NBA Tier-II accreditation demands CO-PO-PSO attainment for every course, every semester. Most institutions compute it manually in spreadsheets before each SAR cycle. GradeLab computes it automatically from graded marks - question-to-CO tagging at exam setup, an interactive CO-to-PO mapping matrix, and SHA-256-hashed results for auditability. Hundreds of hours become zero.
See Criterion MappingEach capability mapped to GradeLab's automated pipeline and the manual legacy process - green means covered, red means a gap.
| Capability | GradeLab (Automated) | Manual / Legacy | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Marks Capture | AI Grading | Attainment Engine | Trends & Reports | Audit Trail | Spreadsheets | Faculty Entry | Cohort Sheets | Manual Recompute | Static PDFs | |
| CO-PO-PSO Attainment | Digital + handwritten | Per-question score | Auto from marks | Level 0-3 per CO | SHA-256 hash | Manual formulas | Per-course sheet | Error-prone | Lost on change | No recompute |
| Per-Student Performance | Per-question trace | Per-student score | Class statistics | Rank & history | Reproducible | Manual pivot | Hand-typed | Inconsistent | Tedious | Static list |
| Continuous Improvement | Cycle data | Consistent scoring | Trend per CO | Before / after | Verifiable | No trend | Anecdotal | No loop | Ad hoc | Snapshot |
| Assessment Blueprint Quality | Blueprint-locked | AI answer keys | OBE-aligned | Health score | Validated | Manual plan | Inconsistent | No standard | Not tracked | Word doc |
| Auditability & Reproducibility | Full trace | Per-question reason | Cached snapshots | Recompute on demand | SHA-256 hash | No hash | Post-hoc | Lost on exit | Fragile | Opaque |
| Bloom's Coverage | Bloom-tagged | AI classifies | Distribution | Blueprint vs actual | Count hashed | Manual count | Inconsistent | No standard | Not tracked | Static chart |
| Multi-Source Marks | Both sources | OCR + rubric | Dual-source | Combined trends | Merged hash | Manual merge | Re-keyed | Mismatch | Risky | Split files |
NBA evaluates individual programs against the Self-Assessment Report (SAR) criteria. The mapping below covers the criteria where GradeLab has the strongest evidential footprint.
Estimated ~15-20% of SAR score - the single sharpest GradeLab fit in any framework
| What NBA Evaluates | GradeLab Feature | Evidence Produced |
|---|---|---|
| CO definition & mapping to POs/PSOs | CO CreationCO LinterAI CO Suggestion | COs with action-verb validation, Bloom levels, and unit linkage - no weak-CO flagging at submission |
| CO-PO/PSO mapping matrix | Matrix EditorStrength Scale 0-3Avg PO Strength | Click-to-cycle strength matrix; average PO strength metric as an accreditation artifact |
| CO attainment from assessed marks | CO AttainmentDual-SourceTarget Config | Per-CO: class avg %, % above target, attainment level 0-3; fed from digital tests AND handwritten exams |
| PO attainment rollup | PO AttainmentDeterministic Hashing | Weighted average of CO levels by CO→PO mapping strength; SHA-256 inputsHash for reproducibility |
| Attainment auditability | Cached SnapshotsRecomputeInputsHash | Cached attainment with hash - re-verify on demand; NBA evaluator can confirm numbers derive from actual marks |
Among the heaviest-weighted criteria - per-student, per-course, per-cohort performance data
| What NBA Evaluates | GradeLab Feature | Evidence Produced |
|---|---|---|
| Course-wise success rates & distributions | Exam ReportGrade DistributionClass Statistics | Mean, median, std dev, pass rate, distribution in 10% buckets - per exam, per course |
| Per-student performance tracking | Student PerformanceStudent Detail | Per-student scores and ranks; identity cards with avg score, assigned exams, results history |
| Item analysis & question quality | Item AnalysisPer-QuestionDiscrimination Index | Per-question avg score, attempt rate, discrimination index, difficulty classification |
NBA requires demonstrated closed-loop action - not just measurement, but visible intervention
| What NBA Evaluates | GradeLab Feature | Evidence Produced |
|---|---|---|
| Documented improvement actions | Continuous ImprovementBatch Comparison | Improvement notes per course/CO with action, owner, target cycle, status - structured process record |
| Trend evidence across cycles | CO Attainment TrendsLearning InsightsAI-vs-Teacher | CO attainment across assessment batches; "CO-3 improved from 1.8 to 2.4 after syllabus revision" - visible, auditable |
| Bloom's coverage across assessments | Bloom CoverageBloom Distribution | Question count per Bloom level across digital + handwritten assessments; blueprint vs actual comparison |
NBA values question papers designed against defined outcomes, not improvised per semester
| What NBA Evaluates | GradeLab Feature | Evidence Produced |
|---|---|---|
| Question paper CO/Bloom alignment | Blueprint SystemOBE AlignmentAssessment Health | Blueprint-validated papers; CO/Bloom/difficulty/time scores per paper before finalization |
| Answer key & rubric quality | Answer Key GenGrid RubricWeighted Rubric | AI-generated answer keys with validation; criteria×levels grid rubrics for complex questions |
CO-PO attainment computation that took hundreds of hours per SAR cycle now happens automatically - zero spreadsheets, zero manual formulas
~18% of SAR Score
CO-PO attainment criterion weight covered automatically
SHA-256 hashing proves attainment numbers derive from actual graded marks - recompute on demand, verify on demand
100%
Reproducible on demand
SHA-256 inputsHash
Cached snapshots
Recompute verified
Dual-source marks
CO attainment trends across cycles, Bloom coverage, and documented improvement actions - all visible, all auditable
CO-3: 1.8 → 2.4
Improved after syllabus revision
Questions tagged to COs, Bloom levels, and units - once, at test creation or paper generation time
Rubric-anchored scoring with per-question reasoning - dual-source from digital and handwritten exams
Per-CO attainment computed from marks: class avg %, % above target, attainment level 0-3
Weighted by CO→PO mapping strength. SHA-256 inputsHash for reproducibility and auditability
Attainment CSVs, continuous improvement records, Bloom coverage - formatted for SAR submission
Questions tagged to COs, Bloom levels, and units - once, at test creation or paper generation time
Rubric-anchored scoring with per-question reasoning - dual-source from digital and handwritten exams
Per-CO attainment computed from marks: class avg %, % above target, attainment level 0-3
Weighted by CO→PO mapping strength. SHA-256 inputsHash for reproducibility and auditability
Attainment CSVs, continuous improvement records, Bloom coverage - formatted for SAR submission
GradeLab tags questions to Course Outcomes (COs) at exam creation time. When grading completes - whether digital tests or handwritten exams graded by AI - the system computes per-CO attainment (class average %, % above target, attainment level 0-3) directly from the marks. PO attainment is then rolled up as a weighted average using the CO→PO mapping strength matrix. No spreadsheets, no manual formulas.
Every attainment computation produces a deterministic SHA-256 hash of its inputs - the marks, CO mappings, and weighting configuration. This means an NBA evaluator can recompute attainment on demand and verify the numbers match exactly. It proves the evidence derives from actual graded marks, not post-hoc spreadsheet fabrication.
Yes. GradeLab grades digital tests natively and also grades handwritten answer sheets using AI OCR + rubric enforcement. Both sources feed into the same CO attainment pipeline, giving you dual-source attainment without manual data merging.
GradeLab classifies questions by Bloom level (Remember, Understand, Apply, Analyze, Evaluate, Create) and tracks the distribution across all assessments - digital and handwritten. You get a blueprint-vs-actual comparison showing question count per Bloom level, ensuring your assessment papers cover the cognitive levels NBA expects.
No. GradeLab provides the CO-PO attainment computation infrastructure and the evidence artifacts NBA evaluators assess. Final accreditation decisions depend on peer-team judgment across all SAR criteria, including faculty qualifications, facilities, and research output - areas GradeLab does not touch. We make the attainment evidence piece auditable and automatic.
Let us show you what an NBA-ready attainment export looks like - generated automatically from a live exam cycle, hash-verified and SAR-formatted.
Save 100+ hours per month
AI-powered accuracy & consistency
Instant student feedback
Easy LMS integration
