Ryan Schuetzler

Researcher and techie extraordinaire

Misleading Conclusions

| Comments

There’s a curse that comes with being a doctoral student. The curse is that I can rarely take the results of any study at face value. I suppose this has more to do with learning statistics than being a doctoral student, but nevertheless, I have a problem. So today when I read a recent article from Wired about how apparently opposites don’t really attract, I got a little suspicious about some of the conclusions of one of the studies they mentioned.

The study comes from Angela Bahns, Kate Pickett, and Christian Crandall at Wellesley College and the University of Kansas. In the study, researchers approached pairs of people on college campuses to ask them questions about religion and views on various social issues. The study found that pairs on a large, more diverse college campus were more similar to each other than pairs on smaller, more homogeneous college campuses.

This finding is interesting. Given more opportunity to make diverse friends, we would expect pairs of friends to be more diverse, but that does not appear to be the case. However, the researchers make the conclusion that the increased diversity causes the students to seek out similar-minded friends. I would argue that rather than the diversity causing people to seek out similar friends, that the size of the university allowed students to find similar people.

Merely finding that students at a large university have more similar friends than at a small university is an interesting finding. However, it’s tempting to make guesses as to why that is. Because these researchers were studying diversity, they put the blame on diversity. It may be true that more diverse universities cause students to seek out similar-minded acquaintences. However, this study didn’t look at that specifically. In order to determine that diversity is the causal factor would require the researchers to at least compare universities of the same size. One large diverse university could be compared to a large homogeneous university and the similarity of friendships in that case could be compared. Otherwise, any of a million confounding factors could be causing the similar friends to come together