Abstract
Using a military sample comprised largely of National Guard personnel, zero-inflation negative binomial regression was applied to estimate the effects of indirect, nonface valid indicators of suicide ideation [Thwarted Belongingness (TB), Perceived Burdensomeness (PB), and Hopelessness], in predicting suicide ideation. Data from a sample of 497 military personnel (82.1% male; m
= 27.24; range = 18-59) were analyzed. TB and the interaction of TB with Hopelessness were significant predictors in the logistic regression, and in the negative binomial regression, the main effects of TB and hopelessness, and the interactions of TB with hopelessness and PB with hopelessness were significant. The findings further indicated that approximately 10% of those not reporting ideation would be predicted to be ideators. Clinically, these results indicate that, in samples reluctant to report ideation, the assessment of suicide risk may improve through the use of relevant measures that do not explicitly reference suicide thoughts.