An Investigation of the Application of Machine Learning as a Tool and a Method for Mathematics and Statistics Instruction Within the Context of ODeL

Abstract

The application of machine learning (ML) in mathematics education has increased in recent years, but not without its own challenges. This study investigates how ML can be used as a tool and method in Mathematics and Statistics instruction within the context of Open and Distance e-Learning (ODeL) and unravels the perceived challenges and prospects. Data were collected through email questionnaires with closed and open-ended statements. These questionnaires were delivered to 50 participants selected through random and purpose sampling to include those with experiences of both conventional and ODeL setups. The participants were qualified mathematics teachers and lecturers from secondary schools, teachers’ colleges and universities in Gweru, Zimbabwe. They rated given statements on a Likert scale and stated their own perceived and observed challenges and prospects of applying ML in instruction. Data were analysed using descriptive and inferential statistics and qualitatively through content analysis and by categorizing and interpreting emerging themes. Thus, a mixed method design - using multiple sources, instruments and analysis methods  - was used. Results indicated that most respondents were concordant on the Likert statements and that ML technologies have advantages such as personalised instruction, interactive assessments, and on time feedback, among others. Challenges included lack of creativity by students, lack of the human, moral and sympathetic aspect, data collection and algorithmic biases, among others. It was recommended that ML for mathematical sciences instruction could not be wholly done away with. This study could help in the evaluation and implementation of ML policies in ODeL institutions.