Generative AI for Survey Research

Generative AI is changing survey research. This course shows participants how to use it well: where it adds value, where caution is needed, and how to apply it responsibly in real research settings.

Participants will learn the core concepts behind modern AI, practical frameworks for applied work, and how large language models can function as assistants, tools, and methods across the survey lifecycle. Through hands-on examples and real-world case studies, the course covers a variety of theoretical and practical topics including:

• A technical overview of Large Language Models (LLMs) aimed at applied researchers

• Principles for applied research with LLMs across different usage patterns and phases of survey projects

• The fundamentals of prompt engineering including common pitfalls

• Application of LLMs for survey data collection through AI-assisted conversational interviewing

• Application of LLMs for data processing such as open-ended response coding

• Application of LLMs for estimation via synthetic response and silicon sample generation

• Risks and ethical issues with the usage of AI both for organizations and individuals

• Common validation and evaluation practices for applied researchers

• Common replicability and reproducibility issues

Participants will learn to identify high-value use cases for AI in survey research, design and assess AI-assisted workflows, evaluate output quality and validity, and make more informed decisions about transparency, bias, and fit for purpose. Designed for survey researchers and applied social scientists across domains, this course is led by NORC experts Soubhik Barari and Joshua Lerner who bring their expertise in survey methodology, NLP, machine learning, and the applied use of generative AI to real-world research.