An analysis of DoD Inspector General's statistical sampling plan for Navy repairable item procurements

Abstract

A recent audit by the Department of Defense Inspector General (DODIG) of Navy inventory control points found a high value of purchase requests for repairable items that the auditors labeled as unnecessary or excessive. The dollar figures reported were based on the auditors' use of stratified sampling. This thesis examined the auditors' use of stratified sampling by attempting to replicate the auditors' process of stratifying and sampling. The author then attempted to verf the auditors' claimed confidence level and precision of the final result. This study questions the chosen sample size and sample stratification. In addition, this thesis found that the auditors' actual precision was not as tight as stated in the DODIG audit report. This was caused by the auditors' emphasis on the very high dollar value strata which had only a few purchase requests rather than on the stratum with the largest number of purchase requests. It was this latter stratum which had the highest projected number and total dollar value of excesses.