School analytics and reporting: a practical guide for school leaders.

Complete one metric-definition record for every recurring number.

The phrase school analytics can cover everything from a class attendance summary to a multi-year planning model. That breadth creates a common failure: the school starts with whatever a system can chart, then searches for a decision that might fit it. Reverse that sequence. Begin with a recurring school question and work backwards to the minimum trustworthy evidence.

India’s UDISE+ reporting describes school-level online collection, validation checks and verification across administrative levels. That public system is not the same as a school ERP, but it illustrates a useful principle: credible decisions depend on current records and deliberate validation. Read the official UDISE+ 2024–25 report ↗ for the source methodology; this guide does not claim that Schylva submits, reconciles or certifies UDISE+ data.

Before comparing a class, learner or month, write down what counts as expected attendance and how Present, Late, Absent, Leave and Not marked states affect the calculation. Confirm whether the summary counts school days, scheduled periods or maintained records. A percentage without that denominator is not yet an operational answer.

Use a repeatable attendance review sequence.

1. 01

   Fix the scope.

   Choose the class, section, academic period and valid date range before opening a trend or ranking.

2. 02

   Check completeness.

   Identify unmarked days or records and confirm whether the timetable or calendar changed inside the period.

3. 03

   Read the status mix.

   Review the breakdown behind the average so Leave, Late, Absent and Not marked records are not collapsed into one unexplained signal.

4. 04

   Compare like with like.

   Use comparable periods and definitions. Document any event, closure, admission change or operational disruption that affects the comparison.

5. 05

   Verify before follow-up.

   Open the maintained attendance record, involve the owning role and use the school’s approved process before contacting a learner or family.

Start with one selected exam or assessment and confirm that the result set is the intended version. Then separate complete scored entries from missing, absent, withheld or otherwise unscored rows. Record whether the calculation uses raw scores, percentages, grade bands or another school-defined method before comparing students, subjects or periods.

Ask these questions before a result meeting.

• Is this the intended exam, academic year, class and result-set version?
• How many expected entries are complete, missing, absent, withheld or unscored?
• Do subject summaries use the same maximum marks and grading basis?
• Is the cohort large and stable enough for the comparison being discussed?
• What changed in curriculum coverage, teaching time or assessment design?
• Which source record must be corrected before the summary is discussed again?
• Who may see the result and who owns any learner-level follow-up?

The National Education Policy 2020 ↗ discusses assessment as part of learning and development. Schools should still apply their own board, academic and specialist guidance when deciding how a particular result is interpreted or communicated.

Build a small data-quality rhythm around each report.

1. 01

   Before the review

   Confirm the source owner, reporting cut-off, expected records, allowed statuses and any known gaps or corrections.

2. 02

   During the review

   Keep the filter scope and definitions visible. Record questions against the source instead of editing a conclusion around an unexplained number.

3. 03

   After the review

   Assign corrections, record the decision and set the next review date. Do not preserve unnecessary exports or screenshots as parallel records.

Use a purpose-and-access checklist.

• State the school purpose before opening learner-level data.
• Give access only to the authorised role that needs the view.
• Use class or aggregate views when learner detail is unnecessary.
• Use fictional or redacted examples for vendor demonstrations and support conversations.
• Avoid downloading, messaging or duplicating records without a defined need and approved process.
• Document retention, correction, deletion and exit responsibilities in the school’s operating process and vendor agreement.
• Keep an escalation path for mistaken access, disclosure or an unexplained data change.

A practical 30-day starting plan.

1. 01

   Week 1 · Choose four questions.

   Select two attendance and two selected-exam questions that already recur in school operations. Name the decision and owner for each.

2. 02

   Week 2 · Document definitions.

   Record the source, filter scope, numerator, denominator, statuses, exclusions, cut-off and correction path for every metric used.

3. 03

   Week 3 · Test with redacted examples.

   Run complete, incomplete and disputed examples. Confirm that authorised roles can trace the summary back to the maintained record.

4. 04

   Week 4 · Hold and review one meeting.

   Separate observation, interpretation and action. Capture questions, source corrections, decisions and the next review date.

The goal is not to make every school decision numerical. It is to make the evidence path explicit when a maintained record can help: what the school asked, what the record shows, what it cannot establish and who remains responsible for the next step.