The CAA process:
Standardized paper-and-pencil tests attempt to provide precise measurement over a wide range of achievement levels—this is usually accomplished by having items with a range of difficulty levels. One consequence of this approach is that for any one examinee, as many as half the items may be too easy or too difficult. The purpose of adaptive testing is to present items at the appropriate difficulty level for an examinee at a given achievement level (i.e., to hone in on the examinee’s ability as quickly as possible during the test administration). To achieve this goal, three basic steps are used:

STEP 1: Given an estimate of an examinee’s current achievement level, the computer selects a new item that should be particularly useful for determining the examinee’s true achievement level.

STEP 2: The item is presented and the examinee responds to one item at a time—after each item, the estimate of the examinee’s achievement level is updated.

STEP 3: A decision is made as to whether another item should be selected or not based on a stopping rule for updating the estimate of the student's achievement level.

In adaptive testing, the sequence of items administered is based on an examinee’s performance to the previous items. Items that are maximally informative about the examinee’s achievement level are administered successively. With this sequence of item selection, tests are shortened without loss of measurement precision because examinees are not required to answer items that provide little or no information about their achievement level.