From Curiosity to Control: 8 Prompt Patterns for Reliable AI in Research and Administration

Authors: Yao Yang

Affiliations: Purdue University

This poster introduces eight foundational prompt patterns that help higher education professionals use large language models reliably in research and administrative work. This includes practical techniques such as few-shot prompting, chain-of-thought reasoning, flipped interaction, role specification, and structured output constraints to improve clarity and reduce hallucination. Through real institutional examples—including executive summaries, qualitative coding support, quantitative reporting, and longitudinal trend analysis—viewers will gain structured strategies for producing consistent, audit-ready outputs.

Download the Poster:

Yao Yang_Poster.pdf

Download PDF • 361KB