Cognitive Ability Test

What cognitive ability tests measure

Most tests combine several question types:

• Verbal reasoning: reading a short passage and judging which statements follow from it.
• Numerical reasoning: working with percentages, tables and rates, often in a business context.
• Abstract or logical reasoning: finding the rule in a sequence of shapes or patterns.
• Attention to detail or error checking: comparing records and spotting mismatches, common for data entry and back-office roles.

Scores across these areas tend to correlate, which is why researchers talk about general mental ability (GMA). Tests are usually 10 to 30 minutes, timed and delivered online. Some are adaptive, getting harder or easier based on previous answers.

Cognitive tests are one family of psychometric tests, alongside personality and integrity measures, and one of many options in a pre-employment assessment program.

How predictive are they?

For decades, cognitive ability was cited as the single best predictor of job performance, based on large meta-analyses from the 1990s. A 2022 re-analysis by Paul Sackett and colleagues corrected some statistical assumptions in that earlier work and found lower validity for cognitive tests than previously reported, with structured interviews coming out ahead. Cognitive ability still predicts performance, especially in jobs with heavy learning demands, complex information or frequent problem-solving, but it is better seen as one useful input than as the best tool for everything.

In practice that means: a short reasoning test makes sense for roles like claims processing, technical support or financial operations, where people must learn complex rules quickly. It adds less for roles where reliability, physical work or customer warmth matter more than speed of learning.

Adverse impact and the law

Cognitive tests are the assessment type most associated with disparate impact. Average scores differ across racial and ethnic groups in the research literature, so a high cut score can screen out protected groups at much higher rates.

The landmark case is Griggs v. Duke Power Co. (1971), in which the US Supreme Court held that the company's aptitude test and diploma requirements violated Title VII because they excluded Black applicants at higher rates and had not been shown to be related to the jobs. Under the Uniform Guidelines on Employee Selection Procedures (1978), employers should monitor selection rates, and a rate for any group below four-fifths of the highest group's rate is generally treated as evidence of adverse impact. A test with adverse impact must be shown to be job-related and consistent with business necessity, and the employer should consider less discriminatory alternatives. The Civil Rights Act of 1991 also bans adjusting scores or using different cutoffs by race, sex or other protected group.

Accommodations also matter. Under the ADA, candidates with disabilities such as dyslexia may need extra time or a different format. Rules vary by state and country, so review your use with counsel.

Using cognitive tests responsibly

1. Do a job analysis first. Document which tasks require reasoning and learning speed. That is your evidence of job-relatedness.
2. Choose a validated test whose publisher can share technical documentation, reliability data and group difference data.
3. Set cut scores carefully. A modest minimum screens out people unlikely to cope with training without excluding large numbers of qualified candidates. Avoid simple top-down ranking on the test alone.
4. Combine with other measures. Pairing a cognitive test with structured interviews, a work sample or a situational judgment test tends to improve prediction and reduce adverse impact compared with the test alone.
5. Monitor results by group at least annually and after any change.
6. Watch for cheating in unproctored online tests. Re-test finalists briefly in a supervised setting if stakes are high.

Common mistakes

• Using the same test and cut score for every role in the company.
• Setting the bar at "top 20 percent" because it sounds selective.
• Testing for roles where reasoning speed is not actually needed, such as many warehouse picking jobs.
• Ignoring candidate drop-off. A 30-minute timed test at the start of an hourly application can lose a large share of applicants. Place it after a short screen.