Structured prompt-template and keyword-card interaction design for human-LLM co-creation¶
Design · All designs
Evidence · 6 claims (6 for) · 1 study (1 associational),q2· 1 of 1 report an effect size · 6 claims rest on one study
Description¶
A novel interaction protocol for studying human-LLM foraging: six editable prompt templates divided into Exploration and Exploitation types, plus a shared pool of keyword cards contributed by course participants and categorized into concepts such as 'Actors', 'AI tools', 'Environments', and 'Subjects'. The templates "facilitated the exploration-exploitationbehavioral modes of IFT, standardized the interaction between participants, guided those without prior experience, and enabled systematic analysis of their foraging behavior". Participants crafted prompts by combining templates with guide or GPT-sourced keywords in 45-minute facilitated ChatGPT sessions.
Design Implications¶
Context¶
Requirements¶
- Prompt templates and keyword cards derived from participants' reported interests gathered before the course
- A researcher-mediated 45-minute session protocol with setup, introduction, execution, and conclusion phases
Constraints¶
- The structured design intentionally limits free-form prompting, so results do not reflect typical ChatGPT use
- The specific task complicates comparisons with previous web search studies that formally assess IFT principles
Target Learners¶
- Doctoral-level educational scientists in an AI in Education course
Learning Goals¶
- Pedagogical content creation and study-design scenario development with generative AI
Claims¶
- Prompt Type Keyword Source Split [+M]
- Exploitation Concentrates Selected Scenarios [+M]
- Participants favor socially-sourced others' keywords over their own in exploration prompts, deviating from personalized learning expectations [+W]
- GPT keyword use is negatively correlated with essay topic diversity, suggesting exploitation narrows reflective focus [+W]
- Exploration prompt frequency is positively correlated with keyword topic diversity in structured human-LLM content generation [+W]
- Cooperativity positively correlates with GPT keyword use in younger, less-experienced participants (Group 1) [+W]
Related Designs¶
Examples¶
-
Key Sources¶
- Pablo Flores Romero, Kin Nok Nicholas Fung, Guang Rong & Benjamin Ultan Cowley. (2025). Structured human-LLM interaction design reveals exploration and exploitation dynamics in higher education content generation. npj Science of Learning. https://doi.org/10.1038/s41539-025-00332-3
Page metadata
| Field | Value |
|---|---|
type |
design |
id |
structured-prompt-template-keyword-interaction-design |
title |
Structured prompt-template and keyword-card interaction design for human-LLM co-creation |
description |
A novel interaction protocol for studying human-LLM foraging: six editable prompt templates divided into Exploration and Exploitation types, plus a shared pool of keyword cards contributed by course participants and c... |
status |
review |
generated |
by process:settle-candidates on 2026-10-26 |
trust |
unverified — no one has confirmed this page's content |
sources |
1 — pablo-flores-romero-2025 |
Frontmatter is the page's machine-readable record — see the Schema & Guide.