Abstract

In this article we attend to recent critiques of psychometric applications of life history theory to variance among humans and develop theory to advance the study of latent LH constructs. We then reanalyze data previously examined by Richardson et al., 2017, https://doi.org/10.1177/1474704916666840 to determine whether previously reported evidence of multidimensionality is robust to the modeling approach employed and the structure of LH indicators is invariant by sex. Findings provide further evidence that a single LH dimension is implausible and that researchers should cease interpreting K-factor scores as empirical proxies for LH speed. In contrast to the original study, we detected a small inverse correlation between mating competition and Super-K that is consistent with a trade-off. Tests of measurement invariance across the sexes revealed evidence of metric invariance, consistent with the theory that K is a proximate cause of its indicators; however, evidence of partial scalar invariance suggests use of scores likely introduces bias when the sexes are compared. We discuss limitations and identify approaches that researchers may use to further evaluate the validity of the K-factor and other applications of LH to human variation.