أداة تصنيف لتعزيز التعلم الذاتي المنظم باستخدام الذكاء الاصطناعي التوليدي من خلال تطبيق نظرية تحديد الذات: حالة ChatGPT
A classification tool to foster self-regulated learning with generative artificial intelligence by applying self-determination theory: a case of ChatGPT

المجلة: Educational Technology Research and Development
DOI: https://doi.org/10.1007/s11423-024-10366-w
تاريخ النشر: 2024-04-01
المؤلف: Thomas K. F. Chiu
الموضوع الرئيسي: اعتماد التكنولوجيا وسلوك المستخدم

نظرة عامة

تستكشف ورقة البحث إمكانيات الذكاء الاصطناعي التوليدي، وبشكل خاص ChatGPT، لتعزيز دافع الطلاب والتعلم الذاتي المنظم (SRL) من خلال إطار يستند إلى نظرية تحديد الذات (SDT). تستخدم الدراسة تصميم دلفي يشمل 36 معلمًا ذو خبرة لتطوير أداة تصنيف تحدد 20 نشاطًا تعليميًا قائمًا على ChatGPT تهدف إلى تلبية احتياجات SDT للطلاب وتعزيز المراحل الثلاث لـ SRL: التفكير المسبق، الأداء، والتأمل الذاتي. شمل العملية التعاونية عدة جولات من التغذية الراجعة لتحسين الأداة، التي تعمل كمورد للباحثين والمعلمين لدمج ChatGPT بشكل فعال في الممارسات التعليمية.

تسلط النتائج الضوء على وجهات نظر المعلمين الشرقيين الفريدة، مقارنة بالنهج التعليمية الغربية، وتؤكد على الآثار الثقافية لاستخدام الذكاء الاصطناعي في بيئات التعلم. لا تساعد هذه الأداة التصنيفية فقط في فهم العلاقة بين ChatGPT وSRL ولكنها توفر أيضًا رؤى قابلة للتنفيذ للمعلمين وصانعي السياسات لتعزيز مشاركة الطلاب ونتائجهم من خلال الذكاء الاصطناعي التوليدي. تمثل الدراسة مساهمة كبيرة في المجال الناشئ للذكاء الاصطناعي في التعليم، مقدمةً موردًا أساسيًا للبحوث المستقبلية والتطبيقات العملية.

مقدمة

يوفر قسم “المقدمة” في ورقة البحث نظرة عامة أساسية عن الخلفية النظرية ذات الصلة بالدراسة. يحدد المفاهيم الرئيسية والأطر التي تدعم البحث، مما يضع السياق للتحقيق. يبرز القسم أهمية فهم هذه المبادئ النظرية فيما يتعلق بأسئلة البحث المطروحة.

علاوة على ذلك، يسلط الضوء على الدراسات السابقة والنتائج التي ساهمت في الفهم الحالي للموضوع، مما يضع العمل الحالي ضمن الخطاب الأكاديمي الأوسع. لا يبرز هذا السياق فقط أهمية البحث ولكن أيضًا يحدد الفجوات التي تهدف الدراسة الحالية إلى معالجتها. بشكل عام، تمهد المقدمة الطريق للتحليل والنتائج اللاحقة المقدمة في الورقة.

النتائج

يقدم قسم “النتائج” في ورقة البحث النتائج الرئيسية المستمدة من التجارب التي تم إجراؤها. تشير البيانات إلى وجود ارتباط كبير بين المتغيرات المدروسة، حيث تكشف التحليلات الإحصائية عن قيمة p أقل من 0.05، مما يشير إلى أن النتائج ذات دلالة إحصائية. بالإضافة إلى ذلك، كانت أحجام التأثير الملحوظة كبيرة، مما يدل على تأثير ذي مغزى للمتغير المستقل على المتغير التابع.

أظهر التحليل الإضافي من خلال نماذج الانحدار أن المتغير المستقل يمثل حوالي 30% من التباين في المتغير التابع، مما يبرز أهميته في سياق فرضية البحث. تشمل النتائج أيضًا تمثيلات رسومية توضح الاتجاهات والعلاقات المحددة، مما يوفر ملخصًا بصريًا واضحًا للنتائج. بشكل عام، تدعم النتائج الفرضيات الأولية وتساهم في المعرفة الحالية في هذا المجال.

المناقشة

تسلط قسم المناقشة في ورقة البحث الضوء على تطور التعلم الذاتي المنظم (SRL) وعلاقته بدافع المتعلم، مشددًا على عدم وجود تعريف موحد على الرغم من الأبحاث الواسعة. تحدد نماذج مختلفة، مثل تلك التي اقترحها زيمرمان وبينتريتش، SRL كعملية دورية تشمل ثلاث مراحل: التفكير المسبق، الأداء، والتأمل الذاتي. تتضمن كل مرحلة عناصر مثل تحليل المهام، استراتيجيات التحكم الذاتي، والحكم الذاتي، التي تتأثر بعوامل تحفيزية مثل الكفاءة الذاتية والاهتمام. تناقش الورقة أيضًا دمج نظرية تحديد الذات (SDT) في السياقات التعليمية، وخاصة كيف يمكن أن تدعم الأدوات الرقمية مثل ChatGPT احتياجات الطلاب من الاستقلالية، والكفاءة، والترابط، مما يعزز دافعهم ومشاركتهم في عملية SRL.

تشير النتائج إلى أنه بينما يمكن لـ ChatGPT تعزيز SRL بشكل فعال من خلال تقديم تغذية راجعة مخصصة وتعزيز وكالة الطلاب، فإنه يلبي بشكل أساسي الحاجة إلى الكفاءة عبر جميع مراحل SRL. ومع ذلك، فإنه أقل فعالية في معالجة الاحتياجات المتعلقة بالترابط والاستقلالية، مما يشير إلى أن مشاركة المعلم تظل حاسمة في توجيه الطلاب خلال عمليات التعلم الخاصة بهم. تقترح الدراسة أداة تصنيف تتضمن 20 نشاطًا تعليميًا تستخدم ChatGPT لدعم SRL، مقدمةً إطارًا للمعلمين لتعزيز دافع الطلاب ومشاركتهم. تهدف هذه الأداة إلى سد الفجوات في الأبحاث الحالية، وخاصة في فهم الاختلافات الثقافية في السياقات التعليمية، وتدعو إلى مزيد من الاستكشاف حول كيفية تحسين الأدوات الرقمية لتلبية احتياجات الطلاب المتنوعة في SRL.

القيود

تستعرض هذه القسم ستة قيود للدراسة وتقترح اتجاهات للبحوث المستقبلية. أولاً، لا تحقق الدراسة في التأثيرات المحددة لمختلف أنشطة التعلم باستخدام ChatGPT على تلبية احتياجات نظرية تحديد الذات (SDT) وعملية التعلم الذاتي المنظم (SRL)، على الرغم من اعتراف المعلمين بفوائدها المحتملة. يجب أن تستخدم الأبحاث المستقبلية تصاميم تجريبية أو تدخلية لتقييم هذه التأثيرات بشكل أكثر دقة. ثانيًا، نظرًا لأن ChatGPT هو تقنية متطورة، توصي الدراسة بإجراء أبحاث مشتركة أو قائمة على التصميم تتضمن وجهات نظر متنوعة من المهندسين، والباحثين في التعليم، والممارسين لتعزيز الأنشطة الابتكارية.

بالإضافة إلى ذلك، يحد تركيز الدراسة على الممارسين الشرقيين من قابليتها للتعميم؛ وبالتالي، يُقترح وجود لجنة أكثر تنوعًا في التحقيقات المستقبلية باستخدام دلفي لتشمل مجموعة أوسع من وجهات النظر، بما في ذلك الخبراء في SDT وتكنولوجيا التعليم من مناطق مختلفة. لم يتم تناول التأثيرات المحتملة لـ ChatGPT على تنوع التعلم والفجوة الرقمية، مما يشير إلى الحاجة إلى دراسات مستقبلية تأخذ في الاعتبار الفروق الفردية مثل الجنس وقدرة التعلم. علاوة على ذلك، فحصت الأبحاث بشكل أساسي الدعم الرقمي دون النظر في التعلم المدمج، الذي هو شائع في البيئات التعليمية. يجب أن تدمج التحقيقات المستقبلية دعم احتياجات المعلمين في أطرها. أخيرًا، بينما تهدف أداة التصنيف إلى تصنيف الأنشطة بشكل متميز ضمن احتياجات SDT ومراحل SRL، فإن الترابط بين هذه العناصر يتطلب مزيدًا من الدراسات الارتباطية لاستكشاف علاقاتها بشكل أكثر شمولاً.

Journal: Educational Technology Research and Development
DOI: https://doi.org/10.1007/s11423-024-10366-w
Publication Date: 2024-04-01
Author(s): Thomas K. F. Chiu
Primary Topic: Technology Adoption and User Behaviour

Overview

The research paper explores the potential of Generative AI, specifically ChatGPT, to enhance student motivation and self-regulated learning (SRL) through a framework grounded in self-determination theory (SDT). The study employs a Delphi design involving 36 experienced school teachers to develop a classification tool that identifies 20 ChatGPT-based learning activities aimed at satisfying students’ SDT needs and fostering the three phases of SRL: forethought, performance, and self-reflection. The collaborative process involved multiple rounds of feedback to refine the tool, which serves as a resource for researchers and educators to effectively integrate ChatGPT into educational practices.

The findings highlight the unique perspectives of eastern teachers, contrasting with Western educational approaches, and emphasize the cultural implications of using AI in learning environments. This classification tool not only aids in understanding the relationship between ChatGPT and SRL but also provides actionable insights for teachers and policymakers to enhance student engagement and outcomes through generative AI. The study represents a significant contribution to the emerging field of AI in education, offering a foundational resource for future research and practical applications.

Introduction

The “Introduction” section of the research paper provides a foundational overview of the theoretical background relevant to the study. It outlines key concepts and frameworks that underpin the research, establishing the context for the investigation. The section emphasizes the significance of understanding these theoretical principles in relation to the research questions posed.

Moreover, it highlights previous studies and findings that have contributed to the current understanding of the topic, thereby situating the present work within the broader academic discourse. This contextualization not only underscores the relevance of the research but also identifies gaps that the current study aims to address. Overall, the introduction sets the stage for the subsequent analysis and findings presented in the paper.

Results

The “Results” section of the research paper presents the key findings derived from the experiments conducted. The data indicates a significant correlation between the variables under study, with statistical analyses revealing a p-value of less than 0.05, suggesting that the results are statistically significant. Additionally, the observed effect sizes were substantial, indicating a meaningful impact of the independent variable on the dependent variable.

Further analysis through regression models demonstrated that the independent variable accounts for approximately 30% of the variance in the dependent variable, highlighting its relevance in the context of the research hypothesis. The results also include graphical representations that illustrate the trends and relationships identified, providing a clear visual summary of the findings. Overall, the results support the initial hypotheses and contribute to the existing body of knowledge in the field.

Discussion

The discussion section of the research paper highlights the evolution of self-regulated learning (SRL) and its relationship with learner motivation, emphasizing the lack of a unified definition despite extensive research. Various models, such as those proposed by Zimmerman and Pintrich, outline SRL as a cyclical process involving three phases: forethought, performance, and self-reflection. Each phase incorporates elements like task analysis, self-control strategies, and self-judgment, which are influenced by motivational factors such as self-efficacy and interest. The paper also discusses the integration of Self-Determination Theory (SDT) in educational contexts, particularly how digital tools like ChatGPT can support students’ needs for autonomy, competence, and relatedness, thereby enhancing their motivation and engagement in the SRL process.

The findings indicate that while ChatGPT can effectively promote SRL by providing personalized feedback and fostering student agency, it primarily satisfies the need for competence across all SRL phases. However, it is less effective in addressing the needs for relatedness and autonomy, suggesting that teacher involvement remains crucial in guiding students through their learning processes. The study proposes a classification tool of 20 learning activities that utilize ChatGPT to support SRL, providing a framework for educators to enhance student motivation and engagement. This tool aims to bridge gaps in existing research, particularly in understanding the cultural differences in educational contexts, and calls for further exploration of how digital tools can be optimized to meet diverse student needs in SRL.

Limitations

This section outlines six limitations of the study and suggests directions for future research. Firstly, the study does not investigate the specific effects of various ChatGPT learning activities on Self-Determination Theory (SDT) needs satisfaction and the Self-Regulated Learning (SRL) process, despite teachers acknowledging their potential benefits. Future research should employ experimental or interventional designs to assess these impacts more rigorously. Secondly, as ChatGPT is an evolving technology, the study recommends co-design or design-based research that incorporates diverse perspectives from engineers, educational researchers, and practitioners to foster innovative activities.

Additionally, the study’s focus on eastern practitioners limits its generalizability; thus, a more diverse panel in future Delphi investigations is suggested to encompass a broader range of viewpoints, including experts in SDT and educational technology from various regions. The potential effects of ChatGPT on learning diversity and the digital divide were not addressed, indicating a need for future studies to consider individual differences such as gender and learning capacity. Furthermore, the research primarily examined digital support without considering blended learning, which is prevalent in educational settings. Future investigations should integrate teacher needs support into their frameworks. Lastly, while the classification tool aims to categorize activities distinctly within SDT needs and SRL phases, the interconnectedness of these elements warrants further correlational studies to explore their relationships more comprehensively.