Middle-Term Loss Prediction Models for the Air Force's Enlisted Force Management System: Information for Updating
Abstract
This Note describes procedures for updating the middle-term loss equations that are used in the Enlisted Force Management System (EFMS). Updating involves four activities: 1. adding data to the files used to estimate the loss equations; 2. reestimating the current specifications of the equations; 3. exploring possible respecifications of the equations to exploit the additional data or to accommodate new EFMS needs; and 4. testing and evaluating the new versions of the equations intended for use in the EFMS. Adding data to the files used for estimating the models requires understanding the structures of three data files: (1) the Enriched Airman Gain/Loss (EAGL) file, (2) the Year-At-Risk (YAR) file, and (3) the analysis files drawn from the YAR file for use as direct inputs into the estimation programs. Adding data also requires understanding the programs that create the YAR and the analysis files. Reestimating the current specifications of the loss equations only requires understanding the programs that calculate the estimates. Exploring possible respecifications is more demanding. It requires understanding: (1) the statistical strategy underlying the estimation procedures, (2) the perils for estimation inherent in the available data, (3) the uses to which the loss equations will be put, (4) the programs for calculating estimates, and (5) how to adapt the equations in response to information from the testing and evaluation exercise.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1989
- Accession Number
- ADA217309
Entities
People
- Michael P. Murray
Organizations
- RAND Corporation