Williams syndrome, stereotypes, and misleading titles

A paper just published in Current Biology and discussed extensively in the media was titled:

“Absence of racial, but not gender, stereotyping in Williams syndrome children.”

Williams Syndrome results from the accidental loss of several genes on chromosome 7, and it affects about 1 in 2500 newborns. The symptoms include a characteristic facial appearance, heart problems, and an uneven pattern of cognitive abilities, with some impaired more than others. Additionally, children with Williams can be “hyper-social,” being more willing to approach and interact with strangers.

The title of this new study makes an intriguing claim, implying that this genetic change eliminates the intuitive tendency to stereotype people based on race, but not gender. Based on the title, you might assume that the study used comparable measures of racial and gender stereotyping and found only gender stereotyping, right? Nope. For the racial stereotyping task, 20 kids were shown pairs of pictures that were identical except for skin color. They then heard stories that mentioned a positive or negative adjective (kind, pretty, smart vs. bad, ugly, stupid) and were asked to pick which person the story described. The Williams kids were less likely than a control group to pick their own race for the positive adjectives. Okay. So far, so good. That’s interesting. (Though it would also be nice to see whether this held true for a larger sample of Williams kids of different races—all of these kids were white, and the study was conducted in Marseille, France.)

So, did the researchers use the same method to test gender bias? Nope. “Sex role items assessed the child’s knowledge of typical sex-stereotyped behaviours, and provided a control measure of general conceptual development.” Yup. You read that right. The gender stereoptyping wasn’t testing whether the kids had any systematic bias or stereotyped attitudes. It wasn’t testing whether they had negative attitudes about one sex or the other. Rather, it tested whether they were familiar with typical gender-based behavior. That has nothing to do with whether kids are biased or think the opposite sex is worse, just whether they know that boys are more likely to play with trucks and girls are more likely to play with dolls. (Apparently, the authors used a standardized test of racial attitudes and knowledge of sex roles called the PRAM-II. It would be sad if the standardized test made this mistake too.) If kids weren’t picking up on these patterns of behavior in the world around them, that would be suggestive of a broken mechanism for associative learning!

I would bet that if these researchers had measured familiarity with typical racial differences in interests/behavior, they would have found that Williams kids were aware of those roles too. It’s cool that the Williams kids didn’t associate darker-skinned people with negative words. Why equate racial biases to gender biases in the title if that’s not what you actually did?

For a misleading title of what was otherwise an interesting result, I throw the authors 3 smelly, rotten, stereotypical bananas.

Chest thumps to Not Exactly Rocket Science, who covered the study in some depth.

Comments are closed.