Students' patterns of interaction with a mathematics intelligent tutor: Learning analytics application

TL;DR

This study analyzes how students interact with an intelligent math tutor and identifies key predictors of academic success, highlighting the importance of prior knowledge and topic selection strategies.

Contribution

It introduces a predictive model using data logs and surveys to identify factors influencing student success in mathematics tutoring.

Findings

Derived attribute (topics mastered/practiced) predicts final grades.

Sequential topic selection improves retention of mastery.

Prior knowledge and interaction patterns explain 42% of grade variance.

Abstract

The purpose of this paper is to determine potential identifiers of students' academic success in foundation mathematics course from the data logs of an intelligent tutor. A cross-sectional study design was used. A sample of 58 records was extracted from the data-logs of an intelligent tutor, Assessment for Learning using Knowledge Spaces (ALEKS). This data was triangulated with the data collected from surveys. Two-step clustering, regression analysis, paired-sample t-tests were applied to address the research questions. The data-logs of ALEKS include information about number of topics practiced and number of topics mastered. A derived attribute, which is the ratio of number of topics mastered to number of topics practiced is found to be a predictor of final marks. Students' prior knowledge and this derived attribute are predictors of final grades with R square=42%. Students were asked to report their preferred way of selecting topics as either sequential or random. Results of paired sample t-test demonstrated that the students who selected topics in sequential manner were able to retain their mastery of learning after the summative assessment whereas the students who chose topics randomly were not able to retain their mastery of learning.