Reliability, construct validity, and subscale norms of the Brief Symptom Inventory when administered to bereaved...

The Brief Symptom Inventory (BSI) was administered to parents (N = 260; 171 mothers and 89 fathers) whose adolescent and young adult children died unexpectedly and violently by accident, homicide, or suicide. Summary statistics and reliability coefficients (Cronbach's alpha) for the nine subscales and the Global Severity Index were calculated. A comparison of means and standard deviations confirmed the expectation that this sample is dramatically different from the normative American community standard. Raw scores for the subscales were transformed into standardized T scores and critical values for a screening heuristic presented. An attempt to obtain construct validity using factor analysis suggested that a five-factor solution provided a description of this population of bereaved parents that is more insightful than the nine standard subscales of the BSI. Implications for both clinicians and future research are discussed.