Date of Award
3-24-2016
Document Type
Thesis
Degree Name
Master of Science in Cost Analysis
Department
Department of Systems Engineering and Management
First Advisor
John J. Elshaw, PhD.
Abstract
The purpose of this research was to investigate the flattening effect at tail end of learning curves by identifying a more accurate learning curve model. The learning curve models accepted by DOD are Wright’s original learning curve theory and Crawford’s Unit Theory. The models were formulated in 1936 and 1944 respectively. This analysis compares the conventional models to contemporary learning curve models in order to determine if the current DOD methodology is outdated. The results are inconclusive as to if there is a more accurate model. The contemporary models are the DeJong and S-Curve and they both include an incompressibility factor, which is the percentage of the process that includes automation. Including models that incorporate automation was important as technology and machinery plays a larger role in production. Wright’s model appears to be most accurate unless incompressibility is very low. A trend for all models appeared. The trend is Wright’s curve was accurate early in production and the contemporary models were more accurate later in production. Future research should have an objective of finding a heuristic for when the models are most accurate or comparative studies including more models.
AFIT Designator
AFIT-ENV-MS-16-M-162
DTIC Accession Number
AD1054102
Recommended Citation
Johnson, Brandon J., "A Comparative Study of Learning Curve Models and Factors in Defense Cost Estimating Based on Program Integration, Assembly, and Checkout" (2016). Theses and Dissertations. 400.
https://scholar.afit.edu/etd/400