The concepts of universe of admissible observation and universe of generalization from the generalizability theory were applied to calculate the intraclass correlation coefficient of a licensure test. The internal consistency coefficient of a dichotomously scored test is identical to the intraclass correlation coefficient of a two-facet design. The intraclass correlation coefficients for both norm- and criterion-referenced decisions were smaller in the nested than in the simple two-facet design and they were also smaller than the internal consistency coefficient. The proper design for most licensure tests and for tests using a design where the items are selected from content areas specified in the blueprint rather than from the universe of possible items must use a nested design to calculate the stability of test results rather than the internal consistency coefficient. (Author)