Forecasting U.S. Marine Corps Reenlistments by Military Occupational Specialty and Grade

摘要

Each year, manpower planners at Headquarters Marine Corps must forecast the enlisted force structure in order to properly shape it according to a goal, or target force structure. Currently the First Term Alignment Plan (FTAP) Model and Subsequent Term Alignment Plan (STAP) Model are used to determine the number of required reenlistments by Marine military occupational specialty (MOS) and grade. By request of Headquarters Marine Corps, Manpower and Reserve Affairs, this thesis and another, by Captain J.D. Raymond, begin the effort to create one forecasting model that will eventually perform the functions of both the FTAP and STAP models. This thesis predicts the number of reenlistments for first- and subsequent-term Marines using data from the Marine Corps Total Force Data Warehouse (TFDW). Demographic and service-related variables from fiscal year 2004 were used to create logistic regression models for the FY2005 first-term and subsequent-term reenlistment populations. Classification trees were grown to assist in variable selection and modification. Logistic regression models were compared based on overall fit of the predictions to the FY2005 data. Combined with other research, this thesis can provide Marine manpower planners a means to forecast future force structure by MOS and grade.