Cost estimation models (1)
     
Date   27 January 1997
Leader   Mary Shaw
     
Papers   Barry W. Boehm.
Software engineering economics.
IEEE Transactions on Software Engineering SE-10:1, January 1984.

Lawrence H. Putnam.
A general empirical solution to the macro software sizing and estimating problem.
IEEE Transactions on Software Engineering SE-4:4, July 1978.

Allan J. Albrecht and John E. Gaffney, Jr.
Software function, source lines of code, and development effort prediction: A software science validation.
IEEE Transactions on Software Engineering SE-9:6, November 1983.

     
Other Material   There are several web sites dedicated both to software cost estimation in general and to particular estimation models:

General.
- The COCOMO2 Team has assembled a general bibliography of cost estimation literature.
- Alan E. Giles and Dennis Barney wrote an article in June 1995 on software cost estimation for CrossTalk, the Journal of Defense Software Engineering.

COCOMO.
- Nasa has created a web form for calculating estimations using Basic COCOMO.
- The COCOMO2 project focusses on modernizing Barry Boehm's COCOMO.
- The International Forum on COCOMO and Software Cost Modeling held its 11th annual meeting in October.

SLIM.
- Putnam's company, Quantitative Software Management, produces a variety of SLIM-based tools.

Function points.
- The International Function Point Users Group (IFPUG) is an advocacy group for the function points model.
- Bruno Peeters has assembled an impressive bibliography of function points literature, which covers publications in several languages.
     
Post Mortem   Unfortunately for our purposes, the canonical work about COCOMO is Boehm's book Software Engineering Economics (Prentice Hall 1981). Because there was no suitably short subset of this work, the TSE article was selected instead. While the article is perfectly fine in its own right, it really doesn't provide a good introduction to COCOMO per se.
   Putnam's paper concentrates on deriving his "software equation," which underlies the SLIM model, from the Rayleigh curve. For this topic, we are much more interested in how this software equation can then be applied to the cost estimation problem. Further, perhaps due to the proprietary nature of SLIM, there does not seem to be a clear introduction to the SLIM approach.
   Similar to the COCOMO problem, the two canonical function points papers appear in unusual publications, which could not be obtained before the meeting. The chosen paper places too much emphasis on information science as a validation of the approach and is unsuitable as an introduction to function points.
   All in all, there is plenty of worthwhile material here for at least one lecture in a PhD-level course. However, rather than trying to teach through the original papers (often a good approach), what's needed here is some effort to digest the material to produce an overview. Aside from teaching the topic itself, these estimation models provide a good opportunity to discuss how emperical ("curve-fit") models can be used in engineering until a theoretically founded structural model exists.

Mary wrote some email with her reaction to the meeting.

[DeLine 02/20/97]