The paper presents and discusses the results of statistical analysis of differences among scores obtained by students of different faculties of the University of Economics in Prague. The analysed dataset contains the scores for 2256 students that took basic mathematics course during the academic year 2013/2014. A two way analysis of variance was performed with semester and faculty as main factors. The interaction between these two factors was also considered. Students have to take two tests. At first, the sum of the scores obtained from both tests is analysed and then, the two tests are analysed separately. It turns out that the significance of factors is the same in the three analyses. The assumptions of linear models are verified. Due to problem of heteroscedasticity, weighted least squares are used and the possibility of using Box-Cox transformation is also discussed, as the errors are not normally distributed. Finally, the differences between the faculties are described.