Latvia began curriculum reform implementation in 2020. Not only has the learning content changed, but students are also now able to choose from one of three levels to study mathematics in secondary school. This requires new exams at the end of each of these courses. The aim of this study is to investigate student performance and to what extent students can use higher cognitive skills on the new exams. Results were analyzed using Classical Test Theory and Item Response Theory. Comparing the results of the new exams using Wright maps shows a trend that, although there are flaws, they are more representative of the relevant population than the previous exam. A group of experts determined the level of cognition for each of the tasks and analyzed students' performance. The results show that there are tasks where students can demonstrate the highest level of cognitive abilities. However, such tasks are relatively few, and they are mathematical rather than problems in a real context. Performance analysis shows that students have difficulty expressing their thoughts and this is one of the reasons why only a small number of students solve challenging problems

No abstract