A computer programmer productivity prediction model

Abstract

A computer programmer productivity prediction model has been developed. This development was preceded by an investigation of the programmer productivity factors. The research was accomplished in three parts. To identify important productivity factors, a data analysis was performed using a combination of existing data bases with stepwise multivariate regression analysis. The derived equation, with programmer productivity as the dependent variable, is included. To gain additional insight into productivity factors, an alternate non-analytical method was used. Experts on computer programming project management were surveyed. The iterative Delphi survey technique was used to identify important programmer productivity parameters from a panel of experts. A background report on the Delphi procedure and the survey results are included. The importance of programmer interactions was evident in both the data analysis and Delphi studies. ...