Comparing qualitative and quantitative text analysis methods in combination with document-based social network analysis to understand policy networks

Abstract The literature that reflects on the application of Social Network Analysis (SNA) in combination with other methods is flourishing. However, there is a dearth of studies that compare qualitative and quantitative methods to complement structural SNA. This article addresses this gap by systematically discussing the advantages and disadvantages relating to the use of qualitative text analysis and interviewing as well as quantitative text mining and Natural Language Processing (NLP) techniques such as word frequency analysis, cluster analysis, topic modeling, and topic classification to understand policy networks. This method-oriented comparative study features two empirical studies that respectively examine the Employment Thematic Network, established under the aegis of the European Commission, and the intergovernmental cooperation network set up within the Bologna Process. The article compares and discusses the underlying research processes in terms of time, human resources, research resources, unobtrusiveness, and effectiveness toward the goal of telling meaningful stories about the examined networks in light of specific guiding hypotheses. In doing so, the paper nurtures the debate on mixed-methods research on social networks amidst the well-known paradigm war between qualitative and quantitative methods in network analysis.