> Also, someone may argue that women who were actually harassed are probably more involved in these matters, and hence more likely to respond to such surveys.

> Also, someone may argue that women who were actually harassed are probably more involved in these matters, and hence more likely to respond to such surveys. You may argue this, but I wouldn’t say it’s a given. For one thing, especially at the higher end of the scale (“unwanted sexual contact”, or hypothetically even rape), people who have had traumatic experiences may have a strong desire to avoid recalling them – especially for something as (relatively) unimportant as a survey – potentially producing the opposite effect.