It’s been shown that our minds wander 50% of the time – even during important tasks. This study asked, how does it affect our learning? During a live lecture, an EEG was used to capture mind wandering among 16 participants and examined whether it had a negative affect on their learning. Wait… did Jon just say mind watering? His brain must have been somewhere else…
Listen in to see if you can catch it … and of course, to hear the results of the study.
KeyLIME Session 250
Reference
Dhindsa et. al., 2019. Individualized pattern recognition for detecting mind wandering from EEG during live lectures. PLoS One.14(9):e0222276
Reviewer
Jon Sherbino (@sherbino)
Background
Ok KeyLIMErs, remember our tag line. “We read the journals so you don’t have to.” I have received some – quite honestly – hurtful emails and text from the co-hosts about my choice of article. But in the same way a grounding in embryology and organic chemistry makes you a good physician (wait, this just in… old school curricula does not lead to better health care professionals)… Anyways, my point is that this article is a “basic science” article for health professions educators.
If you are a part of the digital movement that documents your heart rate, your sleep quality, your KoM Strava rides, your urine ketones… I could go on, but I fear you’re getting too much information about Jason’s personal life. Nonetheless, wearable technologies are giving us richer insight into a personal well-being. Perhaps, wearables could give us better insight into our teaching effectiveness. If you own a smart watch or if you’ve mortgaged your house buying Garmin gear, this podcast is for you.
Here’s the set up. Our mind wanders up to 50% of the time… even when we’re doing something important. If it wanders during a lecture, studies suggest a negative correlation with learning. However, how can a teacher detect a student (or a class) that is no longer paying attention. Sure, the person in the back row sleeping behind a pair of sunglasses is an easy tell. But until now much of the studies to predict inattention have used surrogate markers with poor ROC curves. Enter the EEG.
Purpose
In this study, we demonstrate, for what we believe to be the first time, machine learning-based detection of [mind wandering] from EEG recorded simultaneously across the entire study sample in a naturalistic educational setting: during live lectures.
Key Points on the Methods
The study took place in a 106 -seat performance space that allows simultaneous recording of 16 audience members via a 16-channel EEG.
Enrollment was by email to all orthopedic residents and medical students on elective to a single university-based residency program.
Two 30 minute lectures were given with a 15 minute break between them. Every 4 minutes a bell rang and participants were prompted to self-report their attention immediately prior to the bell. Participants not undergoing EEG monitors also reported their attention to control for the addition of the EEG.
Immediately after each lecture a recall quiz was administered. Two weeks after each lecture a retention test was administered.
The EEG data was optimized to address artifacts and EEG channels at extreme variance.
A non-linear support vector machine (i.e. machine learning analysis) was used to build a classification model of mind wandering, including three common mind wandering spatial patterns and three common attentive spatial patterns.
Key Outcomes
15 participants provided data EEG and behavioural data, 8 provided only behavioural data.
Self-reported mind wandering occurred during 35% of the probes. There was no influence of EEG on report or at time during the lecture.
There was no association between mind wandering and immediate or delayed tests.
Surprisingly (and conflicting with fMRI data) all channels showed mild to moderate effects of mind wandering.
Mind wandering was associated with specific EEG frequencies. Mind wandering could be predicted for an individual approximately 80% of the time, but the frequencies and spatial patterns were not common across the study population.
Key Conclusions
The authors conclude…
“We were able to accurately detect MW from EEG at the individual level using data-driven feature learning and machine learning. These methods allowed us to show that the neural correlates of [mind wandering] might be more variable than suggested by traditional statistical methods. With further study, these findings may lead to the development of new methods for online [mind wandering] detection.”
Spare Keys – other take home points for clinician educators
Other disciplines, especially psychology, have a much more interesting inclusion of supporting literature in manuscripts. In HPE with some exceptions (AHSE, for example has no word limit) articles are restricted to a 2500 to 3000 word count. This means that the supporting literature is limited and the current study is often superficially situated within an unspoken complex literature base. In psychology, there is no such thing as a three paragraph introduction. Rather, the introduction is often akin to a textbook chapter allowing a reader to get up to speed on an area without assuming prior expertise.
Access KeyLIME podcast archives here
The views and opinions expressed in this post and podcast episode are those of the host(s) and do not necessarily reflect the official policy or position of The Royal College of Physicians and Surgeons of Canada. For more details on our site disclaimers, please see our ‘About’ page
