Prompts to Practice: A Pedagogical Framework for Human-Centered AI Engagement

CRITERION	DEVELOPING	PROFICIENT	EXEMPLARY
Iterative Refinement	Submits a single prompt with little adjustment; accepts the first output.	Revise prompts across two or three attempts in response to output quality.	Systematically refines prompts, articulating why each change improves the response.
Critical Interpretation of Output	Accepts AI output at face value; little evaluation of accuracy or bias.	Identifies some inaccuracies, gaps, or biases in the output.	Rigorously evaluates output for accuracy, bias, and relevance, and explains the reasoning.
Reflective Revision	Uses AI output with minimal personal revision or reflection.	Revise the output and notes on how it shaped the final work.	Integrates and transforms output through substantive revision, reflecting critically on how AI influenced their thinking.

Table 2

Instructional Model for Human-Centered Generative AI Engagement: Phases, Pedagogical Focus, and Developmental Outcomes.

Phase		PEDAGOGICAL FOCUS	DEVELOPMENTAL OUTCOMES
1.	Critical and Ethical Awareness	Understanding Risks and Responsibilities	Ethical Discernment and Risk Awareness
2.	Prompt Literacy	Engage with the Prompt Literacy Cycle	Rhetorical Intentionality and AI Literacy
3.	AI-Supported Learning	Use LLMs for Brainstorming and Drafting	Metacognitive Awareness and Reflective Practice
4.	Reflection and Revision	Critically Evaluate and Refine Work	Authorship Development and Critical Judgment
5.	Independent Application	Apply AI Responsibly in Academic Work	Autonomy, Academic Integrity, and Transfer of Judgment

Table 3

Sample Teaching Applications.

PHASE AND INSTRUCTIONAL GOAL	TEACHING APPLICATIONS
Phase 1: Critical and Ethical Awareness Goal: Introduce LLM basics, limitations, and ethics.	AI Bias Exploration: Students review LLM responses on a social issue to identify potential biases, stereotypes, or omissions. Ethical Case Studies: Use real or hypothetical misuse cases (e.g., plagiarism, bias, misinformation) for group discussions. LLM 101 Mini-Lecture & Guided Discussion: Present how LLMs are trained, and outputs are generated; students reflect on surprises or concerns.
Phase 2: Prompt Literacy Goal: Teach students to engage with LLMs using the Prompt Literacy Cycle.	Prompt Remix Workshop: Students refine a vague prompt, generate LLM output, then refine it through multiple iterations. They compare and discuss differences in output and quality. Prompt + Output Annotation: Students craft prompts, generate LLM output, and annotate for tone, clarity, completeness, and bias. Prompt Literacy Cycle Journaling: After each stage, students journal about what they did, what happened, and what they learned. Compare Prompt Literacy vs. Prompt Engineering: Facilitate a class debate or Venn diagram activity on how prompt literacy differs from technical prompt engineering.
Phase 3: AI-Supported Learning Goal: Use LLMs to enhance, not replace, student work.	AI + Human Draft Comparison: Students write a paragraph, then generate one with an LLM, annotate differences in tone, depth, and authorship, and revise for a final submission. Collaborative Idea Generator: In groups, students use LLMs to brainstorm topics or thesis statements, then evaluate which ideas to keep and why. Socratic Seminar with AI Support: Before discussion, students use LLMs to gather perspectives and critique what is useful or lacking.
Phase 4: Reflection and Revision Goal: Promote metacognition and thoughtful LLM-influenced revision.	Revision Log: Students log LLM-assisted drafts, noting changes, acceptances, rejections, and reasons why. Authorship Workshop: Students assess whether LLM output matches their own authorship and explain their choices. AI-Influence Mapping: Students map idea development from prompt to final product, including LLM contributions.
Phase 5: Independent Ethical Application Goal: Support transfer of skills beyond the classroom.	Personal AI Use Policy: Students write a personal or professional AI use philosophy, including boundaries, goals, and ethical considerations. Field-Based AI Case Study: Students research AI use in their major or intended career and present ethical opportunities and challenges. Policy Co-Design: In groups, students draft guidelines for responsible AI use, then present and justify their policies.

References

Authors

Metrics

Articles in this issue

DOI: https://doi.org/10.55982/openpraxis.18.3.1191 | Journal eISSN: 2304-070X

Journal RSS Feed

Language: English

Page range: 481 - 494

Submitted on: Mar 18, 2026

Accepted on: Jun 7, 2026

Published on: Aug 4, 2026

Published by: International Council for Open and Distance Education (ICDE)

In partnership with: Paradigm Publishing Services

Keywords:

generative AI,

large language models,

human-centered learning,

academic integrity,

instructional framework

© 2026 April Joy Miles, Paige Haber-Curran, Khalid Arar, published by International Council for Open and Distance Education (ICDE)
This work is licensed under the Creative Commons Attribution 4.0 License.

Volume 18 (2026): Issue 3

Prompts to Practice: A Pedagogical Framework for Human-Centered AI Engagement

Figures & Tables

Figure 1

Table 1

Table 2

Figure 2

Table 3

Paradigm

My account