Studies in Educational Evaluation vol:35 issue:4 pages:174-182
The aim of this study was to examine the effect of different standard setting methods on the size and composition of the borderline group, on the discrimination between different types of students and on the types of students passing with one method but failing with another. A total of 107 university students were classified into 4 different types of students and completed two examinations. Nine different methods to determine cut scores were applied. Results show that size and composition of the borderline group differed for the different standard setting methods. Results also indicate that it is difficult or even impossible to point out one method across tests as the most appropriate one in terms of discriminating between different students. Situation characteristics such as type and level of difficulty of the test are important determinants when a particular standard setting method has to be chosen.