ABSTRACT
Researchers have found it difficult to determine the effects of teacher and school factors on schooling outcomes because many policies and practices do not vary much across schools, and because schools that are advantaged in some respects tend to be disadvantaged in others. For example, class size may be an important determinant of academic achievement, but in most schooling systems average class size does not vary substantially across schools. In systems where there is sufficient variation, schools with small class sizes tend to be advantaged also in terms of the experience and training of their teachers (e.g., Rumberger and Willms, 1991). Although it may be easy to show that schools with small class sizes have superior academic achievement it is difficult to discern whether it is class size or teacher characteristics causing the desired effect. Another difficulty is that the variation between schools in their outcomes is usually small compared with the variation within schools (Gray, 1988). Even though a school or teacher variable may have an important effect on schooling outcomes, its effect may be small compared with the effects of all of the other factors influencing pupil outcomes.
