Standard-Setting Using Item Response Theory

Standard-setting is a procedure of providing cut-scores on measurement scales used to separate proficiency levels by classifying the test-takers with scores lower than a cut-score into one level and those whose scores are at or above the cut-score into a higher level. Different standard-setting approaches have been proposed to classify mastery levels of the target population, to find out the norms and passing rates for the population or test-takers, to induct the standards through an empirical way, and to validate the assessment framework developed based on the theories. In the current study, an evidence-based standard-setting approach was proposed to estimate the thresholds on the scales of interested latent traits by using item response theory. An empirical data set for evaluating a teacher education program was used to demonstrate the proposed approach. The strengths and limitations of the proposed method were also discussed.