Development of a steady state model for forecasting U.S. Navy Nurse Corps personnel

Abstract

This thesis developed a deterministic Markov state model to provide the U.S. Navy Nurse Corps a tool to more accurately forecast recruiting goals and future years force structure. Nurse Corps data was provided by the Nurse Corps Community Manager's office covering fiscal years 1990 to 2003. The probabilities used in the Markov model were derived from the fiscal year data. Transitions used in this model were stay at present grade, move up one grade or exit the system. Backward movement was not allowed and individuals could only move up one grade per year. The model was limited to eleven years and focused primarily on the ranks of O-1 to O-3. O-4's and O-5's that appeared in the data were allowed to flow through the system. Logistic regression was then used to investigate the probability of "staying" in the Nurse Corps to certain career decision points. Nurse Corps cohort data files for fiscal years 90 through 94 were merged for analysis, as was cohort data for fiscal year 96 through 98. Results of the markov model show that the O-1's and O-2's reach a steady state at the eight-year mark while the O-3's reach a steady state at the seventeen-year mark (based on provided data). The steady state values are compared to actual Nurse Corps goals. Results of the logistic regression show that Recalls, Medical Enlisted Commissioning Program and Nurse Candidate Program were all significant at increasing the probability of staying in the Nurse Corps. Males were more likely than females to stay in the Nurse Corps and changes in education levels decreased the probability of staying in the Nurse Corps.