RESUMO
In this study, we evaluated the factor structure of the Positive and Negative Suicide Ideation (PANSI) Inventory (Osman, Gutierrez, Kopper, Barrios, & Chiros, 1998) using confirmatory factor analysis (CFA). The PANSI assesses the frequency of negative risk and protective factors that are related to suicidal behavior. Participants (n = 195) were adolescent psychiatric inpatients, ages 14 to 17 years, in the CFA. Results of the CFA supported the fit for the 2-factor oblique model as the best fitting model. The internal consistency reliability estimates of the 2 subscales, the PANSI-Negative (alpha = .96) and the PANSI-Positive (alpha = .89) were high. Scores on the PANSI scales differentiated between suicide attempters and controls and those at severe risk for suicide and controls. Correlational analyses provide strong support for the concurrent validity of the scales. The results of the logistic regression analyses provide support for the use of this new inventory as a risk measure of suicide-related behaviors: Scores on the PANSI scales (n = 54) also showed satisfactory evidence for test-retest reliability over a 2-week period.
