What Is the Predictive Index Cognitive Assessment?

The Predictive Index Cognitive Assessment is a 12-minute, 50-question test of general cognitive ability that more than 10,000 companies use to screen job candidates. Scores fall on a scale from 100 to 450, the average is 250, and reaching that average takes only about 20 correct answers, because almost nobody finishes all 50 questions in the time allowed. The Predictive Index states plainly that the assessment does not measure IQ. It measures the same underlying ability that IQ tests target, speed of learning and comfort with complexity, but reports it on its own scale against a working-adult norm group.

The assessment began life as the Professional Learning Indicator, or PLI, a Danish-built screening test. The Predictive Index acquired it and relaunched it as the PI Learning Indicator in January 2016, later renaming it the PI Cognitive Assessment. Its stated purpose has stayed constant: estimating "how fast an individual can be expected to acquire new knowledge" in a job, which is why employers pair it with the separate PI Behavioral Assessment rather than using either alone.

• Three content areas: The 50 items split into 16 verbal, 18 numerical, and 16 abstract-reasoning questions, mixed together rather than presented in blocks.

• A hard 12-minute limit: Extended-time versions of 18 or 24 minutes exist as accommodations, but the standard sitting is 12 minutes.

• Only correct answers count: There is no penalty for wrong answers or unanswered questions, and the number attempted is irrelevant to the score.

• Different questions for every candidate: The test is not "computer-adaptive"; it uses a "linear-on-the-fly" engine that assembles a unique but statistically equivalent item set for each administration, which is an anti-cheating design rather than a difficulty adjustment.

The scaled score, from 100 to 450, is the only number The Predictive Index intends employers to use. The average of 250 corresponds to roughly 20 correct answers, and the norm group behind the percentiles contains more than 288,000 assessment results from a working adult population.

Two details surprise most candidates. First, the assessment is "criterion-referenced": your score is compared with a cognitive target set for the specific job, which can sit above or below 250, so there is no universal passing score. Second, the verbal, numerical, and abstract subscores are explicitly described by the publisher as not valid or reliable enough to support hiring decisions, so the total is the only number that should matter to anyone.

The Predictive Index says no, and on the specifics it is right: the assessment reports no IQ score, and its 12-minute speeded format is far shorter than a full intelligence battery. At the same time, PI's own technical documentation describes the test as a measure of "g", the general cognitive ability factor that IQ tests also measure, and reports correlations of .29 to .44 with job performance. That is consistent with a century of research showing general mental ability to be among the strongest predictors of how quickly people learn and perform at work.

The practical difference is precision. A 12-minute screener estimates general ability coarsely, which is acceptable when employers combine it with other evidence. A full-length test such as the RIOT IQ test (the Reasoning and Intelligence Online Test) measures the same ability with a much smaller margin of error and reports it on the familiar IQ scale. If you want to know the number a screener is approximating, you can take the free RIOT IQ test and see where you stand.

The Predictive Index publishes official sample questions so candidates can learn the item types in advance, with the caveat that practice results do not predict real scores. Because scoring counts only correct answers, the sensible strategy is steady speed: answer what you can confidently, skip what stalls you, and remember that the average candidate answers about 20 items correctly. Retakes are controlled by the hiring company's administrator, and a retake presents different items drawn from the same banks. The preparation logic is similar to other hiring screeners like the Wonderlic and the CCAT.

There is no universal passing score. The average is 250, and each employer sets a job-specific target that can be higher or lower, so a score that misses one role's target can clear another's comfortably.

As many as you can answer correctly. The average test-taker gets about 20 right in 12 minutes, and unanswered questions carry no penalty.

No. Each candidate receives a unique, statistically equivalent set of items assembled in advance. Difficulty does not shift in response to your answers.

Only if the employer's administrator re-sends it, which publishers reserve for cases like technical problems or misunderstood instructions. There is no candidate-initiated retake.

Not on an IQ scale. It measures general cognitive ability, the same construct IQ tests measure, but reports a scaled score from 100 to 450 built for hiring decisions rather than an IQ number.

1. Schmidt, F. L., & Hunter, J. E. (1998). The validity and utility of selection methods in personnel psychology: Practical and theoretical implications of 85 years of research findings. Psychological Bulletin, 124(2), 262-274. doi.org/10.1037/0033-2909.124.2.262

2. Salgado, J. F., & Moscoso, S. (2019). Meta-analysis of the validity of general mental ability for five performance criteria: Hunter and Hunter (1984) revisited. Frontiers in Psychology, 10, 2227. pmc.ncbi.nlm.nih.gov

3. The Predictive Index. (2015). The science behind The Predictive Index [Technical summary]. predictivesuccess.com

4. The Predictive Index. (n.d.). Reliability and validity of the PI Cognitive Assessment. predictiveindex.com

5. The Predictive Index. (n.d.). Introduction to the PI Cognitive Assessment. predictiveindex.com