Cognitive Limits of Software Cost Estimation

This paper explores the cognitive limits of estimation in the context of software cost estimation. Two heuristics, representativeness and anchoring, motivate two experiments involving psychology students, engineering students, and engineering practitioners. The first experiment, designed to determine if there is a difference in estimating ability in everyday quantities, demonstrates that the three populations estimate with relatively equal accuracy. The results shed light on the distribution of estimates and the process of subjective judgment. The second experiment, designed to explore abilities for estimating the cost of software-intensive systems given incomplete information, shows that predictions by engineering students and practitioners are within 3-12% of each other. The value of this work is in helping better understand how software engineers make decisions based on limited information. The manifestation of the two heuristics is discussed together with the implications for the development of software cost estimation models in light of the findings from the two experiments.