Abstract
This paper introduces an empirical study testing three kinds of bias in higher education student assessment. All of them are connected to the repetitive use of the same test questions which may facilitate academic cheating. The ‘same tests effect’ may appear if two or more groups of students are writing the same test one after the other and, as a result, a statistically significant improvement is detectable in the test scores of the second student group. The ‘revealed sameness effect’ is the impact of informing the students in some way that the test questions will be repeated. The ‘self selection effect’ arises when the students choose their examination turn themselves and this boosts their measured performance. The present study examines the three effects with independent t-tests and linear regression models on samples of 1221, 235, and 201 students (in this order), from four business courses in six academic semesters. The results do not support the ‘same test effect’, but support the ‘revealed sameness effect’ and the ‘self selection effect’.
Acknowledgements
The author would like to thank the help given by those who contributed to the data collection and to the discussion of the research problems and results: Edit Barizsné Hadházi, Éva Farkas, Anna Kapitány, and Ágnes Kotsis.