Spearman's hypothesis not supported? Three meta-analyses of Black and White prisoners, Northeast Asians, and Arabs and Jews

•Spearman's hypothesis: group differences are a function of g-loadedness of subtests.•Spearman's hypothesis is tested on prisoners, Northeast Asians, and Jews and Arabs.•Three meta-analyses do not support Spearman's hypothesis.•Spearman's hypothesis appears to be supported half of the time at level of scales.

Spearman's hypothesis states that differences between groups on the subtests of an IQ battery are a function of the g loadings of these subtests, and it is virtually always supported. There are anomalies: comparisons for which Spearman's hypothesis shows no supporting evidence. What they all have in common is an unusual Verbal/Performance profile. We carried out meta-analyses for these comparisons: Black and White prisoners, Asians and Whites, and Arabs and Jews in Israel. We carried out three meta-analyses based on, respectively, 2 data points and a total N = 575; 15 data points and a total N = 8100; 5 data points and a total N = 730. In all three meta-analyses Spearman's hypothesis showed no supporting evidence with sample-size weighted correlations with values of, respectively, −.18, .01 and −.24. Correcting for reliability led to similar outcomes. Bypassing the unusual V/P profile by testing Spearman's hypothesis separately on Verbal subtests and Performance subtests led to a confirmation in approximately half of the comparisons. So, it could be that the unusual V/P profile obstructs support for Spearman's hypothesis in approximately half of the cases. We conclude that the anomalies appear to be more in line with Spearman's hypothesis than previously thought.