A program manager's guide for software cost estimating

Abstract

This thesis will assist current and future program managers by outlining a process to ensure the software cost estimates developed for a system will be credible and supportable throughout the life of the program. This thesis also identifies many of the problems associated with software cost estimating and recommends potential solutions. One of the critical parameters for estimating software cost is the quantity of source lines of code (SLOC) required in the program. Therefore, this thesis examines the software cost implications of improperly estimating SLOC and function points. Some of the other parameters required to estimate the software cost include language, functionality, application, software processes maturity, programmer skill level, design and reuse, productivity factors, complexity, utilization and schedules. Many of these parameters overlap. For example, both the complexity of the code and skill level of the programmer directly impacts the productivity and schedule of the program. This thesis provides a broad view of the software cost estimating process. In the reference and appendix section, a list of valuable resources including commercial estimating models is provided for further assistance.