Why do performance and subjective workload measures dissociate?

Abstract

A set of three experiments are described that examine the sources of information processing that produce a dissociation between subjective workload measures and performance. The results support a theory of the dissociation. Subjective measures are driven more by the number of tasks currently performed. Subjective measures are also less sensitive to resource competition than performance measures are. Factors that demand more resource investment improve performance, but these factors also increase subjective ratings of workload.