Eg. Carbone et al., Predicting outcome of military basic training for individuals referred forpsychological evaluation, J PERS ASSE, 72(2), 1999, pp. 256-265
We examined outcome data for 632 U.S. Air Force personnel who were referred
for psychological evaluation during Basic Military Training (BMT) but who
were subsequently returned to BMT to determine what proportion graduated. W
e analyzed motivational, biographical, and psychological testing variables,
using logistic regression to develop a model predictive of training outcom
e. The results demonstrated that a relatively small number of variables cou
ld predict outcome with close to 70% accuracy. Level of optimism regarding
training, history of physical abuse, and frequency of visits to the trainee
health clinic were major contributors to the model. MMPI-2 (Hathaway & McK
inley, 1989) Scales D and Sc also remained in the model but added little to
its power. The findings are generally consistent with prior research on no
rmal military populations, except that some factors previously linked to ad
justment, such as sex and ethnicity, were found to be unrelated to training
outcome in this population.