Psychometric assessment of the Multidimensional Observation Scale for Elderly Subjects (MOSES).

The factor structure of the MOSES proposed by Helmes et al. (1987) was tested using both exploratory and confirmatory factor analysis procedures. Results indicated that although the number of factors suggested was appropriate, the manifest variables proposed for each latent factor were not stable. A modified 5-factor model using 24 of the original 40 variables was proposed and tested. Confirmation of the modified model suggests factorial invariance across two independent samples. Internal consistency for the five scales measured by coefficient alphas ranged from 0.62 to 0.92. Difficulties in scoring MOSES items are discussed, and solutions offered for alleviating the problems. The 5-factor MOSES model is suggested as a useful research and clinical tool.