Computer adaptive tests (CATs) are computer-administered tests that tailor the selection of test items during the administration of the assessment based on the responses of examinees. By adapting the difficulty of the test to the ability level of the examinee, CATs generally provided greater precision and/or enable shorter testing time, when compared to non-adaptive tests.
CAT revolutionized assessment by giving educators a more precise, individualized, and efficient way to measure student achievement and growth over time. By adjusting the difficulty of questions based on the student’s responses, adaptive tests provide a better test-taking experience and more accurate scores across the achievement continuum.
Now, we’re taking CAT to the next level with ShadowCAT, a new flagship adaptive software engine and cloud-based service built on the Optimal Test Design framework. This design uses the “shadow testing” technique to optimize the adaptive experience and guarantee 100% content validity for every student for the first time ever.
The ShadowCAT Difference
ShadowCAT is lightning fast! It moves the “shadow test” concept from theory into practice, as it anticipates the items needed to meet all requirements of the test, and builds a complete and optimal test form for the student at every adaptive decision point.
From that fully compliant form, the item that will provide the most information about the student’s ability is presented to the student. The student sees only that item, while the rest of test form remains hidden—hence the name shadow test. The creation of the shadow test ensures items are available to meet all subsequent test requirements, guaranteeing 100% blueprint compliance for every test taker.
Adaptive algorithms that select one item at a time cannot make the same guarantee. The more adaptive a test is, the more efficient the estimate of the test taker’s abilities. ShadowCAT provides the most adaptive, most efficient, most precise way to assess your students, and ensures all test requirements are met in every test administered. It also shortens testing time and can reduce student test anxiety.
On a typical computer adaptive test, a student who correctly answers a question will receive a more difficult item, while an incorrect answer will generate an easier item.
The ShadowCAT Benefits
Provides maximum information for high precision
Each test constructed with the ShadowCAT engine automatically optimizes the selection of items to maximize information throughout a student’s testing experience.
Offers high scalability for states, districts, and schools
ShadowCAT can be used to deploy computer adaptive assessments to one student or to many, providing maximum scalability and performance.
Guarantees compliance with test blueprints and psychometric requirements
ShadowCAT guarantees that the test will adhere to all test requirements for every test taker. No other adaptive algorithm can make that guarantee. Unlike typical CAT solutions, which consider only one item at a time, the ShadowCAT engine evaluates the test selection against all requirements simultaneously. Typical tests include complex and multivariate requirements, such as those associated with reading passages. ShadowCAT prevents the selection of an item that may meet a single requirement early on the test, but that will prevent all requirements from being met at the conclusion of the test.
ShadowCAT offers flexible and configurable delivery options
The ShadowCAT engine uses CTB’s Optimal Test Assembler to configure the tests, and supports multiple adaptive delivery approaches for the same set of test requirements and blueprints, including:
To assess each student’s ability beyond a shadow of a doubt, try ShadowCAT. To find out more visit www.shadowtest.com or call 800.538.9547.