Predicting Attrition in the Military Occupational Specialty Phase of an Army Specialty-Training Program
Abstract
Identifying personnel with the right attributes to perform high-risk specialized military duties is critical for individual, team, and unit effectiveness. A large body of research has identified a core set of characteristics common among high-risk personnel and special operators who successfully complete rigorous assessment and selection (A and S) courses (79). However, the attributes required to succeed at later stages of training have not been identified. The present study examined whether demographic, performance, and psychological data collected during A and S could predict training outcome (i.e., graduation and attrition) during the military occupational specialty (MOS) phase of a specialized training course. To address this research gap, data from 1,460 male, active duty, enlisted Army candidates who attended an Army specialty-training program between fiscal years 2014 and 2017 were analyzed using multinomial logistic regression and classification and regression tree (CART) analyses. Study findings revealed that the modeling strategies: [a] identified different variables for each MOS; [b] produced different findings between models for each MOS; and [c] had high rates ofviimisclassification. There are several possible explanations for these findings including: [a] a need for more specific and sensitive assessment instruments, [b] attributes other than those examined are associated with MOS training outcome, and/or [c] the units current MOS classification procedures are as effective as possible. The present study also identified the need for thorough and careful data management strategies. Dedicated database managers, quality assurance procedures, and appropriate statistical analyses are important for meaningful interpretation of the vast amounts of data collected from military personnel.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 24, 2017
- Accession Number
- AD1132863
Entities
People
- Amanda R. Webb
Organizations
- Uniformed Services University of the Health Sciences