CART analysis with unit-weighted regression to predict suicidal ideation from Big Five traits

This study examines the Big Five traits as multiple predictors of suicidal ideation. In addition to using multiple regression, this study introduces the use of classification and regression tree (CART) analysis to identify cutoff scores that can be used with unit-weighted regression; the purpose of this approach is to offer a method of analysis that may be useful in a public health approach to suicide prevention. The models were developed with a derivation group (N = 299) and applied to a validation group (N = 175). Both models performed well in the validation group, with multiple regression correlating at 0.47, and unit-weighted regression at 0.48. The models agreed in suggesting that increased suicidal ideation is associated with high Neuroticism, low Extraversion, and low Agreeableness. Low Conscientiousness was correlated with suicidal ideation in unit-weighted regression. (C) 2002 Elsevier Science Ltd. All rights reserved.