Testing telephone and web surveys for studying men's sexual assault perpetration behaviors.

Three pilot studies were conducted to obtain information about sexual assault perpetration from adult men in the United States. Each used the same random digit-dial sampling and recruitment strategy. One pilot was administered by telephone and two via the Web. Response rates in all pilots were low. Although results cannot be generalized beyond the sample, of the 97 men who completed the surveys, approximately 40% in each pilot reported some sexual assault behavior, broadly defined. Contrary to expectations, telephone respondents reported a wider range of sexual assault behaviors. Suggestions for improving response rates include offering a choice of participation methods.