Effective evaluation requires evaluators and users to be clear about tasks they are performing in terms of purpose for evaluation, type of data-gathering procedures to be used, and the audience to which results are to be directed. Four things are required if the art-science of educational evaluation is to reach its full potential: (1) an explicit comprehensive model of the evaluation process, reflecting the complexity of the educational ecosystem; (2) explicit, articulated, and valid pedagogies available for application and use; (3) available valid testing and assessment procedures; and (4) theoretical and philosophical clarity. A democratic evaluation process concerns itself with the active, critical initiative of the learner and not simply with the academic outcomes of the learning process as reflected in standardized achievement tests. Three recent, unique approaches to evaluation are discussed: (1) interaction between child and teacher as the primary unit of analysis; (2) interaction between teachers and individual classrooms as the primary unit of analysis; and (3) using the whole school as the primary unit of analysis. Reactions to this article are offered by Daniel L. Stufflebeam, Robert J. Munnelly, and Gilbert N. Garcia. (JMK)