A combined multitasking performance measure involving sequential and parallel task executions

Abstract

Research on human multitasking suggests several measures to evaluate performance. However, the suggested measures evaluate performance either when tasks are performed sequentially, or when tasks are performed in a parallel manner. There is a lack of models with performance measures that consider concurrently sequential and parallel task execution. This paper aims to develop a measure of human performance that considers both sequential and parallel execution of tasks in multitasking conditions. First, the literature is reviewed to select a taxonomy to model the features and execution of tasks during multitasking. Task features include a list of tasks, task demands (in terms of physical, psychological, and emotional loads), and coordination between tasks (in terms of priorities, similarities, dependence, and time constraints). Task execution is represented as a network of sequential and overlapping tasks. Second, a set of measures are identified to evaluate human performance in multitasking conditions. The analysis of literature suggests a task switching cost model for sequential task execution and a task interference ratio when tasks are executed in parallel. To enable combining switching cost (i.e., sequential execution) and interference ratio (i.e., parallel execution) in multitasking conditions, a classification scheme based on tasks’ modalities is utilized. Finally, the developed model is applied to different scenarios.