InterActions: UCLA Journal of Education and Information Studies

This article first explains how to estimate the most common value-added models and discusses the assumptions underlying each model. Second, we compare how the models differ in estimates of effectiveness for individual teachers, emphasizing the large differences found for some teachers from one model to another.  Finally, our study illustrates how policies for the use of value-added models could mitigate the implications of these large differences and capitalize on the strengths of these models.