The primary aim of this study is to examine, in the context of Alzheimer's disease and related disorders (ADRD) problem behaviors, the relationships between caregiver controllability attributions, affect, the relationships between caregiver response behaviors. Weiner's attribution- affect-helping model forms the theoretical framework for the study. Independent variable model forms the theoretical framework for the study. Independent variable derived from Weiner's theory include controllability, attributions for care recipient problem behaviors, caregiver anger, number of care recipient problem behaviors, and selected demographic variables. Caregiver response behavior is the dependent variable for the multivariate analyses. This cross-sectional, correlational, multivariate study will be a secondary analysis of a random sample (n=130) of data collected in the National Caregivers Study (Buckwalter et al. 1992-1997). Caregiver attributions recorded in the baseline interviews will be identified, extracted and coded for degree of controllability using the Controllability Rating and coded for degree of controllability using the Controllability Rating and coded for degree of controllability using the Controllability Rating Scale (CRS). Caregiver anger and care recipient problem behaviors will be measured by standardized instruments using in the NCS. Two potential measures for caregiver response behavior will be piloted.