Bridging Disciplinary Boundaries (January 11 - 14, 2007)

Purpose: Research on sexual assault perpetration consists primarily of information gathered from victims, incarcerated perpetrators, or convenience samples of college men. Despite extensive research over the last 25 years, apparently no study has documented the prevalence of sexual assault perpetration by asking a representative sample of adult men in the U.S. about their life experiences, attitudes, and sexual assault behaviors and circumstances. We tested three methods for doing this.

Methods: The primary variable of interest consisted of a continuum of sixteen behaviors, ranging from trying to talk a woman into having sex to legal definitions of sexual assault, also accounting for whether men and/or the women were using alcohol at the time. Additionally, our survey included items about men's health, mental health, substance abuse, religiosity, and relationships with and attitudes toward women. We conducted three pilot studies using a nationally representative sample of men recruited by telephone using random digit dialing. In the first pilot, the survey was administered by telephone at the time of the call. In the second, men were given a web address and asked to take the survey on line. In the third, men were asked for their email addresses; the survey web address was emailed to them and a follow up email message was sent as a reminder. A university-based survey research center recruited participants in all three studies and conducted the telephone interviews in the first pilot. No incentive was offered in the first pilot; in the second and third, survey completers were offered a $20 gift card.

Results: Across pilots, refusal rates were high. In total, after 2,614 attempts to households with potentially eligible respondents, 97 men completed surveys, and 23 provided partial responses. There was some variation in demographic characteristics by pilot study. Cost per completed survey was lower in pilots one ($114) and three ($115) than pilot two ($160). Of survey completers, the percentage that reported engaging in at least one the 16 behaviors was similar for all pilots (38%-41%). The items endorsed most frequently were “engaged in sex play by overwhelming her with arguments and pressure” and “engaged in sexual intercourse by showing displeasure.” Some men endorsed items that included using their position of authority to engage in “sex play” or attempted intercourse, or giving the woman alcohol or drugs to engage in attempted intercourse or intercourse. No respondents endorsed four of the 16 items, including engaging in intercourse by threatening or using some degree of physical force. Of those who endorsed at least one of the 16 behaviors, 38% said they were drinking at the time. Other findings and methodological issues will be discussed.

Implications: This work has implications for conducting a survey of men's sexual assault behavior using a larger, nationally representative sample in order to help practitioners prevent sexual assault and treat men who engage in it. Furthermore, this study addresses important issues for social work researchers about the benefits and disadvantages of various survey methodologies and provides useful methodological tips for future researchers investigating sensitive topics.