Using machine learning to identify learners at risk, and develop targeted interventions

Education is linked to higher salaries, increased job satisfaction, and better health outcomes. It prepares learners to tackle complex societal problems and can address regional skills’ gaps. Thus, being able to identify leaners at risk of not progressing on their studies, or even dropping out of their courses, is of critical importance for the learners themselves, as well as for society more generally. 

For the past decade, I have been supporting a fantastic entrepreneur who, frustrated with the problems that he, himself, experienced at school, set out to develop a range of interventions to identify those learners at risk, and develop targeted interventions. Tej Samani developed a tool that uses machine learning to assess learners along the following dimensions:

Metacognitive domainsPerceptions
Sleep & Tiredness
Retaining & Recalling Information
Test Anxiety
Focus & Motivation
Organisation & Time Management
Stress & Strain
Confidence
Towards Effort & Determination to Work
Towards Learning
Readiness to Learn
Towards Subject Demands
Towards the Institution
Towards Self
Towards Showing Up to Learn
Towards Tutors/ Teachers/ Lecturers

The “metacognitive domains” items are hygiene factors which can prevent or enable academic success. In order to scaffold the learners’ academic success, these barriers must be identified and removed. However, removing these barriers will not be enough to guarantee academic success. For that end, it is also necessary to improve learners’ perceptions. The “perceptions” items are motivating factors, which drive academic success. Once the scaffolding has been established by removing the metacognition barriers, the negative perceptions must be identified and addressed, in order to build the walks of the learners’ academic success. The combination of these two types of factors impacts on learning, as depicted in Figure 1.

Figure 1. Impact of metacognition and perceptions on learning

Last week, Tej was at the IHIET-AI 2021 conference, where he presented our paper, outlining the tool, and providing evidence of how it has been used by a further education college in England, to identify learners at risk and to develop targeted interventions which addressed the needs of different groups of learners. For instance, with some learners there was a need to focus on self-confidence, whereas with other learners attention had to be directed to attitudes towards self and readiness to learn. Based on these interventions, the college reported lower program withdrawals, improved learner outcomes, increased overall satisfaction, and substantial savings.

I am really proud to be involved with such an initiative, which helps individuals, educational institutions, and society. If you would like to know more about this work, check Performance Learning’s website

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s