Blog

Adaptive Assessment: a tailored evaluation for each student

Knowledge assessment is a crucial stage in the learning journey. It allows the teacher to make sure the student is on the right track for the assimilation of concepts. Education sector has faced quite a lot of technical transformations in the past decades, from the digital whiteboard and the online classroom to virtual reality applied to education. Evaluation has also transformed thanks to new technologies and adaptive assessment solutions have appeared. Domoscio uses this technology in its Adaptive Assessment module, allowing teachers to continuously assess students’ level of proficiency thanks to a skills-based approach.

Adaptive assessment: what is it?

Traditional assessment methods such as one exam fits all are limited. Same questions, same duration for each student… lead to limited information on the level of proficiency and the progress of each one. The final score doesn’t take into account the level of difficulty of each question or the progress of the student compared to their initial state.

Currently, adaptive assessment solutions based on artificial intelligence are considered^1,2 among the best tools to assess student’s performance. These tests automatically select the questions used to assess the level of proficiency of the student. They are efficient and accurate. That’s why a lot of international exams use them, such as TOEFL, TOEIC, SAT, GMAT…

An adaptive test starts with a question with an average difficulty level. The questions come from bank of questions whose level of difficulty is known. According to the student’s answer, the level of difficulty of the next questions is increased or decreased. If the student correctly answers to one particular element, the next will be slightly more difficult. On the opposite, if they fail, the next will be easier. Thus, the test adapts to the cognitive ability of the student. It stops when the accuracy of the estimated level of proficiency is high enough.

Benefits of adaptive assessment compared to classic methods

Domoscio Spark Adaptive Assessment module allows to continuously evaluate students thanks to a skills-based approach. Student and teacher are able to take full benefits of this method.

Benefits for the student

Adaptive evaluations provide students with questions based on their knowledge and understanding of a concept. Thanks to Adaptive Assessment, they can quickly identify their strength and weak points and focus on the last ones in their study. Thus, education is tailored to meet the needs of each individual and focus on the real level of the student to build the pedagogic answer. According to studies^3,4, adaptive learning systems increase the motivation of the students. By providing them with questions adapted to their level of proficiency, it stimulates the students without frustrating them. In other words, questions are hard enough to challenge them but easy enough not to scare them.

Benefits for the teacher

Adaptive assessment provides the teacher with an accurate picture of the student’s progress: thanks to report and analysis, they understand better the student’s evolution over time and level of proficiency. Thus, they can quickly identify weak points of each of them and build an adapted educational answer.
These technologies almost completely prevent from cheating. Each student answers a personalized test with different questions. It is then almost impossible to cheat. This is why more and more evaluation and exams providers choose this solution.

Other assessment methods

Domoscio Spark provides its user with a 360° approach based on artificial intelligence. Adaptive assessment algorithms are used in the Adaptive Assessment module but it is not the only way to assess a student’s level of proficiency on a concept. Domoscio Spark allows to combine several evaluation methods in a complex assessment system.
For example, self-positioning allows the student to decide themselves their level. They choose how well they know and have assimilated a concept. Combined with Adaptive Assessment, self-positioning allows to identify students with an unrealistic self-image, the ones who lack self confidence in the classroom or the ones who think they master a subject better than they actually do.
External assessment from a teacher or peer can also be taken into account in Domoscio Spark. This external data provides the system with precious information from an expert or another student.

Conclusion

Adaptive assessment, combined with other assessment techniques, provides the teacher with a good understanding of the level of proficiency of their students and allows them to adapt to pedagogic answer to the need of each of them. This adapted answer can be automatized thanks to artificial intelligence. This is the promise of another module of Domoscio Spark: Smart Tutoring.

¹ Jiao, H., & Lissitz, R. W. (Eds.). (2020). Application of artificial intelligence to assessment. Greenwich, CT: Information Age Publishing.

² Thompson, G. (2016). Computer adaptive testing, big data and algorithmic approaches to education. British Journal of Sociology of Education, 38(6), 827–840. doi:10.1080/01425692.2016.1158640

³ The Utility of Adaptive Testing in Addressing the Problem of Unmotivated Examinees (2014). In Journal of Computerized Adaptive Testing (Vol. 2, Issue 1). International Association for Computerized Adaptive Testing. https://doi.org/10.7333/1401-02010001

⁴ Martin, A. J., & Lazendic, G. (2018). Computer-adaptive testing: Implications for students’ achievement, motivation, engagement, and subjective test experience. Journal of Educational Psychology, 110(1), 27–45. https://doi.org/10.1037/edu0000205

---