Good question. This is always a potential concern. While the risk is never going to be zero, I would suggest that the attitudinal questions for both women and men, but particularly men, be placed with other attitudinal questions about women's and men's roles so that they do not stand out in anyway. One thing to always remember, and something we have learned by talking to women informally, is that violent men can be violent for any reason. Women could be subjected to violence not for the questions that were asked but because the husband feels that she "wasted time" answering a survey. We must do everything we can to minimize risk, but we are unlikely to be able to eliminate it.