Computational Social Science meets Qualitative Research
in collaboration with Dan de Kadt (LSE).
Computational social science (CSS) is typically perceived as an offshoot of traditional quantitative research methods with a greater emphasis on large-scale data and computationally-intensive statistics and models. However, new types of scholarship challenge this perspective, arguing that CSS should be (or already is) closer to qualitative research methodology than most researchers assume (Däubler & Benoit, 2022; Meng, 2021; Rodriguez & Storer, 2020; Tanweer et al., 2021; Theobold et al., 2024). These debates will likely intensify as CSS leans heavily into new developments in generative artificial intelligence (AI) which appear to offer great promise for processing and analysing and collecting new forms of data, and scaling the types of analyses typically found in qualitative work (e.g. Cirone & Spirling, 2021; Chopra & Haaland, 2023; Davidson 2024; Geiecke & Jaravel, 2024; Wheeler, 2025).
This project will bring together scholars whose work addresses these emerging issues, moving beyond field-specific questions to develop a new methodological community with a broader shared research agenda.
We aim to focus on the following themes :
- How innovations from CSS and the closely associated recent developments in AI can improve qualitative research methodology, for example via scaling up and automation ;
- How qualitative research methodology can help address the limits of CSS, for example through research standards usually identified as qualitative such as ethics, reflexivity and contextualisation ;
- The potential for combining CSS and qualitative research methods within mixed-methods research designs ;
- The potential for computational qualitative methods as a new type of methodological family, for example computational discourse analysis ;
- What building the bridge between CSS and qualitative research does to the question of data: from “BIG Qual” to data quality ;
- The difficulties and challenges of collaborating across the CSS / qualitative divide.
The first event out of this project is a workshop organised at the London School of Economics and Political Science and hosted by the Department of Methodology and the Data Science Institute.