Artificial Intelligence Training as a Catalyst for the Transformation of Self-Regulated Learning: A Mixed-Methods Study with University Students

European Conference on Educational Research, Tampere, Finlandiya, 18 - 21 Ağustos 2026, ss.1-2, (Tam Metin Bildiri)

Yayın Türü: Bildiri / Tam Metin Bildiri
Basıldığı Şehir: Tampere
Basıldığı Ülke: Finlandiya
Sayfa Sayıları: ss.1-2
TED Üniversitesi Adresli: Evet

Özet

Seda Aydan

TED University, Turkey (Türkiye)

Presenting Author: Aydan, Seda

Artificial intelligence (AI) tools increasingly shape how university students plan, monitor, and evaluate their learning processes, yet empirical evidence on how AI training influences self-regulated learning (SRL) remains scarce. This study investigates the effect of structured AI training on students’ SRL skills and explores how SRL strategies evolve in AI-mediated learning environments. Grounded in Zimmerman’s (2002) social-cognitive model of SRL, the study employed a mixed-methods design consisting of a pre–post intervention with a validated SRL scale and follow-up qualitative interviews. The research was conducted with 302 undergraduate students aged 18–24 from a mid-sized university during an ELT course. At the beginning of the semester, students completed the Self-Regulated Learning Skills Scale developed by Erdoğan and Senemoğlu (2016), a 67-item instrument with factor loadings ranging from .47 to .91 and high internal consistency (Cronbach’s α = .91). Its reliability and comprehensive coverage of SRL components made it an appropriate measure for the present study.

Students received eight weeks of AI training designed to teach how AI tools can be used across Zimmerman’s (2000) three SRL phases: forethought (goal setting, strategic planning), performance (task strategies, self-monitoring), and self-reflection (self-evaluation, causal attribution). Training modules covered effective prompt writing, the use of generative AI for planning complex assignments, developing task strategies, generating study materials, practicing skills, retrieving tailored explanations, and receiving formative feedback. The intervention explicitly emphasized AI not as a shortcut but as a cognitive partner supporting metacognitive control and strategic learning.

Paired-samples t-tests were used to compare pre- and post-test scores. Findings showed statistically significant improvements in self-monitoring (t(301) = 4.72, p < .001), self-evaluation (t(301) = 2.89, p < .01), and self-efficacy (t(301) = 3.11, p < .01). However, no significant differences emerged in environmental structuring (t(301) = 0.84, p = ns), resource finding (t(301) = 1.02, p = ns), repetition and memorization (t(301) = 0.57, p = ns), or task strategies (t(301) = 0.93, p = ns). These results suggest that while AI training directly strengthened students’ metacognitive dimensions of SRL, behavioral and environmental aspects showed no improvement when measured through traditional SRL constructs.

To better understand these discrepancies, 10 semi-structured interviews were conducted. The content analysis revealed that although quantitative scores did not reflect change in certain SRL categories, students had in fact adopted new AI-mediated strategies not captured by the scale. For instance, instead of structuring their physical study environment, students structured their cognitive environment through personalized AI-generated summaries, practice exercises, and reinforcement materials. Many fed course content directly into AI systems to generate quizzes, explanations, and step-by-step problem-solving guides, which they printed and archived. Resource finding was transformed from library-based searches to real-time AI-supported information retrieval. Repetition and memorization were replaced by interactive AI dialogues, conversational practice, and iterative questioning. Students noted that asking “embarrassing questions” felt safer with AI than with peers or instructors, which strengthened persistence and confidence.

These qualitative insights point to an ongoing transformation of SRL in AI-enabled learning ecosystems. Behavioral SRL dimensions may manifest differently in AI environments, rendering traditional SRL constructs insufficient. AI-assisted SRL involves hybrid strategies which can also be referred as personalized knowledge reinforcement, dialogic practice, dynamic resource generation that existing scales do not measure. Consequently, the study calls for an updated SRL theoretical model that integrates AI-mediated cognitive scaffolding, as well as the development of a new measurement tool capturing emerging AI-assisted SRL behaviors. Overall, the study provides empirical evidence that structured AI training enhances core metacognitive SRL skills while simultaneously reshaping how students plan, manage, and evaluate their learning. These findings have significant implications for curriculum design, teacher training, educational policy, and future SRL theory development.

Methodology, Methods, Research Instruments or Sources Used
Method
This research adopted a mixed-methods sequential explanatory design to examine the impact of AI training on students’ self-regulated learning skills and to capture emerging AI-mediated SRL behaviors not reflected in existing measurement tools. Quantitative data were collected first, followed by qualitative interviews that provided interpretive depth.
Participants
Participants consisted of 302 undergraduate students aged 18–24 enrolled in various departments at a mid-sized university. Participation was voluntary and approved by the institutional ethics committee. Students represented diverse academic areas, which allowed for broad insights into AI-supported learning behaviors.
Instruments
The quantitative component employed the Self-Regulated Learning Skills Scale (Erdoğan & Senemoğlu, 2016). The scale includes 67 items rated on a Likert scale and covers multiple SRL domains including goal setting, planning, task strategies, resource finding, repetition and memorization, environmental structuring, self-monitoring, self-evaluation, and self-efficacy. Factor loadings range from .47 to .91 and internal consistency is high (Cronbach’s α = .91). Given its reliability and comprehensive coverage, the scale was selected to assess SRL before and after the intervention.
Intervention
The intervention consisted of an eight-week AI training program designed according to Zimmerman’s (2000) SRL model. Weekly modules provided instruction and hands-on practice in prompt engineering, AI-supported goal setting, strategic planning, task decomposition, self-monitoring through iterative dialogue with AI, and self-evaluation using AI-generated feedback. Students practiced integrating AI with their own academic tasks, creating customized study materials, and applying SRL strategies with AI assistance.
Data Collection
The SRL scale was administered at the beginning and 8 weeks later after the begining of the semester. Ten students were then selected for semi-structured interviews based on variation in academic major and initial SRL scores to ensure maximum variety. Interview questions explored students’ use of AI tools, changes in study habits, perceptions of SRL, and strategies developed throughout the training.
Data Analysis
Paired-samples t tests were conducted to examine mean differences between pre-test and post-test scores. Significant increases were observed in self-monitoring (t(301) = 4.72, p < .001), self-evaluation (t(301) = 2.89, p < .01), and self-efficacy (t(301) = 3.11, p < .01). No statistically significant differences emerged for environmental structuring, resource finding, repetition and memorization, or task strategies (all p > .05).
Qualitative data were analyzed using content analysis. Codes were developed inductively, categorized, and grouped into themes. Reliability was ensured through researcher triangulation and iterative coding.

Conclusions, Expected Outcomes or Findings
This study provides empirical evidence that structured AI training has a measurable and meaningful influence on university students’ self-regulated learning skills. While quantitative results showed significant improvements in metacognitive dimensions such as self-monitoring, self-evaluation, and self-efficacy, the qualitative findings revealed that AI training also reshaped behavioral and strategic components of SRL in ways not captured by traditional scales. Students increasingly adopted AI-mediated strategies such as generating personalized practice materials, using AI to retrieve sources efficiently, engaging in conversational practice, archiving AI-generated summaries, and using AI as a non-judgmental support tool for clarifying difficult concepts. These strategies suggest a shift from physical or environmental regulation to cognitive-technological regulation, where AI acts as both a resource and a metacognitive partner.
The discrepancy between quantitative and qualitative results highlights a conceptual limitation: existing SRL frameworks and scales were developed before the emergence of generative AI and thus are insufficient for measuring AI-assisted learning strategies. AI transforms how students plan, enact, and reflect upon learning processes, necessitating theoretical adaptation. The findings indicate a need for a revised SRL model that incorporates AI-supported metacognitive scaffolding, dynamic resource generation, and interactive learning mechanisms. Furthermore, a new measurement scale is required to capture the unique characteristics of AI-mediated SRL.
Overall, the study underscores that AI is not merely a technological tool but a catalyst for the evolution of self-regulation. Educators and curriculum designers must reconsider how SRL is taught, supported, and assessed in AI-rich learning environments. The results strongly advocate for integrating AI literacy into higher education curricula to empower learners as strategic, reflective, and autonomous users of emerging technologies.

References
Azevedo, R., & Hadwin, A. F. (2005). Scaffolding self-regulated learning and metacognition: Implications for the design of computer-based scaffolds. Instructional Science, 33, 367–379.
Erdoğan, S., & Senemoğlu, N. (2016). Self-regulated learning skills scale: Development, validity, and reliability. Education and Science, 41(183), 1–24.
Lan, M., & Zhou, X. (2025). A qualitative systematic review on AI empowered self-regulated learning in higher education. npj Science of Learning, 10(1), 21.
Molenaar, I., de Mooij, S., Azevedo, R., Bannert, M., Järvelä, S., & Gašević, D. (2023). Measuring self-regulated learning and the role of AI: Five years of research using multimodal multichannel data. Computers in Human Behavior, 139, 107540.
Panadero, E. (2017). A review of self-regulated learning: Six models and four directions for research. Frontiers in Psychology, 8, 422. https://doi.org/10.3389/fpsyg.2017.00422
Pintrich, P. R. (2000). The role of goal orientation in self-regulated learning. In M. Boekaerts, P. R. Pintrich, & M. Zeidner (Eds.), Handbook of self-regulation (pp. 451–502). Academic Press.
Schunk, D. H., & Greene, J. A. (Eds.). (2018). Handbook of self-regulation of learning and performance (2nd ed.). Routledge.
Zimmerman, B. J. (2000). Attaining self-regulation: A social-cognitive perspective. In M. Boekaerts, P. Pintrich, & M. Zeidner (Eds.), Handbook of self-regulation (pp. 13–39). Academic Press.
Zimmerman, B. J. (2002). Becoming a self-regulated learner: An overview. Theory Into Practice, 41(2), 64–70.
Intent of Publication