In a context where human–machine interaction is growing, understanding the limits between automated and human-based methods may leverage qualitative research. This paper aims to compare human and machine analyses, highlighting the challenges and opportunities of both approaches.
This study applied qualitative secondary analysis (QSA) with machine learning-based text mining on qualitative data from 25 interviews previously analyzed with traditional qualitative content analysis.
By analyzing both techniques' strengths and weaknesses, this study complements the results from the original research work. The previous human model failed to point to a particular aspect of the case, while the machine analysis did not recognize the sequence of time in the interviewee's discourse.
This study demonstrates that combining content analysis with text mining techniques improves the quality of the research output. Researchers may, therefore, better handle biases from humans and machines in traditional qualitative and quantitative research.
