With AI Studying Students' Brains, Teachers Can Improve Their Lessons

Everyone has a different way of learning, some learn better with pictures and videos, some learn better when reading, listening to lectures and others learn better when they practice. Some people will quickly understand the concept faster than others. Therefore, the task of standing in front of the class to deliver lessons and trying to make students them is a very difficult job.

Teachers do not always know if their teaching methods and strategies are effective. Tests and questions for students can be a way to figure that out, but if their results are not good, we cannot know who is to blame. Are the lectures not difficult enough for receptive students or are students too lazy to pay attention to the lecturer?

Researchers from Dartmouth College have developed a machine learning algorithm that measures activity in the human brain. From there, it can determine the extent to which someone understands a certain concept.

The study began with a standard test and picture-related questions for freshmen and intermediate technical students while they sat in the fMRI scanner. The algorithm then generated "neural score", giving predictions on a student's performance.

However, this system still has certain limitations. Typically, we cannot scan each student's brain in the middle of each class. Currently, Dartmouth researchers focus on determining whether the student's brain works the same way when studying STEM and studying literature. The neural scores for this system are still based on a limited amount of knowledge.

However, if this system is further developed, it will be extremely useful for teachers and educators. They can try to adjust the lessons and improve them to make sure the students get the best concepts. It seems to bring potential about using technology to reform our education system.