A Psychometric Investigation of the Suicide Opinion Questionnaire

Abstract

The factor structure of the Suicide Opinion Questionnaire (SOQ) was investigated in 2 studies. In the first study, the SOQ was administered to 568 participants to investigate the relative strength of 3 competing latent structure models previously identified in the literature. Confirmatory factor analyses provided no support for any of the prior models. An exploratory factor analysis of these data resulted in a weak 2-factor structure accounting for only 15.33% of the common variance. In the second study, this 2-factor structure was evaluated using SOQ data from 288 college students. Confirmatory factor analysis results failed to support the 2-factor model. Based on these results, the authors recommend that future research efforts should focus on developing a more contemporary and psychometrically sound measure of attitudes toward suicide.

Notes

Note. ∗indicates statistically significant values at p < .001. Statistically significant chi square values indicate poor fit to the data, but with large samples, the χ²/df ratio of 2 or less is recommended for interpreting good fit. Cut-off values for interpreting good fit for the NFI and the TLI are .90. The recommended cut-off value for the PNFI is .80. Recommendations for interpreting fit for the RMSEA are < .05 for good fit and values between .05 and .08 as adequate fit.

NFI: Normal Fit Index; TLI: Tucker Lewis Index; RMSEA: root mean square error of approximation; CI: confidence interval.

Note. ∗reverse scored items.

Note. ∗indicates statistically significant values at p < .001. Although statistically significant χ²values indicate poor fit to the data, with large samples, the χ²/df ratio of 2 or less is recommended for interpreting good fit. Cut-off values for interpreting good fit for the NFI and the TLI are .90. The recommended cut-off value for the PNFI is .80. Recommendations for interpreting fit for the RMSEA are < .05 for good fit and values between .05 and .08 as adequate fit.