A worker-based approach for modeling variability in task completion times

Three sources of variability in task completion times are identified: the task itself, the worker performing the task, and the environment where the task is performed. Although all three sources might play a role, for practical purposes researchers seek a parsimonious model of variability in task completion times which identifies the most significant source. It is typically assumed that the most significant sources are the task itself or the environment where the task is performed. In this paper we investigate the notion that the worker performing the task may be the most significant source of variability in task completion times, and propose a modeling approach for this situation. We also present the results of a field experiment that support the proposed modeling approach.