Research & Case Studies
Explore in-depth research and comprehensive case studies analyzing student performance, problem-solving trends, and the educational impact of our international mathematical contest.
Article
Summary
Author
Publication
Published
Success of AI math solver tool in solving non-standard mathematics competition problems
Artificial intelligence is increasingly shaping how students learn and solve math problems. This paper evaluates the AI Math Solver on tasks from the 2024 Math Kangaroo competition across three grade levels. Problems were submitted as images in Serbian and English.
Results include qualitative and quantitative analyses: out of 84 tasks, the solver answered 24 correctly (just under 30%), with performance varying by language and difficulty level.
Marko Stanković, Aleksandar Milenković, Marina Svičević, Nemanja Vučićević
Facta Universitatis
2025
Evaluating the Success of AI Tools in Supporting Student Performance in Mathematical Kangaroo Competition
This study examines how generative AI tools support students preparing for math competitions, using Mathematical Kangaroo tasks in Serbian and English. Tools including ChatGPT, AI Math Solver, Math Mentor, and o1-preview were tested across formats and domains.
Results show higher accuracy on text-based problems, especially with o1-preview, while performance was weaker in geometry and logic. The findings highlight AI’s potential in math education and the need for better visual processing and language support.
Marina Svičević, Aleksandar Milenković, Nemanja Vučićević, Marko Stanković
Computer Applications in Engineering Education
2025
The impact of mathematics competitions on teachers and their classrooms in Puerto Rico, Switzerland and UK
This article reports a small comparative study on how math competitions affect teachers and classroom practice in Puerto Rico, Switzerland, and the UK. Surveys explored benefits, challenges, and preparation methods, as well as the role of competitions in developing students’ mathematical talent.
Although limited in scale, the study suggests directions for future research on the impact of math competitions in education.
Meike Akveld, Luis Caceres, David Crawford, Ferney Henao
ZDM, Mathematics Education, Springer (Vol.54, Nr.5, pp.941-959)
2022
The Road from Submission to Perfection
This article describes the problem-selection process for the Mathematical Kangaroo competition, from initial proposals to the final national version. It explains how problems are reviewed and modified, illustrating the typical dynamics of selecting competition tasks.
Robert Geretschläger, Lukas Donner
Mathematics Competitions (Vol.35, Nr.,1, pp.30-48 )
2022
Writing and choosing problems for a popular high school mathematics competition
This paper examines how competition math problems are created and how they differ from classroom tasks. It discusses curriculum links, sources of problem ideas, student motivation, difficulty levels, and mathematical precision in competitions.
Robert Geretschläger, Lukas Donner
ZDM-Mathematics Education , Springer (Vol. 54, Nr. 5, pp.971-982)
2022
Kangaroo on the chessboard as a didactical tool.
Mathematical literacy can be improved through board games such as chess. This paper studies mathematical chess problems for students aged 6–11 using a special piece called the kangaroo, introduced in the Mathematical Kangaroo competition. It examines domination, independence, and tour problems on a 5×5 chessboard, chosen for its suitability for younger students.
Karel Pastor
ELEMENTARY MATHEMATICS EDUCATION JOURNAL (Vol 3, Nr. 2, pp. 33-39, ISSN 2694-8133)
2021
Which test-wiseness based strategies are used by Austrian winners of the Mathematical Kangaroo?
This study examines test-wiseness strategies used in the Mathematical Kangaroo competition. Using the KATS questionnaire with Austrian winners (grades 3–13), it identified common test strategies and how students prepare for the contest.
Lukas Donner, Jakob Kelz, Elisabeth Stipsits and David Stuhlpfarrer
Mathematics Competitions (Vol 34, Nr. 1, pp. 88–101)
2021
