Machine Learning to Cluster Medical Student Data in a Flipped-Learning Course

Flipped learning is an educational approach where students learn new material outside of class using online resources and class time is used for activities such as problem-solving and discussion. Its aim is to make the most effective use of class time by giving educators the opportunity to work with students on higher-order thinking skills. However, it is crucial for students to study before attending class. In a large group, how can we easily differentiate students who studied and who did not before the class?

A group of researchers from Turkey used machine learning techniques to profile students in a flipped classroom considering their data on a learning management system.

The flamingo is ready to present the article.

First of all, it would be better if I provide some information about the clustering technique in order to make sure that each one of you is on the same page.

Clustering is a type of unsupervised learning that groups together similar data points and separates them from dissimilar ones. It is often used to classify entities based on their characteristics and is commonly applied in education to create profiles for students based on their online interactions. It can discover patterns in large datasets without the need for human guidance.

By using clustering, the study investigated medical student profiles in a flipped classroom based on their learning management system interactions.

In the flipped classroom, the small-group practical activities under the guidance of faculty members were arranged to teach 11 clinical skills such as measuring blood pressure and external cardiac massage. Before these practices, students accessed practice checklists and videos through Moodle learning management system. The content was shared weekly before attending each clinical skill laboratory.

The researchers exported the learning management system interaction data generated by three hundred seventy-five students. By analyzing the data using a machine learning technique called clustering, they found a two-cluster structure. It meant that there were two groups of students: Low interaction group and high interaction group.

So what? What is the practical value of this information?

It can be utilized to pinpoint students who are not actively participating in pre-class interactions and provide them with specific feedback that addresses their individual needs. Identifying low-engaged students early on can also allow for interventions to be implemented before their lack of engagement becomes a more significant problem.

If you want to read more, you can find the link to the article at the description below the video: https://youtu.be/pQA1dcSpBf8

See you and adios para amigos.

And also, don’t forget the flamingo.

Follow the Flamingo on Twitter: https://twitter.com/MedEdFlamingo