Date of Award


Document Type


Degree Name

Master of Science in Cost Analysis


Department of Mathematics and Statistics

First Advisor

Edward D. White, PhD.


Cost growth is a problem DoD wide. Cost Estimators attempt to remedy this problem by accounting for uncertainty in the estimates they complete. They use tools such as Engineering Change Orders (ECO’s) to account for the uncertainty, by applying a percentage to the final amount estimated. The following research gives the acquisition community a more precise tool to predict whether a DoD Acquisition Contract will have an Engineering Change Order, which can then be used also during programmatic cost estimating, and also a method for predicting the proper amount of ECO to apply when certain variables are present. The study used both logistic and multiple regression to accomplish this. For both types of regression a stepwise approach was adopted for the response. For the Logistic Regression the Y variable was that an ECO was present and the significant predictor variables were: UAV, >500M (dollars), Navy, Army, Aircraft, Firm Fixed Price (FFP), Cost Plus Fixed Fee (CPFF) and <5M (dollars). The final model was 85% predictive. The multiple regression modeled the expected ECO percent change (less than 100% of baseline). Predictive variables included: <5M, FFP, Munition, Electronics and Missiles, along with a base amount of 22% ECO. This model was more exploratory in nature due to the extreme variability present in ECO percent changes.

AFIT Designator


DTIC Accession Number