Predicting task-general mind-wandering with EEG

Jin, Christina Yi; Borst, Jelmer P.; van Vugt, Marieke K. · 2019 · OpenAlex-citations

DOI: 10.3758/s13415-019-00707-1

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the challenge of tracking the dynamics of mind-wandering, a phenomenon defined as self-generated, task-unrelated thought. Traditional methods rely on experience-sampling probes, which interrupt cognitive processes and prevent continuous monitoring. To overcome this limitation, the authors aimed to develop a machine-learning classifier capable of distinguishing between on-task and mind-wandering states using electroencephalography (EEG) data. A key objective was to determine whether such a classifier could generalize across different cognitive tasks, thereby identifying task-general neural markers of mind-wandering. The researchers employed two distinct experimental paradigms: a Sustained Attention to Response Task (SART), which is easy and prone to mind-wandering, and a Visual Search Task, which relies more heavily on external stimulus processing. Thirty participants underwent EEG recording while performing these tasks. Mind-wandering labels were derived from self-report probes inserted randomly during the tasks; the six trials preceding a probe were classified as either mind-wandering or on-task based on the participant’s response. The study utilized specific EEG markers as features for classification, including single-trial event-related potentials (P1, N1, and P3) and time-frequency measures (theta and alpha power and coherence) at frontal, parietal, and occipital electrodes. A support vector machine (SVM) algorithm was trained on these features to predict mental states, with performance evaluated using leave-one-out cross-validation and across-task prediction tests. The results demonstrated that the SVM classifier could distinguish between on-task and mind-wandering states with accuracy ranging from 0.50 to 0.85, significantly above chance levels. Crucially, the classifiers exhibited task generality: models trained on data from one task successfully predicted mental states in the other task with an average accuracy of 60%. Behavioral analysis confirmed that mind-wandering was associated with reduced accuracy and slower response times, particularly in the visual search task. Among the various EEG features, alpha power (8.5–12 Hz) emerged as the most predictive marker for mind-wandering. The study also noted that classifier sensitivity was positively correlated with individual mind-wandering rates, while specificity was negatively correlated. These findings provide evidence that mind-wandering is associated with specific, detectable electrophysiological signatures, particularly reduced alpha power, which likely reflects the "perceptual decoupling" from external stimuli. The ability to classify mind-wandering states with above-chance accuracy across different tasks suggests that the neural correlates of this state are robust and not strictly dependent on specific task demands. This work supports the feasibility of using non-invasive EEG and machine learning to monitor spontaneous thought processes in real-time, offering a method to study the dynamics of attention without the disruptive effects of self-report probes.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• mind wandering