Results
Per-student and per-question analytics for online exams — class average, accuracy heat maps, Bloom/topic breakdowns and comparative views.
What this screen shows
The Results tab turns submissions into insight: per-student scores, per-question accuracy, class average, high and low, and topic / Bloom heat maps. Objective answers update as students submit; subjective answers appear once graded — each can carry AI feedback (a marks-based verdict with partial credit, the mistakes found, improvement tips and a handwriting transcription) that is held until you review it (the student sees "Waiting for teacher review" until you confirm or override, then the feedback unlocks with a "Teacher reviewed" badge). A Likely Misconceptions panel flags questions where one wrong option trapped much of the class, a Time-vs-Score quadrant plots effort against accuracy, and a ranked student list benchmarks each student against peers. When an exam allows multiple attempts, a Counting selector chooses whether the highest, latest or first attempt feeds every figure and the heat map. You can drill from the class view down to a single student’s attempt, grade subjective answers, and compare across batches.
When you open a question’s solution while reviewing an attempt, an Interactive Simulation button offers an AI-generated, explorable visual of the underlying concept — a quick way to show why an answer works. See Interactive Simulations for how they work.
Common workflow
Pick the exam whose results you want to inspect (search or filter the list).
Read the class summary — average, high, low, completion.
Open per-question analytics to find the items most students missed.
Switch to the topic / Bloom heat map to see where the cohort is weak.
Drill into a student to review their full attempt and grade any subjective answers.
Use comparative views to benchmark batches, then export.
Controls on this screen
Find the exam whose results you want to analyse.
Scope by class, batch, subject or date.
Average, high, low and completion for the selected exam.
Accuracy per item — find the questions that tripped the cohort up.
Per-chapter / per-topic accuracy across the cohort.
Performance split by Bloom level and difficulty.
Drill into one student’s full attempt and grade subjective answers.
Benchmark students, classes and batches side by side.
On multi-attempt exams, choose which attempt counts — Highest (default), Latest or First. The whole page and the heat map recompute attempt-accurately.
MCQs where a single wrong option captured ≥35% of responses — a shared wrong idea worth reteaching, ranked by how many it trapped.
A quadrant chart of each student’s minutes spent against their percentage — spot fast-and-accurate vs slow-and-struggling at a glance.
The student list ranks every attempt against its peers (with medals for the top three) so each result reads in context.
Surface AI-suggested next steps for weak topics.
Export results as CSV for record-keeping.
Exam pattern & percentile
A mark out of 300 tells a student very little on its own. What they actually want to know is where do I stand — and for anyone preparing for a competitive exam, that means a percentile worked out the way the real exam works it out. So an exam carries an Exam Pattern, chosen with the pattern picker when you build it, and that choice decides the arithmetic used on every result it produces. The patterns are JEE Main, JEE Advanced, NEET and CUET, plus General — which is not a card you click but what every other exam falls back to, including ones built on the CBSE Term or Custom presets. General, JEE Main and CUET use the NTA method — your percentile is the share of candidates who scored at or below you, so the topper lands on exactly 100 and tied students share a value. NEET uses its own rank-based formula. JEE Advanced deliberately has no percentile at all, because the real exam publishes ranks only; those results show rank instead, and nothing is invented to fill the gap.
Once an exam has a pattern, the figure follows the result everywhere it goes: - The student’s result page — a Class rank tile with the percentile beneath it and a "Top x%" reading, a Class score distribution bar chart with their own score band highlighted so they can see the shape of the cohort, and — once they have sat two or more exams — a Percentile trend sparkline next to their marks trend. On JEE Main and NEET exams a ≈ AIR estimate appears too: what their All-India Rank would be if this percentile held in the real exam. - Section tables — multi-section exams add a Percentile column so a student can see they are strong in Physics and weak in Chemistry within the same paper. - Teacher leaderboards — under Compare & Ranks, the per-exam Leaderboard ranks every counted attempt with its percentile. - Overall ranking — the same screen’s cross-exam view ranks students across every exam in scope, using one counted attempt per exam. - Parent surfaces — the parent dashboard shows the child’s percentile and class rank, and the weekly digest email reports their cohort percentile plus the change since last week. - Result PDF and email — the exported report and the emailed result both carry rank and percentile, marked (final) once the exam has closed. - Offline OMR — scanned bubble-sheet results are ranked with the same percentile method, so paper and on-screen sittings read alike.
Set when you build the exam, in Settings → General → Exam Pattern. Picking JEE Main, JEE Advanced, NEET or CUET switches in that exam’s own formula. There is no "General" card to click — General is simply what you get when you pick none of those, or pick CBSE Term or Custom.
The student’s position in the cohort with their percentile to two decimals, plus a plain-English "Top x%".
A histogram of the whole class with the student’s own band highlighted — "Highlighted bar = your score band".
A sparkline of percentile across the student’s recent exams, shown once at least two of them have one.
On JEE Main and NEET patterns, an estimated All-India Rank derived from the percentile and the real exam’s candidate pool. Present it as motivation, not a forecast — the app labels it "(est.)" for exactly that reason.
Leaderboard & Overall ranking
In Compare & Ranks: one exam’s ranked table, and a cross-exam ranking that averages a student across every exam in scope.
Percentile is always relative to who has actually sat the exam so far, so it moves while an exam is still open and settles when it closes — that is why the reports mark the closed figure (final). With only a handful of attempts in, the number is technically correct but not yet meaningful; wait for the cohort before you share it.
Pick the pattern to match the exam you are simulating, not the one that flatters. A NEET-pattern percentile on a class test of 30 students is arithmetically sound and educationally misleading — students read it as their standing among lakhs of candidates.