A framework for Army Reserve recruiting analysis : enlistment to initial training

Abstract

Analysis of U.S. Army Reserve recruiting is conducted across the U.S. Army with data from the Recruit Quota System (REQUEST). A combination of partial manual data entry and a decided lack of tools for large scale data extraction make REQUEST difficult to use for analysis without an extensive knowledge of the system. In this thesis, I develop a process for screening, preparing, and evaluating REQUEST data for subsequent analysis. This process uses data mining software to progressively work through a series of rules that outline data inconsistencies, mark these records for exclusion and later investigation, and generate a "clean" dataset for analysis. I examine enlistments over a four year period with respect to Military Occupational Specialty and training program structure. Data from the Army Training Requirements and Resource System (ATRRS) are used to provide an overview of Initial Entry Training seat quotas and usage, and to confirm and/or update training dates in the REQUEST dataset. The joint examination of enlistments and training seats provides new insights into enlistment patterns. Additional analysis is possible using demographic data provided by the U.S. Army Recruiting Command. I provide summaries of a few key demographic variables for various subsets of the enlistees, and discuss how similar analyses might prove useful for targeting recruiting efforts and incentives more effectively. Good decisions require good data. This thesis is a start in providing a framework for generating quality USAR accession data for analysis.