Enlistment and retention in the 1979 and 1988 Navy cohorts

Abstract

This thesis examines the 1979 and 1988 cohort files. The cohort files are maintained by the Defense Manpower Data Center, and contain a total of 365 variable fields to record information about personnel who enlist in a given year. Significant variables among the total are identified, and frequency analysis is performed. Additionally, the significant variable category values are cross.tabulated with reenlistment statistics, to obtain information about reenlistment characteristics. Finally, a LOGIT regression analysis model is constructed and run for each of the cohort files, to establish each variable category value's effect upon the probability for reenlistment. The results of analysis for each cohort year provide a mechanism for viewing the changing characteristics of cohort groups. The thesis also contains a review of previous literature related to the subject of analyzing and/or predicting reenlistment. The results of the analysis performed in the present work are compared to the results of previous studies.