The object of research is a statistics exam that contains problem tasks. One examiner performed two exam evaluation methods to repeatedly evaluate the exam. The goal was to compare the methods for objectivity. One of the two exam evaluation methods we call a serial evaluation method. The serial evaluation method assumes evaluation of all exam tasks of an individual student in sequential order, so evaluation "student by student". Unlike that, a parallel evaluation method assumes the "task by task" evaluation of exams for the whole group of students. A paired samples analysis of exam results indicates a statistically significant difference between methods. Further analysis showed a statistically significant difference in results obtained by the serial evaluation and the repeated serial evaluation, while the difference in results obtained by the parallel evaluation and the repeated parallel evaluation turned out not to be statistically significant. Furthermore, our research gave evidence that the repeated parallel evaluation changes result significantly less than the repeated serial evaluation. Consequently, we consider the parallel evaluation a more objective tool in exam evaluation than the serial evaluation. [For the complete proceedings, see ED631021.]