Browsed by
Month: November 2022

Overthinking the question

Overthinking the question

“Tell me if you agree or disagree with this statement:  Most men are better suited emotionally for politics than are most women.”

That’s one of the questions on the General Social Survey. In 1974, when it was first asked, 47% of respondents said they agreed. In 2018, the most recent survey where this question was asked, that percentage was down to 14%.

Last week I posed the same question to an audience at PyData NYC. Of 56 people who responded, none of them agreed with the statement (0%).

However, if we take the question literally, the correct answer is “I don’t know”. In fact, it is almost certain that either

  1. Most men are better suited emotionally for politics than are most women, or
  2. Most women are better suited emotionally for politics than are most men.

The reason is that saying that “most A are better than most B” is equivalent to saying that the median in group A is higher than the median in group B.

Here’s the proof: If one median is higher than the other, we can find a value between them where most members of group A are higher (by the definition of median) and most members of group B are lower. So unless the medians are identical, either “most A are better than most B” or the other way around.

And what does it mean to be “suited emotionally for politics”? I don’t know, but if we had to quantify it, it would be some combination of emotional factors. Men and women differ in many ways, on average; the difference is small for some factors and bigger for others. So it is likely that a combination of factors also differs between the groups. And even if the difference is very small, the medians are unlikely to be identical.

So, if we take the question literally, the best answer is “I don’t know”. But it’s clear that people don’t take the question literally. Instead, they are probably answering something like “Do you think men are so much better suited emotionally for politics that they should hold a large majority of positions of power?”

I think that one’s easier to answer.

Chasing the Overton Window

Chasing the Overton Window

On November 9 I presented a talk at PyData NYC with the title “Chasing the Overton Window“.

The video from this talk is now available:

This talk is based on a chapter of my forthcoming book called Probably Overthinking It that is about using evidence and reason to answer questions and make better decisions. If you would like to get an occasional update about the book, please join my mailing list.

Here are the slides for the talk.

The results I reported are from 16 questions from the General Social Survey (GSS). If you would like to see the text of the questions, and answer them yourself, you can

Fill out this survey.

I summarized the results from the survey in these slides.

If you would like to read more about the topic of the talk, I’ve written two related blog posts: