Glenn Kessler at the Washington Post said:

[T]his oft-cited statistic comes from a Web-based survey of two large universities, making it misleading to suggest that it is representative of the experience of all college women.

In addition, the statistic has other problems. Among other problems, the study asked questions “about events that you think (but are not certain) happened.” There were other issues with wording noted at the link. What I find most interesting is the comparison of statistics from the survey to others from other surveys. In the cited study:

Nineteen percent of women who responded to the survey said they had experienced some kind of sexual assault, either attempted or completed (12.6 percent reported attempted sexual assaults,13.7 percent reported completed sexual assaults and some women reported both).

But the National Institute of Justice, which conducted the study, has admitted:

Unfortunately, researchers have been unable to determine the precise incidence of sexual assault on American campuses because the incidence found depends on how the questions are worded and the context of the survey. For example, researchers did two parallel surveys of American college women during the same time and came up with very different results. The surveys, conducted between February and May 1997, asked only about sexual assaults that had taken place “since school began in fall 1996.”

One survey found a completed rape rate of 1.7 percent, while the other study found a 0.16 percent rate. Similarly, one study found an attempted rape rate of 1.1 percent, while the other study found a rate of 0.18 percent. Thus, the percentage of the sample that reported experiencing a completed rape in one study was 11 times the percentage in the other study. Researchers believe the disparity arises from the way the survey questions are worded.

Isn’t that often the case with surveys?

Back to the post.