Christopher A. Bail

Sayash Kapoor, Emily Cantrell, Kenny Peng et al. · princeton

Machine learning (ML) methods are proliferating in scientific research. However, the adoption of these methods has been accompanied by failures of validity, reproducibility, and generalizability. These failures can hinder scientific progress, lead to false consensus around invalid claims, and undermine the credibility of ML-based science. ML methods are often applied and fail in similar ways across disciplines. Motivated by this observation, our goal is to provide clear reporting standards for ML-based science. Drawing from an extensive review of past literature, we present the REFORMS checklist ($\textbf{Re}$porting Standards $\textbf{For}$ $\textbf{M}$achine Learning Based $\textbf{S}$cience). It consists of 32 questions and a paired set of guidelines. REFORMS was developed based on a consensus of 19 researchers across computer science, data science, mathematics, social sciences, and biomedical sciences. REFORMS can serve as a resource for researchers when designing and implementing a study, for referees when reviewing papers, and for journals when enforcing standards for transparency and reproducibility.

Petter Törnberg, Diliara Valeeva, Justus Uitermark et al.

Social media is often criticized for amplifying toxic discourse and discouraging constructive conversations. But designing social media platforms to promote better conversations is inherently challenging. This paper asks whether simulating social media through a combination of Large Language Models (LLM) and Agent-Based Modeling can help researchers study how different news feed algorithms shape the quality of online conversations. We create realistic personas using data from the American National Election Study to populate simulated social media platforms. Next, we prompt the agents to read and share news articles - and like or comment upon each other's messages - within three platforms that use different news feed algorithms. In the first platform, users see the most liked and commented posts from users whom they follow. In the second, they see posts from all users - even those outside their own network. The third platform employs a novel "bridging" algorithm that highlights posts that are liked by people with opposing political views. We find this bridging algorithm promotes more constructive, non-toxic, conversation across political divides than the other two models. Though further research is needed to evaluate these findings, we argue that LLMs hold considerable potential to improve simulation research on social media and many other complex social settings.

Nicolò Pagan, Petter Törnberg, Christopher A. Bail et al.

Large language models (LLMs) are increasingly used in the social sciences to simulate human behavior, based on the assumption that they can generate realistic, human-like text. Yet this assumption remains largely untested. Existing validation efforts rely heavily on human-judgment-based evaluations -- testing whether humans can distinguish AI from human output -- despite evidence that such judgments are blunt and unreliable. As a result, the field lacks robust tools for assessing the realism of LLM-generated text or for calibrating models to real-world data. This paper makes two contributions. First, we introduce a computational Turing test: a validation framework that integrates aggregate metrics (BERT-based detectability and semantic similarity) with interpretable linguistic features (stylistic markers and topical patterns) to assess how closely LLMs approximate human language within a given dataset. Second, we systematically compare nine open-weight LLMs across five calibration strategies -- including fine-tuning, stylistic prompting, and context retrieval -- benchmarking their ability to reproduce user interactions on X (formerly Twitter), Bluesky, and Reddit. Our findings challenge core assumptions in the literature. Even after calibration, LLM outputs remain clearly distinguishable from human text, particularly in affective tone and emotional expression. Instruction-tuned models underperform their base counterparts, and scaling up model size does not enhance human-likeness. Crucially, we identify a trade-off: optimizing for human-likeness often comes at the cost of semantic fidelity, and vice versa. These results provide a much-needed scalable framework for validation and calibration in LLM simulations -- and offer a cautionary note about their current limitations in capturing human communication.

Graham Tierney, Srikar Katta, Christopher Bail et al.

Many social science questions ask how linguistic properties causally affect an audience's attitudes and behaviors. Because text properties are often interlinked (e.g., angry reviews use profane language), we must control for possible latent confounding to isolate causal effects. Recent literature proposes adapting large language models (LLMs) to learn latent representations of text that successfully predict both treatment and the outcome. However, because the treatment is a component of the text, these deep learning methods risk learning representations that actually encode the treatment itself, inducing overlap bias. Rather than depending on post-hoc adjustments, we introduce a new experimental design that handles latent confounding, avoids the overlap issue, and unbiasedly estimates treatment effects. We apply this design in an experiment evaluating the persuasiveness of expressing humility in political communication. Methodologically, we demonstrate that LLM-based methods perform worse than even simple bag-of-words models using our real text and outcomes from our experiment. Substantively, we isolate the causal effect of expressing humility on the perceived persuasiveness of political statements, offering new insights on communication effects for social media platforms, policy makers, and social scientists.

Graham Tierney, Christopher Bail, Alexander Volfovsky

Analysis of short text, such as social media posts, is extremely difficult because of their inherent brevity. In addition to classifying topics of such posts, a common downstream task is grouping the authors of these documents for subsequent analyses. We propose a novel model that expands on the Latent Dirichlet Allocation by modeling strong dependence among the words in the same document, with user-level topic distributions. We also simultaneously cluster users, removing the need for post-hoc cluster estimation and improving topic estimation by shrinking noisy user-level topic distributions towards typical values. Our method performs as well as -- or better -- than traditional approaches, and we demonstrate its usefulness on a dataset of tweets from United States Senators, recovering both meaningful topics and clusters that reflect partisan ideology. We also develop a novel measure of echo chambers among these politicians by characterizing insularity of topics discussed by groups of Senators and provide uncertainty quantification.