Towards measuring cognitive load through multimodal physiological data

Cognitive load plays an important role during learning and working, as it has been linked to well-functioning cognitive processes, performance, burnout and depression. Nonetheless, attempts to assess cognitive load in real-time by means of physiological data have been proven difficult, and interpreting these data remains challenging. The aim of this study is to examine whether and how well experienced cognitive load can be measured through psycho-physiological data. The approach of this study is rather unique, for a combination of reasons. First, this study takes a multimodal approach, monitoring EDA (electrodermal activity), EEG (electroencephalography) and EOG (electrooculography). Second, this study is based on a relatively intensive data collection (N = 46) in a controlled lab setting in which varying cognitive load levels are deliberately induced. Finally, not only focussing on statistical significance but also on the size of the association gives insights into how suitable physiological markers are to measure cognitive load. Results from a multilevel analysis suggest that the following physiological markers might be related to cognitive load, for example, in an industrial context: the rate and the duration of skin conductance responses, the alpha power, the alpha peak frequency and the eye blink rate. About 22.8% of the variance in self-reported cognitive load can be explained using these five measures.Cognitive load plays an important role during learning and working, as it has been linked to well-functioning cognitive processes, performance, burnout and depression. Nonetheless, attempts to assess cognitive load in real-time by means of physiological data have been proven difficult, and interpreting these data remains challenging. The aim of this study is to examine whether and how well experienced cognitive load can be measured through psycho-physiological data. The approach of this study is rather unique, for a combination of reasons. First, this study takes a multimodal approach, monitoring EDA (electrodermal activity), EEG (electroencephalography) and EOG (electrooculography). Second, this study is based on a relatively intensive data collection (N = 46) in a controlled lab setting in which varying cognitive load levels are deliberately induced. Finally, not only focussing on statistical significance but also on the size of the association gives insights into how suitable physiological markers are to measure cognitive load. Results from a multilevel analysis suggest that the following physiological markers might be related to cognitive load, for example, in an industrial context: the rate and the duration of skin conductance responses, the alpha power, the alpha peak frequency and the eye blink rate. About 22.8% of the variance in self-reported cognitive load can be explained using these five measures.A