But if we can’t measure the cultural factors and account for them
We can’t directly measure them, but we can get an idea of how large they are and how they work.
For example, the gender difference in empathic abilities. While women will score higher on empathy on self report tests, the difference is much smaller on direct tests of ability, and often nonexistent on tests of ability where it isn’t stated to the participant that it’s empathy being tested. And then there’s the motivation of seeming empathetic. One of the best empathy tests I’ve read about is Ickes’, which worked like this: two participants meet together in the room and have a brief conversation, which is taped. Then they go into separate rooms and the tape is played back to them twice. The first time, they jot down the times at which they remember feeling various emotions. The second time, they jot down the times at which they think their partner is feeling an emotion, and what it is. Then the records are compared, and each participant receives an accuracy score. When the test is run is like this, there is no difference in ability between men and women. However, a difference emerges when another factor is added: each participant is asked to write a “confidence level” for each prediction they make. In that procedure, women score better, presumably because their desire to appear empathetic (write down higher confidence levels) causes them to put more effort into the task. But where do desires to appear a certain way come from? At least partly from cultural factors that dictate how each gender is supposed to appear. This is probably the same reason why women are overconfident in self reporting their empathic abilities relative to men.
The same applies to math. Among women and men with the same math ability as scored on tests, women will rate their own abilities much lower than the men do. Since people do what they think they’ll be good at, this will likely affect how much time these people spend on math in future, and the future abilities they acquire.
And then there’s priming. Asian American women do better on math tests when primed with their race (by filling in a “race” bubble at the top of the test) than when primed with their gender (by filling in a “sex” bubble). More subtly, priming affects people’s implicit attitudes towards gender-stereotyped domains too. People are often primed about their gender in real life, each time affecting their actions a little, which over time will add up to significant differences in the paths they choose in life in addition to that which is caused by innate gender differences. Right now we don’t have enough information to say how much is caused by each, but I don’t see why we can’t make more headway into this in the future.
We can’t directly measure them, but we can get an idea of how large they are and how they work.
For example, the gender difference in empathic abilities. While women will score higher on empathy on self report tests, the difference is much smaller on direct tests of ability, and often nonexistent on tests of ability where it isn’t stated to the participant that it’s empathy being tested. And then there’s the motivation of seeming empathetic. One of the best empathy tests I’ve read about is Ickes’, which worked like this: two participants meet together in the room and have a brief conversation, which is taped. Then they go into separate rooms and the tape is played back to them twice. The first time, they jot down the times at which they remember feeling various emotions. The second time, they jot down the times at which they think their partner is feeling an emotion, and what it is. Then the records are compared, and each participant receives an accuracy score. When the test is run is like this, there is no difference in ability between men and women. However, a difference emerges when another factor is added: each participant is asked to write a “confidence level” for each prediction they make. In that procedure, women score better, presumably because their desire to appear empathetic (write down higher confidence levels) causes them to put more effort into the task. But where do desires to appear a certain way come from? At least partly from cultural factors that dictate how each gender is supposed to appear. This is probably the same reason why women are overconfident in self reporting their empathic abilities relative to men.
The same applies to math. Among women and men with the same math ability as scored on tests, women will rate their own abilities much lower than the men do. Since people do what they think they’ll be good at, this will likely affect how much time these people spend on math in future, and the future abilities they acquire.
And then there’s priming. Asian American women do better on math tests when primed with their race (by filling in a “race” bubble at the top of the test) than when primed with their gender (by filling in a “sex” bubble). More subtly, priming affects people’s implicit attitudes towards gender-stereotyped domains too. People are often primed about their gender in real life, each time affecting their actions a little, which over time will add up to significant differences in the paths they choose in life in addition to that which is caused by innate gender differences. Right now we don’t have enough information to say how much is caused by each, but I don’t see why we can’t make more headway into this in the future.