Teacher Evaluation: The Limits of Looking

Evaluation of teachers rests on the assumption that the character-istics of "good " or effective teach-ers are known and recognizable. Ef-fective teaching has been conceived as generally present or absent in a particular individual. In order for a teacher evaluation procedure to be credible and fair, stability and con-sistency of teacher behavior is assumed. Observations of teachers in their classrooms are often an important part of teacher evaluation systems and procedures. In this paper, I will present both conceptual and empir-ical arguments to challenge the va-lidity of most teacher evaluation procedures that use observations of teachers. It will be argued that ef-fective teaching is not characterized by one set of behaviors that are con-text-free. Rather, teaching must be seen as context-bound. The stabili-ty and consistency of teachers ' be-havior in elementary schools will also be examined because if obser-vations of teachers are used for evaluative purposes, the degree to which teachers are consistent across teaching situations and occa-sions must be known. I will proceed by reviewing cur-rent practices in teacher evaluation with a particular focus on the use of observation. In order to address the