Criterion-Referenced Tests

A criterion-referenced test (CRT) measures a student's performance with respect to a well-defined domain (Anastasi, 1988; Berk, 1988). While norm-referenced tests discriminate between the performance of individual students on specific test items, criterion-referenced tests provide a description of a student's knowledge, skills, or behavior in a specific range of well-defined instructional objectives. This specific range is referred to as a domain. Criterion-referenced tests, instead of using norms, provide information on the performance of a student with respect to specific test items. The results of criterion-referenced testing are not dependent on the performance of other students, as with a norm-referenced test.

There are several characteristics that distinguish CRTs from norm-referenced tests. One of these is mastery. Performance on CRTs provides information on whether students have attained a predetermined level of competence or performance, called mastery. Performance can be interpreted as mastery, nonmastery, or intermediate mastery (Anastasi, 1988). While it is possible to construct a test that is both norm-referenced and criterion-referenced, teachers and other professionals must use caution when interpreting the results of these tests because it is difficult to combine both types of tests in one instrument.

Another distinction between criterion-referenced and norm-referenced tests is the breadth of the content domain that the test covers (Mehrens & Lehmann, 1991). Typical norm-referenced tests survey a broad domain, while CRTs usually have fewer domains but more items in each domain. CRTs typically sample the domain more thoroughly than norm-referenced tests (Mehrens & Lehmann, 1991).

CRTs can also be very useful in helping to make instructional planning decisions. Since they frequently cover a more restricted range of content than norm-referenced tests, they can provide more information about a student's levels of performance.