Because of inadequacies in the methods used to measure sexual assault, national crime statistics, criminal victimization studies, convictions, or incarceration rates fail to reflect the true scope of rape. Studies that have avoided the limitations of these methods have revealed very high rates of overt rape and lesser degrees of sexual aggression. The goal of the present study was to extend previous work to a national basis. The Sexual Experiences Survey was administered to a national sample of 6,159 women and men enrolled in 32 institutions representative of the diversity of higher education settings across the United States. Women's reports of experiencing and men's reports of perpetrating rape, attempted rape, sexual coercion, and sexual contact were obtained, including both the rates of prevalence since age 14 and of incidence during the previous year. The findings support published assertions of high rates of rape and other forms of sexual aggression among large normal populations. Although the results are limited in generalizability to postsecondary students, this group represents 26% of all persons aged 18-24 in the United States.The Federal Bureau of Investigation (FBI) defines rape as "carnal knowledge of a female forcibly and against her consent" and reports that 87,340 such offenses occurred in 1985 (FBI, 1986). However, these figures greatly underestimate the true scope of rape because they are based only on instances reported to police. Government estimates suggest that for every rape reported, 3-10 rapes are committed but not reported (Law Enforcement Assistance Administration [LEAA], 1975). Likewise, it is difficult to obtain realistic estimates of the number of men who perpetrate rape because only a fraction of reported rapes eventually result in conviction (Clark & Lewis, 1977).Victimization studies, such as the annual National Crime Survey (NCS), are the major avenue through which the full extent of the crime is estimated (e.g., Bureau of Justice Statistics [BJS], 1984). In these studies, the residents of a standard sampling area are asked to indicate those crimes of which they or anyone else in their household have been victims during the previous 6 months. These rates are then compared with official crime statistics for the area and the rate of unreported crime is estipervised the sampling procedures, and Ann Maney of the National Institute of Mental Health who provided technical advice.

Many rape studies use judicial records or crisis center files to recruit research participants. Recent studies, however, have suggested that reported rape rates greatly underestimate the number of rapes that occur each year, that the conviction rate for rape is low, and that few victims utilize rape crisis centers. The Sexual Experiences Survey, a self-report instrument designed to identify hidden rape victims and undetected offenders among a normal population, is an alternate approach to sample selection. In the present study, reliability and validity data for the Sexual Experiences Survey are described.Much rape research relies on judicial identification of offenders or recruitment of victims through rape crisis centers. However recent studies have suggested that only 40%-50% of rapes are reported (Law Enforcement Assistance Administration [LEAA], 1975) and that the conviction rate for rape is 16% of reported crimes (LEAA, 1977). Furthermore, the number of victims who utilize rape crisis center services may be as low as 4% (Koss, in press). These data suggest the existence of many hidden victims and undetected offenders in the general population and emphasize the need to go beyond police and crisis center files to obtain individuals for participation in rape research.Additionally, judicial or crisis-center-based methods of sample selection rest on a typological approach to rape. Thus, an individual is labeled either a rapist, a rape victim, or a comparison subject. Several writers have suggested that a dimensional view of sexual aggression and victimization be adopted in rape research. In this framework, rape represents an extreme behavior but one that is on a continuum with normal sexual behaviors within the culture. The Sexual Experiences Survey (Koss & Oros, 1982) has been described as a self-report instrument that is designed to reflect various degrees of sexual aggression and victimization and is capable of identifying hidden rape victims and undetected offenders for participation in research. This article reports reliability and validity data to support the survey. MethodThe original items (Koss & Oros, 1982) were reworded slightly to increase clarity, improve con-

Men and women living in randomly selected 1st-year dormitories participated in tailored single-sex sexual assault prevention or risk-reduction programs, respectively. An evaluation of the men's project is presented (N = 635). The program incorporated social norms and bystander intervention education and had an impact on self-reported sexual aggression and an effect on men's perceptions that their peers would intervene when they encountered inappropriate behavior in others. Relative to the control group, participants also reported less reinforcement for engaging in sexually aggressive behavior, reported fewer associations with sexually aggressive peers, and indicated less exposure to sexually explicit media.

No abstract

Research has indicated that sexual assault on college campuses is pervasive (Koss, Gidycz, & Wisniewski, 1987) and that a history of sexual victimization serves as a risk factor for future victimization (Gidycz, Coble, Latham, & Layman, 1993). The purpose of the present investigation was to extend the findings of Gidycz et al. (1993) in examining the link among sexual victimization experiences. College women were evaluated for child and adolescent sexual victimization, family adjustment, alcohol use, psychological adjustment, interpersonal functioning, and sexual behavior. Women were reevaluated at 3, 5–6, and 9 months for adult victimization, psychological adjustment, interpersonal functioning, and sexual behavior. Loglinear analysis indicated that chances of being victimized in one time period increased with greater severity of victimization in the preceding time period. The path analysis assessing the mediating effects of these variables on victimization experiences was partially supported. Implications for future research and preventative efforts are discussed.