Standard-setting is necessary when we need a cut score to discriminate among individuals with varying capabilities on the construct we measure. All standards basically depend on expert judgments on what is meritorious performance by typical examinees in a given domain (Jaeger, 1989). Standards set by experts can be empirically validated with real score distributions. This chapter presented methods for setting standards based on judgments about item-content, the content of an overall assessment, or score distributions of typical examinees using the contrasting groups method and the ogive method.

Page 3