The Propriety of Generative AI in Sociological Research

• In Event: Implications of AI for Qualitative, Quantitative, and Computational Sociological Research

Abstract

Amidst a growing debate about the propriety of generative artificial intelligence (GenAI) in sociological research, we present a survey of judgements regarding a range of use-cases. Prior research finds that, to the extent sociologists and their collaborators use GenAI, they typically use it for writing tasks or writing and debugging code for data analysis. While this research also demonstrates that sociologists are very concerned by the social and environmental consequences of GenAI in general, less is known about whether sociologists systematically consider some uses of GenAI as inappropriate in research scenarios. Our study focuses on authors of articles published in 50 sociology journals in the last five years. We present authors with a range of tasks divided into planning, writing, data collection, data analysis, and peer review. We ask respondent how appropriate or inappropriate the use of GenAI is for each task. For those who find a particular use inappropriate, we further probe to determine which aspect the respondent finds objectionable.

Authors

• Dustin S. Stoltz, Lehigh University

• AJ Alvero, Cornell University

• Oscar Stuhler, Northwestern University

• Marshall A. Taylor, New Mexico State University