Methods and Results in the IEA Studies of Effects of School on Learning

In this paper, I will first examine the methods used in analysis of school and home effects in three recently-published IEA studies—in science education (Comber & Keeves, 1973), in read ing comprehension (Thorndike, 1973), and in literature (Purves, 1973). The results of this examination lead to measures of effects different from those used in these studies, or in other studies of school and home effects. Using these measures, I come to some what different conclusions than the authors do about the effects on achievement of variations in home and variations in schools. One's first response to the IEA publications must be one of amazement and respect: amazement that such a massive set of studies of cross-national achievement of children could be suc cessfully carried out, and respect for those who have done so. It is important to record this first impression, because the studies constitute the best models in existence for cross-national re search on social institutions and social behavior. This fact should not be lost in detailed comparisons, secondary analyses, and critiques of the three studies. With their publication, the com parative study of the functioning of different societies has made an important advance. I am grateful to Gilbert Peaker, who had much to do with the statistical design and analysis of the IEA studies, for extended comments on an earlier draft of this paper, without, of course, implying his agreement with the methods I propose