Richard Juster points us to this article, “Vocal characteristics predict infidelity intention and relationship commitment in men but not in women,” where by “men,” they meant 88 male college students, and by “women,” then meant 128 female college students, and by “predict,” they meant not very well. The story was featured in various classy news outlets. It hasn’t made it to NPR yet, but give it time.
P.S. Nothing uniquely bad about social psychology. If anyone wants to find a silly statistics paper and post under the heading, “Don’t ever change, statistics! You’re perfect just the way you are,” go for it.
I was listening to NPR this morning, they talked about it. Bad research travels at the speed of light, apparently.
Graham:
NPR will never let us down.
To be fair, I heard it on “Wait, wait, don’t tell me” and they were having a good time with it!
I love “Wait, Wait…” but I’m pretty sure that’s not what Andrew means by NPR.
I totally get that social psychology has p-hacking researchers, crappy topics, and crappy studies. But why go for the jugular on the entire discipline — aren’t there decent areas in social psychology (e.g., persuasion, social norms)?
Also, forgive me if your blog gives air time to this topic: are there areas and researchers in your home area of statistics in which you equally critical (if not, why not)? Thanks
Frenetic:
I wouldn’t call this “the jugular.” I’m just describing the article accurately. And I’m sure there’s great stuff done in social psychology. Feel free to share some of it in the comments.
And, yes, I have lots of problems with statistics papers too, but usually the problem is not so much that the paper is wrong but that it seems pointless.
So this is personality psychology which is actually the opposite of social psychology, which focuses on the power of social situations not individual differences.
Kp:
I noticed this was in a personality psychology journal, but “Vocal characteristics predict infidelity intention and relationship commitment in men but not in women” has a bit of a social-psychology feel to it. To me, hidden signals don’t seem so different from hidden influences. And the methods used in the article are right out of the social psychology playbook.
Title translates as: “men are more easily caught in their lies than women are” ?
If (p = .096) is a “marginally significant” result, doesn’t that make their (p = .044) a “marginally insignificant” result?
At least the last sentence of the abstract is refreshingly honest: “The current findings have important implications for research on voice in the mating-related domain.” In other words, it doesn’t have implications for anything real, just whether or not more people can do research on the topic, squeezing out more meaningless papers.
Yeah. This sort of thing is not great in terms of convincing “hard science” people to take “social sciences” seriously as scientific/rigorous fields…
And yeah, hard sciences have bad statistics too. But perception-wise…
Hey — it’s published in “Personality and Individual Differences”! What did you expect? The social/personality psych equivalent of rocket science?
“The story was featured in various classy news outlets. ”
I’m having a hard time with “classy” as a modifier for “news outlets”.