Kenneth Benoit. July 16, 2019. “Text as Data: An Overview.” Forthcoming in Cuirini, Luigi and Robert Franzese, eds. Handbook of Research Methods in Political Science and International Relations. Thousand Oaks: Sage.
This chapter thoroughly describes the idea of analyzing text “as data” with a social science focus. It traces a brief history of this approach and distinguishes it from alternative approaches to text. It identifies the key research designs and methods for various ways that scholars in political science and international relations have used text, with references to fields such as natural language processing and computational linguistics from which some of the key methods are influenced or inherited. It surveys the varieties of ways that textual data is used and analyzed, covering key methods and pointing to applications of each. It also identifies the key stages of a research design using text as data, and critically discusses the practical and epistemological challenges at each stage.
Patrick Perry and Kenneth Benoit. October 26, 2017. “Scaling Text with the Class Affinity Model.” London School of Economics and New York University manuscript.
Probabilistic methods for classifying text form a rich tradition in machine learning and natural language processing. For many important problems, however, class prediction is uninteresting because the class is known, and instead the focus shifts to estimating latent quantities related to the text, such as affect or ideology. We focus on one such problem of interest, estimating the ideological positions of 55 Irish legislators in the 1991 Dail confidence vote. To solve the Dail scaling problem and others like it, we develop a text modeling framework that allows actors to take latent positions on a “gray” spectrum between “black” and “white” polar opposites. We are able to validate results from this model by measuring the influences exhibited by individual words, and we are able to quantify the uncertainty in the scaling estimates by using a sentence-level block bootstrap. Applying our method to the Dail debate, we are able to scale the legislators between extreme pro-government and pro-opposition in a way that reveals nuances in their speeches not captured by their votes or party affiliations.
Thomas Däubler and Kenneth Benoit. February 13, 2017. “Estimating Better Left-Right Positions Through Statistical Scaling of Manual Content Analysis.”
Borrowing from automated “text as data” approaches, we show how statistical scaling models can be applied to hand-coded content analysis to improve estimates of political parties’ left-right policy positions. We apply a Bayesian item-response theory (IRT) model to category counts from coded party manifestos, treating the categories as “items” and policy positions as a latent variable. This approach also produces direct estimates of how each policy category relates to left-right ideology, without having to decide these relationships in advance based on out of sample fitting, political theory, assertion, or guesswork. This approach not only prevents the misspecification endemic to a fixed-index approach, but also works well even with items that are not specifically designed to measure ideological positioning.