| Autor: |
Kuzmenko, R. V., Moiseev, S. I., Zolototrubov, D. Yu., Sannikov, V. G., Shcheka, A. A. |
| Předmět: |
|
| Zdroj: |
AIP Conference Proceedings; 11/1/2022, Vol. 2647 Issue 1, p1-5, 5p |
| Abstrakt: |
The paper analyzes the methods of processing expert information obtained under heterogeneous conditions. The methods are described on the example of serial tests of students, when it is impossible to create the same conditions in each series. Two mathematical models for obtaining a generalized assessment of the subjects are proposed. The first model is based on the fact that atypical particular results associated with the influence of external factors will be taken into account in the overall assessment, depending on how much they correlate with the assessments of the entire group of subjects. The second model suggests introducing a test complexity indicator, which is determined based on the results shown by the entire group of subjects. Depending on the complexity of the tests, its weight and contribution to the overall result are formed. As the conducted computational experiments have shown, both models can significantly smooth out the influence of external factors and internal conditions on the test results. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
