PI Cognitive Assessment: 50 Questions in 12 Minutes Explained

The PI Cognitive Assessment gives you 50 questions and 12 minutes. That works out to 14.4 seconds per question if you tried to answer every one. Almost nobody does. The average candidate finishes between 20 and 25 questions, and the median raw score sits around 20 out of 50. The honest answer to "how long is a PI question" is: as long as you can afford to spend on it before the timer takes it away.

This article walks through what the 50-questions-12-minutes format actually feels like in practice, how the scoring converts to the 100 to 450 scaled score employers see, how many questions you really need to finish to land in the target range for a typical role, and the pacing rules candidates use to get there.

Quick takeaways

• 50 questions, 12 minutes, no calculator, three question types mixed in random order (numerical, verbal, abstract reasoning).
• The math says 14.4 seconds per question, but only the top 2 percent of candidates finish all 50.
• Raw score maps to a scaled score from 100 to 450. The general-population average is 250.
• Most employers set their Cognitive Job Target between 200 and 310. General management roles sit around 280.
• No penalty for wrong answers. A blank costs the same as a wrong, so guessing on a question you cannot solve is free.
• Reaching 40 out of 50 correct puts you in the top 2 percent. Reaching 30 out of 50 already clears the target for most roles.
• The biggest skill is recognising which questions to skip in the first 3 seconds, before they steal the next question's time.

What the 50 questions in 12 minutes actually looks like

The PI Cognitive Assessment (PICA), also called the PLI (Predictive Learning Indicator) when administered through some employers, is built around nine question subtypes split across three categories: numerical reasoning, verbal reasoning, and abstract or spatial reasoning. The 50 questions are not blocked by category. They appear in random order, so a series-completion question can be followed by an antonym, followed by a shape-progression. You cannot warm up on one type before the next one lands.

Every question has five answer choices. There is exactly one correct option. The interface shows you the question, the choices, a "next" button, and the timer counting down from 12:00. You cannot skip and come back. Once you move past a question, it is locked in, scored or unscored based on whatever you selected before clicking next. You also cannot flag a question, change your answer later, or review.

This locked-forward design is the structural reason the test is hard. On the SAT or GMAT you can park a brutal question, attack the rest, and circle back if you finish early. Here, every question is a fork: spend more seconds and risk running out of time, or guess and move on. The candidates who score well have already decided, before the test starts, which question types are worth a 30-second fight and which get an instant 5-second guess.

How scoring works (raw to scaled to job target)

You get one point per correct answer. There is no fractional credit, no negative marking, and no per-section weighting. Five wrong and five right in the first 10 questions is exactly the same as five wrong at the start and five right at the end. Unanswered questions count as zero, the same as wrong answers, which is the entire reason guessing the last 25 questions in the final 10 seconds is a defensible pacing strategy.

The raw score (out of 50) is then converted to a scaled score from 100 to 450, normed against the general population. A raw of 17 to 20 is roughly average and converts to a scaled score near 250. A raw of 25 lands you near a scaled 270, which clears the Cognitive Job Target for most individual-contributor and mid-management roles. A raw of 40 puts you in the top 2 percent of scorers, at a scaled score around 320 or higher.

Employers do not see your raw out of 50. They see your scaled score and a "match" indicator that compares it against the Cognitive Job Target they set when posting the role. The target is a number on the same 100 to 450 scale. The Predictive Index publishes default targets per job family, and the hiring manager can move the target up or down by 30 points. The match is binary on the high side: a candidate scoring at or above the target gets a match of 10 out of 10, with no extra credit for going much higher.

These are approximate conversions reconstructed from candidate reports and Predictive Index documentation. Your exact scaled score depends on which form of the test you took, since multiple forms are norm-equated against each other.

The atmosphere of a PI Cognitive sitting

The image below shows the structural shape of the test.

What surprises most candidates is not the difficulty of any single question. The math questions are pre-algebra and basic series, the verbal questions are antonyms and analogies, and the abstract questions are 2D pattern progressions. Each one, in isolation, is something a numerate adult can solve in 30 seconds.

The trap is that 30 seconds is twice the budget. Every question you "actually solve" is two questions you do not even see. By question 15, candidates who set out to think carefully are already behind the pace curve. By question 25, they are skimming. By question 40, they are random-clicking the last 10 just to lodge an answer before the timer hits zero.

The candidates who score well solve this not by being faster on individual questions but by being more aggressive about skipping. The standard pacing benchmark is to give yourself a hard 14-second cap per question and a 2-second budget for the skip decision. If a question is not clearly tractable within 2 seconds, you guess immediately and move on. The cost of the wrong guess is one point. The cost of spending 40 seconds working it out is three subsequent questions you never see, which is a worst-case three points.

What candidates actually score

Public data on PI Cognitive distributions is thin, since the Predictive Index does not publish norm tables to candidates. The pattern stitched together from candidate self-reports and prep-platform aggregations is consistent:

• About half of candidates submit a raw score between 17 and 23.
• About 15 percent submit below 12.
• About 10 percent submit above 30.
• About 2 percent submit above 40.
• Almost no first-time, unprepared candidate finishes all 50.

This is why scoring 30 out of 50 already gets you into the 80th percentile band and clears the default target for most roles. You do not have to be exceptional. You have to be efficient.

Sample questions: the three formats

The fastest way to understand the time pressure is to attempt three questions cold. Give yourself 45 seconds total.

Numerical (series). Find the next number in the series: 3, 6, 12, 24, ?

A. 30 B. 36 C. 42 D. 48 E. 60

Verbal (antonym). Choose the word most OPPOSITE in meaning to CAUTIOUS.

A. Careful B. Wary C. Reckless D. Prudent E. Vigilant

Abstract (analogy). TREE is to FOREST as SOLDIER is to:

A. Battle B. Uniform C. Army D. Weapon E. Officer

Answers and reasoning appear in the FAQ at the end of this article, where they will not interrupt your pacing on the cold read.

If those three felt slow, that is the signal. You are working at roughly 15 seconds per question pace at best, and the actual test runs faster because the screen interaction itself eats a second or two.

How many questions you actually need to finish

The instinct on a 50-question test is to finish 50 questions. On the PI Cognitive that instinct is wrong. You do not need to finish. You need to clear the job target.

The chart below shows raw-to-scaled conversion and which job targets each band reaches.

Working backwards from a target raw score of 25 (which clears most individual-contributor roles), you do not need to attempt 50 questions. You need to attempt around 35 and get 25 of them right. That math allows for a roughly 70 percent accuracy rate, which is achievable when you are not panic-skimming the back half. If you attempt 50 and your accuracy drops to 50 percent under time pressure, you also land at 25 correct. Same score, more stress, more chance of error.

The pacing target most candidates train towards is:

• Question 1 to 15 in the first 4 minutes (16 seconds each, careful pace).
• Question 16 to 30 in the next 4 minutes (16 seconds each, sustained pace).
• Question 31 to 45 in the final 4 minutes (16 seconds each, fast pace, more guessing).
• Last 5 questions: guess-and-submit in the final 30 seconds. No penalty for being wrong.

This gets a candidate to 45 attempted and 25 to 30 correct, which lands them in the 270 to 285 scaled range and clears the default target for general management and most individual-contributor roles.

Why the test still trips strong candidates

Three predictable failure modes show up in candidate post-mortems.

The first is the "I know I can solve this" trap. A strong mathematician sees a series question and refuses to abandon it. They spend 90 seconds, get it right, and lose the next five questions. The cost of being correct on the stubborn question is worse than the cost of being wrong on six others.

The second is the abstract-reasoning freeze. The pattern-progression questions look unfamiliar to candidates whose schooling did not include them. They stare for 20 seconds before realising they should have guessed. The fix is to recognise the abstract format from the first second of the question screen and apply the 2-second skip rule even more aggressively than to the numerical and verbal questions.

The third is the calculator instinct. The PI Cognitive does not allow a calculator. Candidates who reach for one mentally (long division, fraction arithmetic) lose 10 seconds per question even when they get the answer right. The defence is to estimate, eliminate, and pick the closest option, not to compute exactly.

How to prepare in a week

A 7-day plan is enough to move most candidates from a raw 18 to a raw 28. The plan is not "study harder questions" because the questions are not the bottleneck. The plan is to install pacing reflexes:

Day 1: Take one full 12-minute practice test cold. Record raw score and how many questions you attempted.

Day 2: Take a second full test. The score will not have moved much. The point is to start internalising the timer.

Day 3: Drill numerical-series questions for 30 minutes. Aim for a 12-second average per question on this single type.

Day 4: Drill verbal antonyms and analogies for 30 minutes. Aim for 8 seconds per question.

Day 5: Drill abstract pattern progressions for 30 minutes. Aim for 15 seconds per question.

Day 6: Take a third full test with a deliberate 14-second cap. If you do not see the answer in 2 seconds, guess and move on. Compare to days 1 and 2.

Day 7: One light practice round. Sleep early. Take the real assessment with the pacing reflex installed.

The candidates who go from 18 to 28 in a week are not learning new math. They are unlearning the habit of finishing every question.

Related on PrepClubs

• Pillar. PI Cognitive Assessment overview. The full breakdown of the test, scoring, and who uses it.
• Deep practice. PI Cognitive practice pack with Pass Guarantee. Full timed mock tests with answer walkthroughs. $39 one time. Pass Guarantee.
• Practice test. PI Cognitive Assessment Practice Test: Free Walkthrough. Free 10-question walkthrough with full reasoning.
• Format. Predictive Index Test: Format, Time Limit, Sample Questions. Companion article on the cognitive vs behavioral split.
• Disambiguation. The Predictive Index in 2026: Cognitive vs Behavioral, Score, and How to Pass. Which one are you taking, and what each measures.

Practice on PrepClubs

Train the pacing reflex on real 50-question timed mocks.

The single highest-leverage prep activity for the PI Cognitive is not learning new question types. It is sitting through enough timed 12-minute mocks that the 14-second-per-question cap becomes automatic. PrepClubs runs full-length PI Cognitive practice tests with the same question mix, the same locked-forward interface, and the same timer behaviour as the real test. Every question comes with a worked walkthrough, so you can see exactly where you burned time. $39 one time. Pass Guarantee.

Start free PI Cognitive practice

FAQ

How many questions are on the PI Cognitive Assessment?

50 questions in 12 minutes. The questions are split across numerical reasoning, verbal reasoning, and abstract or spatial reasoning, with no separate sections. They appear in random order.

What is a good score on the PI Cognitive Assessment?

A raw score of 25 out of 50 maps to a scaled score near 270, which clears the default Cognitive Job Target for most individual-contributor and mid-management roles. A raw of 30 clears the 280 target used for general management. A raw of 40 puts you in the top 2 percent of scorers.

Is there a penalty for wrong answers?

No. An unanswered question and a wrong answer both score zero. Because guessing is free, candidates who run out of time should always lodge an answer on every remaining question, even at random, rather than leave them blank.

What were the answers to the three sample questions?

The numerical series 3, 6, 12, 24 doubles each step, so the next term is 48 (D). The opposite of CAUTIOUS is RECKLESS (C). A SOLDIER belongs to an ARMY just as a TREE belongs to a FOREST (C).

How fast does the real test feel compared to practice?

Faster. The on-screen interaction (clicking the answer, confirming, advancing) adds 1 to 2 seconds per question that does not exist in untimed practice. Candidates who hit 14-second pace in practice typically run at 16 seconds in the live environment, which is why training for 12 seconds in practice is the safer target.

Can I take the PI Cognitive more than once?

Retake policy depends on the employer, not on the Predictive Index. Some employers allow a single attempt per role posting. Others allow a retake after 6 or 12 months. The PI score itself does not expire for the candidate, but the employer's match calculation only uses the most recent submission for the relevant Job Target.

Is the PI Cognitive the same as the PLI (Predictive Learning Indicator)?

Yes. PLI is an older name some employers still use in their application portals. The questions, time limit, scoring, and Cognitive Job Target system are identical. If your application says PLI, the test you sit is the PI Cognitive Assessment.

Does the test allow a calculator?

No. The numerical questions are deliberately designed to be solvable in 15 seconds without a calculator. Candidates who rely on long arithmetic lose time even when they get the answer right. Estimation and answer-choice elimination are the intended solving strategies.