Norm-Based Assessments

The interaction of technology and norm-based, or norm-referenced, assessment provides opportunities for the collection of interesting information related to student learning and development as well as introducing several issues and concerns. Reviewing the purpose of assessment is critical to the use and interpretation of assessments. Categorizing assessments can clarify purposes and help educators and policy makers address issues. This entry first discusses the benefits and disadvantages of norm-based assessment and its use in computer-based and computer adaptive testing. It then discusses some of the uses and potential uses of norm-based assessments.

A common type of assessment involves the comparison of scores to a standard— criterion-referenced —or to those from a group of similar individuals— norm-referenced. Royal Van Horn argued for the use of improvement-referenced assessment (criterion- referenced measures taken multiple times), but Peter Behuniak proposed that assessment should be consumer-referenced. When analyzing the efficacy of norm-referenced assessment, two critical questions become (1) what are the disadvantages and benefits of norm-referenced assessment that cannot be addressed with other approaches, and (2) if norm-referenced assessment is warranted, how can technology be used to make the process efficient and effective?

Dylan Wiliam stated that scores derived from norm-referenced testing are relatively insensitive to instruction; this is the primary challenge for their use in assessment of academic knowledge and skills. As a result, a shift of classroom- and state-based standardized assessments in the United States has occurred toward the emphasis of criterion-referenced. Nevertheless, there are several areas where an effort to use norm-referenced assessment might be justified: (1) increasing the efficiency of norm-referenced assessments when comparisons are made on academic content or skills where there is a lack of coherence of expected standards or taught content, (2) providing learners and other stakeholders information about how students compare with large groups of similar individuals, (3) providing important information for program evaluation if learner input characteristics among the comparison groups are matched, and (4) collecting data on important domains of human development and behavior that do not as yet have established standards and benchmarks.