The impact of AI Chatbot-based blended learning on self-directed learning, academic engagement, and critical thinking disposition

Mehri, A.; Sadipour, E.; Dortaj, F

doi:10.22061/tej.2025.11481.3152

Articles in Press

Document Type : Original Research Paper

Authors

Department of Educational Psychology, Faculty of Psychology and Educational Sciences, Allameh Tabatabaei University, Tehran, Iran

https://doi.org/10.22061/tej.2025.11481.3152

Abstract

Background and Objectives: Artificial intelligence has significantly changed the perspectives and methods of education and learning as a new educational technology. A segment of artificial intelligence includes natural language processors that can produce human-like texts, such as AI chatbots. Due to their abilities to understand human language and answer users' questions, AI chatbots are recognized as important tools for education and learning. However, because the use of AI chatbots in education is still a recent development, there are currently very few studies on the effectiveness of this technology in learning. Accordingly, this research aims to examine the effects of AI chatbot-based blended learning in psychology on self-directed learning, student engagement, and critical thinking disposition among 11th-grade female students in the humanities in Zanjan.
Methods: This research was applied in purpose and quantitative in the nature of the collected data. It employed a quasi-experimental method with a pre-test and post-test design, involving a control group. The statistical population for this study comprised all 11th-grade female students in the humanities in Zanjan. A multi-stage cluster sampling technique was used, wherein one public girl’s high school in Zanjan, specifically Roghani Zanjani School, was randomly selected for the study, and two classes of students were assigned to the experimental and control groups. The sample consisted of 45 participants, divided into 21 in the experimental group and 24 in the control group. Data collection tools included the Self-Directed Learning Questionnaire by Fisher et al. (2001), the Academic Engagement Questionnaire by Reeve and Tseng (2011), and the Critical Thinking Disposition Questionnaire by Ricketts and Rudd (2005). The educational intervention spanned 8 sessions, each 90 minutes long, using the Sider software that offers and supports the ChatGPT AI chatbot. Given the multi-dimensional nature of all three dependent variables and the implementation of pre-tests, Multivariate Analysis of Covariance (MANCOVA) was utilized, and the pre-test and post-test data were analyzed with SPSS version 26.
Findings: The analysis of the data using statistical measures such as mean and standard deviation showed that the mean scores of the post-test for the experimental group increased in comparison to the control group across all three dependent variables when compared to the pre-test. Furthermore, the results from the multivariate analysis of covariance indicated a significant difference in self-directed learning, academic engagement, and critical thinking disposition between the experimental and control groups at the post-test (p < 0.01). Thus, given the higher mean scores of the experimental group in the post-test relative to those in the control group, it can be concluded that the AI chatbot-based blended learning approach in the psychology course proved effective in increasing self-directed learning, academic engagement, and critical thinking among 11th-grade female students in the humanities in Zanjan.
Conclusion: Drawing on the findings from this research, we can conclude that the use of AI chatbots in conjunction with the formal psychology curriculum has great potential to enhance self-directed learning, academic engagement, and critical thinking disposition. However, there is a concern that dependence on the AI chatbot could result in diminished academic engagement, with learners potentially accepting the output data without critical evaluation. Therefore, it is recommended that educators and learners utilize AI chatbots as educational complements and learning aids while considering ethical educational implications, stressing the importance of teacher supervision and instructing adolescents in self-regulation. It should also be noted that this study had limitations, including a lack of resources and prior research, restrictions on access to AI chatbots for IPs within Iran, and various hardware and software challenges in installing the AI chatbot software.

Keywords