The Adaptive Nature of MAP Tests

A computer-adaptive test (CAT) is a method for administering tests that adapt to a student's ability level. The psychometric methodology that allows scores to be computed across different sets of items is called Item Response Theory (IRT). From the student's perspective, the difficulty of the test seems to tailor itself to their ability level. The first time a student takes a computer-adaptive test, the pool of available items is searched for an item based on the student's current ability estimate. Correct responses are followed by more difficult items and incorrect responses are followed by easier items. After only a few questions, the item difficulties are close to a student's true ability and a reasonable estimate can be reported.

There are several advantages to computer-adaptive testing:

  • Tests are given on demand and scores are available immediately
  • Tests are individually paced, potentially reducing one source of test anxiety
  • Test security may be increased because hard copy test booklets are never compromised
  • Seat time is reduced because significantly fewer items are needed to provide uniformly precise scores
  • Accurate scores across a wide range of abilities are provided while traditional tests are usually most accurate for average students

As a student takes a MAP test, he or she is presented with items of varying RITs, or levels of difficulty. Once the MAP system determines the difficulty level at which the student is able to perform and the system collects enough data to report a student's abilities, the test ends and the student is assigned an overall RIT score. In survey with goals tests, the student also receives RIT range scores for the goal strand components.

MAP assessment results are intended to be interpreted and used to guide appropriate instruction to improve student learning. Due to the difficulty in capturing data about various test environment variables, there is not yet data available suggesting that MAP assessments improve student learning inherently.

Related Support Articles Categories:

(Clicking on one of these links will take you to a list of support articles matching that category)