Not sure if anyone pointed this out here before, but asking for a participant's gender in a questionnaire before any of the questions are actually asked skews the results. It's got to do with stereotype threat and is scientifically proven.