Assessment platform

Useful now. Defensible later.

Built for training providers, awarding bodies and employers. Criterium covers the whole path, from question bank to released result, and keeps a record that still answers questions years later: which paper, which answers, who marked it, and when it was released.

See what it does Sign in
Candidates measured against a fixed standard the criterion: a fixed standard candidates performance meets the standard not yet, with evidence of why
Each result is a judgement against the standard, independent of how anyone else performed. Results stay comparable across sessions, cohorts and time.

The platform

What Criterium does

Authoring, delivery, marking and reporting in one place. Everything below is included in the Standard edition.

  1. Authoring and versioning

    Fixed question lists or randomised draws from banks and pools. Every edit makes a new version; old versions are retired, never deleted.

  2. Seven question types

    From single best answer to ordering and matching, with partial credit, numeric tolerances and units, and mathematical notation.

  3. Sessions and delivery

    Papers freeze before delivery and stay pinned to the version sat. Preview a session exactly as a candidate would see it.

  4. Candidate experience

    Entry by access code, one question at a time, autosave, a help button, and full keyboard alternatives to drag and drop.

  5. Marking and results

    Queues for human marking, automatic marking elsewhere, and a governed release path: approve, release, withhold, recall. Every step is audited.

  6. Live invigilation

    A live dashboard of candidate state, elapsed time, help requests and time adjustments. Supervision tooling, not automated proctoring.

  7. Reporting and analysis

    Live reports recompute on every view. When a figure needs quoting later, freeze it: a snapshot records exactly what the report said. Every chart sits beside a table with the same figures.

  8. Confidence-aware marking

    Candidates can rate each answer Certain, Unsure or Guessing. Reports turn this into review signals that show where to teach next. It never changes a mark.

  9. Pre and post measurement

    Cohorts carry durable pre-course and post-course roles, so improvement is measured on matched candidates, with honest caveats built in.

  10. Imports

    Bring existing question stock in from CSV or Word, with a preview stage that reports warnings and errors before anything is saved.

Governance

Built to answer for itself

An assessment result should answer one question: can this person do this thing, to this standard? Everything about how Criterium is governed exists so that the answer still holds up later, for a candidate, an appeal panel or a regulator.

The server decides

Every screen is set by the user's role and the state of the assessment, enforced on the server. Marks, timings and access never depend on trusting the browser.

Evidence is written once

Responses, marks and moderation decisions are recorded with full provenance. A correction never overwrites history; it is added to it, attributed and timestamped.

Explicit states

What a candidate sits, and what a result says, is always a published, governed version. Nothing reaches a candidate from draft.

Ready for the audit

Who marked which response, when, under which version of the mark scheme, and what changed afterwards. The record is there before anyone asks.

The lifecycle the system enforces

  1. Draft

    Authored and revised freely. Nothing in draft reaches a candidate.

  2. In review

    Checked against the standard. Any change sends it back to draft.

  3. Approved

    Fixed and versioned. What was approved is what will be delivered.

  4. Published

    Live for delivery. Scripts, marking and results reference this exact version, permanently.

The same discipline runs downstream. Scripts, marking, moderation and result release each have their own governed states, so a session can always answer: where is this, and who moved it there?

Accessible and accountable by default

Accessibility

Built to WCAG 2.2 AA and held to AAA where practical for candidates. Multiple display themes available.

Privacy

At runtime the platform names no external origin at all; nothing phones home. Subject-access and erasure tooling are standard.

Security

Two-step verification, encrypted access codes, and a full audit trail across authoring, delivery, marking and release.

Access

Four roles, four bounded views

Access is role-based and strictly scoped. Each person sees exactly what their role needs, and no more.

Administrator

Authors assessments, governs their lifecycle, sets up sessions, and manages the organisation's question banks and reporting.

Invigilator

Runs live sessions: admits candidates, monitors attempts in progress, and handles interruptions with a recorded, auditable trail.

Marker

Marks responses against the published mark scheme, records feedback, and takes part in moderation. Every judgement is attributed.

Candidate

Joins with a short access code, sits the assessment in a focused environment, and receives results when they are formally released.