Use of a Binary Choice Model to Determine Marine Officer Attrition

摘要

The ability to predict Marine officer attrition continues to be a significant obstacle to accurate manpower planning at Headquarters Marine Corps. This thesis expands on a previous study which used unemployment in a linear regression model to predict attrition. Through the use of a binary choice model, this study incorporates a number of additional variables, principal of which is a performance index score to measure promotion potential. The logit function was employed to evaluate the inherent causal relationships and to further define the ability of statistical analysis techniques to predict officer attrition.