I mentor senior theses, independent studies, and other student projects connected to my areas of research expertise. You can get a sense of my interests via this website, by taking classes with me, or by coming to speak with me.  If you’d like to carry out research with me, it’s best to approach me early. This can be as early as your first or second year, before you even have a firm idea of what you’d like to research.

In addition to general mentoring, I’ve launched a research “lab” for students to conduct email-based experiments studying discrimination and social exclusion. This is an excellent opportunity for well-organized and dedicated students to gain experience using an interesting and high-impact research method. Online experiments yield clear results that speak to important questions, and are highly adaptable depending on your specific interests.

An example of the online experimental approach is this article, where a colleague and I created several different fake identities with either white or Arab American female names. We then sent hundreds of well-written, clear email replies to “roommate wanted” ads on Craigslist, varying only the sender’s name, and measured the gap in response rates. We found a large difference in treatment of the two identities, suggesting significant social exclusion of Arab women.

Email-based audit studies have been used to examine racial and/or gender discrimination in housing markets, roommate searches, job markets, treatment by social service agencies and elected officials, graduate school interviews, and more. They have been used to examine whether employers discriminate against mothers (they often do) and whether landlords discriminate against against different family styles (sometimes they do). They can be used to compare different geographical locations or types of discrimination, as well.

If you’re interested in this type of research, browse the above studies to get more of a feel for them. Then come speak with me.

You might also enjoy the recent pop social science books Everybody Lies and Dataclysm, both in some ways connected to this approach.