Date of Award

3-2023

Document Type

Thesis

Degree Name

Master of Science

Department

Department of Systems Engineering and Management

First Advisor

Edward D. White III, PhD

Abstract

Time-phasing describes the process of estimating expenditure profiles for multi-year RDT&E acquisition programs. The AFCAH states that one approach to producing a time-phased estimate is with the use of S-curve distributions. These S-curves are most typically modeled as Weibull, Rayleigh, or Beta distributions. Many prior studies have been completed to better understand which of the three distributions is best suited to time-phasing RDT&E programs, and what program characteristics significantly explain variance in the distributional parameters. Those prior studies utilized either contractor-supplied data or SARs, and typically found the Weibull to be the best performing distribution. This paper makes novel use of CCaR to access more granular schedule and expenditure data than prior studies. Utilizing this more granular dataset, this thesis evaluates the performance of the three distributions in time-phasing and assesses program characteristics, including contract expenditure presence and propensity, for significance in altering distributional shape. Utilizing stepwise linear regression, this thesis indicates several significant program characteristics, but finds no evidence to indicate a difference in time-phasing based on contract expenditures. Additionally, two-way ANOVA tests find significant evidence that each distribution varies in predictive strength across schedule deciles, but no significant indication of a difference in performance between the three distributions.

AFIT Designator

AFIT-ENV-MS-23-M-168

Comments

A 12-month embargo was observed.

Approved for public release: 88ABW-2023-0183

Share

COinS