Preparación SAT

What Is Adaptive Testing?

Adaptive tests pick later questions from how you answered earlier ones. Item-level tests re-pick every answer; the digital SAT routes you into a second module.

Adaptive testing

Adaptive testing is computer-based testing in which the difficulty of later questions is chosen from how the candidate answered earlier ones, so two candidates sit different questions.

También llamado: Computer-adaptive testing, CAT, Multistage adaptive testing
Dónde lo encuentra el alumnado: Digital SAT preparation in Grades 10 to 12, where each section runs as two modules and the second is chosen by performance on the first, and in the online screening tests some schools use for setting.

La respuesta corta
Adaptive testing means the test changes shape as you sit it: software chooses later questions using how you answered earlier ones. Item-level adaptive tests re-choose after every single question. Multistage tests, including the digital SAT, route you into a whole second module. A harder second module is usually good news.

Un ejemplo

Module 1: easy, medium and hard mixed → Module 2: the harder set

On the SAT's Maths section you answer a first module built from a broad spread of difficulty. How that goes decides whether the second module is the harder or the easier one. Being sent to the harder module is the outcome you want: because difficulty is part of the scoring, a correct answer on a hard question carries more than a correct answer on an easy one, and the top of the scale is only reachable through the harder route.

Two ways a test can adapt, and the difference matters

Item-level adaptive testing re-chooses after every response. Answer correctly and the next question is harder; answer wrongly and it is easier. Because the next item depends on the answer you just gave, these tests normally refuse to let you skip a question or go back and change one — the machine has already used your answer to pick what comes next. The GMAT is the best-known example.

Multistage adaptive testing works in blocks. You get a whole module of questions, you can move around inside it, flag things and change your mind, and the routing decision is only taken when that module is submitted. The digital SAT is built this way: two modules per section, and the second module is more or less difficult depending on the first.

Why the number you got right stops meaning anything

On a fixed paper, everyone answers the same questions, so counting correct answers is a fair comparison. On an adaptive test it is not, because two candidates with the same raw count may have been answering questions of very different difficulty. Scoring therefore uses the difficulty of the items you actually saw as well as whether you got them right, and the result is placed on a common scale so that scores from different routes can be compared.

This is why practice materials that promise to convert a raw score into a scaled score are approximations even when they come from the test maker. It is also why chasing a raw count in practice is the wrong target: two practice sittings with the same number correct are not the same performance.

What it changes about how you sit the test

The first module carries more weight than its share of the questions suggests, because it decides the route. That is an argument for steady, careful work at the start rather than the common instinct to warm up slowly and push in the second half. It is not an argument for panic — the first module is deliberately a mixture, so hard questions in it are not a sign that anything has gone wrong.

Two practical points follow. On a multistage test you can review within a module but never return to a module you have submitted, so any checking has to happen before you move on. And where there is no penalty for a wrong answer, as on the SAT, leaving a question blank gains nothing at all.

Item-level adaptive: no skipping, no going back, one question at a time
Multistage adaptive: review and change answers freely inside the current module only
Difficulty counts towards the score, so raw correct counts are not comparable
A harder second module means the first module went well

Preguntas frecuentes

Is a hard second module a bad sign?

It is the opposite. The harder module is where the higher scores live, and you are only routed there by doing well on the first module. Students walk out of the exam convinced they failed because the second half felt brutal, when the difficulty itself was the evidence that the first half went well.

Can I go back and change my answers?

On the digital SAT, yes, within the module you are working on — you can flag questions and revisit them until you submit that module. On item-level adaptive tests such as the GMAT you generally cannot, because each answer determines the next question. Check the rules for the specific test before relying on a review strategy.

Does everyone sit a different test?

In an adaptive test, largely yes. Candidates see different questions depending on how they perform, which is the whole point of the design. Scores are still comparable because the scoring accounts for the difficulty of the questions each candidate actually answered and places every result on one common scale.

Why do exam boards use adaptive tests at all?

Precision per minute. A question far above or far below a candidate's level tells you very little about them, so a test that steers towards their level learns more from fewer questions. That is why adaptive tests tend to be shorter than fixed papers measuring the same thing to the same accuracy.

Fuentes

How the SAT Is Structured — College Board

Última actualización 21 de agosto de 2026