Our platform’s proprietary mathematical model is a key feature that accurately estimates both student abilities and question difficulties. Our team of experts assigns difficulty levels to each question, and students’ abilities are estimated based on their performance in questions of varying difficulty levels.
It’s important to note that abilities and difficulties are measured on a scale from 0 to 10, with 0 representing the lowest possible level and 10 representing the highest.
Once the model estimates student abilities, it goes through a re-calibration process to adjust the difficulties of questions based on the newly estimated abilities of students, and vice versa. This iterative process continues until the values converge to stable values, which ensures the final values of question difficulties accurately represent the difficulty of each question and the final values of student abilities accurately represent each student’s true ability.
By regularly re-calibrating the model, we account for actual variations in students’ abilities over time, and these variations are documented in our platform to provide insights into student progress and track trends.
Our model’s ability to accurately estimate student abilities and question difficulties makes it an invaluable tool for educators and students alike.