You'll probably have some sample size issues given how predominately male the pilot population is. Unless you're doing some pretty fancy ANOVA-style statistics, you're going to have a hard time whittling down to a gender difference in accidents with all the other variables floating around (type, experience, time in type, conditions of flight, maintenance, airspace, IFR/VFR, etc., etc.) Still, the NTSB is the place to go for the raw data, after which you may want to interview a male and female CFI for some anecdotal and qualitative analysis to flesh things out.
What might be easier to research and still of relevance for a women's studies class would be a survey of the perceptions of male and female pilots by the pilot/aviation community at-large (heck, you could do it right here at jetcareers!), especially if you had survey takers identify their own gender when filling out their responses. Might produce some interesting data.