Discusses ways of resolving anomalies in the evaluation of human resources training programs. Interpretations of reliability; Respondent rift; Different responses to questions focused on the same general factors; First impressionism; Collision of group responses.