Costs and Errors in Survey Sample Design: An Application to Army Prospect and Recruit Surveys.

Abstract : This report discusses the implications of survey costs and sampling errors for the design for marketing, program evaluation, and sales satisfaction surveys used in Army recruiting. The first section gives an overview of relevant research on the recruitment process, describes survey populations at different stages of recruiting, and identifies major variables that affect sample design and survey content. The second section examines the effect of seasonality on survey response and sampling and presents results of data analyses. The third section presents a general model of costs and errors in survey sampling for four stages of the recruiting process: initial appointments, applications, contracts, and accessions. The fourth section discusses application of the model to ongoing surveys at all four stages of the recruitment process. In particular, the costs, benefits, errors, applicability, and practicability of alternative sampling plans are described. This report is the first of two reports on this subject.