The philosophy and assumptions of the Multi-Attribute Utility-Bayesian Decision Theoretic model (MAUT-Bayesian model) are presented. The evaluator uses the MAUT-Bayesian model along with the knowledge of the decision-maker's, and perhaps the evaluator's own values to decide what data should be collected. Appropriate data are presented to the decision-maker, who weighs the alternatives suggested by the multiple aspects of the data. Using an educational policy question as an example, each step of this decision-making process is described. Implications for formative and summative evaluations and post-assessment organizational change are offered. (Author/MV)