“There ya go: preregistered, within-subject, multilevel”

Kevin Lewis points to this article, Probing Ovulatory-Cycle Shifts in Women’s Preferences for Men’s Behaviors, by Julia Stern, Tanja Gerlach, and Lars Penke:

The existence of ovulatory-cycle shifts in women’s mate preferences has been a point of controversy. There is evidence that naturally cycling women in their fertile phase, compared with their luteal phase, evaluate specific behavioral cues in men as more attractive for sexual relationships. However, recent research has cast doubt on these findings. We addressed this debate in a large, preregistered, within-participants study using salivary-hormone measures and luteinizing-hormone tests. One hundred fifty-seven female participants rated the sexual and long-term attractiveness of 70 men in dyadic intersexual interactions in natural videos. Multilevel comparisons across two ovulatory cycles indicated that women’s mate preferences for men’s behaviors did not shift across the cycle for either competitive or courtship behavior. Within-women hormone levels and relationship status did not affect these results. Hormonal mechanisms and implications for estrus theories are discussed.

I added the bold font above to emphasize the good things in their research design.

But . . . the paper has no scatterplots. No graphs at all! Also I didn’t see any raw data. But we’re moving in the right direction. I’m glad.

8 thoughts on ““There ya go: preregistered, within-subject, multilevel”

  1. On the one hand, seeing a paper with a negative finding get published is a good thing. On the other hand, as a retired guy, I typically don’t breach paywalls even for interesting discoveries, why would I for a negative finding?

    Makes me wonder if that is one of the reasons so few papers like this get published.

  2. Don’t blame the missing graph on the authors. Psych Science poses severe limits on article length, and for a negative finding in particular they may not want to waste more pages than absolutely necessary. I don’t know for sure, but I could very well imagine that Penke and his team DID have graphs attached to the manuscript, but that they had to be edited out during the review and editorial decision process.

    In any event, I am glad that rigorous research like this is being conducted and published. The field of human behavioral endocrinology is riddled with too many far-out claims and false-positive publications (in part due to salivary hormone assays of dubious validity), and publications like this can help to clean things up quite a bit. Kudos to the authors of this excellent paper!

Leave a Reply to jrc Cancel reply

Your email address will not be published. Required fields are marked *