Headquarters: Los Angeles, United States
Assessed: Cognitive, personality and skills testing with AI scoring of structured video interviews.
Rubric: version 1. Search protocol applied and all sources accessed: 2026-09-22.
Documented 2, partial 3, absent 3, of eight criteria. This records what is published about the validity and fairness of the scores. It is not an assessment of the product, the model or the underlying science.
| # | Criterion | Level |
|---|---|---|
| C1 | Technical manual or validity report publicly available without a sales conversation | D |
| C2 | Validity coefficients reported with sample size and uncertainty | P |
| C3 | Adverse impact reported by named subgroup | A |
| C4 | Scored construct named and defined, not described only by outcome | P |
| C5 | Evidence of comparability across languages, versions or delivery modes | A |
| C6 | Human oversight documented, with a candidate route to contest a score | P |
| C7 | Independent review named and the reviewer identified | A |
| C8 | Reliability reported, and reported by group | D |
C1. Technical manual or validity report publicly available without a sales conversation
Documented.
Per-test information briefs are published on the open web as direct downloads, for example the Criteria Cognitive Aptitude Test brief at criteriacorp.com/files/Criteria-InfoBrief-CCAT.pdf, alongside a scientific validation page. The company states it "publishes a technical manual covering several hundred validation studies".
Source: https://www.criteriacorp.com/science/testing-validation (accessed 2026-09-22).
Per-test information briefs download without a form. The full technical manual was not reached.
C2. Validity coefficients reported with sample size and uncertainty
Partial.
The method is described: "collect CCAT scores from a sample of incumbents, collect supervisor performance ratings, compute correlation, correct for restriction of range and measurement error".
Source: https://www.criteriacorp.com/science/testing-validation (accessed 2026-09-22).
A validation methodology is published, including the statistical corrections applied. No individual coefficient with its sample size and an interval was found on the pages examined.
C3. Adverse impact reported by named subgroup
Absent.
No public source was found for this criterion by the search protocol.
Pages examined include https://www.criteriacorp.com/science/testing-validation (accessed 2026-09-22).
C4. Scored construct named and defined, not described only by outcome
Partial.
Instruments are named, including the Criteria Cognitive Aptitude Test and Emotify, with content and construct validity described as part of the validation process.
Source: https://www.criteriacorp.com/science/testing-validation (accessed 2026-09-22).
Instruments are named and the kinds of validity evidence are listed. Individual constructs are not defined on the pages examined.
C5. Evidence of comparability across languages, versions or delivery modes
Absent.
No public source was found for this criterion by the search protocol.
Pages examined include https://www.criteriacorp.com/science/testing-validation (accessed 2026-09-22).
C6. Human oversight documented, with a candidate route to contest a score
Partial.
"You get the final say in which candidates progress to the next stage of your hiring process."
Source: https://www.criteriacorp.com/interview/AI-Scoring (accessed 2026-09-22).
A reviewer route. No candidate-facing route to query or contest a score was found.
C7. Independent review named and the reviewer identified
Absent.
No public source was found for this criterion by the search protocol.
Pages examined include https://www.criteriacorp.com/science/testing-validation (accessed 2026-09-22).
C8. Reliability reported, and reported by group
Documented.
"Based on a sample of 97,929 test takers, the Cronbach’s alpha for the CCAT is 0.86."
Source: https://www.criteriacorp.com/science/testing-validation (accessed 2026-09-22).
A named coefficient, for a named instrument, with its sample size, on an open page. It is not broken out by group or by language.
Absent records the outcome of the published search protocol on 2026-09-22. Where a source is supplied that the search did not reach, the cell is re-scored under the corrections policy, with the date of the change recorded and the previous level retained.
Back to the index, the rubric and the methods note