ABSTRACT
In order to make inferences about properties of a test-taker based on his test results, it is necessary to interpret them. Two basic approaches to interpreting individual results on psychological tests are criterion-referenced assessment and norm-referenced assessment. Aside from these two approaches, cohort-referenced assessment, construct-referenced assessment and curriculum-based assessment are also described in the literature. In the scope of criterion-referenced assessment, individual results are interpreted by comparing them to a certain criterion, while norm-referenced assessment is based on comparing results of a test taker to that of a reference group called the normative sample. In psychological practice, this is done by using the norms – a document describing the results of the normative sample. In the process of norm-referenced assessment, raw test scores are often converted into fractile (usually percentile) ranks or to a standard scale. Popular standard scales include the z scale, IQ scale, T scale and C scale. Norms used to interpret tests can be universal norms, but also national norms, language norms, local norms, age norms, class norms, school norms, occupational norms and gender norms are all in use, depending on the purpose of the test. Properties of the population norms can change over time and with regard to this test authors may opt for periodic re-creation of norms or for a system of scoring based on a fixed reference group (norm “freezing”). After converting raw scores into fractile ranks or into standard scores, their interpretation can be done using a dimensional approach, typological approach or profile analysis approach.
