Uncovering themes in personalized learning: Using natural language processing to analyze school interviews

Abstract

Using a natural language processing tool, this study examined participant discourse in personalized learning schools to better understand what personalized learning looks like in practice. Term frequency-inverse document frequency (tf-idf) was used to identify the significant words and potential emergent themes for 134 interview transcripts. This tool provided a way to swiftly explore the structure of the data, revealing distinctions in the vocabulary students and teachers use as well as a potentially meaningful set of themes. This method provided a valuable lens with which to validate or surface new areas for investigation. By applying this tool to interviews from personalized learning environments, we were able to identify ways educators and students talk differently about project-based learning environments, revealing that tools like tf-idf can be effectively used to quickly provide a preliminary look at large amount of interview data.