Problems with Value-Added Evaluations of Teachers? Let Me Count the Ways!

Abstract

In the United States, but not only here, the movement to evaluate teachers based on student test scores has received powerful political and parental support. The logic is simple. From one testing occasion to another students should show growth in their knowledge and skill. Similar types of students should show similar patterns of growth. Those students that show more growth than the average must have the better teachers, while those that show less growth than the average must have poorer teachers. If the value added by teachers to students' knowledge and skill, as measured by the students' test score growth is quite large, those teachers can be rewarded. If the value added is quite small, those teachers should be fired and replaced by better teachers. This simple and logical approach to teacher evaluation is, however, impossible to do fairly, reliably, and validly, but is nevertheless increasingly being used throughout the country. Here are my views on what is wrong with value-added models of teacher evaluation.