Self-insight assessment compares outcomes from two model-recovery exercises: a statistical exercise to infer a judge's (implicit) policy and an elicitation exercise whereby the judge describes his or her (explicit) policy. When these policies are mismatched, limited self-insight is not necessarily implied: Shortcomings in either exercise could be implicated, whereby Newell & Shanks' (N&S's)relevanceorsensitivitycriteria for assessing awareness may not be met. Appropriate self-insight assessment requires that both exercises allow the original processes to be captured.

Psychologists' lack of methodological uniformity reflects their greater breadth of enquiry than experimental economists. The need for a theoretical understanding of one-shot decisions validates research undertaken without the repetition of trials. Theories tested only with financial incentives may not reliably predict some classes of decision such as those involving health. Undue emphasis on the importance of replication risks the proliferation of theories with limited generalizability.