Cognitive workload across subjective and electrodermal measures in relation to previous-night sleep
Abstract
Cognitive workload is commonly assessed using both subjective and physiological indicators, although these measures may reflect different aspects of task-related strain. Sleep duration may further modify perceived task demand without producing proportional changes in physiological activation. This study examined the association between perceived cognitive workload and electrodermal task reactivity, and explored whether previous-night short sleep was associated with higher subjective workload without increased electrodermal activation. A total of 52 adults were initially assessed; two were excluded due to violation of predefined eligibility criteria, resulting in a final analytical sample of 50 participants. Participants reported previous-night sleep duration and subjective sleep quality, completed a 5-min resting baseline, and then performed the Determination Test from the Vienna Test System. Electrodermal activity was recorded continuously during baseline and task performance, while perceived workload was assessed after the task using the NASA Task Load Index. Task-related electrodermal reactivity was calculated as GSR task–baseline difference, and participants were classified into short-sleep (< 7 h) and normal-sleep (≥7 h) groups. NASA-TLX total workload was not significantly associated with GSR task–baseline difference (rho = 0.126, p = 0.383). In the uncorrected threshold-based comparison, participants with short sleep reported higher perceived workload than those with normal sleep (64.44 ± 8.31 vs. 58.70 ± 9.57; p = 0.027), whereas GSR task–baseline difference was nearly identical between groups (4.57 ± 2.19 vs. 4.61 ± 2.93 μS; p = 0.884). However, the workload difference did not survive FDR correction across exploratory subgroup comparisons and should therefore be interpreted as a nominal finding. Overall, perceived workload and electrodermal reactivity showed limited correspondence, whereas the sleep-related analyses did not provide consistent evidence of an association with either outcome.