The robots are coming for your students' feedback

Abstract

The Data Analytics team at The University of Waikato gathers student feedback (as rich qualitative data) but manual analysis of these comments poses a time challenge for reporting. To address this, we explored the possibility of condensing qualitative information by leveraging natural language processing (NLP) technology, specifcally Google’s NLP sentiment analysis. We employed a robust coding framework to test the validity of NLP-coded student feedback, analysing 1000 comments from the University’s 2021 course evaluations. Results show a statistical correlation between our sentiment analysis and NLP, ofering promising evidence for NLP’s efcacy in providing accurate, high-level insights into student feedback sentiment.