Topic Modeling: Method, Theory, and Application in Management Research

Co-author(s): Dev Jennings (UAlberta), Sarah Kaplan (UofT), Tim Hannigan (UAlberta), Richard Haans (Rotterdam), Vern Glaser (UAlberta), and Milo Wang (UAlberta)

Proposal accepted at Academy of Management Annals, 2018

Download link will be added soon

Abstract: New methods can have profound impact on management scholarship (Arora et al., 2016). They allow scholars to examine new questions that were intractable using existing methods and to address old questions with new approaches. Specifically, the field’s understandings of cognition, meaning, and interpretation have been dramatically reshaped by the emergence of new computer-based natural language processing techniques such as topic modeling. These techniques have amplified the linguistic turn in management research.

The paper will start with a brief review of the methodological and theoretical foundations of topic modeling in content and classification analysis (e.g., Weber, 1990), along the way touching on specific issues, such as establishing corpora and their boundaries, pre-processing textual materials, and applying particular interpretive schemes. We will then focus on how topic modeling has shaped theorizing in management. We will examine subject areas where topic modeling has been applied that are relevant to management theory and some of the derived insights for the management field. Our review will finish with new directions in topic modeling research, both methodological (such as new forms of categorization and natural language processing), and theoretical (such as the application of topic models in the political analysis of organizations and interpretive schemes generated by organizational micro-processes).