Accession Number : ADA132335

Title :   Techniques for Estimating Pay Entry Base Date Enlisted Personnel Force Structures from Data Categorized by Total Active Federal Military Service.

Descriptive Note : Special rept.,


Personal Author(s) : Borack,Jules I

PDF Url : ADA132335

Report Date : May 1977

Pagination or Media Count : 33

Abstract : The purpose of this research effort is to develop and assess the accuracy of mathematical techniques which estimate PEBD force structure data (and, concurrently, obtain cost estimates) from TAFMS force structure data. Statistical techniques involving marginal estimation, cell-by-cell conversion, and lagged correlation and regression models were developed and evaluated in terms of their ability to both accurately estimate individual force structure cells and obtain the cost of the entire force structure matrix. The data base studied consisted of series of annual TAFMS and PEBD force structure matrices for the period FY1966-FY1976. Analysis consisted of applying various statistical techniques to 6-year streams of data and applying the obtained results to convert TAFMS data of the following year to PEBD. In this way, estimates of PEBD force structures for FY1973-FY1976 were obtained and compared to the actual PEBD matrix of the corresponding year. Choice of statistical technique for implementation was based upon both theoretical considerations and the result of these comparisons. A statistical technique based upon apportioned highest lagged correlation linear regression estimates provided highly accurate force-structure cost estimates. A statistical technique based upon apportionment of prior-year conversion factors also provided surprisingly accurate cost estimates. Utilization of either an apportioned linear regression or apportioned prior-year conversion factor technique substantially improved estimates obtained from naive costing of TAFMS data matrices.

Descriptors :   *Mathematical models, Mathematical prediction, Personnel management, Life cycle costs, Grade structure(Personnel management), Statistical processes, Cost estimates, Cost models

Subject Categories : Personnel Management and Labor Relations
      Statistics and Probability

Distribution Statement : APPROVED FOR PUBLIC RELEASE