Accession Number:

ADA275989

Title:

Software Cost Estimating Models: A Comparative Study of What the Models Estimate

Descriptive Note:

Master's thesis

Corporate Author:

AIR FORCE INST OF TECH WRIGHT-PATTERSONAFB OH SCHOOL OF LOGISTICS AND ACQUISITION MANAGEMENT

Personal Author(s):

Report Date:

1993-09-01

Pagination or Media Count:

173.0

Abstract:

This effort developed a consolidated document which highlights and examines differences in definitions, assumptions, and methodologies used by the REVIC, SASET, PRICES, and SEER-SEM cost models. The following research questions were investigated 1 What differences exist between the cost models 2 How do these differences impact the resulting estimates 3 To what degree can we explain and adjust for known differences between the cost models Seven specific areas were addressed 1 software development phases, 2 development activities and cost elements, 3 source lines of code and language differences, 4 key model attributes and key cost drivers, 5 implications of Project size on model output, 6 impact of schedule compression and extensions, and 7 distinctive characteristics of the model data bases. A hypothetical baseline test case was developed to determine if users could explain and adjust for known differences. It is the researchers opinion that the underlying equations and model assumptions are so dissimilar that objective normalization efforts are virtually impossible for the average model user. Cost Estimates, Cost Models, Software Computers, Comparison, Models

Subject Categories:

  • Economics and Cost Analysis
  • Computer Programming and Software

Distribution Statement:

APPROVED FOR PUBLIC RELEASE