The current most emailed headline on the New York Times website is titled “What Women Want,” by op-ed columnist John Tierney. He’s writing about a working paper, “Do Women Shy Away from Competition?”, by Muriel Niederle and Lise Vesterlund, economists at Stanford University and the University of Pittsburgh, respectively. They conducted an experiment where men and women were first paid to add up numbers in their head, earning fifty cents for each correct answer (referred to as the “piece-rate” task). The participants were eventually offered the choice to compete in a tournament where the person who has the most correct answers after five minutes receives $2 per correct answer and everyone else receives zero compensation. One of the main points of the article was that, even at similar levels of confidence and ability, men were much more likely to enter the tournament than women, i.e., women are less willing than men to enter competition. The results of this study yield another possible theory for why there are so few women in top-paying jobs: Even in a world of equal abilities and no discrimination, family issues, social pressures, etc., women might be less likely to end up as tenured professors or CEOs because the jobs are so inherently competitive.

Gender inequality is a hot topic, but my issue with the paper is of a much more mundane statistical nature. For the results of the experiment to be meaningful, you’d want to start with men and women who, aside from the gender difference, are pretty similar; otherwise, something besides gender might explain the women’s lower participation in the tournament. The authors acknowledge this, and conclude that the men and women are comparable because they scored similarly on both the piece-rate and tournament tasks. What isn’t mentioned in the article is any comparison of other potentially relevant characteristics between the men and women in the study, age for example. If, say, men and women scored similarly on the addition tasks but the women were on average twenty years older than the men, you might wonder whether the difference in competitiveness is due to gender or to age. I don’t really think that’s what’s going on here, but I’d be even more convinced if the paper included a table with some background information on the men and women in the study. Actually, even just a sentence stating that the men and women in the study did not differ significantly on age or education (or…) would work. That’s all I want, just a sentence.