Trends in Nursing Research Using Text Mining in Qualitative Data: Focus on Analysis Methods

Abstract

This paper (1) presents an overview of Text Mining (TM), (2) describes trends in nursing research (academic papers) using TM, and (3) discusses typical methods of analysis in nursing research using TM. A subject search was performed with Igaku Chou Zasshi (Ichushi Web) using the keyword “text mining,” finding 57 academic papers in the nursing literature for analysis. Many academic papers using TM tended to be on “nursing education” and were by university researchers. Questionnaires (free description) were the most common data source, but the samples were small. TM is good for big-data analysis, but the possibility of analyzing a large amount of data was low. The most widely used TM software was KH Coder. The main methods of analysis using TM were reference frequency analysis, relationships between words, and multivariate analysis (especially principal component analysis and cluster analysis), which are core TM analytical methods. In particular, multivariate analysis frequently served as a method for extracting categories. Research on “nursing education” and “nursing management” was considered highly compatible with TM.