Posted by

The Statistical Approach for USCM9 CER Development

Methods & Models Track

Downloadable Files:



The Minimum-Unbiased-Percentage Error (MUPE) regression technique was used to develop the Unmanned Space Vehicle Cost Model, Seventh Edition (USCM7) and Eighth Edition (USCM8) Cost Estimating Relationships (CERs). This is because MUPE is a well-established technique to model multiplicative error CERs. To prepare for USCM9 CER development, the Air Force Space and Missile Systems Center (SMC) and Tecolote Research have been evaluating various statistical approaches. This effort includes selecting a regression technique, defining CER selection criteria, and evaluating cost improvement modeling alternatives.

After evaluating multiple regression methods, we have again selected the MUPE method for USCM9 CER development because MUPE meets all USCM9 modeling requirements. We evaluated several statistical measures to determine which ones would be useful in identifying statistically sound and significant CERs. Additionally, we identified the statistics that would be useful in evaluating CERs’ predictive capabilities. We also investigated the feasibility of providing prediction interval estimates for uncertainty analysis when the data set is not available.

This paper also explores issues associated with cost improvement. Historically, USCM recurring methodologies assumed a 95% Cost Improvement Curve (CIC) slope to estimate the theoretical First Unit Cost (T1). Recent published research suggested that using the Quantity As an Independent Variable (QAIV) approach for CER development may eliminate the potential error of using an incorrect cost improvement curve slope to estimate recurring cost. (QAIV calibrates the CER slope objectively from the dataset, rather than relying upon expert opinion.) The pros and cons of using the QAIV approach versus the traditional T1 CERs were examined to aid in determining which approach should be used for USCM9 CERs.


Shu-Ping Hu
Educated at National Taiwan University (B.S., Mathematics) and University of California, Santa Barbara (M.S., Mathematics, and Ph. D., Statistics). Dr. Hu is a Chief Statistician at Tecolote Research, Inc. She joined Tecolote in 1984 and has served as a company expert in all statistical matters. She has over 12 years of experience in Unmanned Space Vehicle Cost Model (USCM) CER development and the related database. She also has 20 years of experience in designing, developing, modifying, and integrating statistical software packages for fitting various types of regression equations, learning curves, cost risk analysis, and other PC-based models.

Michael Pfeifer
Mike Pfeifer is a Senior Analyst with Tecolote Research, Inc where he has been supporting the Los Angeles Division for over 11 years. Mike has supported and led numerous space and launch vehicle estimating efforts for both Space and Missile Systems Center (SMC) and NASA programs. Since 2002 Mike has led the development of the Unmanned Space Vehicle Cost Model (USCM) and is currently working on the 9th Edition of USCM. Mike has a BA in Applied Mathematics from the University of California at Berkeley.

Nick Lozzi
Nick Lozzi is a technical manager with Tecolote Research, Inc. where he has been supporting the Los Angeles Division for 19 years. Nick has participated in and led numerous space system estimating efforts for Space and Missile Systems Center (SMC) programs. Nick currently manages all cost estimating, earned value management, and schedule analysis for the MILSATCOM Wing and frequently leads Independent Cost Assessments (ICA) in support of SMC/FM. Nick has overseen and supported the development of the Unmanned Space Vehicle Cost Model (USCM) since 1989 and authored the publications of both the 7th and 8th Editions. Nick has a B.S. in Economics from the State University of New York (SUNY), Brockport, and an M.B.A. from California State University, Long Beach.