Exploring The Task Space: A Multidimensional Representation of Team Tasks

An effective soccer team does not use the same strategies as an effective quiz bowl team; a team's task powerfully shapes its interaction processes and performance outcomes. Theories about teams should therefore explicitly take task characteristics into account, providing "boundaries" for where a particular theory may generalize. However, current tools make task-sensitive theories challenging to build; frameworks for describing team tasks often rely on conceptual categories that are difficult to compare and quantify. In this work, we present the Task Space, a multidimensional representation of tasks that unifies five common task frameworks. Each team task can be represented as a unique point in the space, allowing researchers to generate a similarity score for any two tasks, systematically sample tasks for experiments, and build task-sensitive models. We also present a fully rated repository of 72 tasks, and build a process for expanding this map to include new tasks and new dimensions.