Statistical Analysis of a Software Department's Project Planning Process
Gary Tracy · PDXScholar (Portland State University) · 1994
This project is a statistical study of the planning process of a software development department. The author describes the process used to plan software engineering projects. Next, he presents the actual results of projects. These are data acquired by analyzing the written records of software projects that followed Sequent's life cycle development model from 1989 until the present. He uses this data for a statistical analysis of the planning process, first developing a control chart to determine if the planning process is in a state of statistical control, then selecting a set of predictor variables and using linear regression analysis to test the null hypothesis that these variables are of no use in predicting variation in the schedule produced by the planning process. The software development planning process has many outputs including: schedule or predicted completion date, product content and functionality, detailed specification of who does what and when, and sales and return on investment analysis.