GITP (PiCompany)

Can You Cheat a GITP Test? Connector Ability

GITP chose detection over cameras: it watches how the test is taken and which answer patterns point to outside help, then verifies with a supervised retest.

Short answer: no.

GITP scores how you take the test as well as what you answer, builds a risk score from answer patterns, and checks a suspicious home result with a supervised follow-up of about the same length somewhere else. A borrowed score flags itself and then has to be reproduced in a room with a supervisor.

No camera

By choice

Detection instead

Risk score

From answer patterns

And how the test is taken

Supervised

Follow-up test

About the same length

Adaptive

Item selection

A personal set per candidate

GITP chose detection over a camera, and said so

Many online tests answer the cheating question with remote proctoring: a webcam, a screen recording, a person watching. GITP wrote in June 2026 that it deliberately does not. It chooses "detectie in plaats van cameratoezicht": it looks at how someone handles the test and at the answer patterns that can point to unauthorised help, turns that into a risk score, and then applies targeted verification, such as a short retest, where the risk is high.

That changes what a shortcut is up against. There is no camera to dodge, and nothing a candidate can see on screen tells them whether they have been flagged. The only way to find out is the verification, and the verification is the part you cannot get help with.

How the test is taken is part of the data

GITP says it looks at how someone handles the test as well as what they answer. It does not publish the details, and we do not guess at them.

Answer patterns feed a risk score

Patterns that suggest outside help raise the risk score. A high score leads to targeted verification rather than to an automatic decision.

The follow-up is supervised, elsewhere, and as long

GITP’s candidate manual: always at another location, always supervised, about as many questions as the first test, and it will soon be clear whether your score differs.

The item set is personal

Connector Ability is adaptive and draws from a large bank, so no two candidates sit the same paper and a leaked item list describes someone else’s test.

What a borrowed score actually buys you

Follow it through. Someone helps you, or you paste items into a chatbot, and the home score comes out higher than you could produce alone. That is exactly the kind of result the risk score exists to catch, and exactly the kind the employer is advised to verify. The follow-up is about as long as the first test, supervised, and at a place of the employer's choosing. If your score differs, it shows at once, and the difference is now the most interesting thing about your application.

The chatbot route fails before it gets that far. GITP tested ChatGPT on Connector Ability items in 2023: it could not read the figure items at all, and on number series it managed easy items but failed once several operations happened at the same time. On an adaptive test, easy items are the ones you would have got right anyway.

What does not work

And why

Getting help at home. The risk score is built from how you took the test, and a flagged result is verified under supervision.

Answer lists from other candidates. Your items are chosen from a large bank by your own answers.

Pasting items into a chatbot. GITP’s own research found it cannot read the figure items and fails the hard number series.

Answering the questionnaires as the ideal candidate. They have no right answers, and a profile that is not you fits no role.

What actually raises the score

In order of effect

Using the time. Minutes per item, speed not scored: check the rule before you commit.

Never leaving an item blank. A guess is right one time in four; a blank never is.

Learning the rule families for each part, the thing the adaptive test is actually measuring.

Sitting practice on a laptop or tablet, the only devices the real test runs on.

How to prepare for a result you can repeat

Find out your norm group

Ask which education level you will be compared with. GITP’s own sample reads one performance as T 50 against Master, T 56 against Bachelor and T 63 against MBO4.

Drill each part on its own

Figure Series, Matrices and Number Series each have their own rule families and their own traps. Work on the weakest one first.

Then sit the full test in one go

The real test takes about an hour with explanations, and part three is where concentration runs out. Practise the whole sitting, with the breaks.

Aim for a score you could reproduce

A home result may be checked with a supervised retest of the same length. The only safe score is the one you can produce twice.

Frequently asked questions

Can you cheat a GITP test?

No. GITP says it chooses detection over camera supervision: it looks at how someone handles the test and which answer patterns suggest unauthorised help, builds a risk score from that, and applies targeted verification such as a short retest. A home result that looks wrong is checked in a supervised sitting you cannot get help in.

What is the GITP follow-up test?

After a home sitting, the employer can ask you to take a follow-up test to check your result. GITP’s candidate manual says it is always somewhere else, always supervised, and contains about as many questions as the first test, and that it will soon be clear whether your score differs.

Are there Connector Ability answers online?

Nothing that would help. Connector Ability is adaptive: which item you get next depends on your earlier answers, and every candidate gets a personal set from GITP’s large item bank. A list of someone else’s items is a list of items you are unlikely to see, and at the same difficulty only by chance.

Can ChatGPT answer Connector Ability questions?

GITP tested exactly that in 2023 and published the result: the chatbot could not decode the figure items at all, and on number series it solved easy items but failed once the items got harder and several operations happened at once. Those are the items that separate candidates.

Should I guess on Connector Ability?

Yes. GITP’s manual says an unanswered item is counted as wrong and a wrong answer is not punished, so a guess among four options is right one time in four and a blank never is. That is a legitimate strategy, not a shortcut.

Can I cheat the Reflector Big Five or Reflector Drives?

There is nothing to cheat. Neither has right answers or a pass mark: the Big Five describes five factors with a named position at both ends, and Drives ranks what matters to you at work. Answering as the person you imagine the employer wants produces a profile that fits neither you nor the role, and GITP only releases the Drives report with your approval.

How do I actually raise a GITP score?

Use the time you are given and learn the rule types. GITP gives minutes per item and does not score speed, so the marks come from finding the rule: movement round a grid on figure series, a rule across the rows and another down the columns on matrices, and changing steps or two woven series on number series. Never let an item expire.

Practise the GITP (PiCompany) assessment instead

Realistic simulations with scored feedback, so you walk in having already done it once. Free to start.

Start Practising Free See the GITP (PiCompany) tests

What to do with the time instead

Practise instead

GITP (PiCompany) practice tests

The Dutch assessment firm that sends the tests PiCompany built: GITP bought PiCompany in 2010 and merged it in 2022, so an invitation may say GITP, PiCompany or Pi Company. Connector Ability (Connector Ability+ with Verbal Analogies) is an adaptive capaciteitentest or intelligentietest in three parts, Figuurreeksen, Matrixen and Cijferreeksen, with a clock of minutes on every item and an unanswered item marked wrong. Connector Verbal Analogies is the optional Dutch-only verbal test. The Reflector Big Five Personality and Reflector Drives are the personality and drijfveren questionnaires, and Talent Match is the online selection assessment that bundles them. Results come back as T-scores, where 50 is the average of the education-level norm group your employer chose (MBO, HBO/Bachelor or WO/Master). Netherlands.

Open

Same question

Can you cheat TestGroup (The Bridge)?

A unique paper every sitting, then a supervised re-test when it matters. The Dutch screen-then-verify model explained.

Open

Same question

Can you cheat Aon / cut-e?

Two candidates sitting at once barely overlap. Why cut-e answer packs describe nothing.

Open

Same question

Can you cheat Korn Ferry / Talent Q?

Every question has its own countdown. Why there is no slack in a Talent Q test to spend.

Open

Browse

Every assessment answered

The same question for every provider we cover, with the mechanism that defeats it.

Open