arXiv cs.CL· Tobias Holtdirk, Georg Ahnert, Joseph W Sakshaug, Anna-Carolina Haensch·· 11 小时前AI 评分31
用大语言模型填补民意调查缺失数据：ICL 方法最多只能与统计基线持平
Imputation of Public Opinion Data with Large Language Models
AI 导读
研究首次按多重插补指标评估 LLM 填补民意调查缺失值的效果，基于四项调查超 500 万条填补回答、覆盖多种缺失机制。在 84 种 ICL 配置和六个开源权重模型上，最佳 ICL 方法优于零样本 persona prompting，但最多只能与 MICE PMM、TabPFN 等统计基线持平，且对调查主题敏感、算力消耗高 2-3 个数量级。
正文 · 原文
Abstract:Large language models (LLMs) are widely evaluated as simulators of survey responses, yet applied survey research mostly faces item nonresponse. Imputation fills in those missing values to restore the overall structure of a partially observed dataset and has its own well-defined evaluation criteria, which differ fundamentally from prediction accuracy. We present the first evaluation of LLM imputation against proper multiple-imputation metrics and statistical baselines on more than 5 million imputed responses from four public opinion surveys under various missingness mechanisms. We evaluate in-context learning (ICL) for this task by investigating 84 ICL configurations and comparing the best-performing configurations on six open-weight LLMs. The strongest ICL approach improves on zero-shot persona prompting but remains, at best, competitive with statistical baselines (MICE PMM, TabPFN). ICL approaches are sensitive to survey topics and consume 2-3 orders of magnitude more compute. Moreover, prediction metrics commonly used in LLM survey simulation (macro-F1, TVD) do not reliably track imputation quality. Our findings highlight the importance of fit-for-purpose evaluation for LLM imputation, where established multiple imputation remains the better default.
| Subjects: | Computation and Language (cs.CL); Methodology (stat.ME) |
| Cite as: | arXiv:2606.09351 [cs.CL] |
| (or arXiv:2606.09351v2 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2606.09351 arXiv-issued DOI via DataCite |
Submission history
From: Tobias Holtdirk [view email]
[v1]
Mon, 8 Jun 2026 11:25:10 UTC (2,278 KB)
[v2]
Fri, 30 Oct 2026 16:50:36 UTC (1,458 KB)
来源：arXiv cs.CL · arxiv.org