Structured question bank
Item ID, domain, competency, intended difficulty, version, answer key and review status.
ASSESSMENT & TEST INTELLIGENCE · BY SIGMALAB
From responses to defensible assessment decisions.
Assessment Intelligence structures the question bank, analyses how each item performs and converts results into auditable recommendations—with a human making the final decision.
Item-level analysis
Human review
Traceable reporting
assessment-intelligence / review-boardSYNTHETIC DATA
ANALYTICAL SAMPLECertification Assessment
Analysis completeBank blueprint306
Analysed sample24
Attempts72
Review signals9
AI-009
REVIEW SIGNALLow discrimination · human investigation
HIGHFROM SCORING TO INTERPRETATION
It does not only show who answered correctly. It shows where the assessment itself requires review.
The value lies in connecting candidate performance with item quality, content coverage and the decision-making process.
Item analysis
Facility, discrimination, omissions, response time and distractor performance.
Cohort view
Score distribution, domain performance, pass-threshold sensitivity and reliability.
Candidate diagnostics
Identify areas for review and create an adjustable study plan.
Recommendation register
Evidence, interpretation, proposed action, owner and verification step.
Human decision
Metrics identify signals. A person approves whether an item is retained, revised or retired.
CONTROLLED METHODOLOGY
A traceable chain from data to recommendation.
No single metric makes the decision. Statistical signals are reviewed alongside content validity, answer-key accuracy and operational context.
- 01
Structure
Audit the bank, metadata, versions and data quality.
- 02
Analyse
Calculate metrics under a pre-approved analysis plan.
- 03
Investigate
Review signals with subject-matter experts and stakeholders.
- 04
Recommend
Document the proposed action and human approval.
- 05
Verify
Monitor the revised item in a subsequent test window.
EVIDENCE TRANSPARENCY
Functional prototype and methodology demonstrator—with explicit boundaries.
Sample bank, diagnostic quiz, mock exam, readiness dashboard, flashcards and study plan.
Item analytics, cohort analysis, KR-20, flags and a recommendation report using synthetic data.
Official EOPPEP commissioning, production deployment, a large real dataset or measured performance improvement.
Institutional clarification: the use case relates to an EOPPEP certification context. This is an independent SigmaLab demonstrator and is not an official EOPPEP system, commission or endorsement.
WORKING DEMO · EL / EN
Explore the bank, indicators and decision report.
The demo uses exclusively fictional questions and deterministic synthetic data for 72 candidates.
Open Assessment Intelligence