It's How You Slice the Pie
This won't be a popular post, but it's a truthful one. So hang on to your hats.
A few months ago, I saw a shocking homemade video where a woman and her daughter pulled all the books from a library shelf that didn't match a pre-determined list of pro-feminist characteristics. The creator wanted books with girl protagonists that weren't in stereotyped roles. After she pulled the books, the bare shelf made the viewer gasp. Could it be possible that of hundreds of books, a bare handful were about girls doing important things?
I gasped, too, but then my professor side kicked in. Yes, I know that publishing favors men. The vast majority of published books are by men and male-sounding names are more than twice as likely to get book contracts than those with female-sounding names (whether the author is male or female). But children's literature has been discussing diversity since before it was trending on Twitter, and, as a mom of two white males, I've struggled to find representative picture books. Girls seem much more common in children's literature. I like my boys to read diverse works, so the problem hasn't bothered me, but it instilled enough doubt in the viral video that I decided to repeat the experiment and see what I found.
First, I picked a random shelf in the children's section. The "L-M" section - mainly because it was far enough away from the librarian's station that no one would see me making a mess of the carefully organized shelf.
Then, I removed all the books about white boys - the books representing the group that supposedly dominates all American culture (and matches my boys), which removed about 27% of the books - actually a proportion below white males in the general population.
Finally, I removed all the books about boys or that had no gender, so that the only books left on the shelf were about females.
The results of my experiment? Well, approximately half of the books on the shelf were about females. That's the appropriate distribution per biological distribution of male versus female biological sex.
But what would happen if I kept slicing that pie? Well, if I used the original question, I had to eliminate books about animals and books that showed princesses doing unusual things. I could do the same with my pile of "boy" books - the more narrowly I defined my acceptable sample, the more books I could eliminate. As I added stipulations, the bookshelf continued to reduce in size until I could prove almost any point.
Here's the point: every time I add a requirement, I reduce my sample size no matter what I'm looking for. If I were to continue slicing this tuxedo cheesecake, I could eventually get a slice that didn't have any chocolate, not because the cheesecake is baked wrong, but because I designed my experiment to have the outcome I wanted.
My point isn't that we shouldn't write more nontraditional books or diverse books - I support those movements completely. No, my point is that we need to stop using outrage as the primary motive of all social experiments. No one is served when we scream and yell and claim there is bias that may not exist. Truth will out. Science is science because it reduces bias, not because it utilizes it. An experiment done with a particular end in mind should be dismissed, not passed around Facebook as the God's honest truth. Because, in science, the question matters as much as the answer.