Generative AI tools such as ChatGPT have rapidly entered classrooms, staff rooms, and professional development conversations. Many educators see their promise—personalised lessons, differentiated materials, faster planning—but also share a common frustration: “The AI doesn’t give me what I need.”

Recent research by Park and Choo (2024) makes a compelling argument that the problem is not the tool itself, but how educators communicate with it. Their study highlights a crucial insight: the educational value of generative AI depends almost entirely on the quality of the prompts educators use.

Why Generative AI Often Disappoints Teachers

Generative AI systems respond to prompts by predicting patterns from vast datasets. When prompts are vague, underspecified, or poorly structured, the outputs are similarly unfocused. Park and Choo found that many educators approach AI with high expectations but limited guidance, leading to frustration when results are too advanced, too generic, or misaligned with students’ needs—especially in special education contexts.

Crucially, most existing prompt engineering frameworks have been designed by computer scientists rather than educators. As a result, they often lack alignment with pedagogical thinking, classroom realities, and learner diversity.

Introducing the IDEA Framework for Educators

To address this gap, the study proposes the IDEA Framework, a practical, educator-friendly model for prompt engineering. Rather than requiring technical expertise, IDEA builds directly on instructional planning habits teachers already use.

1. Include Essential PARTS

Effective prompts must contain five key elements:

  • Persona – Who is the teacher or who should the AI act as?
  • Aim – What is the instructional goal?
  • Recipients – Who are the learners?
  • Theme – What tone, constraints, or conditions apply?
  • Structure – What format should the output take?

This mirrors lesson planning logic and ensures AI outputs are contextually relevant.

2. Develop CLEAR Prompts

The study adapts the CLEAR framework to education:

  • Concise – Avoid unnecessary wording
  • Logical – Sequence instructions clearly
  • Explicit – Specify content, level, and expectations
  • Adaptive – Tailor prompts to learner needs
  • Restrictive – Set boundaries for length, scope, and domain

Teachers who use CLEAR prompts receive responses that are more targeted, usable, and classroom-ready.

3. Evaluate and REFINE Through Iteration

One of the strongest findings of the study is that effective AI use is iterative. Educators should expect to:

  • Rephrase keywords
  • Add examples
  • Ask follow-up questions
  • Challenge or verify responses

This process not only improves output quality but also strengthens teachers’ critical thinking and AI literacy.

4. Apply With Accountability

AI is not neutral. The study cautions educators about hallucinations, bias, privacy risks, and academic integrity concerns. Human judgment remains essential. Responsible use checklists and ethical reflection are not optional extras—they are central to safe and effective implementation.

Why This Matters for Special and Inclusive Education

The implications are particularly significant for special educators. When guided by structured prompting, generative AI can support:

  • Differentiated instruction
  • IEP goal writing
  • Adapted assessments
  • Reduced administrative workload

However, without intentional prompt design, AI outputs frequently fail to address cognitive, linguistic, and accessibility needs. The IDEA framework provides a pathway to make AI genuinely inclusive rather than superficially efficient.

A Shift From Tool Adoption to Skill Development

Perhaps the most important takeaway from the study is this:
AI effectiveness in education is a skill issue, not a software issue.

Training educators to use generative AI responsibly and effectively requires moving beyond tool demonstrations toward prompt engineering as a new form of professional literacy. Just as teachers learned to design good questions, assessments, and learning tasks, they must now learn to design good prompts.

Looking Ahead

As generative AI continues to evolve, its role in education will only expand. Park and Choo’s work offers a timely reminder that technology alone does not transform teaching—teachers do. With structured strategies like the IDEA framework, educators can move from frustration to fluency, ensuring AI serves learning rather than distracting from it.

Used thoughtfully, generative AI is not a shortcut—it is a partner in professional thinking.