Portrait of Brandon M. Stewart, Princeton University

Princeton University

Brandon M. Stewart

Professor of Sociology

Brandon M. Stewart is Professor of Sociology at Princeton University (opens in new tab). He is Co-Editor-in-Chief of Political Analysis (opens in new tab) and Associate Editor at Sociological Methods & Research (opens in new tab). He develops new methods for applications across computational social science.

Download CV bms4@princeton.edu Google Scholar
Cover of Text as Data: A New Framework for Machine Learning and the Social Sciences

Book

Award 2025 ASA Methodology Section Outstanding Publication Award

Text as Data: A New Framework for Machine Learning and the Social Sciences

Justin Grimmer, Margaret E. Roberts, Brandon M. Stewart

A research-design framework for using text and machine learning across representation, discovery, measurement, prediction, and causal inference.

Reviewed in Sociological Methods & Research, New Media & Society, Technometrics, and elsewhere. See all 10 reviews.

About the Book Princeton University Press

Featured Publications

Figure 1 from State Media Control Influences Large Language Models showing the logical flow of six studies from training data to model responses. Nature, 2026 State Media Control Influences Large Language Models Figure 4 from Using Large Language Model Annotations for the Social Sciences comparing bias, coverage, and RMSE when ignoring prediction errors versus using design-based supervised learning. 2025 Gosnell Prize American Journal of Political Science, 2026 Using Large Language Model Annotations for the Social Sciences: A General Framework of Using Predicted Variables in Downstream Analyses Figure 1 workflow diagram for STATe-of-Thoughts showing action templates, output generation, and evaluation. Accepted at the Conference on Language Modeling (COLM), 2026 STATe-of-Thoughts: Structured Action Templates for Tree-of-Thoughts Figure 1 from What Is Your Estimand? showing the path from theory or a general goal to theoretical estimands, empirical estimands, and estimation strategies. 2022 ASA Outstanding Publication Award American Sociological Review, 2021 What Is Your Estimand? Defining the Target Quantity Connects Statistical Evidence to Theory Figure 8 from Embedding Regression showing nearest-neighbor terms that distinguish liberal, shared, and conservative understandings of the most important political problem facing the United States. American Political Science Review, 2023 Embedding Regression: Models for Context-Specific Description and Inference Figure 2 from The Decade-Long Growth of Government-Authored News Media in China under Xi Jinping showing increases in scripted propaganda share, front-page scripted propaganda, and wholesale reprinting over 2012-2022. PNAS, 2025 The Decade-Long Growth of Government-Authored News Media in China under Xi Jinping