Too True to Be Bad: When Sets of Studies With Significant and Non-Significant Findings Are Probably True

doi 10.31234/osf.io/nnkg9