Assessing School Effects: Some Identities

Two methods of assessing school effects-the contextual variables method and the analysis of covariance-are discussed within the framework of a general school effects model. The similarities of the two methods are elaborated and it is shown that in the general case they are identical with respect to the identification and interpretation of school context effects. Problems of measurement error and grouping bias are discussed briefly in connection with the applicability of the model. An example shows how contextual effects in both metric and standard form can be computed. The general usefulness of these methods for studying group effects is emphasized.