Using College GPA and Test Scores in Teacher Licensure Decisions: Conjunctive Versus Compensatory Models

The logic of using a sequential decision-making approach using college grade point average (GPA) and test scores for teacher licensure decisions within the conjunctive decision-making model is discussed and contrasted with the compensatory model for decision making. The major issues to consider in choosing between the conjunctive and compensatory models are (a) whether there is a linear (or at least monotonically increasing) relationship between each predictor variable and the criterion variable and (b) whether an excess of one predictor variable can compensate for a deficiency in another variable. If linearity (monotonicity) and compensation are reasonable assumptions, the compensatory model is generally preferred. However, for nonlinear and noncompensatory situations, such as teacher licensure where the purpose is to set a cutoff score to represent minimal competence, not to predict the degree of success, the conjunctive model is more appropriate. Some have suggested using college GPA and licensure tests in a compensatory model for licensure decisions. However, given the lack of content validity of the GPA for the purpose of licensure, and given the lack of comparability of GPAs across institutions (and indeed within institutions), a compensatory model using test scores and college GPA seems inappropriate.