Elusiveness of menstrual cycle effects on mate preferences: comment on Gildersleeve, Haselton, and Fales (2014)

This comment uses meta-analytic techniques to reconcile the apparent conflict between Gildersleeve, Haselton, and Fales's (2014) conclusion of "robust" effects of menstrual cycles on women's preferences for men of purported genetic quality and Wood, Kressel, Joshi, and Louie's (2014) assessment that the few, limited effects in this literature appear to be research artifacts. Despite these divergent conclusions, the literature in both reviews shows a broad distribution of effects, with fully one third of findings countering evolutionary psychology predictions. We demonstrate that Gildersleeve et al.'s conclusions were influenced by a small minority of supportive studies. Furthermore, we show that in both reviews, these supportive studies used imprecise estimates of women's cycle phase by failing to validate cycle day (e.g., with hormonal tests) or by including a large number of days in the fertile phase. More recently, as published studies have used more precise methods to estimate menstrual phase, the effect has declined to zero. Additionally, publication status proved important in both reviews, with published but not unpublished studies showing the predicted effects. In general, the limited evidence for evolutionary psychology predictions calls for more sophisticated models of hormonal processes in human mating.