Standardization

Standardization of Procedure

Standardization of procedure is highly important in any kind of research design and essentially refers to experimental control. This is relevant for any test or experimental procedure and includes the standardization of instructions, administration (including manipulation), and measurement of variables of theoretical interest. Instructions need to be as clear as possible for the particular population. The presentation order of test material or experimental manipulations needs to be identical between people, trials, or conditions. Time constraints might be specified. In many test designs, instructions might also include ways of handling questions of participants. For some tests, oral instructions might include the consideration of rate of speaking, tone of voice, inflections, facial and bodily expressions, or pauses. For example, correct answers might be given away if smiling or pausing when a critical test item is being read out loud. In essence, this type of standardization tries to reduce the influence of any extraneous variable on the test or experimental performance of participants. A lack of standardization of procedure will lead to unreliable results, which in turn will threaten the validity of the test or experiment. The standardization of procedure as a term is more often used in the context of test development (e.g., test instructions, item order, and time limits) than in experimental design. In experimental designs, researchers are concerned with internal and external validity, which is often treated as a separate topic.

Standardization of Interpretation

The standardization of interpretation refers to the standardized interpretation of obtained scores. This is of particular importance in test administration and interpretation. Here, the concept of so-called norm groups is central. Psychological tests have no predetermined standards against which performance of individuals or groups of individuals can be evaluated. Therefore, scores are typically compared with some norm, which is the performance of others obtained using the same test. Norms imply some form of normal or average performance. For example, if the typical 8-year-old child answers 10 questions of 50 correctly, then this would be the norm for 8-year-olds. Without knowing the average performance of a comparable group, the raw score of 10 correct answers is meaningless. The raw scores on tests could include the number of correct answers (or percentage of correct answers), number of errors or mistakes, reaction times, mean or modal response to some attitude item, or some other objectively measurable indicator relevant for the content of the test. In addition to such age norms, percentile ranks or [Page 1434]grade norms are other typical examples. To obtain these norms, tests are administered to large and representative samples drawn from the population for which the test is being designed. This is an essential element of test standardization. The quality of the test norm depends on this so-called standardization sample chosen for norming the test. For example, if a test is developed to measure mental abilities for managers, it would be meaningless to norm the test on patients who underwent psychiatric treatment.

...