Attitudinal versus ecological approaches to the characterization of institutional treatment environments

The relational and discriminative properties of attitudinal and ecological measures of environments were examined within the context of a state hospital for mentally retarted clients. Three hypotheses were advanced, and the results gave substantial support for one of these and partial support for a second: (a) that the correlations between attitudinal and ecological measures should not differ significantly from zero, and (b) that differences among hospital treatment programs and living units should be reflected in ecological but not in attitudinal measures. However, the results did not provide clear confirmation for a third hypothesis: that differences among staff demographic characteristics and employment history should be reflected in attitudinal but not in ecological measures. It was suggested that these background variables may not be appropriate measures of personality-linked attributes.