Verified by 
A captionless cartoon - not from the study!
You may have seen recent news coverage suggesting that men are funnier than women. There's a new study making this claim that was published in the journal Intelligence. It argues that this mens' increased capacity for funniness is caused by sexual selection. One of my goals for this blog is to encourage critical thinking about science in the media, especially when it's about the science of judgment and decision-making. I therefore present six reasons to doubt the claim that men are funnier than women.
1. How do you measure sense of humor anyway? There's no standard test for measuring a sense of humor so the study authors had to invent one. In this study, subjects were given a three captionless New Yorker cartoons and asked to write as many funny captions as they could for each in 10 minutes. Who says this is a good measure of humor ability? Only the authors of the study. That's unusual - careful psychologists try to make sure a measure is widely validated before they draw conclusions about things like evolution and sexual selection.
2. That's a problem because there is no general agreement on what is funny. This won't make me any friends on the internet, but I just don't enjoy Monty Python. My wife (and the rest of the internet) love it. The authors' strategy for dealing with this problem is to just average peoples' ratings. But that makes as much sense as choosing the funniest TV show based on its viewership. The same logic leads you to conclude that Two and a Half Men is funnier than 30 Rock. And that's just silly.
3. Part of why that assumption is flawed is that different things are funny in different cultures. Jokes rely so heavily on cultural background that many hugely successful comedy films made in the US aren't released to foreign theaters. Almost a third of the students in the study were Hispanic, but who knows what cultural background the judges are (the authors don't say). Even for one person, humor is incredibly context dependent. Think about how tiny things can be hilarious at a funeral. The sterile environment of the science lab skews all sense of what's funny. For me, the relabeled New Yorker cartoons found here are hilarious. But in the lab, I'm sure they would be rated zero.
4. Judges rated captions from 1 (not funny) to 7 (very funny). The vast majority of captions received a score of one. That indicates the experimental subjects weren't well matched to the test - they were just swinging wildly. Maybe it's because so many of these cartoons deal with post-college situations (e.g. the workplace). Or maybe it's because it's hard to be funny on command in a very specific domain.
5. The study authors report that men wrote 16% more answers in the limited time than women did. Since score distributions were highly skewed towards 1's ("not funny at all"), and only highest scores were used to measure humor, subjects were essentially playing a lottery where the more tickets you buy, the more you win. Since the correlation between gender and humor is weak (or at least appears to be: the paper is very vague about how its numbers derive from cartoon ratings), this lottery effect ought to be very large - you would expect male scores to be roughly 16% greater even if the two genders were equally funny. [Another subtle statistical point: if males are more divergent in humor abilities than women, regardless of their means, this particular measure will inflate their scores.]
6. Finally, the judges were 4 male and 2 female students. You say you want to design an unfair study to prove that men are funnier than women? My advice: have twice as many male as female judges and then keep the number of judges low overall. Everyone who has been in middle school knows boys are more willing to laugh at other boys' jokes, especially when they involve bodily functions.
I haven't even touched on the authors' tools for measuring sexual success. I am sure that there's NO problem with a self-reported measure of sexual prowess among college males, right?
Original study:
Greengross and Miller. Humor ability reveals intelligence, predicts mating success, and is higher in males. Intelligence vol. 39 (2011) p 188-192.
[Thanks to invaluable help from Nancy Heilbronner, Sarah Heilbronner, and Alexandra Harwin!]
