Computer Adaptive Test

Definition

A computer adaptive test (CAT) is a test where the questions you see depend on how you’ve answered previous questions. Get a question right, the next question is harder; get one wrong, the next is easier. The SHSAT moved to a CAT format starting with the 2024 administration.

How it actually works

The test starts at average difficulty. After your first answer, the test estimates your ability based on whether you got it right and selects the next question accordingly. As you answer more questions, the ability estimate becomes more precise, and the test selects questions that maximize information about your true ability level — usually questions whose difficulty is close to your current estimated ability.

This is more efficient than a fixed-form test. A traditional fixed-form test gives every test-taker the same questions, many of which are too easy or too hard for any given student. A CAT focuses test-time on questions near the student’s ability level, getting a precise score in fewer questions.

What the SHSAT-CAT looks like to a test-taker

From the student’s perspective:

• You see one question at a time. You can’t skip ahead and come back, and you can’t change your answer once you submit.
• Each section adapts independently. ELA performance doesn’t affect Math question selection or vice versa.
• You can’t tell the test’s difficulty level from your performance feel. A test that feels “easy” might be calibrating to a low ability estimate; a test that feels “hard” might be calibrating to a high one. Trying to infer your score from question difficulty is unreliable.
• Sections still have field questions. About 10 of 57 questions per section are field questions — experimental items not scored. You can’t identify which are which.

How it’s scored

The SHSAT-CAT uses Item Response Theory (IRT) for scoring. Specifically, the 3-Parameter Logistic (3PL) model, which accounts for each question’s difficulty, discrimination (how well it distinguishes ability levels), and guessing probability. Your final scaled score depends not just on how many you got right, but on the difficulty of the questions you answered correctly. Two students with the same number of correct answers can get different scaled scores if one answered harder questions correctly.

What changes for SHSAT prep

The CAT format affects test strategy in a few ways:

• Early questions matter more. Early answers shape the ability estimate, which then shapes which questions you see. A weak start can route you to easier questions whose ceiling is lower than your actual ability.
• You can’t skip and return. Strategic skipping (answer easy questions first, return to hard ones) doesn’t work in a CAT environment.
• Guessing strategy is the same. No penalty for wrong answers means you should answer every question. Random guessing is better than leaving blank.
• Pacing is critical. Spending too long on a single question costs time across many later questions. CAT formats reward steady pacing.

Common misconceptions

“If I get a hard question, I’m doing well.” Possibly, but not reliably. The algorithm balances difficulty with information value, and question difficulty isn’t a clean signal of your current estimated ability.

“CAT means I can’t prepare in advance.” Wrong. The content is the same as the pre-CAT SHSAT — same ELA topics, same math topics. What changes is the format and pacing, both of which can be practiced.

Try our free adaptive practice test for a real CAT experience with full IRT scoring.