Workload is multidimensional, not unitary: What now? gerald matthews

Abstract

It is commonly assumed that workload is a unitary construct, but recent data suggest that there are multiple subjective and objective facets of workload that are only weakly intercorrelated. This article reviews the implications of treating workload as multivariate. Examples from several simulated task environments show that high subjective workload is compatible with a variety of patterns of multivariate psychophysiological response. Better understanding of the cognitive neuroscience of the different components of workload, including stress components, is required. At a practical level, neither subjective workload measures nor single physiological responses are adequate for evaluating task demands, building predictive models of human performance, and driving augmented cognition applications. Multivariate algorithms that accommodate the variability of cognitive and affective responses to demanding tasks are needed.