The manipulation of flow and its effects on neural correlates of precision-weighted auditory prediction error
Date
2024
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Flow is a cognitive state associated with focused, yet effortless attention, a perceived quickening of time, and increased positive affect. The flow state has been observed in various settings in which one’s skill level matches task challenges, such as the workplace or video gaming. Given previous research on flow has relied on subjective measures, such as self-report, the neural mechanisms associated with flow remain unclear. In this EEG study, we investigated the effects of flow on the neural correlates of precision-weighted auditory prediction error, specifically the mismatch negativity (MMN) and P300 event-related potentials (ERPs). We recorded the amplitude of the MMN and P300 during an auditory oddball task consisting of local, global, and local-global deviants while participants simultaneously played a video game. During the task, subjective measures of flow were consistently self-reported by the participants. Based on previous Active Inference modeling (Smith et al. 2022), we hypothesized that local deviants would elicit a biphasic MMN-P300 complex, while global deviants would solely elicit a P300. We also expected P300 amplitude to be negatively correlated with self-reported feelings of flow. Such findings would provide evidence that the induction of flow results in reduced selective attention to task-irrelevant auditory stimuli. From an Active Inference perspective, reduced P300, but not MMN, amplitude during flow represents a reduction in the precision-weighting of higher-level sensory prediction error. To the contrary, lower-level, or pre-attentive sensory prediction error, as indexed by the MMN, may remain unaffected. We found that the environmental flow and non-flow blocks were ineffective in causing a significant difference in reported FSS scores. The reported FSS scores varied greatly, inferring that there was a high dependence on the skill-task difficulty component of flow. Since both the environmental blocks induced flow over the course of the whole task no specific relationships or trends could be defined between flow and the ERPs. There were however trends found with the local, global and local-global auditory paradigm conditions. The local deviant condition was consistent with our hypothesis and elicited an MMN, but not a P300. Contrary to our hypothesis the global deviant condition elicited an MMN in addition to a small P300. The local-global condition also elicited an MMN and a very small P300. In increase in MMN amplitude was observed across the three deviant conditions. The local on average had a smaller MMN amplitude, while it increased in the global condition, and was the most negative in the local-global condition. This may suggest that MMN amplitudes increase as the deviant being presented becomes more complex in detecting. Finally, the P300 was overall absent in all conditions being absent if the majority of the participants. This could have been because of the participants experiencing flow the entire task, but due to a lack of a baseline MMN and P300 for each participant that relationship is unable to be defined.
Description
Keywords
EEG, Flow state, Mismatch negativity, P300, Sensory prediction error