Could AI uncover the secrets of a student’s success or failure?

“If this practice was rolled out more widely, the information would alert universities to who is at risk”

AI could soon be helping universities spot students who are struggling early so they can better support them and prevent them from dropping out, says Fred Singer, CEO of platform Echo360.

Imagine you’re a lecturer in the early days of teaching a large group of first and second-year students. You’re still getting to know your audience, but there aren’t many questions coming from the room, making it hard to confirm students’ understanding of the topics being covered.

You’re confident about the quality of your lecture, but questions linger.

Are some students struggling silently due to a limited grasp of English? Does the lack of questions indicate confusion rather than comprehension? Are students opting out of group discussions because they’re shy or because they don’t understand the concepts being debated?

The early clues a student is at risk of failing are not easily identifiable, even to the most experienced teachers. Before you know it, a student can fall behind and be at risk of failing the course or dropping out altogether.

But according to a new piece of research, universities could soon use AI to monitor early behaviour and predict students’ success or failure.

The study, conducted at the University of Michigan in the United States, showed that behaviours within the first 15 days of starting a course identified with 72% accuracy which students would go on to secure good grades and which would not.

If this practice was rolled out more widely, the information would alert universities to who is at risk and allow them to intervene earlier, preventing students from falling too far behind.

“Before you know it, a student can be at risk of failing the course or dropping out altogether”

Understanding the obstacles

Conducted across seven semesters between 2015 and 2018, the University of Michigan study looked at data from a number of the university’s systems to ascertain the learning patterns and grade point average of a cohort of 1,283 students enrolled in entry-level STEM courses.

It revealed that behavioural signals in the first couple of weeks, such as the number of correct responses to in-class questions, the number of lecture slides students flagged as confusing, and the number of times students viewed slides and video recordings on the university’s learning platform could be used to make accurate predictors of a student’s overall success in a course.

So, how could this AI-focused approach work in the classroom?

Imagine Maryam, a first-year economics student from the Middle East, who has been fairly active in class from day one. She shares her views with classmates and seems to be enjoying her lectures. On the face of it, Maryam is unlikely to give her lecturers any cause for concern.

However, AI could be used to analyse Maryam’s behaviour in the first few days and weeks of her course to reveal that she rarely responds to group questions on specific topics or she’s watched the same piece of video from her last lecture over and over again.

“Being able to flag the issue early could make a real difference”

This might indicate that Maryam is already struggling. Her lecturer, who is focused on explaining the complexities of a socio-economic theory to a class of over a hundred students, may not be in any position to spot the signs.

So, being able to flag the issue early could make a real difference to Maryam’s progress and with a 72% accuracy rate, a quick catch up tutorial might be all that’s needed to get her back on track.

Personalised learning in action

Most universities today use information such as assessment grades to monitor students’ achievement at regular points. Grades can provide an accurate indication of whether a student has understood the topics covered and help lecturers to spot students who could benefit from some study tips or extra support.

What the research shows is the potential for information on learning behaviours to be used – with the relevant permissions – to support students more proactively and effectively.

Giving instructors insight like this makes it easier for them to intervene early and provide the most appropriate help to those who need it, in line with their background, previous achievement and current circumstances.

An AI-enhanced approach to assessment might help students like Maryam – and others who are at risk – to better understand what steps they need to take to study more effectively for success.

Read the full research paper here.

About the author: Fred Singer is CEO of Echo360,  a video platform for education.