PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
March 29, 2026Public Opinion Quarterly0 citationsOpen Access

Scaling Open-Ended Survey Responses Using LLM-Paired Comparisons

View Full Paper
MDMatthew DiGiuseppeLeiden UniversityMFMichael E FlynnKansas State University

Key Points

  • The aim is to develop a method for scaling open-ended survey responses using pairwise comparisons facilitated by LLMs.
  • Developed a pairwise comparison technique for open-ended survey responses.
  • Utilized large language models to identify winning and losing statements.
  • employed a Bayesian Bradley-Terry model to assign scores to statements.
  • Compared performance across different LLM sizes.
  • Evaluated consistency with crowdsourced worker comparisons.
  • Pairwise comparisons demonstrated greater consistency than traditional zero-shot ratings.
  • The Bayesian approach improved item discrimination and uncertainty measurement.
  • Findings indicated reduced anchoring bias in the measurements.
  • Comparison with crowdsourced workers showed strong agreement.

Abstract

Abstract Survey researchers rely heavily on closed-ended questions to measure latent respondent characteristics like knowledge, policy positions, emotions, ideology, and various other traits. Closed-ended questions are easy to analyze and collect, but necessarily limit the depth and variability of responses. Open-ended responses allow for greater depth and variability in responses, but are labor intensive to code. Large language models (LLMs) may help with this problem, but existing approaches to using LLMs have a number of limitations. In this paper, we propose and test a pairwise comparison method to scale open-ended survey responses on a continuous scale. The approach relies on LLMs to make pairwise comparisons of statements that identify which statement “wins” and “loses.” With this information, we employ a Bayesian Bradley-Terry model to recover a “score” on a latent dimension for each statement. This approach allows for finer discrimination between items, reduced anchoring bias, better measurement of uncertainty, and is more flexible than methods relying on Maximum Likelihood Estimation techniques. We demonstrate the utility of this approach on an open-ended question probing knowledge of interest rates in the US economy. A comparison of six LLMs of various sizes reveals that pairwise comparisons show greater consistency than zero-shot 0–10 ratings across a variety of model sizes. Further, comparison of pairwise decisions is consistent with knowledgeable crowdsourced workers.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

DiGiuseppe et al. (2026) studied this question.

synapsesocial.com/papers/69c8c277de0f0f753b39cbf4https://doi.org/10.1093/poq/nfag013
Ask AI
Helpful
Bookmark
Share
View Full Paper