Abstract
Italy and other Western countries face significant challenges in teaching STEM
subjects, particularly mathematics, efficiently. Despite the growing demand for mathematical and quantitative skills, there remains a substantial gap in meeting this need with effective tools and methods. This study, conducted within the RAISE project funded by the PNRR (Robotics and AI for Socio-economic Empowerment, PNRR, Mission 4 Education and Research, Territorial Ecosystem for the Liguria Region), focuses on the development of an adaptive test to evaluate mathematical proficiency in middle school students. The test was
designed in collaboration with teachers to ensure adherence to school curricula, employing expert query techniques and intra-rater reliability analysis to finalize the problem set. The adaptive version of the test was developed using the framework of Knowledge Space Theory. Extensive simulations were conducted to evaluate the test’s efficiency and accuracy, followed by a statistical analysis to identify the optimal assessment parameters for evaluating proficiency. The results highlight the potential of adaptive testing not only as a precise tool
for individualized formative assessment but also as a valuable complement to traditional summative assessments in classroom settings.
subjects, particularly mathematics, efficiently. Despite the growing demand for mathematical and quantitative skills, there remains a substantial gap in meeting this need with effective tools and methods. This study, conducted within the RAISE project funded by the PNRR (Robotics and AI for Socio-economic Empowerment, PNRR, Mission 4 Education and Research, Territorial Ecosystem for the Liguria Region), focuses on the development of an adaptive test to evaluate mathematical proficiency in middle school students. The test was
designed in collaboration with teachers to ensure adherence to school curricula, employing expert query techniques and intra-rater reliability analysis to finalize the problem set. The adaptive version of the test was developed using the framework of Knowledge Space Theory. Extensive simulations were conducted to evaluate the test’s efficiency and accuracy, followed by a statistical analysis to identify the optimal assessment parameters for evaluating proficiency. The results highlight the potential of adaptive testing not only as a precise tool
for individualized formative assessment but also as a valuable complement to traditional summative assessments in classroom settings.
| Original language | English |
|---|---|
| Title of host publication | IES 2025 - Innovation & Society: Statistics and Data Science for Evaluation and Quality |
| Subtitle of host publication | BOOK OF SHORT PAPERS |
| Pages | 156-162 |
| ISBN (Electronic) | 978 88 5495 849 4 |
| Publication status | Published - 25 Jun 2025 |
