Two divergent views are contrasted: the multidimensional and the unidimensional views of faculty evaluation. The views of H. W. Marsh, the most outspoken advocate of the multidimensional view of evaluation based on factor scores, is summarized and a response to his concerns is presented. (Author/MLW)

Limitations of the confirmatory factor analysis conducted by H. W. Marsh to demonstrate the multidimensionality of a rating form to assess student evaluation of teaching effectiveness are discussed. A secondary analysis suggests that dimensionality is not stable across solutions and a global factor can be retrieved. (SLD)

Multidimensional rating forms used for college student evaluation of faculty elicit information that can be used to evaluate the dimensions of instruction either separately or globally. Portions of the information may be useful for formative evaluation, but global items are best for making personnel decisions. (MSE)