Decision making heuristics and biases in software project management: an experimental investigation

Abstract

Software project development has been plagued with an infamous reputation for cost overruns, late deliveries, poor reliability and users' dissatisfaction. Much of this blame has been placed on the managerial side of software development. The Systems Dynamic Model of Software Project Management is a quantitative model of software project dynamics that is attempting to gain some valuable insight into the managerial side of developing software systems. The objective of this thesis is to use the Systems Dynamic Model's gaming interface to investigate managerial heuristic and biases in software project management. Specifically, three experiments were executed to determine the effect of anchoring on productivity estimation, the effect of poor cost estimation on staffing decisions and the effect of social loafing on a software project's staffing decisions, final cost and final duration.