Prompt
You are a data coach for school counseling programs, fluent in how the ASCA
National Model uses participation, Mindsets & Behaviors, and outcome data. Below
is aggregate, de-identified data from one of my interventions.

Intervention: {{intervention_description}}
Data: {{data_table}}
Audience for the report: {{audience}}

Produce:
1. A data summary with percent changes and simple comparisons — show your
   arithmetic in parentheses so I can check every number
2. Three plain-English findings, each tied to a specific number
3. A results narrative under 150 words written for {{audience}}
4. Two honest caveats or limitations, plus one suggestion for what data to
   collect next cycle
5. A one-slide outline: headline, three bullets, and one recommended chart type

Constraints: use only the numbers I provided — never estimate, extrapolate, or
fill in missing values, and if a cell is blank say so. If the data cannot
support a claim (small group size, no comparison group), state that plainly
instead of writing around it. Use "associated with," never "caused," unless I
explicitly describe a comparison group. No education-research jargon.

Fill in your details and the prompt updates live — then copy.

What you get back (excerpt)

**Finding 1:** Average absences among the eight group members fell from 10.75 per quarter (86 total / 8 students) to 7.25 (58 / 8) — a 32.6% reduction — during a quarter when the schoolwide 7th-grade average slightly rose (4.1 to 4.3). **Caveat:** With eight students and no matched comparison group, this change is associated with the intervention but can't be attributed to it; regression toward the mean is plausible for students selected at their attendance low point. **Slide headline:** Attendance group members cut absences by a third while schoolwide absences ticked up.

The full workflow

  1. Aggregate and de-identify in your spreadsheet first — totals and lists of values, never names, IDs, or rows a colleague could trace to a student.
  2. Recompute every figure the model reports; arithmetic slips are the most common failure.
  3. Keep the caveats in the final report — a principal trusts a counselor who names limitations.
  4. Reuse the same structure each cycle so your results reports become comparable year over year.

Watch out for

Small-n data can identify students even without names — a group of 8 with grade level and intervention type attached may be re-identifiable. Share only what you'd be comfortable seeing forwarded, and use district-approved tools for anything finer-grained (FERPA still applies to indirectly identifiable records).

Language models miscalculate more often than they admit. Never publish a percentage you did not verify yourself.

Where this comes from

Every use case on this site is grounded in real reports from working school counselors — not invented by us.

More AI use cases for school counselors

← All 6 use cases: How School Counselors Use AI