-20 on improvement in pedagogic skills under laboratory or simplified training conditions. Criterion Level 6 is concerned with the effects of a training program on improvements in teacher knowledges and understandings. (53:7). Since Criterion Levels 2 and 3 form the basis for this dissertation, they will be elaborated further. Criterion Level 2 consists of two parts by which teachers may be appraised. The first part is observation of the acts or behaviors in which the teacher engages in the classroom. The observations must be conducted with a set of instruments which permit classification of teacher behaviors in both the cognitive and affective domains. The second part is systematic analysis of the level of outcomes achieved by the teacher with the pupils he teaches. This criterion level is identical to Criterion Level 1 except that a shorter performance period is involved. Criterion Level 3 differs from Criterion Levels 1 and 2 in that pupil performance data are eliminated from the criterion. Judgments about competence or proficiency are thus based on the observable behaviors of the teacher rather than on the pupil outcomes associated with these behaviors. Nonetheless, this criterion level is still performance-based in the sense that the teacher actually does engage in teaching and is gauged on the quality of his professional actions. How good or valid this criterion level is depends almost wholly on whether empirical relationships between teacher actions and pupil performance have been established through research or through data obtained by use of Criterion Levels 1 and 2. (53:4-5) Much debate is presently ensuing over the question of whether teacher education will be held accountable by producing effective change in pupil behavior or teacher behavior. The Committee on National Priorities in Teacher Education claim