectional data limited the firm temporal ordering of our variables. Future research should explore the causal relationship between the study variables. Nonetheless, comparisons between our model and the alternate model suggest that religiosity, positive experiences with homosexual individuals, and number of interactions with gay men and lesbian predicted sexual prejudice better than the alternate causal route. Second, financial and resource limitations diminished our ability to collect a larger sample. A larger sample that can be split in random halves may allow for additional modifications to be made on one half of the sample, while testing the fully-modified model in the other random half. Third, study findings may be non-generalizable for Puerto Rican non-college populations. Although participants seemed motivated to participate in the study, previous research found individuals with low educational attainment tend to score higher on sexual prejudice (Herek, 2002). Fourth, we did not ascertain if participants interacted with gay men or lesbians who self-disclosed their sexual orientation or if they assumed some men and women were homosexuals. Fifth, our inability to assess the number of students absent within each classroom during the survey limits the assessment of potential coverage biases. Nonetheless, because the data collection took place early in the semester, we believe the likelihood of dropouts and students not attending lecture may have diminished. Finally, religiosity's effect on sexual prejudice deserves greater exploration. Our study was unable to collect information on participants' religious affiliations. Given the diversity of Christian traditions (85% Roman Catholic and 15% Protestant and other; CIA World Factbook (2005)), future studies should explore whether religiosity's effect on sexual prejudice varies by religious traditions.

The limitations of this study not withstanding builds on the sexual prejudice literature in several ways. First, this study explores the relationship between sexual prejudice and its predictors in a sample of Puerto Rican college students. Aside from a previous study targeting Mexican Americans (Herek & González-Rivera, in press), sexual prejudice has not been studied across different Latino ethnic groups. A deeper understanding of sexual prejudice for each ethnic group may clarify similarities and differences and improve culturally targeted interventions to reduce stigma against gay men and lesbians in these communities. Second, our study is the first to begin the validation of a translated Spanish-version of the ATLG scale for a Puerto Rican population. We hope future studies will continue this validation and incorporate the scale into assessments of sexual discrimination among Puerto Rican populations. Third, while other studies have used telephone interviews as their mode for data collection, the use of a self-reported questionnaire in our study likely decreased participants' tendencies to provide socially desirable answers. Finally, our studies are similar to those of other studies done with English-speaking non-Latino U.S. samples. The consistency across studies suggests attitudes toward homosexual men and women may be similar across races and ethnicities. Nonetheless, the expression of these mechanisms across different populations may differ. Future research should expand the study of sexual prejudice toward homosexual people in the Puerto Rican population, as well as in other Latino ethnic groups.

Acknowledgments

For comments and help in various stages of this research project, the authors would like to thank the anonymous reviewers from the Journal, Dr. Gregory Herek, Dr. Gilbert Gee, and the research faculty from Center for Applied Social Research at the University of Puerto Rico – Mayagüez.

Appendix A. English and Spanish Translation of the ATLG Scale

Attitudes Toward Lesbians (ATL) Scale
1. Lesbians just can't fit into our society.
[Las lesbianas simplemente no pertenecen en nuestra sociedad]
2. A woman's homosexuality should not be a cause for job discrimination in any situation. (R)
[Bajo ninguna circunstancia, la homosexualidad en la mujer debería ser factor para la discriminación en el empleo.]
3. Female homosexuality is detrimental to society because it breaks down the natural divisions between the sexes.
[La homosexulidad en la mujer hace daño porque viola las divisiones entre los sexos establecidas por la naturaleza].
4. State laws regulating private, consenting lesbian behavior should be loosened. (R)
[Las leyes estatales que las regulan conductas privadas y consentidas entre lesbianas deberían ser más flexibles].
5. Female homosexuality is a sin.
[La homosexualidad en la mujer es un pecado].
6. The growing number of lesbians indicates a decline in American morals.
[El aumento en el número de lesbianas indica un debilitamiento en los valores morales puertorriqueños].
7. Female homosexuality in itself is no problem, but what society makes of it can be a problem. (R)
[Lo que la sociedad piensa de la homosexualidad en la mujer es el verdadero problema].
8. Female homosexuality is a threat to many of our basic social institutions.
[La homosexualidad femenina es una amenaza a muchas de nuestras principales instituciones sociales.]
9. Female homosexuality is an inferior form of sexuality.
[La homosexualidad femenina es una forma inferior de sexualidad.]
10. Lesbians are sick.
[Las lesbianas son personas enfermas.]
Attitudes Toward Gay Men (ATG) Scale
11. Male homosexual couples should be allowed to adopt children the same as heterosexual couples. (R)
[A las parejas de hombres homosexuales se les debería permitir adoptar niños].
12. I think male homosexuals are disgusting.
[Considero que los hombres homosexuales repugnantes].
13. Male homosexuals should not be allowed to teach school.
[A los hombres homosexuals no se le debería permitir dar clases en escuelas].
14. Male homosexuality is a perversion.
[La homosexualidad en hombres es una perversion].
15. Just as in other species, male homosexuality is a natural expression of sexuality in human men. (R)
[La homosexualidad en hombres es una expresión natural de la sexualidad en humanos, al igual que en otras especies].
16. If a man has homosexual feelings, he should do everything he can to overcome them.
[Si un hombre tiene sentimientos homosexuals debe hacer lo posible para superarlo].
17. I would not be too upset if I learned that my son were a homosexual. (R)
[Yo no estaría muy molesto/a si descubriera que mi hijo fuera homosexual].
18. Homosexual behavior between two men is just plain wrong.
[Las conductas homosexuales entre dos hombres simplemente están mal].
19. The idea of male homosexual marriages seems ridiculous to me.
[Me parece ridículo la idea de matrimonio entre parejas de hombres homosexuales].
20. Male homosexuality is merely a different kind of lifestyle that should not be condemned. (R)
[La homosexualidad en hombres es meramente un estilo de vida diferente que no debería ser condenado].
Open in a new tab

Note. Items followed by an (R) indicate reversed scoring.

Contributor Information

José A. Bauermeister, University of Michigan's School of Public Health.

Mercedes M. Morales, Ecological-Community Psychology at Michigan State University.

Gretchen Seda, Psychology Department at the Inter American University – San Germán.

Milagritos González-Rivera, University of Puerto Rico – Mayagüez.

References

  1. Allport GW. The nature of prejudice. Reading, MA: Addison-Wesley; 1954. [Google Scholar]
  2. Altemeyer B. Changes in attitudes toward homosexuals. Journal of Homosexuality. 2001;42(2):63–75. doi: 10.1300/j082v42n02_04. [DOI] [PubMed] [Google Scholar]
  3. Badgett MV. The wage effects of sexual orientation discrimination. Industrial and Labor Relations Review. 1995;48(4):726–739. [Google Scholar]
  4. Bedeian AG, Day DV, Kelloway EK. Correcting for measurement error attenuation in structural equation models: Some important reminders. Educational and Psychological Measurement. 1997;57(5):785–799. [Google Scholar]
  5. Bentler PM. EQS Structural Equations Program Manual. Encino, CA: Multivariate Software, Inc.; 2004. [Google Scholar]
  6. Boomsma A. Reporting analyses of covariance structures. Structural Equation Modeling. 2000;7(3):461–483. [Google Scholar]
  7. Byrne BM. Structural equation modeling with EQS and EQS/Windows: Basic concepts, applications, and programming. Thousand Oaks, CA: Sage Publications; 1994. [Google Scholar]
  8. Carballo-Diéguez A. Hispanic culture, gay male culture, and AIDS: Counseling implications. Journal of Counseling & Development. 1989;68(5):26–30. [Google Scholar]
  9. Carballo-Diéguez A. The sexual identity and behavior of Puerto Rican men who have sex with men. In: Herek G, Greene B, editors. AIDS, identity, and community: The HIV epidemic and lesbians and gay men. Vol. 2. Thousand Oaks, CA: Sage; 1995. pp. 105–114. [Google Scholar]
  10. Carballo-Diéguez A. Latinos. In: Smith RA, editor. Encyclopedia of AIDS: A social, political, cultural, and scientific record of the HIV Epidemic. Chicago: Fitzroy Dearborn Publishers; 1998. pp. 321–323. [Google Scholar]
  11. Carballo-Diéguez A, Dolezal C. HIV risk behaviors and obstacles to condom use among Puerto Rican men in New York City who have sex with men. American Journal of Public Health. 1996;86(11):1619–1622. doi: 10.2105/ajph.86.11.1619. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Carrier J. De los otros: Intimacy and homosexuality among Mexican men. New York: Columbia University Press; 1995. [Google Scholar]
  13. Centers for Disease Control. AIDS - United States, 2000. MMWR. 2002;51:592–595. [Google Scholar]
  14. Central Intelligence Agency. The World Factbook: Puerto Rico. 2005 Retrieved October 28, 2005 from the World Wide Web: http://www.cia.gov/cia/publications/factbook/geos/rq.html.
  15. Cole SW, Kemeny ME, Taylor SE, Visscher BR. Elevated physical health risk among gay men who conceal their homosexual identity. Health Psychology. 1996;15(4):243–251. doi: 10.1037//0278-6133.15.4.243. [DOI] [PubMed] [Google Scholar]
  16. Cunningham I. An innovative HIV/AIDS research education program in Puerto Rico. SIECUS Report. 1998;26(3):18–21. [PubMed] [Google Scholar]
  17. Díaz R. Latino Gay Men and HIV: Culture, Sexuality, and Risk Behavior. New York: Routledge; 1998. [Google Scholar]
  18. Díaz R, Ayala G. Social Discrimination and Health: The case of Latino gay men and HIV risk. New York: The Policy Institute of the National Gay and Lesbian Task Force; 2001. [Google Scholar]
  19. Fowler FJ. Some general rules for designing good survey instruments. In: Fowler FJ, editor. Improving survey questions: Design and evaluation. Thousand Oaks, CA: Sage Publications; 1995. pp. 78–103. [Google Scholar]
  20. Garofalo J. Hate crime victimization in the United States. In: Davis RC, Lurgio AJ, editors. Victims of crime. 2nd. Thousand Oaks, CA: Sage Publications; 1997. [Google Scholar]
  21. Grossman AH. Avoiding HIV/AIDS and the challenge of growing up gay, lesbian, and bisexual. In: D'Augelli AR, Patterson CJ, editors. Lesbian, gay, and bisexual identities and youth: Psychological perspectives. New York: Oxford University Press; 2001. pp. 155–180. [Google Scholar]
  22. Herek G. Attitudes toward lesbians and gay men: A factor-analytic strategy. Journal of Homosexuality. 1984;10(1):39–51. doi: 10.1300/J082v10n01_03. [DOI] [PubMed] [Google Scholar]
  23. Herek G. Hate crimes against lesbians and gay men: Issues for research and policy. American Psychologist. 1989;44(6):948–955. doi: 10.1037//0003-066x.44.6.948. [DOI] [PubMed] [Google Scholar]
  24. Herek G. The Psychology of Sexual Prejudice. Current Directions in Psychological Science. 2000a;9(1):19–22. [Google Scholar]
  25. Herek G. Sexual prejudice and gender: Do heterosexuals' attitudes toward lesbians and gay men differ? Journal of Social Issues. 2000b;56(2):251–266. [Google Scholar]
  26. Herek G. Gender gaps in public opinion about lesbians and gay men. Public Opinion Quarterly. 2002;66(1):40–66. [Google Scholar]
  27. Herek G. Beyond “homophobia”: Thinking about sexual prejudice and stigma in the 21st century. Sexuality Research and Social Policy. 2004;1(2):6–24. [Google Scholar]
  28. Herek G, Capitanio JP. “Some of my best friends”: Intergroup contact, concealable stigma, and heterosexuals' attitudes toward gay men and lesbians”. Personality and Social Psychology Bulletin. 1996;22(4):412–424. [Google Scholar]
  29. Herek G, Capitanio JP. Sex differences in how heterosexuals think about lesbians and gay men: Evidence from survey context effects. The Journal of Sex Research. 1999;36(4):348–360. [Google Scholar]
  30. Herek GM, Gills JR, Cogan JC. Psychological sequelae of hate-crime victimization among lesbian, gay, and bisexual adults. Journal of Consulting & Clinical Psychology. 1999;67(6):945–951. doi: 10.1037//0022-006x.67.6.945. [DOI] [PubMed] [Google Scholar]
  31. Herek GM, González-Rivera M. Attitudes toward homosexuality among U.S. residents of Mexican descent. Journal of Sex Research. doi: 10.1080/00224490609552307. in press. [DOI] [PubMed] [Google Scholar]
  32. Hu L, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling. 1999;6:1–55. [Google Scholar]
  33. Kite ME. Individual differences in males' reactions to gay males and lesbians. Journal of Applied Social Psychology. 1992;22(15):1222–1239. [Google Scholar]
  34. Klem L. Structural Equation Modeling. In: Grimm LG, Yarnold PR, editors. Reading and understanding more multivariate statistics. Washington, DC: American Psychological Association; 2000. pp. 227–260. [Google Scholar]
  35. Kline RB. How to fool yourself with SEM. In: Kline RB, editor. Principles and practice of structural equation modeling. New York: The Guildford Press; 1998. pp. 273–281. [Google Scholar]
  36. MacCallum RC, Wegener DT, Uchino BN, Fabrigar LR. The problem of equivalent models in applications of covariance structure analysis. Psychological Bulletin. 1993;114(1):185–199. doi: 10.1037/0033-2909.114.1.185. [DOI] [PubMed] [Google Scholar]
  37. Mays VM, Cochran SD. Mental health correlates of perceived discrimination among lesbian, gay, and bisexual adults in the United States. American Journal of Public Health. 2001;91(11):1869–1876. doi: 10.2105/ajph.91.11.1869. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Meyer IH. Minority stress and mental health in gay men. Journal of Health and Social Behavior. 1995;36(3):38–56. [PubMed] [Google Scholar]
  39. Meyer IH, Dean L. Internalized homophobia, intimacy, and sexual behavior among gay and bisexual men. In: Herek GM, editor. Psychological perspectives on lesbian and gay issues: Volume 4 Stigma and sexual orientation: Understanding prejudice against lesbians, gay men, and bisexuals. Thousand Oaks, CA: Sage Publications; 1998. pp. 160–186. [Google Scholar]
  40. Michener HA, DeLamater JD, Myers DJ. Social Psychology. Fifth. Belmont, CA: Thomson Wadsworth Learning; 2004. [Google Scholar]
  41. Nelson ES, Krieger SL. Changes in attitudes toward homosexuality in college students: Implementation of a gay men and lesbian peer panel. Journal of Homosexuality. 1997;33(2):63–81. doi: 10.1300/J082v33n02_04. [DOI] [PubMed] [Google Scholar]
  42. Nunnally JC, Bernstein IH. Psychometric Theory. Third. New York: McGraw-Hill Inc.; 1994. [Google Scholar]
  43. Peterson JL, Carballo-Diéguez A. HIV Prevention among African-American and Latino Men who have sex with men. In: Peterson J, DiClemente R, editors. Handbook of HIV Prevention. New York: Kluwer Academic Plenum Publishers; 2000. pp. 217–224. [Google Scholar]
  44. Raghunathan TE. What do we do with missing data? Some options for analysis of incomplete data. Annual Review of Public Health. 2004;25:99–117. doi: 10.1146/annurev.publhealth.25.102802.124410. [DOI] [PubMed] [Google Scholar]
  45. Raykov T, Tomer A, Nesselroade JR. Reporting structural equation modeling results in Psychology and Aging: Some proposed guidelines. Psychology and Aging. 1991;6(4):499–503. doi: 10.1037//0882-7974.6.4.499. [DOI] [PubMed] [Google Scholar]
  46. Rosario M, Hunter J, Maguen S, Gwadz M, Smith R. The coming-out process and its adaptational and health-related associations among gay, lesbian, and bisexual youths: Stipulation and exploration of a model. American Journal of Community Psychology. 2001;29(1):133–160. doi: 10.1023/A:1005205630978. [DOI] [PubMed] [Google Scholar]
  47. Russell ST, Seif H, Truong NL. School outcomes of sexual minority youth in the United States: Evidence from a national study. Journal of Adolescence. 2001;24:111–127. doi: 10.1006/jado.2000.0365. [DOI] [PubMed] [Google Scholar]
  48. Sakalli N. The relationship between sexism and attitudes toward homosexuality in a sample of Turkish college students. Journal of Homosexuality. 2002;42(3):53–64. doi: 10.1300/J082v42n03_04. [DOI] [PubMed] [Google Scholar]
  49. Sakalli N, Ugurlu O. Effects of social contact with homosexuals on heterosexual Turkish university students' attitudes toward homosexuality. Journal of Homosexuality. 2001;42(1):53–62. doi: 10.1300/j082v42n01_03. [DOI] [PubMed] [Google Scholar]
  50. Sherrod D, Nardi PM. Homophobia in the courtroom: An assessment of biases against gay men and lesbians in a multiethnic sample of potential jurors. In: Herek GM, editor. Stigma and sexual orientation: Understanding prejudice against lesbians, gay men, and bisexuals. Thousand Oaks, CA: Sage Publications; 1998. pp. 24–38. [Google Scholar]
  51. U.S. Census Bureau. Profiles of General Demographic Characteristics: 2000 Census of Population and Housing. Washington, DC: United States Printing; 2001. [Google Scholar]
  52. Vinokur AD, Pierce PF, Buck CL. Work-family conflicts of women in the Air Force: their influence in mental health and functioning. Journal of Organizational Behavior. 1999;20:865–878. [Google Scholar]
  53. Wager G, Serafini J, Rabkin J, Remien RH, Williams J. Integration of one's religion and homosexuality: A weapon against internalized homophobia? Journal of Homosexuality. 1994;26(4):91–110. doi: 10.1300/J082v26n04_06. [DOI] [PubMed] [Google Scholar]

RESOURCES