The sensitivity of physiological measures to changes in mental workload.

Mental Workload (MWL) quantifies the subjectively experienced mental strain imposed by any task and poses interesting avenues in machine or system optimization. Thus, the physiological measurement of MWL has been of large interest, but despite multiple decades of research, sensitivity to changes in MWL remains inconclusive (Ayres et al., 2021; Charles & Nixon, 2019; Tao et al., 2019). This study therefore re-evaluates sensitivity of six physiological measures to changes in MWL, namely in their ability to distinguish levels of MWL. With increasing MWL, heart rate (HR), blink rate (BR), pupil dilation (PD), skin conductance responses (SCR), and skin conductance level (SCL) were expected to increase, whereas heart rate variability (HRV) was expected to decrease. For MWL induction, an adaptation of the Swaps task (Stankov & Crawford, 1993) was employed, with the aim of inducing four distinct levels of demand. MWL manipulation was successful, as in the sample of 74 university students, response times significantly increased with each level, while accuracy decreased. Subjective ratings of MWL in the NASA-Task Load Index (Hart & Staveland, 1988) reflected this pattern. BR and PD were highly sensitive, increasing with each level, while HR and HRV were sensitive but less able to differentiate fine-grained changes, with the inability being more pronounced for HRV. SCR were sufficiently sensitive, except for very high levels, and SCL showed no sensitivity. These findings provide further basis for research and application of physiological responses to MWL, to represent task demands more comprehensively, as well as mental strain in machine-operator relations.