If you put a group of people in an escape room and give them an hour to solve the riddles and win the game, you may notice, with the passage of time, their progress gradually slowing and their frustration growing until they eventually give in to the body’s signals to throw in the towel.
Thanks to the research of Kristina Lerman, a project leader at the USC Information Sciences Institute and research associate professor in the USC Viterbi School of Engineering’s Department of Computer Science, we now have a clearer understanding about the culprit that may be behind this downward progression when it comes to virtually any test— glucose.
Glucose, for those of us who were keeping count, is the most abundant biological molecule produced on Earth, to the extent of about 50 billion tons a year. All the cells and organs in our body feed on it. But glucose is also a finite cognitive resource. How much glucose we produce and store will determine how well we do on tasks that require mental focus and accuracy.
Lerman uses Big Data, especially from social media sites, to discover how people use and share information. She studies how cognition and perception shapes watching a video, playing an online game, or even reading a tweet — anything that requires mental effort. For example, she studies the relationship between performance and practice in an online video game and the effect it has on players’ attention.
“Our limited attention,” Lerman said, “is the reason why most viral campaigns fail on social media.”
So, what is she doing dabbling in sugar, you may ask?
Lerman’s research group connected a possible link between our test-taking abilities and the amount of cognitive resources of the brain, potentially glucose, which she concluded, will have a direct effect on our test-taking abilities.
In the study, published in the Journal of Computational Social Science in September, 2018, Lerman devised a behavioral model gauging the length of time it takes for an individual to complete a test. She then compared her model to existing behavioral models to measure and predict the levels of cognitive resources an individual had during the test. She looked at the data provided by an online testing website, grockit.com, and analyzed 2.8 million attempts by 180,000 users to answer 6,000 questions. This would eliminate any outliers within the data.
The study looked at the length of time it took each individual to complete the test. She made predictions of the amount of cognitive resources each individual started off with, and by looking at the amount of questions they got wrong, she predicted it was the result of cognitive depletion over the length of time.
Focusing on the dynamics of test performance, her research further revealed that while the performance declined over the course of a test-taking session, it recovered following prolonged breaks between sessions. Basically, the more breaks you take between tests, and the more chocolate you can sneak in, the better your performance.
“This singular model produced evidence that performance on standardized tests is tied to levels of cognitive resources, which are finite, much like research on ego depletion, linked loss of willpower and self-control to diminished levels of glucose in the body,” Lerman said. She showed that a model with two interacting finite resoruces, potentially describing how lactate is metabolized in the brain, better explained the data.
She compared the time it took the testers to read a question to the time it took them to answer each question, while assessing users’ ability to learn from these answers and see if they got it right the next time around. This lead to emerging evidence that suggested that glucose may not be the primary energy source for neurons engaged in intensive activity; instead, lactate metabolism (the production of glucose) may be more important for this function.