A Statistical Comparison of Norm-Referenced Assessment Systems Used in Higher Education in Turkey

The purpose of this study is to identify different norm-referenced assessment systems used in Turkish higher education, and to compare them empirically. Norm-referenced assessment regulations of 70 universities in Turkey was primarily analyzed, and universities were divided into four different groups depending on their norm-referenced assessment systems (only applying T-score conversion, the most commonly used method; applying T-score conversion and quantiles together; applying T-score conversion, quantiles and standard deviation together; applying standard deviation based norm-referenced assessment system). After the algorithms of two universities applying T-score conversion and three universities applying other norm-referenced assessment system were selected, they were used to convert end-of-year grade for each course of 19,574 students in a state university into letter grades and 4-point system. To test the differences of the norm-referenced assessment systems used in these universities, the norm-referenced system of a university were compared with the criterion-referenced system of the same university as well as norm-referenced systems of other universities. The paired t-test was used to identify the difference between norm-referenced and criterion-referenced assessment, while the differences between norm-referenced assessment systems were analyzed through one-way analysis of variance. The findings revealed that the letter grades calculated through the norm-referenced assessment were statistically different than the ones calculated with criterion-referenced; besides, a statistically significant difference was identified between the letter grades obtained using the norm-referenced assessment systems of universities. At the end of the study, the findings were discussed in term of students and instructors