Illness Prediction Studies

Abstract

Fourteen psychosocial and occupational illness predictors were assessed as to their utility in the prediction of subjects’ illness reporting over a six- to eight-month period. Subjects were 4,463 US Navy men aboard six large Navy ships. Eleven of these 14 Illness predictors proved to have significant correlations with the illness criterion. Intercorrelations between the significant illness predictors were presented. The 14 original illness predictors were assessed (validation and cross-validation experiment) by the multiple regression analysis technique to determine the best predictive equation. Six of the 14 predictors, three occupational and three psychosocial predictors, were seen upon cross-validation to establish a multiple correlation of 0.27.