Software Project Control: An Experimental Investigation of Judgment with Fallible Information

Software project management is becoming an increasingly critical task in many organizations. While the macro-level aspects of project planning and control have been addressed extensively, there is a serious lack of research on the micro-empirical analysis of individual decision making behavior. The heuristics deployed to cope with the problems of poor estimation and poor visibility that hamper software project planning and control are investigated, and the implications for software project management are examined. A laboratory experiment in which subjects managed a simulated software development project is reviewed. The subjects were given project status information at different stages of the lifecycle and had to assess software productivity in order to dynamically readjust project plans. A conservative anchoring and adjustment heuristic is shown to explain the subjects' decisions quite well. Implications for software project planning and control are presented. >