Evaluation of the Swedish Trauma Symptom Inventory-2 in a clinical and a student population

Abstract

IntroductionThe Trauma Symptom Inventory-2 (TSI-2) is a broad-spectrum assessment instrument designed to identify symptoms that can appear in the aftermath of potentially traumatic experiences. ObjectiveThis study aimed to evaluate the external and internal validity of this newly reconstructed instrument. MethodIn total, 696 individuals participated in the study, including 83 psychiatric outpatients. Participants answered the TSI-2, together with a trauma history questionnaire, and other questionnaires assumed to correlate with the different scales included in the TSI-2. ResultsValidity was evaluated by correlations between the TSI-2 and the other instruments and by the differences between clinical and non-clinical populations. Reliability was calculated by testing internal consistency and test-re-test reliability. A confirmatory factor analysis (CFA) was computed to test the postulated four-factor structure. Cronbach's alpha was found to be good and ranged from α=.77 to .91 and test-retest reliability was strong. Strong to satisfactory correlations were found between the TSI-2 and the other instruments. The student sample scored significantly lower than the clinical group on all clinical scales. Sensitivity and specificity were calculated with different cut-off scores. ConclusionDespite the CFA demonstrating a questionably good model of fit, most of the scales proved to be sound and the TSI-2 could be recommended as a broad-spectrum assessment instrument.