pp. 273-294

At elementary school age, learning to read takes children varying lengths of time depending on the class the child is in and also to a much lesser degree, on the school itself. Furthermore, the performances from one class to another in the same school are not homogenous ; the differences between schools are due to one or two classes with a particularly good level (or bad level) which tend to raise (or lower) the level within that school and not a global progression of all the classes together. These results, combined with the fact that there is a low level of homogeneity in teaching methods and expectations (testing) within one school, make the idea of the school as a strongly-formed entity difficult to accept. It would seem, given these results, that the school effect should be interpreted more as an aggregation of class effects than as the effect of the school as an entity. A number of elements are put forward in reference to the theories of organization sociology in order to explain the results.