In recent years, evolutionary psychologists have conducted lab-based and naturalistic studies suggesting that naturally cycling women (i.e. women who are not on hormonal contraceptives, such as the pill) experience a suite of behavioral and cognitive changes depending whether they are in the follicular, ovulatory, or luteal phase of their menstrual cycles. During ovulation, when a woman’s chance of conception is highest, she is likely to report higher levels of sexual desire, have a strong preference for masculine-looking men, and wear certain types of clothing—specifically, red clothing.
In 2013, a study conducted by psychologists Alec Beall and Jessica Tracy found that women at high conception risk (women who self-reported being on days 6-14 of the cycle) were over 3 times as likely as women at low conception risk (women who self-reported being on days 0-5 and 15-28 of the cycle) to wear red or pink shirts. Day of the cycle was determined by counting the number of days since women’s last self-reported menses.
There was just one problem—that “day of the cycle was determined by counting the number of days since women’s last self-reported menses.” This counting method is frequently employed in studies relating cycle phase to behavior because of its ease relative to collecting and assaying saliva samples for hormone concentrations. However, prior to when this study was conducted, there were several reasons to doubt its accuracy in classifying high versus low fertility days, which may make results from studies using this method suspect.
Acknowledging this flaw, evolutionary psychologists Adar Eisenbruch, Zachary Simmons, and James Roney conducted similar analyses to Beall and Tracy, but instead of using the counting method, they collected saliva samples (that were then assayed for hormone concentrations) each time women came into the lab. They then also used the counting method, and examined the concordance between the counting and hormonal methods of conception risk classification.
Using the counting method, there was no difference in the percentage of low and high conception risk women who wore red. When using the hormonal method, however, a significantly higher percent of high conception risk women wore red than did low conception risk women. So, while the use Beall and Tracy’s methods resulted in an inability to replicate their own original findings, the use hormonal methods for conception risk classification resulted in support for high conception risk women being more likely to wear red.
Perhaps more interesting, and certainly more worrisome than this central finding, was the lack of concordance between the counting and hormonal methods of classification—the two agreed in a mere 64% of cases. In other words, for more than 1/3 of the time, these two methods classified women as being in opposite conception risk categories. Further, almost half of the days identified as high conception risk by hormonal methods were classified as low conception risk by the counting method.
That the counting method can differ substantially from hormonal methods of conception risk classification challenges the reliability of some prior findings of cycle phase effects. While it is certain that using the counting method is easier, quicker, and less expensive than collecting and assaying saliva samples, it is unclear whether these advantages outweigh the findings of Eisenbruch et al. suggesting that the counting method may be incorrect more than third of the time.
It may be that as evidence of the flaws of the counting method continues to accumulate, its use in evolutionary psychology will become increasingly harder to justify, thus opening the door for broader use of more methodologically-sound research practices.