Algorithmic Decision-Making and Education: The Acceptance of Learning Analytics by Secondary School Students and Parents

Abstract

Algorithmic decision-making systems such as Learning Analytics (LA) are widely used in an educational setting ranging from kindergarten to university. Most research focuses on how LA is used and adopted by teachers. However, the perspective of students and parents who experience the (in)direct consequences of these systems is underexplored. This study employs an adapted UTAUT 3 model and utilizes a double survey design to investigate the acceptance of a LA system predicting school grades by secondary school students and parents. We surveyed 1013 parents and 277 students in Flanders on how they would accept such potential LA system in their (child's) school. Our findings indicate that performance expectancy was the strongest predictor of acceptance for both students and parents. Effort expectancy and social influence were significant predictors of acceptance for parents, but not for students. Algorithmic concern posed as a stronger negative predictor for students and a weaker one for parents, whereas personal innovativeness did not significantly predict acceptance for either group. Overall, our findings show that the UTAUT model can be used to investigate acceptance of those who do not adopt the technology but experience the consequences. Students and parents show both similarities and differences in what explains their acceptance of LA systems. This could be due to their different role as stakeholder in the educational context but also because of their different backgrounds altogether. Further research focusing on different stakeholders than teachers and school management is argued for.