Tasks Beyond Taxonomies: A Multidimensional Design Space for Team Tasks

Abstract

Many experimental studies of team performance involve teams working on one specific task or at most a handful of tasks; thus, it is natural to ask how the results of any such study generalize to tasks other than the one(s) studied. Unfortunately, this simple question is generally impossible to answer due to the absence of commensurability among tasks: although many taxonomies for classifying tasks have been proposed, it is not possible to say comprehensively, consistently, and quantitatively how similar or different any task is to any other. We therefore cannot in general say when we should expect to find different or similar results for teams working on different tasks. Here, we address the commensurability problem by introducing a multidimensional framework that synthesizes existing taxonomies and defines a between-task distance measure in terms of 24 taxonomic dimensions (e.g., “degree of creativity required,” “amount of physical effort required,” “extent of demonstrable correctness”). We then use these dimensions to label 102 tasks sourced from an interdisciplinary social science literature. Finally, we use this repository of tasks to demonstrate some of the benefits of our multidimensional approach: by shedding light on the key underlying dimensions for every team task, our paradigm makes it easier to integrate ostensibly “contradictory” findings, identify contingencies for theories, and efficiently sample tasks for future research experiments.