Standard setting in student assessment: is a defensible method yet to come?

INTRODUCTION: Setting, maintaining and re-evaluation of assessment standard periodically are important issues in medical education. The cut-off scores are often "pulled from the air" or set to an arbitrary percentage. A large number of methods/procedures used to set standard or cut score are described in literature. There is a high degree of uncertainty in performance standard set by using these methods. Standards set using the existing methods reflect the subjective judgment of the standard setters. This review is not to describe the existing standard setting methods/procedures but to narrate the validity, reliability, feasibility and legal issues relating to standard setting.

MATERIALS AND METHODS: This review is on some of the issues in standard setting based on the published articles of educational assessment researchers.

RESULTS: Standard or cut-off score should be to determine whether the examinee attained the requirement to be certified competent. There is no perfect method to determine cut score on a test and none is agreed upon as the best method. Setting standard is not an exact science. Legitimacy of the standard is supported when performance standard is linked to the requirement of practice. Test-curriculum alignment and content validity are important for most educational test validity arguments.

CONCLUSION: Representative percentage of must-know learning objectives in the curriculum may be the basis of test items and pass/fail marks. Practice analysis may help in identifying the must-know areas of curriculum. Cut score set by this procedure may give the credibility, validity, defensibility and comparability of the standard. Constructing the test items by subject experts and vetted by multi-disciplinary faculty members may ensure the reliability of the test as well as the standard.