Mind-wandering is well established as being associated with negative outcomes across contexts
, including productivity and health and safety. However, quantitative investigation of mind-wandering is limited by a pivotal constraint: the inability to detect its onset. Researchers continue to rely on self-report and apply arbitrary timeframes to categorise neurophysiological data as containing mind-wandering or on-task activity, assuming that these timeframes are dominated by a given state. This
proposal seeks to overcome current limitations to mind-wandering research by applying novel methods to identify mind-wandering onset. Study 1 will involve experienced meditators completing an instructed breath-focus/mind-wander task. Their EEG data will be submitted to a machine learning paradigm to identify neural activity most likely to represent mind-wandering and breath-focus. The identified activity will constitute baselines for the subsequent studies. Study 2 will integrate a novel approach building on closed-loop neurofeedback, which is a tool used to train individuals to control their neural activity. However, instead of training individuals, we will use neurofeedback to enable individuals to teach the computer about the link between their neural activity and their mental experience. Specifically, when participants’ neural activity reaches a certain threshold, a question will appear on screen, and their response will inform adjustments to the threshold for the next question.
For a more robust picture, Study 3 will triangulate using pupillometry. Successful identification of mindwandering onset could have far-reaching implications, beyond advancing methodological rigor in research, towards improving diagnostic tools and gaining a cleaner, more precise picture of the
temporal flow of individuals’ attentional states.