In this commentary, we examine Braverman's insights into the trade-offs between feasibility and rigor in evaluation measures and reject his assessment of the trade-off as a zero-sum game. We, argue that feasibility and policy salience are, like reliability and validity, intrinsic to the definition of a good measure. To reduce the tension between…

Discusses the process of conducting a summative evaluation, characteristics of evaluators, and choice of assessment techniques. (SK)

Ethical issues relating to naturalistic evaluation are addressed, focusing on the role of the evaluator, problems of privacy and data gathering techniques, and issues relating to the use/abuse of findings. Benefits and costs of the naturalistic approach to program evaluation are also identified. (EGS)

Triangulation strategy results in evidence characterized by the following: (1) convergence; (2) inconsistency; and (3) contradiction. In order to render the data sensible, the researcher or evaluator must report data collection procedures, as well as the three levels of information from which explanations of social phenomena are constructed. (BJV)