# Beyond Averages: Evaluating LLMs on Human Survey Replication at the Distributional Level

arXiv: https://arxiv.org/abs/2606.09013 · HTML: https://mlsanity.com/paper/2606.09013/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 9 of its 29 references that are in the map.
- First posted on arXiv 8 Jun 2026.

## Builds on (most cited first, 9 of 9)

- [Out of One, Many: Using Language Models to Simulate Human Samples](https://mlsanity.com/paper/2209.06899/) (2022)
- [Using Large Language Models to Simulate Multiple Humans and Replicate Human Subject Studies](https://mlsanity.com/paper/2208.10264/) (2022)
- [Do LLMs exhibit human-like response biases? A case study in survey design](https://mlsanity.com/paper/2311.04076/) (2023)
- [Aligning Language Models to User Opinions](https://mlsanity.com/paper/2305.14929/) (2023)
- [Do Large Language Models Show Decision Heuristics Similar to Humans? A Case Study Using GPT-3.5](https://mlsanity.com/paper/2305.04400/) (2023)
- [Evaluating Cultural Adaptability of a Large Language Model via Simulation of Synthetic Personas](https://mlsanity.com/paper/2408.06929/) (2024)
- [Beyond correlation: The Impact of Human Uncertainty in Measuring the Effectiveness of Automatic Evaluation and LLM-as-a-Judge](https://mlsanity.com/paper/2410.03775/) (2024)
- [LLM-based Human Simulations Have Not Yet Been Reliable](https://mlsanity.com/paper/2501.08579/) (2025)
- [Restoring Heterogeneity in LLM-based Social Simulation: An Audience Segmentation Approach](https://mlsanity.com/paper/2604.06663/) (2026)

## Built on by (most cited first, 1 of 1)

- [Large Language Models Substantially Compress Well-Being Inequality but Largely Preserve Its Socioeconomic Structure](https://mlsanity.com/paper/2609.33055/) (2026)

## Abstract (arXiv)

LLMs are increasingly used to simulate human survey responses, but prior work has mainly evaluated replication using mean-level or aggregate agreement, offering limited insight into whether LLMs reproduce the variability of human behavior. We evaluate LLM-based survey replication at the distributional level using a non-public 2010 consumer choice experiment on Korean instant noodle purchases, a setting unlikely to overlap with model training data. We evaluate three response variables of differing statistical type: binary purchase incidence, categorical brand choice, and count purchase quantity. For each, we compare human and LLM responses at mean-level, pattern, and distributional alignment, and against reference baselines from the human data alone. LLMs reproduce condition-level patterns reasonably well but fail to capture distributional structure: for purchase quantity, no model beats a condition-insensitive baseline that simply matches the pooled human distribution. Because models that match human means well can still produce distributions further from humans than this baseline, mean-based evaluation alone can be actively misleading. Replication also varies with input configuration, with structured personas and multimodal inputs improving alignment while explicit reasoning prompting degrades it monotonically.