Sexual violence is unfortunately a common problem in the United States. The Federal Bureau of Investigation (2000) estimates that one in four women will be sexually assaulted in her lifetime and in a national survey of college women, 53.7% of the participants reported experiencing some form of sexual violence (Koss, Gidycz, & Wisniewski, 1987). Not only are such experiences frequent but they are associated with significant physical and mental health difficulties (Goodman, Koss, & Russo, 1993) and have implications for cultural norms and values. Despite this, we know little about the causes of sexual violence.

It is possible that predominant cultural attitudes, at least in part, facilitate continued tolerance of aggression toward women, and thus the occurrence of sexual violence. The ecological model has been utilized as a framework to understand the multiple factors involved in the occurrence of sexual assault and provides a nice model for considering the role of cultural values (Belsky, 1980; Bronfenbrenner, 1977, 1979; Grauerholz, 2000; Messman-Moore & Long, 2002; Nurius & Norris, 1996; White & Koss, 1993). This integrated model suggests that a person’s behavior can only be understood if the individual is considered in the context of the microsystem or family, the exosystem or larger social system in which the family is embedded, and the macrosystem or the cultural norms (Bronfenbrenner, 1977, 1979). Given that sexual violence continues to occur at high rates in the United States, it is vital that we understand attitudes and cultural norms that serve to minimize or foster tolerance of sexual violence.

Rape myths are a specific set of attitudes and beliefs that may contribute to ongoing sexual violence by shifting blame for sexual assault from perpetrators to victims. A number of studies have illustrated that many people, both women and men, from various backgrounds ascribe to rape myths (for review see Lonsway & Fitzgerald, 1994). Similarly, studies have demonstrated that high rape myth acceptance is associated with perpetration of sexual assaults (e.g., Abbey, McAuslan, & Ross, 1998; Koss & Dinero, 1989; Koss, Leonard, Beezley, & Oros, 1985; Malamuth, Socklosie, Koss, & Tanaka, 1991; Murphy, Coleman, & Haynes, 1986). Thus, both perpetrators of sexual assault and people in the general population report beliefs that tolerate and even support sexual violence. Despite this evidence, most researchers have not gone beyond examination of rape myth acceptance to consider the role of other forms of intolerance (e.g., racism, homophobia, classism, ageism, religious intolerance) in predicting the occurrence of sexual assault (with the exception of sexism).

Thus, another set of variables to consider in studies of rape myth acceptance is oppressive beliefs, such as sexism, racism, homophobia, ageism, classism, and religious intolerance. It may seem intuitive that these constructs are interrelated and a logical extension might suggest that each would be related to rape myth acceptance as well as sexual violence perpetration. In fact, an increasing number of prevention educators have begun to address issues of racism, sexism, homophobia, and other forms of intolerance in their work (Cohen, Parks, Flores, & Culross, 2006; Funk, 1993, 2006; Katz, 2006; Lang & Lee-Pethel, 2006; Wantland, 2005). There are also several psychological theories that would suggest multiple forms of intolerance should be related. As long as 60 years ago, Allport (1954) suggested that individuals who demonstrate prejudice against a group likely have a rigid, intolerant cognitive style that results in prejudice toward multiple groups or topics. Similarly, social dominance theory (Sidanius, Pratto, van Laar, & Levin, 2004), or the idea that certain groups should be dominant over other groups, allows for the logical conclusion that if one form of intolerance is related to perpetration of sexual assault, then other forms of intolerance might also be related to sexual assault. Both social dominance theory and the ecological model more generally point to the need for study of multiple contextual factors and cultural belief systems, but there is a dearth of empirical evidence about the interrelationships of sexual assault related beliefs and other oppressive belief systems. It was the purpose of the present study to consider six such belief systems: sexism, racism, homophobia, classism, ageism, and religious intolerance.

Sexism (as characterized by negative attitudes toward women, their social roles, and their traditional gender roles) is frequently examined in relation to rape myth acceptance. A number of studies have demonstrated that negative stereotypical attitudes toward, and beliefs about, women are associated with greater rape myth acceptance (Lonsway & Fitzgerald, 1994). Specifically, this has been found in college student samples (e.g., Johnson, Kluck, & Schander, 1997; Larsen & Long, 1988; Lonsway & Fitzgerald, 1995) as well as nonstudent samples (e.g., Burt, 1980; Costin & Schwarz, 1987).

Racism has been defined as deeply and emotionally held stereotypes about racial or ethnic groups that persist in the face of social change and affect the behavior of the individuals who hold the beliefs (Kowalewski, McIlwee, & Prunty, 1995). To date, no studies are available in the literature that investigate the interrelationships of racist beliefs and rape myth acceptance. Racism and sexism, however, have a long history of being linked theoretically (Glick & Fiske, 1996; Lewis, 1977; Swim, Aikin, Hall, & Hunter, 1995). Due to the many parallels between the experiences of women and racial minorities, researchers (Glick & Fiske, 1996; Sidanius, 1993; Swim et al., 1995) have begun to explore the co-occurrence of racism and sexism and have revealed that, as would be suspected, greater endorsement of racist beliefs is associated with greater endorsement of sexist beliefs. Given the known ties between sexism and racism, exploration of the relationship between racism and rape myth acceptance appears warranted.

Homophobia, another intolerant belief system, was originally defined as the fear of being near homosexuals (Smith, 1971). More recently, the term has been used to refer to a variety of negative reactions to, and stereotypes about, gay, lesbian, and bisexual individuals (Polimeni, Hardie, & Buzwell, 2000). Analogous to the literature on racism, no research to date has explored the interrelationship of homophobia and rape myth acceptance. Homophobia, however, has also been found to be associated with sexism (Agnew, Thompson, Smith, Gramzow, & Currey, 1993; Campbell, Schellenberg, & Senn, 1997; Polimeni et al., 2000; Thompson, Gristani, & Pleck, 1985). Given the known relationship between sexism and rape myth acceptance, an investigation focused on the association between homophobia and rape myth seems warranted.

Ageism, defined by Butler (1969, 1975, 1978), is considered to be “...institutionalized and individual prejudice against the elderly, stereotyping, myth-making, distaste, and/or avoidance” (Butler, 1978, p. 14). Although the definition of ageism is strikingly similar to those of the aforementioned intolerant attitudes (i.e., sexism, homophobia, and racism), the relationships between these intolerant beliefs systems have yet to be investigated empirically.

Classism is another intolerant belief system. Specifically, Lott (2002) defined classism as an institutional and individual distancing (i.e., discrimination), stereotyping, and prejudice against poor people. As with sexism, homophobia, racism, and ageism, there is the theme of intolerance toward the “other” who is different from the members of a majority group. Thus, classism serves to maintain the status quo by keeping the economically disadvantaged invisible and powerless and the wealthy powerful (Lott, 2002).

Religious intolerance can be conceptualized as stereotypes, prejudice, and discrimination against particular religious groups or individual members of those religious groups (Godfrey, Richman & Withers, 2000; Richman, Kenton, Helfst, & Gaggar, 2004). Despite its similarities to other forms of intolerance, religious intolerance is rarely studied by psychologists. In fact, there is no empirically supported measure of religious intolerance that considers stereotypes about multiple religions, and there is only one measure of intolerant attitudes that includes items regarding religion (Godfrey et al., 2000). Although the concept of religious intolerance is not nearly as advanced as those of other intolerant belief systems (e.g., sexism, racism, homophobia, or ageism), it is likely that religious intolerance is similar to other forms of intolerance and thus worthy of investigation.

Given the theoretical similarities between the aforementioned constructs of sexism, racism, homophobia, ageism, classism, and religious intolerance, an investigation of all factors together in relation to rape myth acceptance appears warranted. Thus, in the present study the collective and unique interrelationships of these factors were explored. Further, given the gendered nature of sexual coercion in our society (i.e., most victims are female, and most perpetrators are male) the impact of gender on the interrelationships of these variables was also explored.

It was hypothesized that greater rape myth acceptance would be associated with greater intolerance in the areas of sexism, racism, homophobia, ageism, classism, and religious intolerance. With regard to the moderating impact of gender, it was hypothesized that men with high levels of sexism would have particularly high levels of rape myth acceptance, the highest rape myth acceptance in comparison to the other four groups. Women with low levels of sexism were expected to have the lowest levels of rape myth acceptance, while men with low levels of sexism and women with high levels of sexism were expected to have moderate levels of rape myth acceptance. No specific hypotheses about the direction of possible moderating effects due to gender were made for other intolerant beliefs. Finally, it was hypothesized that each intolerant belief (i.e., sexism, racism, homophobia, ageism, classism, and religious intolerance) would uniquely predict a portion of the variance in rape myth acceptance.

Method

Participants

Participants were 492 male and 506 female students recruited from a research participant pool for a study on student attitudes. Participants ranged in age from 18 to 55 years, with an average of 20.18 years (SD = 3.24). The majority of individuals reported they had never been married (91.1%; n = 879); 5.0% (n = 48) reported they were married or cohabitating, 1.0% (n = 10) reported they were divorced or separated, and 2.9% (n = 28) reported themselves in the “other” category. The majority of participants were European Americans (83.7%; n = 835); 2.6% (n = 26) were African Americans, 2.2% (n = 22) were Hispanics, 4.4% (n = 44) were Native Americans, 5.3% (n = 53) were Asian/Asian Americans, and 1.8% (n = 18) placed themselves in the “other” category. Socio-economic status (SES) was assessed using the two factor index of social position (Myers & Bean, 1968) and ranged from lower to upper class; the average participant fell into the middle class. The majority of participants were heterosexual (98.4%; n = 977); 0.40% (n = 4) were gay men, 0.20% (n = 2) were lesbians, 0.60% (n = 6) identified as bisexual, and 0.40% (n = 4) were undecided/questioning. Finally, the majority of participants were Protestants (67.3%; n = 671); 13.1% (n = 131) were Catholics, 2.1% (n = 21) were Buddhist/Muslim/Hindu, 3.2% (n = 32) were agnostic/atheist, 0.3% (n = 3) were Wiccan/pagan, 0.1% (n = 1) were Jewish, 9.7% (n = 97) were nonaffiliated, and 4.2% (n = 41) identified themselves as “other.”

Measures

Illinois Rape Myth Acceptance Scale (IRMA)

The IRMA is a 45-item self-report instrument developed to measure the complex set of cultural beliefs that serve to support and perpetuate sexual violence (Payne, Lonsway, & Fitzgerald, 1999). Example items include “Many women secretly desire to be raped” and “Men from nice middle-class homes almost never rape.” Items are responded to on a Likert-type scale that ranges from 1 (totally disagree) to 7 (totally agree). The IRMA provides a mean score; higher IRMA scores indicate higher levels of rape myth acceptance.

Internal consistency for the IRMA total score has been reported to be 0.93 (Payne et al., 1999). Internal consistency for the overall scale was also calculated for this sample and resulted in a α of 0.95. The construct validity of the scale has also been supported, as the IRMA has been found to correlate with measures of gender-role stereotyping, adversarial sexual beliefs, adversarial heterosexual beliefs, hostility toward women, and acceptance of interpersonal violence (Payne et al., 1999).

The Attitudes Toward Women Scale (AWS)

The 15-item short version AWS (Spence & Helmreich, 1972) was developed to measure attitudes toward the rights and roles of women. The AWS is over two decades old, yet continues to be the most commonly used measure of gender-related attitudes toward women (McHugh & Frieze, 1997; Spence & Hahn, 1997), although it has been suggested that the AWS measures old-fashioned sexism rather than subtler, modern sexism (McHugh & Frieze, 1997). The AWS includes such items as “There should be a strict merit system in job appointment and promotion without regard to sex” and “The intellectual leadership of a community should be largely in the hands of men.” Items are responded to on a Likert-type scale from 0 (strongly disagree) to 3 (strongly agree). Items are summed to create a total score that ranges from 0 to 45; higher scores reflect more negative attitudes toward women.

Internal consistency has been demonstrated for the 15-item short version of the AWS. Specifically, Daugherty and Drambrot (1986) found a Cronbach’s alpha of 0.85 for the 15-item version. Internal consistency for the scale was calculated for the present sample and resulted in a α of 0.81. The 15-item version has a 3-week test-retest reliability of 0.82 for men and 0.86 for women (Daugherty & Drambrot, 1986). The validity of the scale has also been supported, as the short form is almost perfectly correlated with the original version (Loo & Logan, 1977; Smith & Bradley, 1980; Spence & Hahn, 1997). In addition, the construct validity of numerous other measures of sexism, attitudes toward women, and attitudes toward gender roles have been established by their strong correlations with the AWS (e.g., Tougas, Brown, Beaton, & Joly, 1995).

The Neosexism Scale (NS)

The Neosexism Scale was developed to measure the construct of modern sexism, or the conflict between negative attitudes toward women and egalitarian values (Tougas et al., 1995). Example items include “Women shouldn’t push themselves where they are not wanted” and “Due to social pressures, firms frequently have to hire underqualified women.” Items are responded to on a scale that ranges from 1 (total disagreement) to 7 (total agreement). Scores are calculated by averaging the ratings of the 11 items; higher scores indicate greater levels of sexism.

The 11-item Neosexism Scale has demonstrated good internal reliability (alpha = 0.81) and corrected item-total correlations range from 0.10 to 0.76 (Campbell et al., 1997; Tougas et al., 1995). Internal consistency for the scale was calculated for the present sample and resulted in a α of 0.82. Furthermore, principal component analysis revealed that the scale is unidimensional (Campbell et al., 1997). The construct validity of the Neosexism Scale has also been supported, as it is correlated with the Modern Sexism Scale, the Attitudes Toward Feminism Scale, and the Women’s Movement Scale (Campbell et al., 1997).

The Modern Homophobia Scale (MHS)

The 46-item MHS (Raja & Stokes, 1998) measures both attitudes toward lesbians and attitudes toward gay men. This is a strength, given that many of the previous homophobia scales do not refer specifically to lesbians or gay men but refer instead to “homosexuals” in general. In addition, the MHS was developed to update existing homophobia scales in an attempt to tap into modern, subtler forms of homophobia. Both lesbian (MHS-L; 24 items) and gay men (MHS-G; 22 items) subscales are scored from the instrument, and each reflects respondents’ institutional homophobia, personal discomfort, and beliefs that homosexuality is deviant and changeable. Example items include “I wouldn’t mind working with a lesbian” and “I welcome new friends who are gay.” Items are responded to on a Likert-type scale that ranges from 1 (do not agree) to 5 (strongly agree). Scores for each subscale are calculated by averaging subscale items; lower scores indicate higher levels of homophobia.

The 46-item MHS has demonstrated good internal consistency with alphas of 0.95 for both the MHS-L and MHS-G subscales (Raja & Stokes, 1998). In addition, internal consistency was calculated for both the MHS-L and MHS-G subscales for the present sample and resulted in alphas of 0.91 and 0.95, respectively. There is also evidence to support the construct validity of the MHS (Raja & Stokes, 1998). For example, the MHS-L and the MHS-G correlated significantly with Hudson and Ricketts’ (1980) Index of Homophobia (Raja & Stokes, 1998).

The Modern and Old Fashioned Racism Scale

This 14-item scale contains two 7-item subscales that measure old fashioned and modern racism (McConahay, 1986). The Old Fashioned Racism Scale contains items that tap into pre-1965 civil rights issues related to equal rights for minorities and stereotypes related to those same issues. The Modern Racism Scale was created in an attempt to measure racial attitudes after 1965 and includes items that are less blatant (McConahay, 1986). In addition, the Modern Racism items tap into the idea that modern racism is founded in abstract principles of justice and generalized negative feelings toward racial minorities that are related to political and racial socialization rather than personal competition or experiences with racial minorities. Old fashioned and modern example items include, respectively, “Black people are generally not as smart as Whites” and “Blacks are getting too demanding in their push for equal rights.” Items are responded to on a Likert-type scale that ranges from 1 (strongly agree) to 5 (strongly disagree). Scores for each scale are calculated by summing the ratings of the seven items in each scale. Scores range from 7 to 35; higher scores indicate higher levels of both modern and old fashioned racism. Although McConahay’s instrument is focused on attitudes toward African Americans, the focus of the present study was racial prejudice against any ethnic minority group. Therefore, “minority” was substituted for “Black” in each item, as per Ducote-Sabey (1999).

The internal consistency of the Modern Racism Scale has been demonstrated with a Cronbach’s alpha coefficient of 0.82 (McConahay, 1986). In addition, internal consistency has been demonstrated for the Old Fashioned Racism Scale; alphas range from 0.75 to 0.79 in various samples (McConahay, 1986). Ducote-Sabey (1999) calculated internal consistency for the “minority” modification to this scale and reported alpha coefficients of 0.77 and 0.63 for the Modern and Old Fashioned scales, respectively. Internal consistency was calculated in the present sample and resulted in a Cronbach’s alpha coefficient of 0.80 for Modern Racism and an α of 0.70 for Old Fashioned Racism.

The Fraboni Scale of Ageism (FSA)

This 29-item scale was developed to measure the affective and cognitive components of ageism (Fraboni, Saltstone, & Hughes, 1990). Sample items include “Complex and interesting conversations cannot be expected from most old people,” “It is best that old people live where they won’t bother anyone,” and “I sometimes avoid eye contact with old people when I see them.” Items are responded to on a Likert-type scale that ranges from 1 (strongly agree) to 4 (strongly disagree). Per scoring instructions from Fraboni et al. (1990), scores for each response are recoded on a Likert-type scale that ranges from 1 (strongly agree) to 5 (strongly disagree); unanswered items are scored as 3 (neutral). Scores for the scale are calculated by summing the ratings of the 29 items. Scores range from 29 to 145; higher scores indicate lower levels of ageism.

Internal consistency has been demonstrated with a Cronbach’s alpha coefficient of 0.86 (Fraboni et al., 1990). The alpha coefficient for the measure in the present sample was 0.84. In addition, a significant negative correlation between the FSA and a measure of acceptance of others supports the construct validity of the FSA (Fraboni et al., 1990).

Modified Economic Beliefs Scale (M-EBS)

This scale is a modified version of the Economic Beliefs Scale (Stevenson & Medler, 1995), which was designed to measure classism (i.e., attitudes toward the economically disadvantaged). The original scale contained eight items. In the version used for the present study the original items were retained and seven additional items were created (see Appendix A). Sample items from the original scale include “People who stay on welfare have no desire to work” and “Equal educational opportunities exist for all people in our society.” Example items that were created include “Poor people are lazy” and “If given the chance, a poor person would be able to keep a job.” Items are responded to on a Likert-type scale that ranges from 1 (strongly agree) to 5 (strongly disagree). Scores for the scale are calculated by summing the ratings of the 15 items. Scores range from 15 to 75; lower scores indicate higher levels of classism. The internal consistency reliability coefficient for the original measure was 0.77 (Stevenson & Medler, 1995) and it was 0.85 for the modified version used in the present study.

Religious Intolerance Scale (RIS)

Because there were no existing measures of religious intolerance that assess attitudes toward multiple religious groups at the time the present study was begun, the RIS was developed. The 9-item scale was developed using five items from the measure of prejudice of Godfrey et al. (2000) and four additional items that were created (see Appendix B). Example items from Godfrey et al. (2000) include “Jewish people are deceitful and money hungry” and “Muslims are more treacherous than other groups of religious people.” Example additional items include “Many of the social problems in the US today are due to non-Christian religious groups” and “Wiccan and pagan people practice thinly veiled evil.” Items are responded to on a Likert-type scale that ranges from 1 (strongly agree) to 5 (strongly disagree). Scores for the scale are calculated by summing the ratings of the nine items. Scores range from 9 to 45; lower scores indicate higher levels of religious intolerance. The alpha coefficient for the measure in the present sample was 0.79.

The Marlowe–Crowne Social Desirability Scale (M-C SDS) Short Form

The M-C SDS was developed to measure the desire of individuals to present themselves in a favorable manner (Reynolds, 1982). It was used in the present study to control for any reporting bias on the part of participants. The M-C SDS Short Form contains 13 true or false items. Example items include “I have never intensely disliked anyone” and “I never resent being asked to return a favor.” Responses are scored as socially desirable (0) or not socially desirable (1), and then summed to result in a total score that ranges from 0 (all socially desirable responses) to 13 (no socially desirable responses). Higher scores indicate lower levels of social desirability.

The internal consistency coefficient of the M-C SDS Short Form with the Kuder–Richardson formula 20 is 0.76 (Reynolds, 1982). Internal consistency for the scale was calculated for the present sample and resulted in a α of 0.70. In addition, the validity of the scale has been supported, as there are statistically significant correlations between the M-C SDS Short Form and the standard version of the M-C SDS as well as the Edward Social Desirability Scale (Reynolds, 1982).

The Life Experiences Questionnaire (LEQ)

For the purposes of this study, the LEQ (Long, 2000, unpublished manuscript) was used to gather demographic information.

Procedure

All participants were recruited from a research participant pool (made up of participants enrolled in Psychology and/or Marketing classes) and all received course credit for their participation. Participants took part in small 1-hour group testing sessions, and all responses were kept confidential and anonymous. After they gave informed consent, participants completed the questionnaire packet, which included all of the measures presented in random order.

For a number of participants, responses to individual items were missing. Values for missing data were imputed using the average response of the entire sample to the missing item. However, if a participant failed to complete a measure entirely, or left more than 25% of the items blank, his or her data for that particular measure were not included.

Results

Preliminary analyses

Prior to examination of the items of interest, a series of t-tests were conducted to replicate previously noted differences between men and women on the various attitudes studied here (e.g., Burt, 1980; Costin & Schwarz, 1987; Johnson et al., 1997; Larsen & Long, 1988; Payne et al., 1999; Sidanius, 1993). Consistent with previous literature, men reported higher levels of rape myth acceptance, modern sexism, old fashioned sexism, modern racism, old fashioned racism, homophobia toward gay men, ageism, classism, and religious intolerance than women did (all ps < 0.05, see Table 1). There were no gender differences evident in level of homophobia toward lesbians.

Table 1 Comparisons of men and women on constructs of interest.
Full size table

Further, in order to examine the intercorrelations between sexism, racism, homophobia, ageism, classism, and religious intolerance, simple Pearson’s correlation coefficients were calculated. Each of these intolerant belief systems was strongly correlated with the other belief systems (all ps < 0.05, see Table 2). In fact, the overlap is quite high in some instances (e.g., modern sexism and modern racism share 30.25% of their variance, and old fashioned sexism and homophobia toward gay men share 37% of their variance).

Table 2 Simple intercorrelations of study variables.
Full size table

Finally, to explore possible associations between the criterion variable (rape myth acceptance), predictor variables (sexism, racism, homophobia, ageism, classism, and religious intolerance), demographic variables (age, race/ethnicity, SES, marital status, religious affiliation, and sexual orientation), and social desirability, a number of t-tests (for categorical variables) and simple correlations (for continuous variables) were conducted. Findings indicated that there were some interrelationships between predictor variables, demographic variables, and social desirability. Although these relationships were statistically significant, the actual correlations were fairly small and may not be particularly meaningful. Nevertheless, we employed a conservative approach; planned analyses were conducted both including and excluding demographic factors as covariates. Results were not different, and thus only analyses without covariates are reported here.

Relationships between constructs and the moderating role of gender

The first purpose of the present study was to examine the interrelationship of rape myth acceptance with modern sexism, old fashioned sexism, modern racism, old fashioned racism, homophobia toward lesbians, homophobia toward gay men, ageism, classism, and religious intolerance. It was hypothesized that greater rape myth acceptance would be associated with greater intolerance in each area. Results of simple correlations supported this hypothesis (see Table 2).

The moderating impact of gender on each of these relationships was also explored. No specific hypotheses about the direction of possible moderating effects were made except in the case of sexism. With regard to this interaction, it was hypothesized that men with high levels of sexism would have particularly high levels of rape myth acceptance, i.e., the highest rape myth acceptance among the other groups. Women with low levels of sexism were expected to have the lowest levels of rape myth acceptance, whereas men with low levels of sexism and women with high levels of sexism were expected to have moderate levels of rape myth acceptance.

To examine moderation, the approach recommended by Baron and Kenny (1986) was employed. Hierarchical regression analyses were conducted to predict rape myth acceptance with (1) modern sexism, old fashioned sexism, modern racism, old fashioned racism, homophobia toward lesbians, homophobia toward gay men, ageism, classism, OR religious intolerance; (2) sex; and (3) the interaction of sex with the respective predictor variable. No moderating effects were found.

Collective impact

The final purpose of the present study was to examine the collective ability of sexism, racism, homophobia, ageism, classism, and religious intolerance to predict rape myth acceptance. To control for multicolinearity, factor scores were created to represent the predictor constructs of sexism, racism, and homophobia. Specifically, a principal components factor analysis with varimax rotation that required a one-factor solution was conducted to create a sexism score based on the combination of individuals’ NS and AWS scores. Similarly, a factor analysis was conducted to create a racism score based on individuals’ Modern Racism and Old Fashioned Racism scores. Likewise, a factor analysis was conducted to create a homophobia score based on individuals’ MHS-G and MHS-L scores.

Results of a multiple regression analysis indicated that sexism, ageism, classism, and religious intolerance each were significant predictors of rape myth acceptance (all p < 0.01; see Table 3). Racism and homophobia, however, failed to enter the model. Sexism, ageism, classism, and religious intolerance accounted for almost one-half (45%) of the variance in rape myth acceptance for the present sample. Sexism accounted for the greatest proportion of the variance (35%). The other intolerant beliefs accounted for relatively smaller amounts of variance beyond that of sexism: classism (2%), ageism (2%), and religious intolerance (1%).

Table 3 Regression analyses collectively predicting rape myth acceptance.
Full size table

Given the high intercorrelations between racism, homophobia, and sexism, a final exploratory analysis was conducted to determine whether racism and homophobia could be considered important predictors of rape myth acceptance if sexism were not considered in the context. We conducted a hierarchical regression analysis including rape myth acceptance as the criterion; predictor variables were entered in two blocks. The racism factor, homophobia factor, ageism, classism, and religious intolerance scores were entered in Step 1. Finally, the sexism factor score was allowed to enter the model in Step 2.

Results from Step 1 of this analysis (see Table 3) indicate that when considered outside of the context of sexism, both racism and homophobia, along with ageism, classism, and religious intolerance, are important predictors of rape myth acceptance. In fact, 36% of the variance in rape myth acceptance can be accounted for by these factors. Racism accounted for 13% of the unique variance, and homophobia accounted for 10% of the unique variance in rape myth acceptance; ageism accounted for 2%, as did classism (2%), and religious intolerance (2%). When sexism entered the model in Step 2, however, it demonstrated a stronger overlap with rape myth acceptance than either racism or homophobia. Racism and homophobia dropped out of the model as statistically significant predictors, leaving the sexism factor to account for 8% of the unique variance after other factors were considered.

Discussion

The purpose of the present study was to investigate the relationship between rape myth acceptance and sexism, racism, homophobia, ageism, classism, and religious intolerance, and to examine the impact of gender on the strength of these associations. As hypothesized, higher levels of racism, sexism, homophobia, ageism, classism, and religious intolerance were each associated with higher rape myth acceptance for both men and women. Correlations also show that all of the aforementioned constructs are related to one another. These interrelationships provide support for the idea that these beliefs are part of a larger belief system that is intolerant. However, it is important to note that, although these constructs are highly interrelated, they are not perfectly correlated, and, therefore, do reflect unique constructs.

Results also point to interesting patterns with regard to gender. Our data are consistent with past findings that showed that men report higher levels of rape myth acceptance, as well as racism, sexism, homophobia, ageism, classism, and religious intolerance than women do. No evidence for a moderating role of gender was found, however. Instead, it appears that oppressive beliefs systems predict rape myths in a manner similar for men and women. These results suggest that, when researchers try to understand rape myth acceptance, gender of participants per se (men will report more rape myth acceptance) and the role of adherence to traditional privileged or intolerant beliefs regarding sex, race, sexual orientation, age, class, or religion each should be considered.

Finally, results indicated that each of these intolerant beliefs systems was an important predictor of rape myth acceptance independently and in the context of each other. It is interesting that, when considered together, sexism, ageism, classism, and religious intolerance accounted for a substantial portion of the variability in rape myth acceptance (approximately 45%). Notably, sexism accounted for much more of the variance in rape myth acceptance than ageism, classism, or religious intolerance did. It was surprising that, racism and homophobia did not enter the initial model. The high overlap of racism and sexism, as well as the overlap between homophobia and sexism, appears to account for this finding, as each of these variables does become an important predictor of rape myth acceptance when considered outside of the context of sexism. In fact, 36% of the variance in rape myth acceptance can be accounted for by racism, homophobia, ageism, classism, and religious intolerance. In sum, each belief system does appear to play an important role in understanding the level of rape myth acceptance reported by men and women, however sexism is the best predictor.

In order to explore fully the implications of the present study, it may be helpful to return to a discussion of the ecological model. Bronfenbrenner’s (1977, 1979) model suggests that in order to understand human behavior we must consider the individual, the microsystem or family, the exosystem or larger social system, and the macrosystem or the cultural norms. Consistent with the ecological model, results here indicate that there are interrelationships between sexism, racism, homophobia, ageism, classism, and religious intolerance and rape myth acceptance at both the individual level and across individuals at the cultural level.

Specifically, these phenomena could be conceptualized as contemporary beliefs about masculinity at the cultural level that have been internalized by individuals. In particular, what it means to be masculine in our society often includes being young, strong, powerful, heterosexual, and a part of the majority group, as opposed to a minority group (e.g., Berkowitz, 1992; Berkowitz, Burkhart & Bourg, 1994; David & Brannon, 1976; Funk, 1993; Katz, 2006). Thus, it is possible that these cultural ideas about masculinity are directly related to rape myth acceptance in individuals who have internalized the cultural message about masculinity as a set of oppressive or intolerant beliefs. For example, masculinity as youth at the cultural level might be expressed as ageism at the individual level. Similarly, masculinity as heterosexuality at the cultural level might be expressed as homophobia toward gay men at the individual level. Likewise, the notion at the cultural level that masculinity requires power is likely expressed as a multitude of oppressive beliefs at the individual level. It is possible that if these beliefs are changed at the individual level, it might impact the family, social, and cultural levels. If oppressive beliefs are targeted in individuals, it is possible that cultural definitions of constructs such as masculinity also might change.

The results of the present study have many implications related to oppression and rape myth acceptance. For example, the intolerant beliefs that we examined predicted almost one-half of the variance in rape myth acceptance. Studying rape myth acceptance is important because it is implicated in actual perpetration behavior and may well be important for prevention of sexual violence. Thus, the current study offers further considerations for interventions. For instance, many sexual violence prevention programs specifically target the reduction of rape myth acceptance (e.g., Koss et al., 1985; Marx, Van Wie, & Gross, 1996). Our findings indicate that interventions focused on diversity or tolerance (of sex, race, sexual orientation, age, social class, or religious affiliation) might have the added benefit of reducing rape myth acceptance. Likewise, programs that directly target rape myth reduction may have the added benefit of reducing other intolerant belief systems, such as racism, sexism, homophobia, ageism, classism, and religious intolerance. However, there are other unexplained factors that contribute to rape myth acceptance. It is important to identify and include these factors in sexual violence prevention programs that focus on reduction of rape myth acceptance.

The results of the present study offer clear contributions to the literature by providing evidence of the interrelationships between racism, sexism, homophobia, ageism, classism, and religious intolerance, with rape myth acceptance. Although sexism has been identified as a predictor of rape myth acceptance (Burt, 1980; Johnson et al., 1997; Payne et al., 1999), ours is the first study to demonstrate the relationship between rape myth acceptance and racism, homophobia, ageism, classism, and religious intolerance. The use of a large sample size and many standardized, reliable, and valid measures for assessment of the constructs of interest represent additional strengths of the current study.

However, there are also limitations to the current study. One limitation is the fact that potential differences due to participants’ race, sexual orientation, marital status, and religious affiliation may have been overlooked. Due to the small numbers of participants of color, gay men/lesbians/bisexuals, and minority religions, differences that might exist between majority and non-majority groups were not tested. With large enough sample sizes it might be possible to see, for example, that homophobia would or would not be associated with rape myth acceptance in participants who were gay, lesbian, bisexual, or transgendered. In addition, the use of a college sample in the present study limits the generalizability of our findings. Only about 27% of the population has obtained a bachelor’s degree while 52.5% of the population has attended some college in the United States (Stoops, 2004), and thus these findings are most relevant for that group of people. However, it is important to point out that college age individuals are at the highest risk for sexual assault, and, therefore, it is important to examine these issues in this particular population (e.g., Marx et al., 1996). Despite these limitations, results from the present study have implications for future research and interventions.

With regard to future research, our results suggest that many forms of intolerant beliefs are associated with rape myth acceptance. Future researchers may benefit from exploring intolerant beliefs as a system rather than considering each type of belief as a fragment (as has been the case historically). There may be some common components, such as intolerance or distrust of others who are different, of various oppressive belief systems. Perhaps it is that common component that is responsible for the strong associations between rape myth acceptance, racism, sexism, homophobia, ageism, classism, and religious intolerance. It may be that each belief system is one facet of a system of intolerance for difference. Future investigators should consider whether other intolerant beliefs (e.g., anti-fat attitudes, intolerance of people with disabilities) are also related to rape myth acceptance. If different types of prejudiced beliefs are facets of a system of intolerance, one would expect associations between racism, sexism, homophobia, ageism, classism, religious intolerance, rape myth acceptance, and other specific prejudiced beliefs. Future researchers may want to explore the basis (e.g., cognitive style, or personality variables) for shared variance between multiple intolerant belief systems. In addition, future investigators may want to determine if differences in race, sexual orientation, and other demographic variables impact the relationship between oppressive beliefs and rape myth acceptance (e.g., exploring whether racism would predict rape myth acceptance in racial minorities).

Finally, future researchers could explore the relationship between intolerant belief systems, rape myth acceptance, and perpetration of violence. We are hopeful that findings from the present study might be helpful in planning sexual violence and other violence prevention programs (e.g., interventions to reduce elder abuse, gay bashing, and other hate crimes). A natural extension of this work is to begin consideration of whether and how intolerant beliefs and rape myth acceptance together may be related to perpetration of violence. More precisely, it may now be the time to ask “How is ending ageism related to ending sexual violence?” or “What similarities are there between perpetration of sexual aggression and stereotypes, prejudice, or discrimination aimed at minority groups?” as we struggle to build communities free from violence and oppression.