According to the researchers, the differences between the two components of the study illustrate how different methodologies can influence estimates of rape. The two studies were conducted at the same time, employing similar samples and interviewing methodologies.
Differences arose when the context under which the surveys were conducted changed, as did the wording of the questions used to screen for victimizations and the wording used to determine the type of incident.
Since rape statistics have been the subject of controversy in the past, pitting feminists and conservatives against each other with different numbers, this study can help show how a range of results come about, the researchers said.
"Our results shed some light on some of these measurement issues," Fisher said.
Estimate Too Low?
According to Kim Gandy, executive vice president of the National Organization for Women, the new government study is likely more reliable than those relying only on rapes reported to the police or campus security.
Still, she said, the report probably underestimates the problem of rape in campus. "I'd say [3 percent] is too low," she said. "It's appallingly high when you think about the number of women on a campus, but at the same time I believe it is substantially higher in reality."
The surveys were conducted between February and May 1997, and were administered to women who were enrolled in college at the start of the 1996 fall semester.
The results are based on telephone surveys of randomly selected national samples of women who attended two- or four-year colleges or universities. The sample sizes were 4,445 respondents for the main component and 4,432 for the companion component.