Application of text analytics in examining students’ qualitative feedback in relation to teaching and learning.

Abstract

Background: With the advancement of machine learning technology and data science in the 21st century, alternative methods of analyzing natural language data have emerged. New text-mining techniques developed in the past decades have enabled automatic analysis of large amounts of text data. A promising direction is the use of co-occurrence information to extract semantic and thematic patterns from qualitative data. The current study aims to demonstrate the application of text analytics in students’ qualitative feedback pertaining to teaching and learning. Method: Undergraduate students at the University of Hong Kong were asked to provide comments on the best aspects and areas of improvement of their learning experience in an annual institution-wide survey. Key themes and concepts from students’ comments were extracted using an automatic text analytic tool LeximancerTM. Results: In the most prominent themes identified, keywords including “students”, “courses”, and “learning” emerged. Key findings from the text analysis were, to a great extent, in accordance with quantitative results from the same survey. Items that received the highest rating from students, like exchange experience for final-year students, were also emerged as one of the key concepts in student comments. Key concepts varied among student cohorts from different study years. Discussion: Using text analytics to extract key concepts and themes from text data, the present study demonstrates an efficient, evidence-based approach of analyzing students’ feedback for informing teaching and learning practice. This approach appears to be promising for educators to turn data into insights for quality enhancement in higher education.