Re-examining cognitive load measures in real-world learning: Evidence from both subjective and neurophysiological data.

Abstract

Background: Cognitive load theory is widely used in educational research and instructional design, which relies heavily on conceptual constructs and measurement instruments of cognitive load. Due to its implicit nature, cognitive load is usually measured by other related instruments, such as commonly-used self-report scales of mental effort or task difficulty. However, these concepts are different in nature, as they emphasize distinct perspectives on cognitive processing. In addition, real-world learning is more complex than simplified experimental conditions. Simply assuming that these variables will change in a monotonic way with workload may be misleading.

Aims: This study aims to examine whether these measures are consistent with each other, and to discover the neurophysiological basis underlying the potential discrepancy.

Sample: The study collected data in both a real-world (Study 1, 22 high school students in 13 math classes) and a laboratory setting (Study 2, 30 students in 6 lab-based math tasks).

Methods: In addition to self-report measures, the study also collected multimodal neurophysiological data, such as electroencephalography (EEG), electrodermal activity (EDA), and photoplethysmography (PPG).

Results: The results show that although the difficulty level can be perceived with difficulty ratings, it does not lead to the corresponding level of mental effort. Only within an appropriate level of load, can we observe a positive correlation between self-report difficulty and mental effort. Neurophysiological evidence also supports the conceptual discrepancies and group differences, indicating distinct neurophysiological mechanisms underlying these 'similar' constructs.

Conclusions: These findings also emphasize the need for combining these concepts to better evaluate students' cognitive load.