Comparing Expert and AI-Based Assessments in Evaluating Learner Knowledge

Λεπτομέρειες βιβλιογραφικής εγγραφής
Τίτλος: Comparing Expert and AI-Based Assessments in Evaluating Learner Knowledge
Γλώσσα: English
Συγγραφείς: Mohamad Al Assaad, Houssam Kanso
Πηγή: International Association for Development of the Information Society. 2026.
Διαθεσιμότητα: International Association for the Development of the Information Society. e-mail: secretariat@iadis.org; Web site: http://www.iadisportal.org
Peer Reviewed: Y
Page Count: 4
Ημερομηνία έκδοσης: 2026
Τύπος εγγράφου: Speeches/Meeting Papers
Reports - Evaluative
Descriptors: Artificial Intelligence, Student Evaluation, Evaluation Methods, Man Machine Systems, Computer Assisted Testing
Περίληψη: The rapid development of large language models (LLMs) has introduced new possibilities for educational assessment, particularly in the creation of learner evaluation tools. Traditionally, expert-designed tests have been the standard for assessing knowledge, relying on subject-matter expertise, pedagogical principles, and alignment with curriculum objectives. In contrast, LLMs such as ChatGPT can now generate assessment items on demand, offering scalability and adaptability but raising questions about validity, reliability, and educational value. This paper compares expert-created assessments with LLM-generated assessments to evaluate their effectiveness in measuring learner knowledge, in the computer science and project management fields. Through this comparison, we aim to highlight both the strengths and limitations of AI-based assessment generation and evaluation. The findings have implications for the future of educational assessment design, teacher practice, and the responsible integration of AI in learning environments. [For the full proceedings, see ED682246.]
Abstractor: As Provided
Entry Date: 2026
Αριθμός Καταχώρησης: ED682314
Βάση Δεδομένων: ERIC
Περιγραφή
Η περιγραφή δεν είναι διαθέσιμη