Social psychology findings have fared poorly in multi-site replication attempts. This article considers and evaluates multiple factors that may contribute to such failures, other than the “crisis” assumption that most of the field’s published research is so badly flawed that it should be dismissed wholesale. Low engagement by participants may reduce replicability of some findings (while not affecting certain others). Incentives differ between original researchers and replicators. If multi-site replications are indeed biased toward failure, this may have a damaging effect on the field’s ability to build correct theories.
