Thursday, May 18, 2017

Ideological filtering

I am enjoying reading Seth Stephens-Davidowitz's recently published Everybody Lies: Big Data, New Data, and What the Internet Can Tell Us About Who We Really Are but I am also deeply frustrated. I'll get into the details in a review later but one of the issues is that Stephens-Davidowitz lets his ideological assumptions color his interpretation of the data. Stephens-Davidowitz sees the world through SJW eyes, postmodernist, critical theory, critical race theory, deconstructionism, etc. On page 134 I came across this example of what I mean.
. . . I was able to use Google searches to find evidence of implicit prejudice against another segment of the population: young girls.

And who, might you ask, would be harboring bias against girls?

Their parents.
Alright. A pretty serious charge. What's the evidence?
It's hardly surprising that parents of young chilren are often excited by the thought that their kids might be gifted. In fact, of all Google searches starting "Is my 2-year-old," the most common next word is gifted. But this question is not asked equally about young boys and young girls. Parents are two and a half times more likely to ask "Is my son gifted?" than "Is my daughter gifted?" Parents show a similar bias when using other phrases related to intelligence that they many shy away from saying aloud, like, "Is my son a genius?"
It is a central claim of Stephens-Davidowitz that why it is so useful to use Google search data is that people are willing to ask on Google that which they would be unwilling to ask in most other contexts.

So the evidence that parents are biased against their daughters is that they ask Google 2.5 times as often whether their son is gifted as they do whether their daughter is gifted.

It would seem obvious that parents are biased against their daughters if boys and girls have identical IQs. Do they? It is well established that boys and girls have near identical mean IQs but that boys have a larger standard deviation. In other words, girls are more tightly clustered around the mean of 100 whereas boys have a flatter distribution. While the male mean of 100 is the same, there are both more boys at the very low end of the distribution curve as well as more at the upper end as well.

The further out on the tail of the distribution you go, the more unequal is the distribution between male and female in absolute terms.

Everyone's definition of gifted varies somewhat, but let's take the Mensa cutoff of the top 2% percent, those in the 98th percentile which is an IQ of 130. There are two times as many males with an IQ of 130 as there are female. In a population of one hundred people with an IQ of 130, 67 will be male and 33 will be female. Correspondingly, at the other end of the curve, there will be 67 males with an IQ of 70 and only 33 females. The imbalance is even greater the further out on the distribution you go. At IQ 145 it is more like 80 males to 20 females.

This equivalence between the sexes of means and mismatch of standard deviations is well known and longstanding. For example, see here, here, here, here, and here.

This has nothing to do with the relative valuation of people as individuals. It is simply an acknowledgement that there are objective differences between people and sometimes those differences follow patterns based on sex.

Depending on whether you mark 130 or 145 or 160 as gifted, you would expect for there to be a 2X, 3X, or 5 or greater multiple ratio of boys to girls. So Stephens-Davidowitz's finding that parents search 2.5 times as often whether their son is gifted as they do whether their daughter is gifted is not surprising. It is what you would expect based on real world conditions.

But let's not forget the other end of the curve as well. You would expect parents to be searching for whether their sons are dull, stupid, behind much more often as they do for their daughters. And that is what Stephens-Davidowitz finds.
Are parents picking up on legitimate differences between young girls and boys? Perhaps young boys are more likely than young girls to use big words or otherwise show objective signs of giftedness? Nope. If anything, it's the opposite. At young ages, girls have consistently been shown to have larger vocabularies and use more complex sentences. In American schools, girls are 9 percent more likely than boys to be in gifted programs. Despite all this, parents looking around the dinner table appear to see more gifted boys than girls. In fact, on every search term related to intelligence I tested, including those indicating its absence, parents were more likely to be enquiring about their sons rather than their daughters. There are also more searches for "is my son behind" or "stupid" than comparable searches for daughters. But searches with negative words like "behind" and "stupid" are less specifically skewed toward sons than searches with positive words, such as "gifted" or "genius."
Stephens-Davidowitz is correct that at younger ages, boy and girl IQ results look much closer to one another than later and he is correct that girls tend to perform better on vocabulary than do boys. The differences show up around puberty. He is also correct that girls more than boys (usually owing to earlier maturation of girls and disruptive boy behavior) are more often enrolled in gifted programs.

But the replicated reality is that males and females have near identical mean IQs but boys have a higher standard deviation and boys therefore are more heavily represented among the population of the gifted by significant multiples.

It almost feels like Stephens-Davidowitz cherry-picked and presented only the evidence to support the assumption he wished to make, that parents favor boys over girls. If you know the full context, it is apparent that the proportion of google searches actually matches the reality but that is not what you would conclude drawing from Stephens-Davidowitz's description.

I keep coming across this tic of letting ideological biases trump the data but this was one of the most blatant ones. The book is interesting but this ideological filtering is an annoying habit.

No comments:

Post a Comment