The Journal of Nervous and Mental Disease · 1980 · 14 citations · 0 references
Social SciencesPsychiatryPsychologyAbnormal PsychologyPsychiatric AssessmentPsychometricsDepressionDepression AssessmentDepressive DisordersMental HealthPsychological TestingSymptom Burden AssessmentDepression Symptom AssessmentDepression Symptom BurdenDepression SymptomatologyDepression SymptomsDiagnostic Assessment Of DepressionSelf-report Depression ScalesSelf-report MeasurementStatistical Discriminant Analysis
The present report describes the initial development of a comprehensive self-report inventory for assessing dysphorias and their response to treatment. A review of the clinical literature and experience with both outpatients and inpatients indicated that numerous dysphoric qualities commonly burden the ongoing lives of depressed patients. A set of 20 dysphorias was considered representative of these burdens, and each was interpreted in the form of three self-descriptive items. The resulting 60-item inventory was administered to 110 newly hospitalized depressed patients and 83 matched controls. The responses of the depressed patients were intercorrelated and submitted to principal component factor analysis with varimax rotation. This analysis produced 16 interpretable factors involving 54 items with significant factor loadings (≥.40). Each factor was clinically recognizable. Fiftythree of the 54 factorially significant items distinguished between the depressed and control groups in the required direction (p<.01). Factor cluster scores were prepared for each factor. Normative distributions for the 16 factor cluster scores are presented for the depressed patients. The ability of the cluster scores to distinguish between the depressed and control samples was examined by a multivariate discriminant analysis. A very high degree of separation was indicated, and the cluster scores of suicidal depression and burdening of others had the greatest discriminating potential. A traditional discriminant classification procedure using the 16 cluster scores indicated that the clusters discriminated between the two groups quite well, with 89.1 per cent of the subjects assigned to their appropriate group.