Synthesize school data for different audiences with AI

Usage context

A principal or instructional leader (curriculum coordinator, head of studies) who arrives at year-end or semester close with a set of institutional data — standardized test results, internal grades by subject and class group, attendance percentages — and needs to communicate them to three different audiences in a short time. The prompt teaches a chained workflow: you define the data once and the model transforms it into three differentiated outputs within the same session, acting as a specialized assistant for each audience: a technical summary for the teachers' council, an accessible letter for families, and a formal report for the school authority (district office, regional authority, or governing body). This is conceptually distinct from batch-generation prompts — which start from templates — and from single-document standardized test analysis: here the value lies in coherence across all three outputs from a single data source.

Originally written for Chilean classrooms. References to "SIMCE" reflect Chile's national standardized assessment; substitute your school's own external assessment. "DAEM", "SLEP", and "corporación" are Chilean school authority structures — substitute your district office or governing body. "PME" (School Improvement Plan), "SEP" (preferential education subsidy), and "IVE-SINAE" (vulnerability index) are Chilean instruments — substitute your school's planning framework, targeted funding program, and student-needs index. "apoderados" means parents/guardians.

Best moment: semester or year-end close, before family meetings and the report to the school authority.

Paste before

Before opening the model, have the following data ready. Everything must be anonymized:

• School description: name or alias, type of school (public, charter, independent), district, number of students, and grade levels served
• Available standardized test results: scores by subject area and comparison with national or regional averages if available (institutional averages only — no individual data)
• Internal grades: averages by subject and grade level or class group (e.g. "Grade 8: Math 4.9 / Language Arts 5.2"). A table copied from a spreadsheet works well
• Attendance: average percentage by grade level or class group
• Relevant context: Is there an active School Improvement Plan? Any targeted funding program? Any specific institutional commitment for this year?
• The three audiences who will receive the documents and what they will use them for

Privacy rule: Do not paste student ID numbers, individual student names, grades linked to identifiable individuals, special-education assessments, or psycho-pedagogical reports. Always use averages and percentages by class group or grade level. This applies even when using a private or local model (check your jurisdiction's data-protection legislation).

Act as a team of three specialized assistants who will work in sequence from the same set of institutional data. I will give you the data once and you will transform it into three differentiated documents, one for each audience. When you finish each document, wait for my confirmation before moving to the next; if I say "continue", move directly to the next one.

**School details:**
- Name or description: {{school_name_or_description}}
- School authority: {{authority, e.g. "public school — Riverside Unified School District"}}
- Grade levels served: {{grade_levels, e.g. "Grades 1 to 8"}}
- Number of students: {{number_of_students}}
- Relevant context: {{context, e.g. "second year of School Improvement Plan implementation, targeted funding program active, high-needs student population 68%"}}

**Institutional data (anonymized — no student names or ID numbers):**
{{Paste here the available standardized test results, internal grade averages by subject and grade level, and attendance percentages. Plain text, a Markdown table, or a list all work — the model will read any format. Do not include individual grades or individual diagnostic reports.}}

**Comparison with previous year (optional):**
{{If you have prior-year data that shows trends, paste it here. If not, write "not available".}}

**Institutional focus for this year:**
{{State the main commitment or goal of your School Improvement Plan or school development plan for the current year, e.g. "improve reading comprehension in Grades 3 and 4" or "reduce irregular attendance rates in the middle school cycle".}}

Generate the three documents following these instructions:

**Document 1 — Summary for the teachers' council**
Audience: the full teaching staff. Tone: technical and direct. Length: three to four paragraphs plus a table of key indicators. Include: three main findings (one positive, one requiring attention, one critical), one hypothesis per finding for the team to discuss — not a school-level conclusion —, and a generative question to open collective reflection.

**Document 2 — Letter for families**
Audience: parents/guardians. Tone: warm, without technical language. Length: three paragraphs. Avoid technical terms such as "standardized score", "percentile", "achievement band", and "School Improvement Plan". Describe results in terms of what students achieved and what the school will do to support them in the next period.

**Document 3 — Report for the school authority**
Audience: district office, regional authority, or governing body. Tone: formal and institutional. Structure: header with school details and period; results section with a comparative table if the data allows; trend analysis; commitments for the next year; and the principal's signature block. Use the name of the school authority I provided.

Do not invent data. If a field in the documents requires information I did not provide, mark it with [TO COMPLETE: what data is missing] instead of inventing it.

Expected output

Real example (trimmed):

Document 1 — Teachers' council (excerpt):

Positive finding: Grade 4 standardized test 2025: Reading 247 pts (+6 vs. 2024), Math 239 pts (+4). Hypothesis for discussion: The School Improvement Plan strategies in their second year may be generating measurable impact, although the gap with the national average remains −23 pts in Reading.

Critical finding: Attendance in the middle school cycle fell from 83% (2025) to 80% (2026); Grade 7 records the minimum (78%), equivalent to approximately 40 days of absence per student.

Document 2 — Letter to families (excerpt):

The external assessments of our Grade 4 students showed improvement in Reading and Math compared to last year. We still have ground to cover, but the direction is right. We are concerned about irregular attendance in the upper grades; if your child has difficulty attending, please let us know so we can find solutions together.

Document 3 — Standardized test table for the school authority:

Subject 2024 2025 Change National average Reading 241 pts 247 pts +6 pts 270 pts Math 235 pts 239 pts +4 pts 261 pts

Watch out for

• Student personal data: do not paste names, ID numbers, individual grades, or individual special-education reports, even when using a model in private mode. All three documents can be generated from averages and percentages aggregated by class group or grade level (check your jurisdiction's data-protection legislation).
• The model may establish causal relationships where only correlations exist. Every finding in Document 1 is a starting point for the teaching team's discussion, not a validated school-level conclusion.
• Verify that Document 3 uses exactly the name of the school authority you provided. If the school belongs to a regional authority, the correct recipient is its executive director, not the local government representative.

Suggested iteration

To adjust the tone of the family letter: "Document 2 sounds too formal. Rewrite it with shorter sentences, no bullet lists, and a tone that acknowledges students' effort before mentioning challenges."

To add a fourth output for social media: "Generate a short message for the school's parent WhatsApp group: maximum five lines, without data that would invite comparisons between class groups, and with a clear invitation to the next parents/guardians meeting."

If the teachers' council needs a presentation: "Convert Document 1 into a five-slide structure: title with the year's focus, main findings, indicators table, hypotheses for discussion, and next steps. Keep the same content and adapt only the format."