Are men funnier? Here are the vital statistics
Gareth Morinan properly crunches the numbers
Chortleís post-Edinburgh article Men Are Funnier Than Women caused me great offence. Primarily because of itís use of the word Ďanalysisí in reference to what consisted of calculating a few averages and whacking them in a bar chart. Itís your classic case of numbers being used to draw conclusions that arenít statistically valid.
So being the fan of data analysis that I am, I took a more detailed look at all the Edinburgh 2010 comedy reviews from Chortle and Three Weeks Ė 629 reviews in total (after one off shows and outliers had been removed). Hereís what I foundÖ
Men got on average 3.30 stars, while women got 3.06 stars. And this difference of 0.24 stars is statistically significant, as determined by an independent two-sample t-test assuming an unequal variance (Google it if you care that much). So men do get better reviews. FACT.
However, many things will affect the outcome of a review other than the quality of the show itself. For instance, looking at the gender of the reviewer we can see that women give on average 3.30 stars, while men give only 3.09 stars. A 0.21 difference that is statistically significant. So women are more generous reviewers than men. FACT.
Average number of stars, broken down by gender:
If we look at each gender of reviewer separately we see that men give on average 0.27 more stars to male performers, a significant difference, while women give on average only 0.10 more stars to male performers, which is not a significant difference. There are two possible explanations for why the bias to men is only significant among male reviewers:
- Men are inherently funnier than women. Male reviewers recognise this but female reviewers do not, because the mind of a women cannot comprehend the finer technicalities of comedy.
- Men are not inherently funnier than women. However men bias towards male performers because sexism is still rife in certain areas of society (e.g. performance comedy).
There is a long history of people claiming that certain groups of society are somehow fundamentally inferior (eg any race not native to Europe), claims that have turned out to have no scientific basis. However there is substantial evidence that people continue to hold bigoted opinions about certain groups of society (eg women). Therefore common sense would dictate that the second explanation is much more plausible.
So we have identified that two of the factors influencing a review (while not influencing the quality of the show) are the gender of the reviewer and the bias of male reviewers. Using the data we can quantify these factors and make adjustments to the data in order to remove their effects, thus giving us a more accurate indication of the actual tangible quality of the show.
Once we have done this the average number of Ďstarsí for men reduces to 3.22, while for women it rises to 3.09; a difference of 0.13 stars that is not significant. Therefore we can conclude that; Men get better reviews than women, but only because men give more favourable reviews to other men. FACT.
I should point out that when I say 'FACT', I mean 'a statistically valid statement, based upon this single study of comedy reviews'. So I canít claim that this definitively answers anything. But I can claim that the conclusions in my paper are more factual than some moron saying, 'Iíve never met a woman who made me laugh, so women canít be funny'.
But regardless of your opinion on this issue, I think we can all agree that the one thing the comedy industry really needs right now is data analysis. FACT.
- Gareth Morinan: Comedian by night, analyst by day. Click here to read his full paper on the subject.
Posted: 30 Sep 2010