The Effectiveness of Artificial Intelligence-Based Recommendation Systems in an E-learning Environment in Developing Scientific Thinking Skills

Sayed, Radwa Amir Salah; Mehany, Eman Mohamed Makram; Yousef, Ahmed Mohamed Fahmy

doi:10.21608/sjseas.2025.406005

The Effectiveness of Artificial Intelligence-Based Recommendation Systems in an E-learning Environment in Developing Scientific Thinking Skills

Authors

¹ مدرس مساعد تكنولوجيا التعليم بكلية التربية النوعية جامعة الفيوم

² استاذ تكنولوجيا التعليم بكلية التربية النوعية جامعة الفيوم

³ استاذ تكنولوجيا التعليم المساعد بكلية التربية النوعية جامعة الفيوم

10.21608/sjseas.2025.406005

Abstract

This study set out to examine the effectiveness of AI-based recommendation systems in an e-learning environment, specifically in developing students' scientific thinking skills at Fayoum University, Egypt. The research explores how both content-based and collaborative recommendation systems can guide students to suitable educational materials that enhance their critical and analytical thinking abilities. Two primary hypotheses were tested:
The first hypothesis posits that there are statistically significant differences between the pre- and post-test scores of the scientific thinking test for the first experimental group (collaborative filtering), with the post-test showing improved results. Data analysis using the "T" test revealed significant differences at the 0.01 level in favor of the post-test, demonstrating the effectiveness of collaborative filtering methods in enhancing students' scientific thinking. These methods improved student interaction, problem-solving abilities, and communication skills, ultimately leading to better performance on the scientific thinking test after the experiment.
The second hypothesis suggests that there are no statistically significant differences between the average post-test scores for the first experimental group (collaborative filtering) and the second experimental group (content-based filtering). Statistical analysis confirmed that no significant differences existed between the two groups, although the content-based filtering group showed slightly higher average scores for educational recommendations than the collaborative filtering group. However, these differences in scientific reasoning scores between the two groups did not yield any statistically significant effects. This suggests that both recommendation systems contributed to enhancing students’ scientific reasoning skills, though the higher volume of recommendations in the collaborative filtering environment led to some confusion for students, making it more challenging for them to engage effectively with the recommended content.

Keywords