Spearman's law of diminishing returns: investigation on large-scales Russian samples

The reduction of average correlations among cognitive tests in groups of individuals of above-average IQ was invented by C. Spearman (1927) and now is known as Spearman’s “Law of Diminishing Returns” (SLODR). Although many studies also founded that inter-correlations between subtests decreased from low to high ability groups, several empirical investigations have challenged the SLODR. In Russia similar studies didn’t carry out by now. We investigated the presence of SLODR in large-scaled dataset consisted of sample of 11335 applicants for joining to military institutes (mean age 18.42±1.29 years). The subjects were tested with the same battery of 10 ability tests. SLODR was assessed (tested) by dividing the sample into low/high ability groups based on the total score. In order to account for criticism of Murray et. al. (2013), test scores in each half-sample were transformed to normalized stens separately to equate variability and minimize skewness. Our findings were fully consistent with SLODR-based predictions, thus refuting suggestions about artifactual genesis of this phenomenon. To cross-validate this result analysis was repeated on second analogous sample after exclusion of all applicants older than 19 years (N = 11934, mean age = 17.91±0.47). And again, all indexes accounted were consistent with SLODR. The results indicate that SLODR effects are emerged on the level of group ability factors, are quite reproducible and relatively independent from test’s g-loading and psychometric properties of their scores. Hypotheses are suggested to account for possible reasons of negative outcomes in some earlier studies.