Social workers responsible for developing rape prevention programs on college campuses must have valid evaluation instruments. This article presents the challenges encountered by the authors when they attempted to keep rape myth measures relevant to student populations by updating the language to reflect the subtleties involved with rape myths. The development of a modified version of the Illinois Rape Myth Acceptance Scale is described. Focus groups were conducted to gather feedback about the language used by college students related to sexual encounters and rape. The instrument was then tested with 951 undergraduate students at a large northeastern university. Exploratory structural equation modeling was used to assess the factor structure of the scale. In addition, multiple-indicators multiple-causes modeling was used to assess the potential differential item functioning of the measure's items by gender, previous experience with sexual assault prevention programming, and knowing someone who was sexually assaulted. A four-factor structure was hypothesized and a five-factor structure supported, indicating a separate factor that looks at alcohol and accountability. Implications for social workers are discussed, including the necessity of continuously updating rape myth measures to ensure validity.
KEY WORDS: measurement; prevention; rape
According to the most recent National Violence Against Women Survey, 17.7 million women in the United States were raped at some point in their lives (Tjaden & Thoennes, 2006). The high number of victims has resulted in the implementation of rape prevention programs in a variety of social work practice settings, especially in high schools and on college campuses, where rape seems to be especially problematic. The primary focus of rape prevention programs in schools is often on changing individuals' beliefs in rape myths, defined as false beliefs about rape shaped by sexism and other prejudices individuals hold (Burt, 1980; Lonsway & Fitzgerald, 1994). Common rape myths cited over time include the belief that the way a woman dresses or acts indicates that she asked for it and that rape occurs because men cannot control their sexual impulses. Researchers have demonstrated that the acceptance of rape myths not only indicates problematic attitudes, but also is an explanatory predictor in the actual perpetration of sexual violence or proclivity to rape (Hinck & Thomas, 1999).
The failure to have psychometrically sound outcome measures has hampered the accumulation of evidence-based practice knowledge in this area (Farmer & McMahon, 2005; Lonsway & Fitzgerald, 1995). One of the major validity problems with rape myth measures hinges on the issue of language. The reliability and validity of surveys depends, in part, on questions being clear and relevant to the respondents, yet the language used in rape myth measures is often outdated, antiquated, and irrelevant to groups such as high school and college students. Many rape education prevention programs use instruments that were developed years or even decades ago, which raises serious concerns about their validity for high school and college students, as the instruments' language and context are not a part of current student culture.
SUBTLE RAPE MYTHS
Many of the measures currently used to assess rape myth attitudes fail to capture the more subtle and covert rape myths that have evolved. As many high schools and colleges have implemented some form of education on issues of sexual violence over the past decade, students have greater awareness that certain traditional rape myths are not socially acceptable (Frazier, Valtinson, & Candell, 1994). However, these myths may exist in various, more subtle and covert forms that are not being accurately assessed because of the phrasing of questions, language used, and subsequent social desirability bias. The complexity of rape myths is especially apparent among college students, who likely received exposure to rape prevention education in some shape or form by the time they graduated from high school. For example, Hinck and Thomas (1999) found that college students with previous rape education had less adherence to rape myth beliefs. They suggested that this may have occurred, in part, because of the "obvious" phrasing of many rape myth measurement items.
Research on subtle sexism provides a framework for studying subtle rape myths. Benokraitis and Feagin (1995) are most often cited for identifying three categories of sexism: overt, covert, and subtle sexism. Overt sexism includes blatant, direct, and observable treatment of women that is unequal and unfair. Covert sexism is unequal treatment of women that is intended but is deliberately hidden (Swim, Mallett, & Stangor, 2004). Subtle sexism is unequal treatment of women that is hidden because it is regarded as normalized behavior (Swim & Cohen, 1997). Research suggests that over time, prejudicial beliefs about women have become more covert and subtle, with less expressed agreement with overtly sexist attitudes and beliefs (Swim & Cohen, 1997). As the newer forms of "modern" sexism have been acknowledged, so too has been the need to develop updated measures such as the Modern Sexism Scale (Swim, Aikin, Hall, & Hunter, 1995).
Under the umbrella of sexist beliefs, violence against women can be regarded as one subset of beliefs. Rape has long been viewed by feminist scholars as a function of sexism and a form of patriarchal control over women (Brownmiller, 1975). Progress by the antirape movement over the past 40 years has included an increased awareness of the problem, proliferation of services for victims, legislative reform, and the creation of policies to address the issue. Just as with sexism in general, one of the results of these efforts has been a cultural shift and change in acceptable discourse regarding women and rape. Arguably, certain overt, victim-blaming, sexist beliefs and attitudes that may have been socially acceptable 40 years ago are often no longer regarded as tolerable, at least overtly. This shift toward less overt victim blaming is illustrated in the judicial system. Over the past few decades, measures to block overt expressions of victim blaming have been implemented, such as rape shield laws that prohibit discussion of rape victims' sexual histories and removal of requirements for alleged victims such as proof of resistance or corroboration by a witness. Despite this progress, research indicates that beliefs in rape myths still influence lawyers' perspectives as well as juries' and judges' decisions in rape cases to rule in favor of alleged perpetrators (Ehrlich, 2001; Krahe, Temkin, Bieneck, & Berger, 2008).
In particular, the shift to more subtle and covert rape myths appears to affect expressions about victim blaming. It can be argued that as overt sexism in general has declined, those rape myths that blatantly blame girls and women for rape have become less acceptable. However, many of the underlying beliefs that the girls and women did something to contribute to the assault and that it is not completely the perpetrator's fault still exist but in more covert expressions. For example, in a study conducted with college student-athletes, McMahon (2005) found that respondents would not directly blame the victim for her assault but expressed the belief that women put themselves in bad situations by dressing a certain way, drinking alcohol, or demonstrating other behaviors such as flirting. In addition, some respondents indicated a belief that rape could happen accidentally or unintentionally and that there are certain situations in which men should not be held entirely accountable for sexual assault.
The Illinois Rape Myth Acceptance Scale (IRMA) is arguably the most reliable and psychometrically demonstrated rape myth scale to date (Payne, Lonsway, & Fitzgerald, 1999). The scale consists of a general rape myth construct and seven subscales: (1) She Asked for It, (2) It Wasn't Really Rape, (3) He Didn't Mean To, (4) She Wanted It, (5) She Lied, (6) Rape Is a Trivial Event, and (7) Rape Is a Deviant Event. The resulting 45-item scale was tested with a sample of 604 undergraduate students. The overall scale reliability was .93, with subscale alphas ranging from .74 to .84 (Payne et al., 1999). The IRMA authors conducted a series of studies to demonstrate the scale's construct validity through the relationship of the IRMA to empirically and theoretically related rape acceptance variables (see Payne et al., 1999). The IRMA has demonstrated predictive validity through its positive correlation with men's actual rape proclivity and sexual aggression (Stephens & George, 2009) and related variables such as hostile sexism toward women (Chapleau, Oswald, & Russell, 2007). Due to the potential challenges of implementing a lengthy scale, a 20-item short form of the IRMA was developed that includes items from all seven subscales (Payne et al., 1999).As with the long version of the scale, individuals responded to the items using a five-point Likert-type scale (ranging from 1 = strongly agree to 5 = strongly disagree). Higher scores indicate greater rejection of rape myths.
Despite the significant progress made by the introduction of the IRMA, issues of validity remain, especially regarding updated language and the ability to capture subtle rape myths. Now that the measure is more than 10 years old, the need for updated language is especially salient. The colloquial phrases and sexual slang that were used a decade ago have likely changed on college campuses. As noted by Payne et al. (1999), rape myth measures "are necessarily time and culture bound. Several items use colloquial phrases that might be unclear to certain people or could quickly become outdated. This problem is not easily avoided, however, as sexual communication relies heavily on slang terminology" (p. 61). In addition, sexual slang...