Can the outside-view approach improve planning decisions in software development projects?

This study empirically tackles the question of whether taking an outside-view approach, recommended for reducing the irrational behaviours associated with the planning fallacy, can also reduce the time underestimation, scope overload and over-requirement problems plaguing planning decisions in software development. Drawing on descriptive behavioural decision theory, this study examines whether the planning fallacy, a cognitive bias referring to the tendency of people to underestimate costs and overestimate benefits in evaluating a task to be performed, can provide a theoretical platform for mitigating irrational behaviours in the planning of software development projects. In particular, we argue that taking an outside-view approach in planning decisions for software development may have the same mitigating effects on time underestimation, scope overload and over-requirement it has been shown to have on cost underestimation and benefit overestimation. In an experiment investigating this argument, participants were randomly assigned to four groups by manipulating two outside-view mechanisms: reference information about past completion times (present/absent) and role perspective (developer/consultant). After being presented with a to-be-developed software project, they were requested to estimate development times of various software features and to recommend which features to include within project scope given a fixed duration for the entire project. The results confirm that the three problems of time underestimation, scope overload and over-requirement are manifested in planning decisions for fixed-schedule software development projects. Moreover, the results show that these problems are mitigated, yet not eliminated, by presenting reference information about past completion times and by having a consultant role.