What we're looking for
Social scientists are incorporating AI into their research pipelines, including designs, data collection, and analyses. Yet, despite this rapid adoption, important methodological questions remain: What are the principled approaches to validating measurements via AI tools? To what extent are AI-generated observations interchangeable with those from human respondents? How can AI tools advance experiments and causal inference? How should replication be understood and achieved in an era when models, prompts, and systems evolve quickly? What new research designs become possible when AI agents organize, execute, or participate in elements of a research project?
Building on previous workshops at Yale, this conference convenes researchers working on the methods of AI-enabled social science. We welcome submissions that develop new methodological frameworks, propose evaluation strategies, introduce tools or benchmarks, or present empirical findings that clarify when and how AI improves (or complicates) social science inference. Critical and negative results are as welcome as new designs. Contributions may focus on established themes, such as measurement and simulation, or on emerging directions, including the development and deployment of agentic systems.
We explicitly invite work that proposes new methodological designs that are only possible because of AI. Illustrative examples include, but are not limited to:
- Measurement from unstructured and multimodal data
- Validation of AI-generated measurements against human coders and ground truth
- Correction for bias in AI-generated labels and implications for uncertainty quantification
- AI agents as participants in experiments, games, or markets
- Hybrid human-AI group and mixed subject designs
- Causal inference leveraging model reasoning or structured interaction
- AI-assisted theory generation or hypothesis refinement
- Validating and auditing agent-executed analyses
- Replication when models, prompts, and systems change
- AI-assisted replication and reanalysis at scale
- AI-assisted content generation, selection, and evaluation
- Personalization and adaptive interventions using AI
- AI-enabled preference elicitation and decision modeling
- Modeling social influence and information diffusion with AI