How valid are assessments of conception probability in ovulatory cycle research? Evaluations, recommendations, and theoretical implications

Abstract

Over the past two decades, a large literature examining psychological changes across women's ovulatory cycles has accumulated, emphasizing comparisons between fertile and non-fertile phases of the cycle. While some studies have verified ovulation using luteinizing hormone (LH) tests, counting methods – assessments of conception probability based on counting forward from actual or retrospectively recalled onset of last menses, or backward from actual or anticipated onset of next menses – are more common. The validity of these methods remains largely unexplored. Based on published data on the distributions of the lengths of follicular and luteal phases, we created a sample of 58,000+ simulated cycles. We used the sample to assess the validity of counting methods. Aside from methods that count backward from a confirmed onset of next menses, validities are modest, generally ranging from about .40–.55. We offer power estimates and make recommendations for future work. We also discuss implications for interpreting past research.