Clustering student errors

This paper gives an in-depth description of the research design in the pursuit of clustering of students’ performance when solving different types of linear equations. The student performances are clustered using data from 457,185 answers to equation tasks, made by 37,585 students, distributed across 3,438 unique linear equations in a digital learning environment. The tasks consist of different categories of linear equations. The clustering analysis contributes to the development of an online tool to provide the teachers with easy accessible formative assessment. At this point, the attempt to cluster the students’ performance have not yet been successful, meaning that no clusters are found. Instead, a description of how the pursuit of these clusters will continue is presented alongside the research design.


Citation style:
Could not load citation form.