Delusions of Gender—I watched a video of the author speaking about her book, and it was interesting, but the same information could be told much quicker than in one hour. So here are some points I remembered:
Selection bias: if you make a study and you don’t find a difference between male and female brain, you don’t write a bestseller. Also, comparing the male and female results is the first obvious idea of any researcher, so given p = 0.05, one research in twenty would publish something about the differences between men and women, even if there was none. So if you want some meaningful results, you need to do the meta-analysis of the published studies—and they often look just like they would if the difference wouldn’t really exist: larger samples have smaller differences, and almost half of them shows the difference in the opposite direction.
Some differences are exaggerated and misinterpreted. For example, there is a picture of a brain showing that in these little areas women had more signal than men (or vice versa) when solving a maze. First, many popular authors will interpret it as “women only used these parts, and men only used those parts”, while in reality it means that both men and women used their whole brains, but in the most of the brain their activity was the same, and only in these little parts a difference was found (which is likely a random noise that will appear differently if the study will be replicated). Second, there will be a huge generalization in popular books, from solving the maze to… pretty much any mental activity.
The rest of the video is mostly talking about how stereotypes are bad and self-fulfulling; with some examples of how e.g. the way a question was asked has influenced the results.
...will ya look at that. Some jerk silently downvoted a book recommendation without comment, just because the book being praised seems vaguely feminist and/or anti-pop-evopsych.
I second Ben’s recommendation for How to Think Straight about Psychology from last month.
Delusions of Gender—I watched a video of the author speaking about her book, and it was interesting, but the same information could be told much quicker than in one hour. So here are some points I remembered:
Selection bias: if you make a study and you don’t find a difference between male and female brain, you don’t write a bestseller. Also, comparing the male and female results is the first obvious idea of any researcher, so given p = 0.05, one research in twenty would publish something about the differences between men and women, even if there was none. So if you want some meaningful results, you need to do the meta-analysis of the published studies—and they often look just like they would if the difference wouldn’t really exist: larger samples have smaller differences, and almost half of them shows the difference in the opposite direction.
Some differences are exaggerated and misinterpreted. For example, there is a picture of a brain showing that in these little areas women had more signal than men (or vice versa) when solving a maze. First, many popular authors will interpret it as “women only used these parts, and men only used those parts”, while in reality it means that both men and women used their whole brains, but in the most of the brain their activity was the same, and only in these little parts a difference was found (which is likely a random noise that will appear differently if the study will be replicated). Second, there will be a huge generalization in popular books, from solving the maze to… pretty much any mental activity.
The rest of the video is mostly talking about how stereotypes are bad and self-fulfulling; with some examples of how e.g. the way a question was asked has influenced the results.
...will ya look at that. Some jerk silently downvoted a book recommendation without comment, just because the book being praised seems vaguely feminist and/or anti-pop-evopsych.
Typical LessWrong.