Confirmatory Factor Analysis of the Center for Epidemiological Studies DepressionScale (CES-D) in Military Recruits

摘要

Personnel morale is important for military units, and surveys to assesssatisfaction and well-being are commonplace. These surveys frequently rely on measures of psychological status originally developed and validated in civilian populations. One advantage of this approach is that evidence of construct validity that has accumulated in the development of the measures can be invoked as a basis for claiming validity in the military survey. This generalization is legitimate provided conditions which would affect the measures are comparable in the two populations, but the generalization can be misleading if the demands of training and operational conditions in the military alter the significance of the indicators comprising the survey measure. In Meehi's (1991) terminology, the ceteris taribus (i.e., all other things equal) condition for valid hypothesis tests would not hold in this case. Tests of the validity of ceteris paribus assumptions are needed to determine whether it is legitimate to compare results obtained in military populations to norms developed in civilian populations. In the case of military personnel, the demands of operational or training conditions could selectively alter scores on some indicators of morale (e.g., fatigue symptoms). Comparing scores under these circumstances to civilian norms might convey a misleading impression that morale was suffering when all that was actually observed was a normal response to the demands of the situation.