# Adaptive political surveys and GPT-4: Tackling the cold start problem with simulated user interactions

arXiv: https://arxiv.org/abs/2503.09311 · HTML: https://mlsanity.com/paper/2503.09311/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 17 of its 71 references that are in the map.
- First posted on arXiv 12 Mar 2025.

## Builds on (most cited first, 17 of 17)

- [Out of One, Many: Using Language Models to Simulate Human Samples](https://mlsanity.com/paper/2209.06899/) (2022)
- [From Pretraining Data to Language Models to Downstream Tasks: Tracking the Trails of Political Biases Leading to Unfair NLP Models](https://mlsanity.com/paper/2305.08283/) (2023)
- [Systematic Biases in LLM Simulations of Debates](https://mlsanity.com/paper/2402.04049/) (2024)
- [Large Language Models are Competitive Near Cold-start Recommenders for Language- and Item-based Preferences](https://mlsanity.com/paper/2307.14225/) (2023)
- [Collaborative Large Language Model for Recommender Systems](https://mlsanity.com/paper/2311.01343/) (2023)
- [LLM Voting: Human Choices and AI Collective Decision Making](https://mlsanity.com/paper/2402.01766/) (2024)
- [Assessing Political Bias in Large Language Models](https://mlsanity.com/paper/2405.13041/) (2024)
- [Can Small Language Models be Good Reasoners for Sequential Recommendation?](https://mlsanity.com/paper/2403.04260/) (2024)
- [Intelligent Model Update Strategy for Sequential Recommendation](https://mlsanity.com/paper/2302.07335/) (2023)
- [Text-Based Ideal Points](https://mlsanity.com/paper/2005.04232/) (2020)
- [Representation with Incomplete Votes](https://mlsanity.com/paper/2211.15608/) (2022)
- [Large Language Model Simulator for Cold-Start Recommendation](https://mlsanity.com/paper/2402.09176/) (2024)
- [Survey of Computerized Adaptive Testing: A Machine Learning Perspective](https://mlsanity.com/paper/2404.00712/) (2024)
- [Aligning Large Language Models with Diverse Political Viewpoints](https://mlsanity.com/paper/2406.14155/) (2024)
- [Dynamic Question Ordering in Online Surveys](https://mlsanity.com/paper/1607.04209/) (2016)
- [Adaptive questionnaires for facilitating patient data entry in clinical decision support systems: Methods and application to STOPP/START v2](https://mlsanity.com/paper/2309.10398/) (2023)
- [Fast and Adaptive Questionnaires for Voting Advice Applications](https://mlsanity.com/paper/2404.01872/) (2024)

## Built on by (most cited first, 3 of 3)

- [Persona-driven Simulation of Voting Behavior in the European Parliament with Large Language Models](https://mlsanity.com/paper/2506.11798/) (2025)
- [AI and Collective Decisions: Strengthening Legitimacy and Losers' Consent](https://mlsanity.com/paper/2604.05368/) (2026)
- [IXPLORE: Bounded Ideal Point Estimation with Grid-Based Uncertainty Quantification](https://mlsanity.com/paper/2609.06018/) (2026)

## Abstract (arXiv)

Adaptive questionnaires dynamically select the next question for a survey participant based on their previous answers. Due to digitalisation, they have become a viable alternative to traditional surveys in application areas such as political science. One limitation, however, is their dependency on data to train the model for question selection. Often, such training data (i.e., user interactions) are unavailable a priori. To address this problem, we (i) test whether Large Language Models (LLM) can accurately generate such interaction data and (ii) explore if these synthetic data can be used to pre-train the statistical model of an adaptive political survey. To evaluate this approach, we utilise existing data from the Swiss Voting Advice Application (VAA) Smartvote in two ways: First, we compare the distribution of LLM-generated synthetic data to the real distribution to assess its similarity. Second, we compare the performance of an adaptive questionnaire that is randomly initialised with one pre-trained on synthetic data to assess their suitability for training. We benchmark these results against an "oracle" questionnaire with perfect prior knowledge. We find that an off-the-shelf LLM (GPT-4) accurately generates answers to the Smartvote questionnaire from the perspective of different Swiss parties. Furthermore, we demonstrate that initialising the statistical model with synthetic data can (i) significantly reduce the error in predicting user responses and (ii) increase the candidate recommendation accuracy of the VAA. Our work emphasises the considerable potential of LLMs to create training data to improve the data collection process in adaptive questionnaires in LLM-affine areas such as political surveys.