Understanding student emotions when completing assessment: technological, teacher and student perspectives

Abstract

ABSTRACT Student emotions influence assessment task behaviour and performance but are difficult to study empirically. The study combined qualitative data from focus group interviews with 22 students and 4 teachers, with quantitative real-time learning analytics (facial expression, mouse click and keyboard strokes) to examine student emotional engagement in an online Data Science assessment task. Three patterns of engagement emerged from the interview data, namely whizz, worker and worrier. Related emotions for these were discerned in the real-time learning analytics data, informing interpretations of associations between emotional and other forms of engagement. Instead of displacing human insights, learning analytics used alongside student self-report and teacher professional insights augment our understanding of student emotions in a naturalistic school assessment setting.