The IDeaS group offers workshops that provides management doctoral students and faculty members working in the areas of data science, data analysis, Big Data, and artificial intelligence an opportunity to discuss and advance their ongoing research.

2020 PDW at the Academy of Management

Our PDW’s goal is to engage participants from OMT, RM, STR and SAP in big text analysis and theory generation with some of the latest advances in rendering with topic modeling (Hannigan et al., 2019 for a review). This includes methods of curating large textual corpora, modified versions of the core (Latent Dirichlet Allocation – LDA) topic modeling techniques, and refined visualizations for building theory.

Topic modeling is at the interface of method and theory, big and little data, quantitative and qualitative analysis, induction and deduction (DiMaggio, 2015; Grimmer and Stewart, 2013; Mohr and Bogdanov, 2013; Mützel, 2015; Nelson, 2017). Participants (who need no more than basic programming knowledge and an interest in theory building) will be guided through pre-built programs in Python and R, which will be available on GitHub. The programs will be based off of the standard unstructured topic modeling routines (LDA), but include structured topic modeling (STM) and also bi-term (BT) analysis. We will also explore topic mapping via LDAvis and Gephi network software in order to help build conceptual relationships.

The analyses will be guided by co-leads who have published and have an ongoing topic modeling research program: Tim Hannigan, Richard Haans, Dev Jennings, Hovig Tchalian, and Rodrigo Valadao. Academics with experience in theory building from mixed methods, Tim Hannigan, Vern Glaser, and Dev Jennings, will discuss different methods and examples for moving from analysis to theory in article writing. The PDW is targeted to researchers in OMT and STR, along with methodologists from RM.