The Extended Computational Case Method: A Framework for Research Design

Abstract

This paper considers the adoption of computational techniques within research designs modeled after the extended case method. Echoing calls to augment the power of contemporary researchers through the adoption of computational text analysis methods, we offer a framework for thinking about how such techniques can be integrated into quasi-ethnographic workflows to address broad, structural sociological claims. We focus, in particular, on how this adoption of novel forms of evidence impacts corpus design and interpretation (which we tie to matters of casing), theoretical elaboration (which we associate to moving empirical claims across scales and empirical domains), and verification (which we see as a process of reflexive scaffolding of theoretical claims). We provide an example of the use of this framework through a study of the marketization of social scientific knowledge in the United Kingdom.

Sociolinguistic Perspectives on Machine Learning with Text Data

“…There are many specific methods to consider, and here I show how CTM can be useful in how it models topical information in ways that can then be used in other statistical frameworks. I also show how embedding regressions can be used to compare meaning across groups in ways that could pair nicely with qualitative methods and analysis, something noted previously by sociologists (Nelson, 2020;Pardo-Guerra & Pahwa, 2022). The findings of the analytical example show how intracategorical variation observed in a way similar to how sociolinguists would describe it can point to important forms of social difference that traditional sociological categories might not yield.…”

Machine Learning and the Analysis of Culture

“…Engaging with social contexts using qualitative research methods, research is able to develop "insight [s] about what to look for in the data and how to theorize what is being observed" (Grigoropoulou & Small, 2022, p. 905). Alternatively, "results from computational workflows" can be used as "prompts shared with informants in ethnographic research;" in turn, insights from informants can be used to design computational analyses (Pardo-Guerra & Pahwa, 2022, p. 1828. Using abductive logics of inquiries between computational analysis and qualitative interpretation, cultural sociology is able "to locate surprising empirical findings" and to discover "innovative and creative theoretical insights" (Brandt & Timmermans, 2021, p. 192).…”

Start Generating: Harnessing Generative Artificial Intelligence for Sociological Research

“…The applications of LLMs for computational text analysis extend well beyond classification tasks. The flexibility of the approach dovetails with the way that computational sociologists are increasingly conducting what Bonikowski and Nelson (2022) term "methodological bricolage," concatenating multiple techniques into their analyses (e.g., Nelson 2017;Knight 2022;Pardo-Guerra and Pahwa 2022). Conventional machine learning methods are often dichotomized into supervised or deductive models, like text classifiers, and unsupervised or inductive models, like topic modeling (Molina and Garip 2019).…”