My friend recently sent me a link about how a new “study” shows “Hormones drives sexy women to infidelity.” Of course I clicked the link to find out what all the hoopla was about. The idea is that women who have this hormone are more likely to be attractive and thus more likely to “trade up” to find a better partner and commit infidelity.
My first reaction was to consider the possibility that this was true. So I imagined a ravishing woman with long dark hair wearing a classy low-cut dress, hair flowing freely in the wind, walking seductively down the street. She stops, flips opens her phone, “We’re done, I’ve found someone else.” All you hear on the other end is a grown man sobbing pleading her not to leave. She casually hangs up. Walks over to the corner, gets into a red Ferrari and drives off with her new boyfriend.
Then I come back to reality. I realize I’m sitting at my computer, staring at my monitor with my legs crossed wondering how I veered off my original train of thought. This is when I asked myself: why was I just imagining a poorly written movie? Something was off here. My intuition isn’t great but it usually steers me in the right direction. There had to be something wrong with this article. I quickly skimmed the article to find exactly what I was looking for — they only used 52 women.
Well, that’s not even the flaw in the study exactly but it’s part of it. More than likely, the women were not picked “randomly”. If you want to get a statistician excited, just start talking about things being random, because if they were truly random (assuming that’s possible), 52 samples would be more than enough to justify the sort of conclusion they make. But even then, what population does this represent? All women? Women aged 17-33 (the range they used)? Women applicants they had to this study?
Besides the obvious flaws in sampling, their methodology for determining the likelihood of infidelity was to ask the participating women questions because people never lie about things. Just ask 52 people what kind of coffee they want (hint: the answer would be “I want a dark, rich, hearty roast.”). Now wouldn’t it also be possible to draw another conclusion from the study? Women who are more attractive are more likely to lie about how often they would cheat on their partners. In fact, this may even be a more plausible explanation.
To be fair, the original authors of the paper most likely did not make such strong conclusions that the author of the article inferred (I don’t blame the author of the article, he did a good job in writing something interesting — albeit misleading). Still, I bet a lot of people believed what they read without thinking because it was a “study”. It seems like that could get you into trouble — to not think carefully. But then again, I was surprised when I found out that 1 out of every 8 adult women in America is a prostitute. Go figure.