Exploring a Text-Mining Approach as Rapid Prototyping Tool for Formative Assessments in Inquiry-Based Online Learning

Abstract

We make a preliminary case for a computational method intended to facilitate real-time formative assessment in online inquiry-based learning environments. With a focus on talk and text as disciplinary, we aim to address how learning analytics, in this case, text mining, can provide learners and instructors with meaningful information in rapid and real-time to support learning and engagement. Our results show that in measuring the distance between the expert and learners’ discourse from forum posts and verbal discussions, resulting similarity values can offer stakeholders evidence of student learning trajectories. Moreover, similarity values provide teachers with an automated measure of students’ progress toward disciplinary discourse, and also reveal critical moments during the collaborative activity where more alignment with disciplinary ways of talk are being enacted.