Oral

Mental health measurement is based primarily on subjective judgment and classical test theory. Typically, impairment level is determined by a total score, requiring that all respondents be administered the same items. An alternative to full scale administration is adaptive testing in which different individuals may receive different scale items that are targeted to their specific impairment level. This approach to testing is referred to as computerized adaptive testing (CAT) and is immediately applicable to mental health measurement problems. The net result is that an individual is administered a small, optimal number of items from a much larger “bank” of items, without loss of measurement precision. Rather than fixing the number of items and allowing measurement precision to vary, we fix measurement precision and allow the items to vary. The statistical contribution is the development of CAT based on multidimensional item response theory (IRT), which is necessary to incorporate the multidimensionality of mental health constructs. The methodology is illustrated using the recently developed CAT-Depression Inventory (CAT-DI) which requires an average of 12 items but maintains a correlation of r=0.95 with the entire 400 “item bank” score. Sensitivity of 0.92 and specificity of 0.88 are obtained for predicting a clinician-based DSM diagnosis of major depressive disorder. Applications in the areas of screening depression in primary care, child psychiatry, global health, randomized controlled trials, molecular genetics, and psychiatric epidemiology are described.

Session Objectives: 1. Describe item response theory and the bi-factor model 2. Describe computerized adaptive testing 3. Discuss improved mental health measurement

CE Credits: Medical (CME), Health Education (CHES), Nursing (CNE), Public Health (CPH)