Proper Prompting for School Psychologists, Part 2: A Practical Prompt Template for Better Results

Key Points (For Readers on the Go)

• AI prompting for school psychologists is not about secret phrases or technical tricks. It is about clear professional communication.

• Better prompts include the task, audience, context, example, rules, clarification request, steps, format, and review expectations.

• Examples are especially powerful because they show AI what “good” looks like.

• AI tools should be asked to clarify before starting when important information is missing.

• AI output still requires human review, especially when privacy, bias, eligibility, diagnosis, or student support decisions are involved.

Introduction

In the first post in this series, I described AI prompting as something closer to supervising an intern than “engineering” a machine. That framing matters. AI can be fast, flexible, and surprisingly useful, but it does not know your school, your students, your professional standards, or your ethical obligations unless you clearly explain them.

That is where prompting comes in.

AI prompting for school psychologists is simply the practice of giving AI clear instructions, useful context, and enough structure to produce a better first draft. It is not magic. It is not coding. It is not about finding the perfect phrase that unlocks the system.

It is about communication.

This post builds on Part 1 by offering a practical prompt template you can adapt for school-based work, consultation, communication, planning, and professional writing.

Why AI Prompting for School Psychologists Needs Structure

A vague prompt usually leads to a vague answer.

For example, imagine typing:

Write an email to a parent.

You might get something usable. You might also get something too formal, too vague, too long, too casual, or too confident about information you did not provide.

Now compare that with this:

Help me draft a warm, parent-friendly email from a school psychologist to a caregiver about scheduling a meeting to discuss evaluation results. Do not include confidential details. Do not imply eligibility has already been determined. Use two short paragraphs and end with a clear next step.

That second prompt gives AI more to work with. It explains the task, the audience, the tone, the boundaries, and the format.

That is the point of prompting. You are reducing ambiguity.

If AI is the intern, the prompt is the assignment sheet.

The 9-Part Prompt Template for Better AI Results

A useful prompt does not need to be complicated. In most cases, it just needs to include the right ingredients.

Here is a simple 9-part structure.

1. Task: What Do You Want AI to Do?

Start by clearly naming the task.

Weak prompt:

Help with this behavior data.

Better prompt:

Summarize the following behavior data for a school team. Focus on observable patterns, not causes.

The task tells AI what job it is doing. Is it drafting? Summarizing? Rewriting? Brainstorming? Organizing? Reviewing? Creating questions?

The more specific the task, the better the output.

2. Audience: Who Will Read or Use the Output?

AI needs to know who the final product is for.

A parent email should sound different from a staff consultation note. A district training handout should sound different from a technical report. A teacher-facing recommendation should be more practical than a research summary.

For example:

Audience: A caregiver with no background in special education or assessment.

Or:

Audience: A general education teacher looking for practical classroom strategies.

This helps AI adjust language, tone, depth, and format.

3. Context: What Background Information Does AI Need?

Context is the information AI needs to complete the task well.

In school-based work, context might include:

• grade level

• referral concern

• setting

• what has already been tried

• relevant strengths

• relevant constraints

• the purpose of the communication

• what should not be assumed

For example:

Context:The student is in 4th grade and has difficulty with written expression. The teacher reports challenges with organization and task initiation. The team is looking for practical classroom supports.

Context helps prevent generic output. It also reduces the chance that AI will fill in gaps with assumptions.

4. Example: What Does a Good Version Look Like?

This is one of the most useful prompting habits.

If you want AI to write an email, give it an example of an email you previously wrote and liked. If you want a parent-friendly explanation, provide a sample paragraph that has the tone and level of detail you want.

You can say:

Here is an example of the style I want. Do not copy the content. Use it only as a guide for tone, structure, warmth, and level of detail.

Examples are especially helpful when the task is nuanced. AI may not know what “warm but professional” means to you. But if you show it a strong example, it can usually get much closer.

This is not cheating. It is good supervision.

A supervisor would not simply tell an intern, “Write a good email.” A supervisor might say, “Here is the kind of email we usually send. Notice the tone, length, and structure.”

That same logic applies to AI.

5. Rules: What Should AI Do or Avoid?

Rules are the boundaries of the task.

For school psychologists, educators, and related service providers, this section is essential because many AI tools are designed to be helpful, but not necessarily cautious.

Useful rules might include:

• Do not invent information.

• Do not include confidential or identifying student information.

• Do not diagnose.

• Do not imply eligibility decisions have already been made.

• Use plain language.

• Keep the tone supportive and collaborative.

• Flag anything that requires professional review.

• Do not cite laws, policies, or research unless source material is provided.

Rules help keep the output safer, more accurate, and more aligned with professional expectations.

6. Clarify Before Starting: What Should AI Ask Before It Begins?

For important tasks, ask AI to clarify before drafting.

A useful instruction is:

Before you begin, ask any clarifying questions you need to complete this well. Do not guess.

This is especially helpful when the task depends on missing information. For example, if you ask AI to draft a meeting summary but do not provide the audience, purpose, or tone, it should ask before proceeding.

This does not mean AI needs to ask questions every time. For simple tasks, that can become annoying. But for higher-stakes communication, unclear data summaries, or sensitive school situations, clarification can prevent a poor first draft.

7. Steps: What Process Should AI Follow?

Sometimes the order matters.

For example, if you are asking AI to summarize behavior data, you may want it to first describe what is observable, then identify missing information, and only then suggest possible next steps.

A step-based prompt might say:

First, summarize the concern. Second, identify what information is missing. Third, draft possible next steps. Fourth, flag anything that requires professional review.

This helps AI slow down and organize the response in a way that supports better professional judgment.

8. Format: What Should the Final Answer Look Like?

AI often gives better responses when you tell it exactly how to organize the output.

For example:

• Write two short paragraphs.

• Use bullet points.

• Create a table.

• Use headings.

• Keep it under 300 words.

• Write at an 8th-grade reading level.

• Separate “Draft Language” from “Professional Review Notes.”

Format matters because a technically accurate answer may still be unusable if it is too long, too dense, or poorly organized.

9. Review: What Should AI Flag for Human Judgment?

The final part of the prompt should remind AI that some things require human review.

For example:

Flag any assumptions, missing information, privacy concerns, ethical concerns, or areas that require professional review.

This is important because AI output can sound polished even when it is incomplete. A review instruction encourages the tool to identify limitations instead of hiding them behind confident language.

The goal is not to make AI the decision-maker. The goal is to make the draft easier for the professional to evaluate.

Copy-and-Paste Prompt Template

Here is a reusable version of the template.

Task:

[What do you want AI to do?]

Audience:

[Who will read or use the output?]

Context:

[What background information does AI need?]

Example:

[Paste a sample of the style, structure, or level of detail you want.]

Rules:

[What should AI do or avoid?]

Clarify Before Starting:

Before you begin, ask any clarifying questions you need to complete this well. Do not guess.

Steps:

[What process should AI follow?]

Format:

[What should the final answer look like?]

Review:

[What should AI flag for human judgment?]

You do not need to use every section every time. For quick tasks, a shorter prompt may be enough. But when the task involves communication, interpretation, student support, or professional documentation, structure helps.

Practical Examples

Example 1: Drafting a Parent Email

Weak prompt:

Write an email to a parent about an evaluation.

Better prompt:

Task: Draft a warm, professional email from a school psychologist to a caregiver.

Audience: A parent or caregiver with no background in special education.

Context: I need to schedule a meeting to discuss evaluation results.

Example: Use the same tone as the sample email below, but do not copy the content.

Rules: Do not include confidential details. Do not imply eligibility has already been determined. Use plain language.

Clarify Before Starting: Ask any clarifying questions needed before drafting. Do not guess.

Format: Two short paragraphs with a clear next step.

Review: Flag anything that may require professional review before sending.

Why this works: The prompt gives AI the purpose, audience, tone, boundaries, and format. It also keeps the professional in control.

Example 2: Summarizing Behavior Data

Weak prompt:

Summarize this behavior data.

Better prompt:

Task: Summarize the following behavior data for a school team.

Audience: Teachers, caregivers, and school-based team members.

Context: The team is reviewing patterns in classroom behavior and considering next steps.

Rules: Focus only on observable patterns. Do not infer causes beyond the data provided. Do not diagnose.

Steps: First summarize the main patterns. Then identify missing information. Then suggest possible questions for the team to discuss.

Format: Use three brief sections with headings.

Review: Flag any limitations in the data.

Why this works: The prompt keeps AI from jumping too quickly to explanations. It encourages a careful summary that can support team discussion.

Example 3: Generating Teacher Consultation Ideas

Weak prompt:

Give me recommendations for this student.

Better prompt:

Task: Generate possible classroom supports for a student who has difficulty starting independent writing tasks.

Audience: A general education teacher.

Context: The student often understands the assignment but delays starting. The teacher is looking for practical strategies that can be used during independent work. The teacher and I have had awkward interactions lately, and I want to avoid making them defensive.

Rules: Do not diagnose. Do not claim these supports are required. Keep suggestions feasible in a general education classroom.

Steps: Provide 3-5 possible supports. For each one, explain when it would be used and what the adult would do.

Format: Use a table with columns for Strategy, When to Use It, and Adult Role.

Review: Flag any information that would be helpful to know before selecting an intervention.

Why this works: The prompt asks for practical, classroom-ready ideas while making clear that the output is for discussion, not a final decision.

Ethical Considerations

Several cautions are especially important.

First, be careful with student privacy. Do not enter identifiable student information into tools that are not approved for that purpose. Never enter student information (even redacted) into a model that isn’t FERPA or HIPAA compliant.

Second, watch for bias. AI-generated language may unintentionally frame students or families in deficit-based ways. It may also make assumptions based on incomplete information. Review outputs carefully for fairness, cultural responsiveness, and accuracy.

Third, verify facts. AI can produce incorrect citations, outdated policy summaries, or confident-sounding claims that are not supported by the information provided. If the output references law, policy, research, or district procedures, check the original source.

Finally, remember that a polished draft is not the same as a correct draft. AI can make weak ideas sound professional. Expert review is not optional.

Final Takeaways

• Better prompting starts with clarity, not clever wording.

• Examples help AI match your preferred tone, structure, and level of detail.

• Rules and review criteria reduce the risk of unsupported or inappropriate output.

• Asking AI to clarify before starting can prevent avoidable errors.

• You remain responsible for the final product.

Internal Links

For more on this topic, see:

Part 1: Better AI Prompting for School Psychologists and Educators https://lockwoodconsulting.net/blog/m5psrgic276lswt8hs37raj4sckgxj

AI Is Like an Intern: Useful, Uneven, and in Need of Supervision https://lockwoodconsulting.net/blog/ai-is-like-an-intern-useful-uneven-and-in-need-of-supervision

• More AI guidance and blog posts from Lockwood Educational & Psychological Consulting https://lockwoodconsulting.net/blog

Call to Action

If your school, district, professional association, or clinical organization is trying to make AI feel more practical, ethical, and manageable, I provide training and consultation focused on responsible AI use in education and psychological practice.

Visit Lockwood Educational & Psychological Consulting at https://lockwoodconsulting.net/ or sign up for updates at https://lockwoodconsulting.net/blog.

AI Use Disclosure

Portions of this post were drafted with the assistance of an AI writing tool and revised by the author for accuracy, clarity, and professional judgment.