* I have long wondered whether psychologists’ ideological refusal to recognize group differences undermined the bulk of these studies. They rarely identify the race or ethnicity of subjects and blindly assume that these factors would not have any effect on the outcome. It seems very unlikely. The replication failure may be linked to this, at least on some significant level.
* I think two other major factors are affecting the reproducibility of psychology studies, and social sciences studies generally. First, there’s an enmeshment between popular outlets that publicize these studies’ results, and the scientists who perform the studies. The scientists want the publicity of having their study referenced in some magazine, and the magazine writers love having the veneer of Science fronting for their biases. This puts pressure on the research to find “interesting” results, rather than particularly honest ones – and this is especially true if the study can be related to political alignments, however spuriously (see e.g. “Conservatives scare easily”). The second factor is the biases of the scientists themselves. It’s been repeatedly shown that social science peer groups have remarkably similar worldviews. This of course pushes results away from the unfettered truth, toward an ideologically preferable one. There is research about this that has been published, particularly by Haidt et al. Although who knows if it will prove reproducible.
