CAT-II IQ Test Explained: How Adaptive Testing Works
CognitiveIndex’s CAT-II is a computer-adaptive IQ-style assessment. Instead of giving every person the same fixed sequence, it updates an ability estimate after each response and selects another unused item that should be informative near that estimate. The implemented test administers between 30 and 60 questions.
What CAT-II Means Within CognitiveIndex
CAT-II is the product name used by CognitiveIndex for its current adaptive reasoning assessment. The repository does not define a longer official expansion of the full “CAT-II” label. In testing generally, CAT refers to computerized adaptive testing: software selects questions during the session rather than relying on one predetermined form.
The CognitiveIndex implementation uses a 60-item pool and three reported content areas:
- Quantitative knowledge
- Verbal comprehension
- Fluid reasoning
It produces an IQ-style estimate, an internal percentile estimate, section percentages, a model-based precision value, and an estimated standard error. These describe the website’s algorithm; they do not establish independent clinical validation or professional norm equivalence.
How the Next Question Is Selected
CAT-II begins with a provisional ability estimate. After a response, the code:
- Records whether the submitted answer was correct.
- Updates the current ability estimate.
- reviews unused questions in the pool.
- Prefers an item expected to provide useful information near the new estimate.
- Keeps the three content areas reasonably balanced.
This is more precise than saying every correct answer simply causes a harder question and every incorrect answer causes an easier one. Estimated difficulty matters, but content balance and the information provided by an item also affect selection.
Submitted answers are final because each response changes the next selection. If an item timer expires, the unanswered response is recorded and the test advances.
Why Users May Receive Different Questions or Test Lengths
Two users can take CAT-II and see different paths through the same pool. Their early responses can produce different ability estimates, which can make different unused items more informative.
The stopping rule also allows different lengths:
- The test always administers at least 30 questions.
- It may stop after that point when the code’s precision threshold is reached.
- It continues when more information is required.
- It stops at 60 questions at the latest.
Because each item allows either 60 or 75 seconds, CAT-II has no single guaranteed total duration. A 30-item path can finish much sooner than a 60-item path, and answering before each timer expires shortens the session further.
What the Three CAT-II Areas Mean
These descriptions refer to the implemented item pool, not every test that uses similar labels.
Quantitative knowledge
Quantitative items include arithmetic, rates, percentages, algebra, number sequences, and short applied problems. They receive a 75-second item limit.
Verbal comprehension
Verbal items sample vocabulary, analogies, sentence completion, classification, and word relationships. They receive a 60-second item limit.
Fluid reasoning
Fluid-reasoning items include deductions, ordering, symbolic rules, letter or number patterns, and other novel relationship problems. They also receive a 60-second limit.
The category percentages summarize correct responses among the items actually administered in each area. Because adaptive paths differ, they should not be treated as if every user completed the same section form.
Adaptive Test vs. Fixed-Form Test
| Feature | Adaptive test | Fixed-form test |
|---|---|---|
| Question selection | Chosen during the session from the current estimate and item rules | Predetermined before the session |
| Test length | May vary when a stopping rule is used | Usually the same for every user |
| Difficulty | Can shift with the estimated ability level | Same planned sequence or form for users |
| User experience | Answers are often final; users may see different items | Backtracking may be allowed; users usually share the same structure |
| Score interpretation | Depends on the adaptive model, calibrated pool, and stopping rule | Depends on the form’s raw-score conversion and norms |
Neither format is automatically better. A well-designed adaptive test can estimate efficiently with fewer items, while a fixed test can be easier to review and compare because the form is stable. Quality depends on the item pool, calibration, reliability, norms, administration, and evidence supporting the intended use.
Is an Adaptive IQ Test Easier or Harder?
Adaptive tests often feel harder because they try to avoid giving a long run of questions that are obviously too easy or impossibly difficult. The aim is not to punish correct answers; it is to gather information efficiently near the current estimate.
Difficulty also feels different because CAT-II does not allow backtracking. A fixed-form test may let someone skip an item and return later, while CAT-II requires a decision within the item timer. That difference affects the experience even when the underlying reasoning demands are similar.
Can CAT-II Be Compared With a Traditional IQ Test?
Only cautiously. Similar-looking IQ numbers are not proof that two tests are interchangeable. A defensible comparison requires evidence about:
- The population used for norms
- Reliability and score precision
- The item calibration process
- The abilities represented in the item pool
- Administration conditions
- The score scale and ceiling
- Correlations with the specific comparison test
CognitiveIndex’s CAT-II reports a model-based internal estimate. The repository does not contain an independent representative norm study or validation study showing equivalence to a named professional intelligence test.
Use the result to review performance on this adaptive form, not to make a clinical, educational, or employment decision.
What to Do Before Starting CAT-II
- Choose a quiet place where item timers will not be interrupted.
- Allow enough time for a path of up to 60 questions.
- Review basic arithmetic, ratios, sequences, analogies, and short deductions.
- Do not plan to return to earlier questions; submitted answers are final.
- Avoid calculators, search, AI tools, or help from another person.
- Use a stable connection and do not refresh during the adaptive sequence.
The CAT-II landing page explains the current format and starts the assessment.
Want to experience the adaptive format? Take the CAT-II IQ Test.
Frequently Asked Questions
Is CAT-II the same as the school Cognitive Abilities Test, or CAT4?
No such equivalence is supported by this repository. CAT-II is CognitiveIndex’s product label for its own adaptive IQ-style assessment and should not be assumed to be another publisher’s test.
How many questions are on CAT-II?
The test administers at least 30 and no more than 60 questions from a 60-item pool. The exact length depends on the implemented stopping rule.
Does CAT-II get harder after every correct answer?
Not necessarily in a simple step-by-step way. The response updates the ability estimate, and the code then considers item information and content balance when choosing an unused question.
Can I change an answer?
No. Submitted answers are final because each response affects the next selected item.
Is the CAT-II result a clinical IQ score?
No. It is an informational, model-based estimate from this online implementation, not a diagnosis or a professionally administered intelligence assessment.
Sources
- TOEIC Link Assessments Technical Manual, Educational Testing Service, for general computerized-adaptive testing mechanics.
- ITC/ATP Guidelines for Technology-Based Assessment, International Test Commission and Association of Test Publishers.
- Standards for Educational and Psychological Testing, AERA, APA, and NCME.