Skip to content
← All Assessments
Cognitive

Cognitive Ability Test for Hiring

The single strongest predictor of job performance. HeyHRM's cognitive assessment tests four domains — numerical, verbal, matrix, and spatial reasoning — with 40 questions from a validated 120-item bank.

What It Measures

## What This Assessment Measures Cognitive ability is still the cleanest single predictor of job performance we have. Schmidt and Hunter's classic meta-analysis (1998), later extended by Schmidt, Oh, and Shaffer (2016), showed that general mental ability consistently predicts learning speed, training success, and on-the-job performance better than most of the nonsense companies still rely on. In plain English: smarter candidates ramp faster, solve harder problems, and need less managerial babysitting. HeyHRM's Cognitive Ability Assessment measures four domains because hiring teams do not need one vague “IQ-ish” number. They need a working map of how a candidate processes information. **Numerical reasoning** measures quantitative judgment: reading tables, checking assumptions, calculating changes, and spotting bad math before it becomes an expensive decision. **Verbal reasoning** measures written comprehension and inference: understanding policy language, making sense of dense documentation, and identifying what is actually being said versus what someone wishes had been said. **Matrix reasoning** measures abstract pattern recognition and fluid intelligence. This is the strongest signal for learning novel systems quickly, especially in roles where the playbook changes every quarter. **Spatial reasoning** measures mental manipulation of objects, systems, or structures. Lubinski and colleagues' longitudinal work on spatial ability showed that this dimension matters far more for technical and STEM performance than many employers realize. ### Why these four domains matter together A lot of hiring teams over-index on one cognitive trait and call it a day. That is lazy. A RevOps analyst may need strong numerical reasoning but only moderate spatial reasoning. A product designer may need high spatial and matrix reasoning. A customer-facing manager may lean harder on verbal reasoning and judgment. By separating the domains, HeyHRM gives employers a role-relevant profile instead of a blunt score. ### Research-backed interpretation - **General mental ability** predicts job performance across occupations, with stronger effects in more complex jobs. - **Structured, job-related testing** tends to outperform unstructured interviews because it reduces impression bias. - **Multi-method assessment batteries** perform best when cognitive ability is combined with personality and skills evidence. - **Benchmarking matters**. A score only becomes useful when compared against similar roles and seniority bands. ### What employers actually learn from the results A high cognitive score does not mean “auto-hire.” It means the candidate is likely to learn faster, process complexity better, and handle ambiguity with less friction. A lower score does not mean “bad candidate.” It means the role fit needs more scrutiny, especially if the job requires fast pattern recognition, analytical judgment, or learning dense systems under pressure. ### Detailed scoring methodology HeyHRM reports four domain scores plus a composite cognitive index. Each section is percentile-normalized against relevant peer groups. The recommended default weighting for broad knowledge-work roles is: - **30% numerical reasoning** - **30% verbal reasoning** - **25% matrix reasoning** - **15% spatial reasoning** For engineering, analytics, and technical design roles, the system can increase matrix and spatial weighting. For management, consulting, or policy-heavy roles, verbal reasoning can carry more weight. Speed and accuracy are scored separately before being blended, because a candidate who is fast but sloppy creates different risk than one who is careful but slow. ### When to use this assessment Use cognitive ability testing early when the role requires complex learning, high-stakes decisions, data interpretation, or cross-functional ambiguity. Use it alongside skills and personality data when you want the highest signal. Do not use it like a magic oracle. Use it like a sharp instrument.

How It Works

## How It Works **Step 1: Start with the role, not the test.** Paste the job description and HeyHRM identifies the cognitive demands of the position. A controller role gets heavier numerical and verbal weighting. A UX role gets more spatial and matrix emphasis. A general leadership role blends all four. **Step 2: Candidates complete a browser-based assessment.** No downloads, no friction, no consulting project. Candidates move through a calibrated mix of reasoning questions designed to surface both baseline capability and upper-bound potential. **Step 3: HeyHRM scores performance in real time.** Recruiters see four domain scores, a composite percentile, speed-accuracy interpretation, and guidance on how to use the result in the rest of the hiring process. **Step 4: Results feed the FIT Score.** Cognitive output is combined with [personality](https://heyhrm.com/assessments/personality), skills, and custom AI scenarios so teams can rank candidates with a full-picture view rather than one isolated number. ### Six sample questions with explanations 1. **A revenue dashboard shows monthly growth of 8%, 12%, and -4%. What is the main conclusion?** Best answer: growth is positive overall but volatile, so trend quality matters more than one headline number. Explanation: strong candidates distinguish pattern from snapshot. 2. **A policy says managers must approve all expenses above $1,000. An employee submits three receipts of $450, $420, and $250 from the same trip. What should happen?** Best answer: aggregate the trip spend and review the intent of the policy rather than treating each line item in isolation. Explanation: verbal reasoning is about meaning, not just words. 3. **If sequence A rotates 90 degrees clockwise each step, what should the fourth image show?** Best answer: the correctly rotated continuation of the pattern. Explanation: matrix and [spatial reasoning](https://heyhrm.com/assessments/spatial-reasoning) detect structural consistency. 4. **A team of 5 completes 60 tickets in 3 days. At the same rate, how many tickets will 8 people complete in 5 days?** Best answer: 160. Explanation: proportional reasoning surfaces quickly under light time pressure. 5. **A memo says churn improved from 8% to 6%. What is the best interpretation?** Best answer: churn declined by 2 percentage points, not 2 percent. Explanation: precision beats hand-wavy math. 6. **A candidate solves easy questions rapidly but slows sharply on harder pattern problems. What does that suggest?** Best answer: solid baseline processing with a possible ceiling on abstract complexity. Explanation: score patterns matter as much as totals. ### Competitor comparison: HeyHRM vs SHL, TestGorilla, Criteria Corp, HireVue, Testlify **SHL** has depth and legacy credibility. It also tends to feel heavier, more enterprise-first, and less flexible for mid-market teams. **TestGorilla** is fast and easy to deploy, but fixed libraries can make differentiation and role calibration thinner. **Criteria Corp** is respected for aptitude testing, especially for standardized cognitive batteries, but many teams still need separate workflow, personality, and AI interpretation layers. **HireVue** is more interview-centric. Useful in process design, but video interview data is not a replacement for cognitive measurement. **Testlify** is broad and accessible, though breadth alone does not create a better prediction model. **HeyHRM** wins on one thing buyers actually care about: integrated interpretation. The platform blends cognitive ability, personality, skills, and role-specific AI scenarios into a single decision workflow, instead of forcing hiring teams to juggle five score reports and pretend that counts as rigor.

Sample Questions

1. A report says revenue increased from $5.0M to $5.8M. What is the percentage increase?

2. A policy requires approval for any vendor spend above $2,500 per quarter. A team submits two invoices for $1,600 in the same quarter. What should happen?

3. If a visual sequence rotates 90 degrees clockwise and adds one dot each step, what should the next figure show?

4. A team of 4 completes 48 tasks in 6 days. At the same rate, how many tasks can 6 people complete in 5 days?

5. A memo says “customer complaints dropped from 12% to 9%." Is that a 3% decrease?

6. Why pair a cognitive test with a structured interview?

Frequently Asked Questions

What is a cognitive ability assessment?
A hiring assessment that measures how well a candidate reasons through numerical, verbal, abstract, and spatial problems tied to workplace performance.
Is this the same as an IQ test?
No. It is designed for hiring, benchmarking, and role relevance rather than clinical classification.
How long does it take?
Usually 20-35 minutes depending on whether you deploy the full battery or selected domains.
Which roles benefit most?
Roles with high learning demands, analytical complexity, or ambiguous decision-making benefit most.
Is it legal in hiring?
Yes, when it is job-related, consistently administered, and used as part of a structured process.
Can I test one domain only?
Yes. Employers can deploy numerical, verbal, matrix, or spatial reasoning separately.
How do you score it?
Candidates receive domain scores, percentiles, and a composite index with role-aware weighting.
Can candidates prepare?
They can practice reasoning formats, but randomized and calibrated items reduce rote memorization gains.
How does HeyHRM compare to SHL or Criteria?
Those tools are credible and established. HeyHRM adds integrated workflow, AI interpretation, and role-specific weighting.
Why combine this with interviews?
Because the test shows likely strengths and risks, while the interview validates them in context.

Ready to assess candidates?

Start screening with Cognitive Ability Assessment today. Free to get started.

Get Started Free

Related Assessments

Browse All Assessments →

From the Blog