Structural Topic Models for Open‐Ended Survey Responses

TLDR

Open‑ended survey responses are rarely analyzed, and when they are, it is almost always done by human coding. This article introduces the structural topic model (STM) as a semiautomated method to aid survey researchers and experimentalists in analyzing open‑ended responses. STM leverages machine‑learning text analysis and incorporates document metadata such as author gender, political affiliation, and treatment assignment to model topics. Using STM, researchers can more easily analyze open‑ended data, gain richer insights, and estimate treatment effects.

Abstract

Collection and especially analysis of open‐ended survey responses are relatively rare in the discipline and when conducted are almost exclusively done through human coding. We present an alternative, semiautomated approach, the structural topic model (STM) (Roberts, Stewart, and Airoldi 2013; Roberts et al. 2013), that draws on recent developments in machine learning based analysis of textual data. A crucial contribution of the method is that it incorporates information about the document, such as the author's gender, political affiliation, and treatment assignment (if an experimental study). This article focuses on how the STM is helpful for survey researchers and experimentalists. The STM makes analyzing open‐ended responses easier, more revealing, and capable of being used to estimate treatment effects. We illustrate these innovations with analysis of text from surveys and experiments.