Inspect the methodology behind your assessment results

Read how Codility assessments are designed, scored and checked for fairness, and have our assessment scientists validate them on your data.

Validated assessments SOC 2 Type II ISO 27001

2009
The year Codility started assessing engineers. Almost two decades of assessment science since.
1,300+
Validated tasks, each scored against a documented method.
9 in 10
Candidates who say the content is fair. 48,000+ candidate responses, October 2025 to April 2026.

Trusted by teams at

Citi Tesla Accenture SAP Deloitte BMW Group SpaceX Deutsche Bank GitHub Bain & Company American Express McLaren Racing LSEG Microsoft KPMG Halliburton EY Barclays Skyscanner ABB Vinted HSBC PwC Palo Alto Networks TD Broad Institute Publicis Groupe Zalando ING NIQ Nissan Pearson Korn Ferry

What the methodology covers

Occupational psychologists design every task and subject-matter experts review it. The 92-page Technical Manual documents the method, and fixed rules score the work.

  • The Technical Manual follows the Standards for Educational and Psychological Testing, the benchmark psychologists use to judge whether a test is valid and fair
  • Engineering Skills Model 2.1: 30 skills and 180 subskills across five categories, validated by engineering leaders
  • Mapped to SWEBOK v4, SFIA v9, O*NET, the NIST AI Risk Management Framework and ISO/IEC 42001
  • Every score traces back to a named skill in that model
  • Coding tasks run against test cases, and multiple-choice items score against a key
  • Business tasks score against an expert-defined answer space
  • In a live interview, the interviewer scores against a rubric
  • Each expert review is recorded, and no AI model grades the response

Source: Codility Technical Manual, current edition. Engineering Skills Model 2.1, 2026.

Read the Technical Manual

Skills Intelligence proficiency profile for a sample employee showing assessed engineering skills rated from novice to expert.

Fairness is measured at every cut score

For the groups we measure, adverse impact ratios meet the EEOC four-fifths threshold at every measured cut score. The Technical Manual documents the method and analysis for legal review.

Language is checked as a source of unfairness in its own right. An independent linguistic audit by cApStAn found 65% of tasks at or below level B1 on the Common European Framework of Reference, the CEFR scale, so candidates working in a second language spend their time on the skill being assessed.

Customer-specific analysis uses demographic data your organization supplies. Codility never infers demographics from names, locations or other signals.

Source: Codility Technical Manual, adverse impact analysis. cApStAn independent linguistic audit of the Codility task library.

Validate the assessment for your roles and your pass mark

Standard content validation documents each task’s design, expert review and skill mapping in the Technical Manual, available before purchase.

A local validation study tests whether an assessment supports your decision for your population and pass mark. Professional Services and our assessment scientists run the study with your team and document the pass-mark rationale for candidate, regulator or client review.

Professional Services also covers taxonomy mapping, rollout design, pass-mark governance and end-of-campaign analysis of adverse impact and assessment health.

Recognized by the assessment profession

The International Personnel Assessment Council’s Innovations in Assessment Award 2026 went to the Skills Intelligence proficiency scale and scoring methodology, for the paper “Applying Psychometric Science for Precise Technical Assessment.” James Meaden received the award at the council’s annual conference on July 14, 2026. The award attaches to the scoring methodology: the scale that places an engineer’s assessed skill on a proficiency level, from candidate to employee.

Source: International Personnel Assessment Council, Innovations in Assessment Award 2026, presented July 14, 2026.

International Personnel Assessment Council

Innovations in Assessment Award 2026

Applying Psychometric Science for Precise Technical Assessment

Received by James Meaden, Assessment Scientist

Presented July 14, 2026

Candidates rate the assessments as fair

91%

of candidates say the assessment content is fair

48,000+ candidate responses

90%

agree the assessment gave them a fair chance to show their abilities

46,204 responses

93%

of women candidates rate the assessment content as fair

9,007 responses

83%+

rate the content as fair in every ethnic group surveyed

40,688 responses

Source: Codility post-assessment candidate feedback survey, October 2025 to April 2026.

Questions your reviewers ask

Review the methodology with your assessment, legal and hiring leads

We bring the assessment science team.