CAAT Exam Risk Dashboard

Marigold Aviation

Academic Performance Management System

APMS · CAAT Exam Risk — Phase 1 (interim build, exam outcomes only)

Pass mark %

Protected data loaded

Hardest papers

First-cut targeting for tutorials & curriculum review — lowest pass rates first.

Does resitting help?

If the resit bar isn't clearly higher than first-attempt, remediation between attempts isn't working.

MIS weekly summary

Auto-generated management note for the weekly academic / MIS review.

Top 3 cadets to address — MIS priority

Highest-priority support cases from the intervention model. Review these first in the weekly MIS / academic meeting.

Priority	Cadet	Batch	Progress	Open	Days	Action focus

Risk distribution

How cadets fall across the RAG bands under the current pass mark. Bands are rule-based and explainable — see method below.

How a cadet's risk band is computed

Computed from exam data only (this is Phase 1). For each subject a cadet has sat, the latest-dated attempt decides current status — a subject failed once but later passed counts as cleared.

Red — 5+ papers ever failed, or 5+ resits.
Orange — 3–4 papers ever failed, or 3–4 resits.
Green — fewer than 3 papers ever failed and fewer than 3 resits.
Awaiting exams — no results recorded yet; deliberately not flagged as failing.

This is a transparent index, not a calibrated probability. Treat a band as a prompt for a conversation, not a verdict.

Subject league table

Pass rate, mean score and volume per paper. n shown so you can judge how much weight to put on each.

Subject	Pass %	Mean	Results	Clear ≥80	Marginal 75–79	Fails

Subject performance across batches

Pass rate for every paper in every batch — read across a row to compare how the batches do on that subject. Hover a cell for the sample size; cells under n=5 are dimmed (too few to trust).

≥85%70–84% 60–69%<60% dimmed = n<5

Cadets flagged for support

Click any cadet for the full breakdown. Sorted by risk, then papers ever failed. Cadets who have cleared all papers are excluded here and listed on the Completed tab.

Risk	Progress	Trend	Cadet	Batch	Failed	To clear	Cleared	Resits	Pace	Why flagged	Next best action

Hard-stop watch — deadline risk

Separate from academic Red. Flags cadets who still have uncleared exams and whose CAAT hard-stop is near or already passed — a scheduling / regulatory risk, not an academic-performance one. A cadet can appear here while being academically Green, and vice-versa.

Flag if hard-stop is within days (or already passed) and ≥ 1 exam is uncleared

Cadet	Batch	Hard-stop	Days	To clear	Cleared	Ever failed	Academic band

Intervention tracker

Prioritised action list generated from progress, trend, open papers and hard-stop pressure. Use it as the weekly support meeting queue.

Priority	Cadet	Batch	Progress	Trend	Open	Days	Support action	Review trigger

Progress distribution

How examined cadets are moving through the programme, independent of their historical risk band.

Trend quality

Trend uses repeated sittings only. A cadet with no resits yet shows as no trend.

Completed cadets — cleared all papers

Cadets who have cleared (passed) every CAAT paper. They are removed from the At-risk list. Any Red/Orange shown here is retained from their historical failure / resit record — it does not mean an open academic risk.

Risk	Cadet	Batch	Ever failed	Resits	Marginal	Cleared	Flag history

Batch comparison

Pass rate by batch. Small batches swing hard — n is shown; don't over-read a batch under ~40 results.

Within a batch — strongest to weakest paper

Pick a batch to see which papers it handles well and which it struggles with, ranked best first.

Failure depth by batch

Cadet–subjects grouped by how many sittings were failed, worst first. A "4 failed" case means all four permitted sittings were failed — the critical tier. Counts are cadet–subject combinations, not cadets.

4 failed3 failed 2 failed1 failed

Cadet detail

Per-subject current status, full attempt history, and pace against the hard-stop.

Cadet scorecard

Standardised printable scorecard correlating the cadet's latest CAAT exam record (live, from this dashboard's data) with their static ADAPT/TNA markers. Pick a cadet, review the draft verdict/actions, and print to PDF.

Load ADAPT/TNA capture .xlsx ADAPT markers not loaded

How at-risk cadets are identified

A plain-English account of the method behind every flag on this dashboard. Read it before acting on a band — the flag is a prompt for review, not a verdict.

The pipeline. Exam scores and dates are entered in the live sheet. APMS flattens that into one record per cadet–paper–attempt, then for each paper it takes the latest-dated attempt to decide that paper's current status — so a paper failed earlier but later cleared no longer counts against the cadet. From each cadet's papers it computes four signals, applies the fixed rules below to assign one colour band, and lists the reasons in plain words. It recomputes whenever the data refreshes or the pass mark is changed.

The signals

Paper ever failed ("failure") — a paper the cadet has failed at least once, whether or not it was later cleared. The count of these is a primary band driver.
Resit — any paper sat more than once. The count of these is the other band driver.
Outstanding paper — a paper whose latest attempt is still below the pass mark (currently 75%). Shown as current status; does not itself set the band.
Marginal pass — a pass within 5 marks of the line (e.g. 75–79). Context only.
Behind pace / hard-stop — time elapsed to the hard-stop vs progress. Context only for the academic band (the separate Hard-stop watch tab handles deadline risk).

What each band triggers

Red Education Council review and root-cause analysis.
Orange Academic intervention plan within 48 hours.
Green Continue monitoring.
Awaiting No exams sat yet — deliberately not flagged.

How at-risk cadets are calculated

For each cadet the model reads every CAAT paper they have sat and takes the latest-dated attempt as that paper's current status. It then counts two things across all the cadet's papers:

• Papers ever failed ("failures") — papers with at least one attempt below the pass mark, including papers later cleared on a resit.
• Resits — papers sat more than once.

The academic band is set from whichever of those two counts is higher — first rule met wins:

Band	Assigned when
Red	5 or more papers ever failed or 5 or more resits.
Orange	3 or 4 papers ever failed or 3 or 4 resits (and not already Red).
Green	2 or fewer of each.
Awaiting	No exams sat yet — kept separate, not counted as Green (no data to judge).

This band reflects academic performance only. Because a "failure" counts a paper failed at any point, a cadet who failed several papers but recovered them can still be Red. Deadline risk is handled separately on the Hard-stop watch tab. A cadet is "flagged" when Red or Orange.

How the scorecard is generated & propagated

The pipeline. The scorecard is assembled live when a cadet is selected — nothing is pre-stored. Its CAAT side reuses the very same cadet model that drives every other tab on this dashboard (one record per cadet–paper–attempt, latest-dated attempt sets each paper's status), so a scorecard always agrees with the cohort view and moves automatically whenever the CAAT data is refreshed or the pass mark is changed. Its ADAPT side is read from the separate ADAPT/TNA capture workbook you load once on the Scorecard tab — that file is static, since the underlying reports don't change. The two sources are joined on Batch + Name (the only shared key; the CAAT sheet carries no A-number), the matched values are dropped into the fixed four-section template, and Print scorecard renders that single card to PDF with the rest of the interface hidden.

Section 1 — CAAT parameters (live, computed)

Derived from the cadet's exam record at render time, using the pass mark in the top bar:
• Cleared / Outstanding — papers whose latest attempt is at/above vs below the mark.
• 1st-attempt pass % — share of sat papers passed on the first sitting.
• Resit recovery — of papers re-sat after a first-attempt fail, how many later cleared (papers failed once but never re-sat are excluded from the denominator).
• Mean latest — average of each paper's latest-dated score.
• Risk band — the same red / orange / green / grey band and reason chips used across the dashboard.
• Per-paper table — first score, resit history (e.g. 60 → 70 → 73), latest score, and status (Pass · Marginal · Fail · Pending), with the hard-stop date.

Section 2 — ADAPT / TNA parameters (static, captured)

Read verbatim from the capture workbook row for that cadet — no re-interpretation:
• Bands — Overall suitability, Learning (FAST-derived), Flight Test & Performance, FAST, Coordination, Anxiety, Safety Culture, Abnormal Traits, Culture, Motivation.
• Objective scores — Maths, Physics, Cognitive (+ six sub-scores), English / ICAO.
• Silhouette — target vs candidate code, fit band and %.
• Flags & notes — Flight-Test scenario-adherence warning, coordination score, FAST level, and the TNA flagged development areas.
Anything not captured shows as “—”; a cadet absent from the workbook shows a “load / no match” prompt.

Sections 3–4 and the cadet identity

• Verdict and Actions are auto-drafted from the data (current risk band and reasons; cognitive vs readiness; the failed/pending papers as target list) and are editable in place before printing — they are prompts for the reviewer, not a clinical judgement.
• The card is identified by the cadet's MGA code (from the capture sheet) and batch; the ADAPT A-number is held internally only and is never printed, since it is restricted vendor material.
• Because the join is on name, a scorecard only attaches markers when the cadet's name matches exactly between the CAAT data and the capture sheet — the dropdown selection is also the visual confirmation that the right cadet is being printed.

What this method is — and is not

• This is a transparent, rule-based index, not a statistical probability. It does not predict a percentage chance of passing or failing. A band is a trigger for a conversation.
• It uses exam outcomes only (Phase 1). Attendance, mock tests, study hours and instructor input are not yet captured and so are not part of the score.
• Small samples swing hard — a band built on very few sittings should be read with caution.
• The pass mark is configurable (top bar). All thresholds above move with it; 75% is assumed pending confirmation against the current CAAT regulation.

Source: CAAT_Exam_Results_Live_Sheet.xlsx · Result Sheet · parsed to one row per cadet–subject–attempt. Interim build reads a static export; the live Google Sheet link is the next step once parameters are finalised.