MetaClassroom: A New Paradigm and Experience for Programming Education

Abstract

This study explored the impact of MetaClassroom, a virtual immersive programming learning environment designed based on the three-dimensional learning progression (3DLP) concept, on students’ multidimensional development. Utilizing a quasi-experimental research design, this study compared students’ programming learning achievements (PLA), self-regulated learning (SRL) skills, learning beliefs, and learning motivation in MetaClassroom with those in traditional classroom settings. The findings revealed that MetaClassroom significantly enhanced students’ PLA and SRL skills, particularly in subdimensions including Metacognitive Skills, Persistence, and Seeking Help. Additionally, MetaClassroom positively impacted students’ learning beliefs and learning motivation, demonstrating its potential in optimizing knowledge acquisition, application processes, and fostering students’ higher-order thinking skills. By integrating the 3DLP concept into targeted programming learning environments, MetaClassroom created an innovative immersive learning ecosystem, that bridged theory and practice, offering students a comprehensive and engaging platform to develop both foundational knowledge and practical skills. This study not only validated the effectiveness of MetaClassroom in improving students’ programming learning performance and learning experience, but also introduced a new paradigm that integrated teaching, learning, and assessment. The paradigm offers new insights and directions for both programming education and broader educational practices, paving the way for future developments in educational technology.

Keywords

three-dimensional learning progression programming education self-regulated learning learning belief learning motivation metaclassroom

Introduction

With the continuous advancement of globalization and technological innovation, the importance of programming education is rapidly increasing. It has gone beyond skill training, becoming a core element of the 21st-century education system (Chen et al., 2020; Jing et al., 2024b). In many countries and regions, programming education has been integrated into national education policies to cultivate a new generation of innovators and problem solvers (Tellhed et al., 2022). Since 2013, the UK government has included computer science in the national curriculum, requiring all students to learn programming fundamentals (Department for Education, 2013). In 2016, Australia officially incorporated programming into the national curriculum and, in 2018, released the National STEM Strategy 2019–2023 (Engineers Australia, 2018). This strategy aimed to advance children’s programming education, so as to cultivate innovative future technology professionals. Around 2017, the United States became involved in the initiative to develop students’ programming skills (The White House, 2017). The Federal Government’s five-year strategic plan for STEM education, for example, emphasized the vital role that programming education played in furthering the country’s technological agenda (The White House, 2018). Despite these progressive policies, little research has examined how these national-level initiatives were translated into classroom practices and whether programming education effectively addressed the developmental needs of students in diverse educational settings.

With the emergence of generative AI technology, the critical role that programming education plays at the foundational education stage has been re-emphasized globally (Jin et al., 2025; Liu & Li, 2024; Wang et al., 2025). Organizations such as UNESCO and EDUCAUSE released a series of educational development guidelines, notably including Generative AI and the Future of Education (UNESCO, 2023) and 2023 EDUCAUSE Horizon Report: Teaching and Learning Edition (EDUCAUSE, 2023). These guidelines emphasized the development of learners’ programming skills, computational thinking, and higher-order critical thinking skills through scenario-based and practice-oriented learning. Programming education has undoubtedly become a key route to cultivating high-quality talent and satisfying the demands on human resources required for the advancement of society. The review of the aforementioned policies and strategies highlighted a significant shift in the perception of programming education: once the exclusive domain of a few tech elites, it is now globally recognized as critical to the development of future citizens’ core competencies. The implementation of these policies has further elevated the role of programming education within the educational system. Consequently, for the new generation of learners, learning programming has become an inevitable choice in keeping with the current trends (Sung et al., 2024). However, current guidelines primarily focused on policy-level advocacy, with insufficient attention to the challenges learners face in acquiring programming skills or how to create effective, scalable programming instruction environments.

A brief review of the development of programming education revealed remarkable changes over the past 20 years. Innovated teaching models that integrate information and communication technologies (ICT) such as virtual campuses, Moodle platforms, multi-core processors in laptops, and programming modes have been proven to effectively enhance learners’ programming performance (Capel et al., 2017). More specifically, Chen et al. (2020) developed EdCode, a platform that enabled students to seek remote learning support directly from the Integrated Development Environment while coding. This platform provided personalized instructional support and lowered the risk of plagiarism that was often associated with traditional forums. Furthermore, web technologies facilitated community content sharing and real-time interactive experiences by enabling learners to retrieve step-by-step guides organized by other users that emphasized logical structuring and computational thinking (Hwang et al., 2008; Kazemitabaar et al., 2024; Lye & Koh, 2014). In particular, the quickly evolving 3D virtual learning environments—which were once intended for gaming and entertainment—are increasingly being utilized for educational purposes. This technology is considered to have a significant impact on learners’ cognitive development because it enables users to engage in creative development and social interactions, which facilitates their collaborative learning ability (Chau et al., 2013; Jing et al., 2024a, Jing et al., 2024a; Reisoğlu et al., 2017). The integration of these technologies can improve programming instruction and provide a broader platform for developing learners’ innovative capabilities and problem-solving skills, indicating that future education will be more intelligent and interactive. While technological advances offer promising avenues for programming education, gaps remain in understanding how these technologies can holistically address learners’ cognitive, social, and emotional needs in programming education environments.

Despite the numerous cutting-edge technologies that have significantly enhanced instructional strategies and supported innovative learning environments in programming education, there are still several critical issues that require immediate attention. For example, many educational technologies ignored educational goals, in favor of concentrating on flimsy technological displays (Messer et al., 2023), which can stunt learners’ ability to think critically and stimulate their innovative spirit. Additionally, some educational technologies overly emphasized learning outcomes while neglecting students’ learning experience and satisfaction (Messer et al., 2023). These issues were further evidenced by the commercialization of learning content, a decline in social presence (Cloete, 2017), and increased feelings of loneliness (Kryeziu et al., 2021), all of which may weaken learners’ intrinsic motivation and self-confidence, ultimately impacting their long-term interest and engagement during their learning process.

To investigate a potential solution to the aforementioned issues, this study created a virtual programming learning environment called MetaClassroom based on the Three-Dimensional Learning Progression (3DLP) concept (He et al., 2024; Kaldaras et al., 2021). By integrating concepts such as personalized adaptive learning, developmental differences theory, and situated learning theory, MetaClassroom was developed to support students’ learning progression. The study specifically examined programming learning achievements (PLA) as they directly reflect students’ mastery of programming concepts and skills. Additionally, we investigated self-regulated learning because it plays a crucial role in virtual learning environments where students need to manage their own learning process. Learning beliefs were also studied as they significantly influence students’ engagement and persistence in programming education, particularly in novel learning environments. Through examining these interconnected factors, we aimed to comprehensively understand how MetaClassroom impacts both learning outcomes and the learning experience. Specifically, this study explored the following research questions:

RQ1: How can MetaClassroom be designed and implemented based on the 3DLP concept to support students’ learning progression?

RQ2: How does programming instruction conducted in MetaClassroom impact students’ PLA?

RQ3: How does programming instruction conducted in MetaClassroom impact students’ self-regulated learning (SRL) skills?

RQ4: How does programming instruction conducted in MetaClassroom impact students’ learning beliefs and learning motivation?

Literature Review

Concept and Value of 3DLP (RQ1)

3DLP builds upon the concept of learning progression that describes how students’ understanding of a given concept evolves and becomes sophisticated over time (Kaldaras et al., 2021). Learning progression aims to depict the developmental process of students’ understanding of scientific concepts and their disciplinary logic, taking into account both their current knowledge level and intended learning objectives (He et al., 2024; Kaldaras et al., 2023). In disciplinary education, learning progression can be described as the developmental process of learners’ ability to apply knowledge in practice, where knowledge becomes more “complex,” enabling learners to effectively transfer their understanding to new contexts.

3DLP expands on learning progression, incorporating the concept of “three-dimension” described in the National Research Council’s report A Framework for K-12 Science Education (National Research Council, 2012). This framework included three dimensions: Disciplinary Core Ideas (DCI), Science and Engineering Practices (SEP), and Crosscutting Concepts (CCC). Therefore, 3DLP is an educational framework that integrates these three dimensions to assess and enhance learners’ understanding and application of scientific concepts (Kaldaras et al., 2021, 2023). Unlike traditional learning concepts that prioritize rote memorization of concepts, 3DLP focuses on helping learners understand and master a set of DCI and CCC, as well as how to integrate these concepts into SEP.

When applied to programming education, 3DLP provides a unique value in fostering students’ deeper comprehension and practical application of core programming concepts. For example, DCI in programming education corresponds to foundational programming concepts and algorithms, while SEP emphasizes practices such as debugging, coding, and computational problem-solving. CCC encourages students to identify patterns, understand abstraction, and transfer programming knowledge across various scenarios. By integrating these dimensions, 3DLP supports students in mastering not only the theoretical aspects of programming but also their ability to solve real-world problems and create functional programming solutions.

3DLP emphasizes problem-solving by leveraging the three dimensions to explain complex phenomena and address real-world issues. It asserts that learning should not be confined to isolated dimensions but rather integrate these three dimensions to understand and explore complex problems. In programming education, this approach helps students build programming fluency and develop transferable skills for applying coding knowledge across diverse contexts. Educators can leverage 3DLP to facilitate students’ coherent, progressive development of deeper and more applicable knowledge of programming concepts over time (Alonzo & Gotwals, 2012). Additionally, 3DLP provides a design framework for guiding the development and use of curriculum materials, instruction, and assessment to support the advancement of students’ programming knowledge application (He et al., 2024).

Therefore, in this study, a three-dimensional, practice-oriented virtual learning environment called MetaClassroom was developed and implemented based on the 3DLP concept. This environment enabled an immersive learning experience and aimed to facilitate students’ mastery of new programming concepts and the application of interdisciplinary knowledge. It provided specific opportunities for students to practice programming in realistic, problem-based scenarios, thereby supporting them in achieving their desired programming learning performance and cultivating essential computational thinking skills.

Research on Technology-Enhanced Programming Education and Learning Achievements (RQ2)

With the rapid advancement of information technology, programming education is undergoing significant transformations in its structure, logic, and paradigms. Emerging technologies are reshaping the programming education landscape in profound ways (Su et al., 2022; Wang et al., 2023; Wei et al., 2024). For example, the proliferation of web technologies has facilitated a shift from traditional classroom settings to online learning platforms (Çetin & Demircan, 2020), and on this basis, scholars have also explored how to redesign publicly available online documentation to help novice programmers conduct more efficient informal learning and improve their PLA (Zhang et al., 2025). The incorporation of ICT has further introduced new instructional tools and high-quality resources, thereby enhancing both the efficiency of knowledge dissemination and skill development in programming education (Chen, Xiao, et al., 2024). Furthermore, mobile application technologies have extended the spatial and temporal boundaries of programming instruction, allowing learners to practice and apply programming concepts across different contexts (Ma et al., 2023; Vinnervik, 2023). The integration of AI technologies is also playing a crucial role in advancing modern programming education, offering new opportunities for personalized and adaptive learning experiences (Kao et al., 2022; Sobral, 2021). Research has examined the practical applications of generative AI in programming education, with empirical studies validating its effectiveness in enhancing students’ learning achievements and teacher’s instructional methodologies (Jing et al., 2024c). It is evident that incorporating these technologies not only enhances interactivity and engagement in programming education but also encourages the diversification of instructional approaches (Lee, Lin, et al., 2023; Sorva et al., 2013).

Despite significant advancements brought by technology in programming education, research on context-based programming education remains in its early stages. Context-based programming education aims to embed real-world complex problems into teaching to help learners better accommodate practical programming environments (Coşkunserçe, 2023). This approach is intended to improve learners’ practical programming skills. However, current research indicates that the effectiveness of context-based programming education has not yet met expectations (Abdüsselam et al., 2022). For example, while some attempts have been made to integrate real-world scenarios into programming instruction, designing effective and engaging context-based learning tasks continues to pose a significant challenge (Trontl, 2010). The application of VR and AR technologies in programming education is unlocking new opportunities. These technologies can create immersive learning environments, enabling learners to engage in hands-on practice within simulated programming scenarios. As a result, VR and AR hold great potential to provide innovative perspectives and methodologies for enhancing programming education (Yolcu & Demirer, 2023; Çetin & Demircan, 2020). However, current research primarily focused on the implementation of the technologies and theoretical discussions, leaving challenges and issues in practical applications largely unresolved (Lukkarinen et al., 2021). Therefore, this study aimed to further explore how context-based programming instruction within a VR environment can improve students’ programming performance and learning experiences (Seralidou & Douligeris, 2021).

SRL Skills in Programming Education (RQ3)

SRL refers to the way learners use effective skills and strategies to manage their learning progression. This typically involves goal setting, strategy selection, cognitive monitoring and regulation, feedback adjustment, and behavioral motivation management (Li et al., 2024). This concept was originally introduced by American psychologist Albert Bandura in the 1970s, describing how learners actively motivate themselves and use appropriate learning strategies during their learning progression (Bandura, 1986; Zimmerman, 2002). SRL represents an active and constructive learning process in which learners enhance their learning by setting goals and by monitoring, regulating, and controlling their cognition, motivation, and behavior.

Despite the widespread popularity of SRL research in the educational field (Schunk, 2008), there is relatively limited attention to SRL in programming education (Loksa et al., 2022). Furthermore, existing studies on SRL in this context often misunderstand the concept. For instance, Cheng et al. (2019) used the Motivated Strategies for Learning Questionnaire to assess students’ use of SRL strategies in programming education. This approach overlooked the existence of well-established SRL questionnaires (e.g., Jansen et al., 2017; Li et al., 2020) and narrowed the concept of SRL. SRL goes beyond simply emphasizing learners’ motivation; it focuses more on their organic monitoring, self-assessment, and self-regulation of their earning progression—key aspects that are particularly crucial in programming education. Given the current misconceptions about SRL in the field of programming education and the importance of SRL skills for programming learning (Loksa et al., 2022), this study aimed to explore the dynamic interaction between SRL and programming education by incorporating SRL subdimensions, such as Metacognitive Skills, Time Management, Persistence, and Seeking Help.

Previous research has shown that learners’ SRL skills are not static but evolve throughout their learning progression (Panadero et al., 2023; Xu et al., 2023). Therefore, this study approached SRL as a dynamic process and explored the bidirectional empowerment between SRL and programming learning. By investigating the dynamic changes in both students’ PLA and SRL skills, this study aimed to offer insights for the design and organization of programming educational activities and to provide insights for creating effective programming learning environments.

Learning Beliefs and Learning Motivation in Programming Education (RQ4)

Learning beliefs and learning motivation are crucial factors influencing learners' learning progression and programming performance. In educational settings, both learning beliefs and learning motivation are integral components that contribute to achieving educational goals and fostering students’ continuous progress and development (Morales-Navarro et al., 2023). Learning beliefs encompasses learners’ perceptions of intelligence, learning goals, and their own abilities, which in turn shape their learning behaviors and outcomes (Hidayatullah & Csíkos, 2024). Learning motivation refers to the internal force that drives learners to engage in learning activities (Lin et al., 2023; Vu et al., 2022). Scholars often categorize motivation into intrinsic motivation and extrinsic motivation (Magen-Nagar & Cohen, 2017). Deci and Ryan (2013) defined intrinsic motivation as the interest in an activity for its own sake, where behaviors are performed out of enjoyment and interest, with no external rewards involved. In contrast, extrinsic motivation is driven by external values and needs, where learners participate in activities controlled by social demands and rewards (Deci et al., 1991). In summary, intrinsic motivation stems from curiosity and personal interest, while extrinsic motivation is more about obtaining rewards or avoiding punishment.

Prior research has shown that intrinsic motivation is more closely associated with learning achievements and positive psychological outcomes compared to extrinsic motivation (Gottfried et al., 2007). According to research, positive learning beliefs (such as the belief that intelligence can be improved with effort, a focus on mastering knowledge, and confidence in one’s learning ability) and strong intrinsic motivation are frequently associated with better learning outcomes. Previous studies in programming education have also examined the impact of learning beliefs and motivation on PLA and learning experience and found a positive correlation between the two (Lee, Liang, et al., 2023). However, the majority of the current research focused on how factors like gender affected the learning beliefs and motivation of learners, particularly beginners (Tellhe et al., 2023; Zdawczyk & Varma, 2022), without exploring deeply how different instructional contexts and learning environments might influence these factors. Undoubtedly, learning beliefs and motivation are critically important for novice programmers. Learners who have not had systematic exposure to programming tend to have an inherent fear of this subject. Therefore, exploring ways to help learners build positive learning beliefs and motivation to reduce their fear and anxiety toward programming is of great significance and value (Kong et al., 2018; Tellhe et al., 2023).

Learning beliefs and motivation can be cultivated by emphasizing the importance of effort, designing instructional tasks that are appropriately challenging, providing constructive feedback, and stimulating learners’ interest. It is also worth further exploring whether creating a supportive learning environment can support learners’ learning beliefs and motivation. This study aimed to investigate whether constructing a learner-centered environment—MetaClassroom—could offer students a more immersive and context-rich learning experience, thereby effectively enhancing their learning beliefs and motivation.

Methodology

Quasi-Experimental Research Method

The research adopted a quasi-experimental research strategy, selecting two parallel classes as experimental subjects (specific participant information can be provided in the “Participants and Research Context” section below). The entire study lasted approximately 3 months (12 weeks), with one week each allocated to pre-test and post-test, and 10 weeks of formal instruction. The core reason for choosing a quasi-experimental design was that for high school students, conducting a fully randomized double-blind controlled experiment over ten weeks would be extremely challenging. The existing teaching classes provided a ready-made research foundation, facilitating the study’s implementation. Moreover, at the time of the research, students in both teaching classes had just begun learning programming, ensuring the similarity of programming abilities across the two classes. Simultaneously, the research team conducted comprehensive pre-testing to eliminate potential inter-class differences that might interfere with the research results.

All procedures conducted in this study complied with the ethical standards of the 1964 Helsinki Declaration and its subsequent amendments or equivalent ethical guidelines. The study’s procedures and instruments were reviewed and approved accordingly. Informed parental consent was obtained for all participants prior to their involvement in the study (see Appendix A for the consent form). The authors declared that they had no conflicts of interest.

Participants and Research Context

The participants of this study were 70 first-year high school students (33 males and 36 females) from two parallel classes at a key high school in Eastern China. They had already studied basic programming knowledge for about one and a half semesters and had a certain level of understanding of Python. Specifically, they had a solid grasp of fundamental concepts such as functions and packages. The students in the two classes were highly similar, as both classes had been taught programming by the same teacher over the past one and a half semesters. The class sizes and gender ratios were also nearly identical, and their overall performance in previous tests was quite similar. This provided a solid foundation for the implementation of this study.

To ensure the authenticity of the classroom experience, the study was conducted in the spring semester of 2024 using a quasi-experimental design. The participants were initially grouped according to their existing class divisions and then further divided into an experimental group (n = 35) and a control group (n = 35). However, a participant in the control group missed part of the experiment due to illness, so the final valid data for the control group included 34 participants. The study was conducted over a 10-week Python programming course, with two 40-min sessions per-week. All participants were required to learn the knowledge structure presented in Figure 1.

This study selected “Python Extended Modules” of the Python course as the learning target based on several considerations. First, Python is an elective subject in the college entrance examination for the Zhejiang Province, making it a widely required and relatively challenging course at the high school. Additionally, according to the TIOBE Index (TIOBE, 2023), Python is currently the most popular programming language and is also the predominant language in many research fields, such as artificial intelligence (Jing et al., 2024c). Therefore, mastering Python not only benefits students academically but also significantly contributes to their future career development. Given these considerations, Python is undoubtedly the most appropriate choice of programming language for this research.

Learning Scenes: MetaClassroom Constructed Based on the 3DLP Concept

While fully practice-oriented programming teaching models remain challenging in traditional offline classrooms, the development of virtual immersive learning environments now offers a feasible approach to achieving seamless integration between concept learning and practical application. Therefore, guided by the 3DLP concept, our research team organically integrated the core concept of Python, interdisciplinary concepts, and authentic programming tasks when designing the Python learning environment. To further enhance this design, we incorporated principles from personalized adaptive learning, developmental differences theory, situational learning theory, cognitive learning theory, and scenario-based assessment concepts. Through this holistic approach, the advanced virtual programming learning environment known as MetaClassroom was developed. The overall design of this environment was shown in Figure 2.

The development of MetaClassroom integrated front-end technologies such as Babylon. js, HTML, CSS, and JavaScript with back-end technologies including MySQL, Java, and Docker. This integration enabled authentic simulation of real-world scenarios, creating a three-dimensional representation of the learning process and learning objectives. This immersive environment fused the learning process with practical applications and featured a procedural, panoramic, and embedded real-time assessment model.

MetaClassroom consisted of eight primary learning scenes, with four dedicated to practice-oriented knowledge learning (see Figure 3). These included Knowledge Corridor, Video Paradise, Textbook Library, and Website Bookshelf. The design of these four scenes focused on improving students’ programming problem-solving skills by offering a wealth of learning resources and flexible learning environments that encouraged personalized and adaptive learning, driven by real-world, practical challenges. The functions and design concepts of each scene were detailed in Appendix B.

The other four learning scenes—Game Corner, Virtual Classroom, Memory Tree, and Data Dashboard,—were designed to emphasize context-based teaching practices and assessment (see Figure 4). These scenes functioned both as platforms for teachers to deliver instruction and as spaces for students to consolidate and expand their knowledge application. To enhance the teaching practices, MetaClassroom incorporated an embedded assessment mechanism, which collected learning data from students across all scenes, allowing for formative evaluations of students’ implicit knowledge and skills. This assessment approach aligned with the 3DLP concept’s emphasis on the dynamic nature of learning progression, standing in contrast to traditional, one-way, and static assessment methods. The detailed functions and design concepts of these scenes could be referenced in Appendix B.

Our research team successfully designed and developed MetaClassroom, a virtual immersive programming learning environment that integrated teaching, learning, and practice. This platform bridged theoretical learning with hands-on practice and incorporated a real-time formative assessment mechanism, reflecting the core principles of problem-solving and personalized learning emphasized by the 3DLP concept. The platform is currently open-source on GitHub.

Instruments

The following research instruments were utilized in this study:

SRL Questionnaire

The SRL questionnaire used in this study was adapted from the validated instruments developed by Jansen et al. (2017) and Li et al. (2020), with modifications made to suit the study’s specific context. The final version of the questionnaire employed a 7-point Likert scale and included four subdimensions: Metacognitive Skills (8 items, α = 0.72), Time Management (3 items, α = 0.81), Persistence (4 items, α = 0.80), and Seeking Help (4 items, α = 0.70). The Cronbach’s alpha coefficients for all four dimensions met or exceeded the recommended threshold of 0.7, indicating acceptable reliability. We also conducted a structural validity test on the questionnaire. First, the results of the principal component analysis showed that four factors had eigenvalues greater than 1. Subsequently, we performed a complete exploratory factor analysis and found that the questionnaire had a clear structure. The rotated matrix revealed four factors, with a one-to-one correspondence between items and factors. This ensured the structural validity of the questionnaire. The full list of items can be found in Appendix C.

Learning Beliefs and Learning Motivation Questionnaire

The learning beliefs and motivation questionnaire used in this study was adapted from the validated instrument developed by Duncan and McKeachie (2005). Items from the belief and motivation dimensions were carefully selected to evaluate students’ learning beliefs and motivation in programming education, with minor modifications to suit the study’s specific focus. The learning belief section of the questionnaire used a 7-point Likert scale with 4 items, achieving a Cronbach’s alpha coefficient of 0.914. The learning motivation section, also using a 7-point Likert scale, was divided into two subdimensions: Intrinsic Goal Orientation (IGO, 4 items, α = 0.914) and Extrinsic Goal Orientation (EGO, 4 items, α = 0.781). Similarly, we conducted a structural validity test on this questionnaire. Exploratory factor analysis revealed a clear structure, with the rotated matrix showing three factors and a one-to-one correspondence between items and factors. This indicates that the questionnaire has good structural validity. The full list of items was available in Appendix D.

Python Fundamentals Test (Pre-test)

The pre-test consisted of 20 multiple-choice questions, each worth 5 points, with a total score of 100. It primarily assessed students’ comprehension of Python fundamentals as well as their preparedness for the upcoming content on the “Python Advanced Modules.” The full test was provided in Appendix E.

Post-test

The post-test included 4 main questions, each containing 5 sub-questions, with each sub-question worth 5 points, for a total score of 100. It primarily assessed students’ comprehension of the learning content covered during the 10-week instructional period. The post-test was built upon the pre-test and mid-term test, with a moderate increase in difficulty. It is also worth noting that although the post-test and pre-test contained different questions, they assessed the same core knowledge points. Furthermore, these tests were evaluated by experienced frontline educators, ensuring that the assessment methods and their effectiveness were appropriate and suitable. This ensures that changes in scores across these tests can reliably reflect changes in students’ programming knowledge and abilities. The full post-test was provided in Appendix F.

Experiment Design and Data Collection

This study adopted a quasi-experimental research design at the classroom level drawing on the methodologies outlined by Mo et al. (2022) and Jing et al. (2024c), with modifications tailored to the specific objectives of this research. The experiment was structured into three main phases: preliminary preparation and pre-test, formal experiment, and post-test, as illustrated in Figure 5. The primary objective was to evaluate the impact of programming instruction within MetaClassroom on students’ PLA and to examine the effects on their SRL skills, learning beliefs, and motivation. Additionally, this study sought to optimize programming learning strategies by analyzing participants’ performance and experiences within MetaClassroom, and to provide practical insights for the development and implementation of virtual learning environments in educational settings.

Preliminary Preparation and Pre-test

Before the study began, the research team developed the instructional materials for a 10-week course, along with questionnaires (including the SRL questionnaire, Learning Beliefs and Learning Motivation questionnaire, as detailed in Appendixes C and D) and test materials. The test materials included a Python fundamentals test (see Appendix E), two regular assignments, and a post-test (see Appendix F). These materials were developed through multiple rounds of communication with two frontline teachers to ensure that the difficulty was appropriate and that sufficient time was allotted for completing them. Once preparations were finalized, a pre-test was conducted. Participants first completed the Python fundamentals test to assess their prior knowledge of the course content for the upcoming 10 weeks. Simultaneously, their SRL skill levels were measured.

Formal Experiment

The formal experiment spanned 10 weeks, with a weekly learning duration of 80 minutes, divided into two 40-min sessions. At this phase, both the experimental group and the control group received instruction on the “Python Extension Module”. However, the experimental group engaged in programming learning within the MetaClassroom environment, while the control group followed a traditional classroom setting. It is important to note that, aside from the learning environment, the teaching content and instructors were identical for both groups. This controlled other teaching-related variables, ensuring the scientific, rigorous, and reliable nature of the research conclusions. Throughout the experiment, participants in both groups took two mid-term tests at the end of the third and sixth weeks. These tests were designed to provide periodic insights into the participants’ learning progression.

Post-test

After completing the 10-week learning period, participants took a post-test consisting of four challenging questions to be answered within 40 minutes, aimed at evaluating their PLA. After the test, the SRL, learning beliefs, and learning motivation questionnaires were administered to evaluate participants’ SRL skill, learning beliefs, and motivation levels following the intervention.

Data Analysis

This study used IBM SPSS (version 26) and Python for data analysis. Before conducting the formal analysis, the K-S test was employed to verify the normality of the data distribution, ensuring that the assumptions for parametric testing were satisfied.

RQ1 was omitted as it did not require statistical knowledge. And in order to examine changes in students’ PLA (RQ2), an independent samples t test was first utilized to verify that there were no significant differences between the experimental and control groups at the pre-test stage. Following this, independent samples t-tests were performed to compare performance differences between the two groups during the two mid-term tests and the post-test. To enhance confidence in multiple hypothesis testing, the Benjamini-Hochberg test was employed (implemented in Python, as SPSS does not support this functionality). This test adjusted the significance levels based on the p-value sequence to control the False Discovery Rate, ensuring the stability and reliability of the conclusions (Haynes, 2013).

To examine changes in students’ SRL skills (RQ3), a paired samples t test was conducted to assess whether there was a significant improvement before and after the intervention. Subsequently, one-way ANCOVA was used to examine differences between the control and experimental groups, with the pre-test SRL scores included as a covariate to control for any pre-existing differences between groups. To explore changes in students’ learning beliefs and motivation (RQ4), independent samples t-tests were applied to determine whether significant differences existed between the control and experimental groups.

Findings

Designing and Applying MetaClassroom: Based on 3DLP (RQ1)

MetaClassroom was designed with the aim of applying the 3DLP framework to support students’ learning progression in programming education. By aligning the design with the three core dimensions, MetaClassroom aimed to provide an immersive and engaging learning environment that fosters deeper comprehension and practical application of programming concepts. The design of MetaClassroom integrated these dimensions through both content and interactive features, ensuring that the students could not only grasp the theoretical aspects of programming but also develop the essential problem-solving skills required for real-world programming tasks.

The first dimension, DCI, in the context of programming education, was represented by the foundational programming concepts and algorithms. MetaClassroom focused on introducing students to core programming principles such as variables, control structures, and functions, while emphasizing the importance of algorithmic thinking and computational logic. The design of the environment allowed student to interact with these concepts through coding exercises and problem-solving tasks. For example, students were tasked with building algorithms and debugging code, engaging directly with the programming concepts they were learning. By integrating DCI within an interactive, virtual environment, MetaClassroom provided opportunities for students to experience the application of these concepts in simulated real-world scenarios, thereby helping to bridge the gap between theory and practice.

The second dimension, SEP, was incorporated into MetaClassroom by embedding practices such as debugging, coding, and computational problem-solving throughout the learning experience. In traditional classroom settings, students often encounter programming as a theoretical subject with limited hands-on practice. However, MetaClassroom was designed to emphasize the process of coding, debugging, and refining solutions in an authentic, iterative way. Interactive exercises within the platform encouraged students to engage with real programming tasks, such as troubleshooting code errors, optimizing algorithms, and improving the efficiency of their solutions. By doing so, the environment not only improved their technical skills but also promoted a deeper understanding of the engineering practices that are central to programming, such as testing, debugging, and adapting solutions based on feedback. This active participation in SEP activities allowed students to refine their problem-solving abilities and to see firsthand the iterative nature of the programming process.

The third dimension, CCC, was integrated into MetaClassroom by emphasizing the transferability of programming knowledge across different scenarios. CCC encourages learners to identify patterns, apply abstraction, and transfer their knowledge to new contexts. To achieve this, MetaClassroom included a variety of diverse programming challenges that required students to recognize patterns in problems and apply their learning in novel contexts. For example, students were tasked with developing programs for different types of problems—such as games, simulations, and data processing tasks—requiring them to abstract concepts and adapt their solutions based on the specific context of the task. This approach not only deepened their understanding of core programming concepts but also encouraged them to think creatively about how to apply programming to solve real-world challenges.

Furthermore, the MetaClassroom environment facilitated a continuous feedback loop, allowing students to monitor their own progress through frequent formative assessments. These assessments, which included quizzes, coding challenges, and peer-reviewed assignments, were designed to align with the three dimensions of 3DLP, offering insights into students’ understanding of the material and providing opportunities for improvement. Through iterative assessments and feedback, students could track their own learning progression, adjusting their approach as needed to enhance both their technical skills and problem-solving abilities. The incorporation of these assessments ensured that MetaClassroom was not only a tool for learning programming but also a platform that supported students’ ongoing development, helping them move progressively toward mastering complex programming concepts.

In summary, the design of MetaClassroom, based on the 3DLP framework, focused on providing students with a balanced learning experience that integrated foundational programming knowledge (DCI), hands-on practices (SEP), and transferable problem-solving skills (CCC). By aligning with these three dimensions, MetaClassroom was able to support students’ learning progression in programming, promoting not only their technical competence but also their ability to apply their knowledge creatively and effectively across various contexts. MetaClassroom aimed to cultivate students' computational thinking, enhance their problem-solving skills, and help them foster a deeper understanding of programming, all of which are essential for success in the rapidly evolving field of computer science.

Evaluation of the Impact of MetaClassroom on Students’ PLA (RQ2)

To analyze whether MetaClassroom could effectively improve students’ PLA, an independent samples t test was conducted to assess whether there was a difference in Python fundamentals between the two groups before the experiment. The results presented in Table 1, showed that there was no significant difference in pre-test scores between the experimental and control groups (t = 1.220, p > .05).

Table 1.

Independent Samples t Test for Pre-test.

Group	N	Prior Knowledge Level	t-value	df	p
EG	35	20.71 ± 9.56	1.220	67	0.227
CG	34	18.09 ± 8.26	1.220	67	0.227

Note. EG: Experimental Group; CG: Control Group.

Building on this, this study further conducted three rounds of independent samples t-tests to compare differences in students’ PLA between the experimental and control groups across the two mid-term tests and the post-test. The results were shown in Table 2.

Table 2.

Independent Samples t Test for Mid-term Tests and Post-test.

Test	Group	N	Mean Test Scores	t-value	df	p
Mid-term test 1	EG	35	47.29 ± 10.53	2.851**	67	.005
Mid-term test 1	CG	34	39.56 ± 11.96	2.851**	67	.005
Mid-term test 2	EG	35	64.43 ± 9.91	4.045***	67	<.001
Mid-term test 2	CG	34	54.26 ± 10.95	4.045***	67	<.001
Post-test	EG	35	78.14 ± 9.56	2.636*	67	.010
Post-test	CG	34	70.15 ± 15.10	2.636*	67	.010

Note. *p < .05; **p < .01; ***p < .001. The same applies to the below table.

To ensure the reliability of multiple hypothesis test results, the Benjamini-Hochberg Test (code detailed in Appendix G) was applied to adjust the significance levels of each test. The results confirmed that all initially significant findings retained their significance, validating the persistence of our hypotheses. As shown in Table 2, the experimental group demonstrated significantly higher PLA compared to the control group from the first mid-term Test onward (t = 2.851, p = .005 < .05). Significant differences were also observed in the second mid-term test (t = 4.045, p < .001) and the post-test (t = 2.636, p = .010 < .05). To provide a clearer view of the achievement changes between the experimental and control groups across the four tests, a visualization was presented in Figure 6. This visualization offered an intuitive depiction of students’ learning progression over the 10-week learning period.

Evaluation of the Impact of MetaClassroom on Students’ SRL Skills (RQ3)

In addition to focusing on students’ PLA, this study also examined the improvement in students’ SRL skills during the learning process. Paired samples t-tests were conducted for analysis (see Table 3), revealing that both the experimental and control groups experienced significant improvements in SRL skills over the 10-week learning period.

Table 3.

Paired Samples t Test for Mean SRL Scores before and after Intervention.

Group	N	Mean SRL Score (Pre-Test)	Mean SRL Score (Post-Test)	t-value	df	p
EG	35	2.52 ± 0.48	3.72 ± 0.48	−9.080***	34	<.001
CG	34	2.43 ± 0.29	3.23 ± 0.43	−8.920***	33	<.001

To further investigate changes in each SRL subdimension over the 10-week learning period, paired samples t-tests were conducted, with results shown in Table 4. The analysis revealed significant improvements in all SRL subdimensions, except for Seeking Help in the experimental group (t = −1.694, p = .099 > 0.05).

Table 4.

Paired Samples t Test for Different SRL Subdimensions before and after Intervention.

Group	N	SRL dimensions	Mean (pre-test)	Mean (post-test)	t-value	df	p
EG	35	Metacognitive skills	2.66 ± 0.73	4.53 ± 0.76	−10.267***	34	<.001
	35	Time management	2.11 ± 0.91	3.28 ± 1.16	−4.196***	34	<.001
	35	Persistence	2.29 ± 0.71	3.71 ± 1.25	−5.507***	34	<.001
	35	Seeking help	3.01 ± 0.87	3.35 ± 0.92	−1.694	34	.099
CG	34	Metacognitive skills	2.37 ± 0.41	4.09 ± 0.67	−12.025***	33	<.001
	34	Time management	2.41 ± 0.61	2.99 ± 1.02	−2.801**	33	.008
	34	Persistence	2.41 ± 0.77	2.94 ± 0.95	−2.690*	33	.011
	34	Seeking help	2.54 ± 0.48	2.89 ± 0.58	−2.520*	33	.017

To explore differences in SRL subdimension improvements between groups, one-way ANCOVA was performed using the SRL pre-test scores as a covariate. This approach aimed to mitigate the impact of sample differences. Levene’s test confirmed that the assumption of homogeneity of variances was not violated, making ANCOVA analysis appropriate. The results of the analysis were shown in Table 5.

Table 5.

One-Way ANCOVA Results for SRL Post-test Scores.

Group	N	Mean	Adjust M	F	p	η2
EG	35	3.72 ± 0.48	3.73 ± 0.08	21.594***	<.001	.249
CG	34	3.23 ± 0.43	3.23 ± 0.08	21.594***	<.001	.249

Note. Adjusted M using pretest scores as a covariate.

Table 5 showed a significant difference in SRL post-test scores between the experimental and control groups (F = 21.594, p < .001). After the 10-week learning period, participants in the experimental group demonstrated significantly higher SRL skills compared to those in the control group.

To identify specific SRL subdimensions showing significant differences between groups, one-way ANCOVA was conducted for each subdimension, with results presented in Table 6.

Table 6.

One-Way ANCOVA Results for SRL Subdimensions.

SRL subdimensions	Group	N	Mean	Adjust M	F	p	η2
Metacognitive skill	EG	35	4.54 ± 0.76	4.55 ± 0.13	7.094*	.010	.098
Metacognitive skill	CG	34	4.09 ± 0.67	4.06 ± 0.13	7.094*	.010	.098
Time management	EG	35	3.28 ± 1.16	3.22 ± 0.18	.723	.398	.011
Time management	CG	34	2.99 ± 1.02	3.00 ± 0.19	.723	.398	.011
Persistence	EG	35	3.71 ± 1.25	3.70 ± 0.19	8.054**	.006	.110
Persistence	CG	34	2.94 ± 0.95	2.93 ± 0.19	8.054**	.006	.110
Seeking help	EG	35	3.35 ± 0.92	3.32 ± 0.14	5.579*	.021	.079
Seeking help	CG	34	2.89 ± 0.58	2.84 ± 0.15	5.579*	.021	.079

Table 6 showed significant differences between the experimental and control groups in Metacognitive Skills (F = 7.094, p = .010 < 0.05), Persistence (F = 8.054, p = .006 < 0.05), and Seeking Help (F = 5.579, p = .021 < 0.05), with participants in the experimental group scoring significantly higher than those in the control group. However, no significant difference was found between the two groups in Time Management (F = 0.723, p = .398 > 0.05). To visualize these differences in SRL post-test scores across various subdimensions, a graphical representation was provided in Figure 6. The values in the visualization represented adjusted means, and the error bars indicated adjusted variances (Figure 7).

Evaluation of the Impact of MetaClassroom on Students’ Learning Beliefs and Motivation (RQ4)

In addition to investigating students’ PLA and SRL skills, this study also focused on students’ learning experiences, specifically analyzing students’ learning beliefs and motivation during their learning progression. An independent samples t test was first conducted to compare students’ learning beliefs and motivation in the two groups (see Table 7). The results indicated that students in the experimental group exhibited significantly higher learning beliefs than those in the control group (t = 8.456, p < .001).

Table 7.

Independent Samples t Test for Learning Beliefs.

Group	N	Mean (pre-test)	t-value	df	p
EG	35	4.92 ± 1.55	8.456***	48.064	<.001
CG	34	2.47 ± 0.71	8.456***	48.064	<.001

The study further examined students’ learning motivation in both groups (see Table 8). The results demonstrated that the experimental group had significantly higher learning motivation than the control group (t = 14.300, p < .001). Since the scale used in this study divided learning motivation into EGO and IGO, further analysis was conducted to assess differences between the two groups in these subdimensions. As shown in Table 9, significant differences were observed in both subdimensions (t_EGO = 5.463, p < .001; t_IGO = 15.359, p < .001), with the experimental group outperforming the control group in both aspects. For clearer data presentation, we also provide a visualization, see Figure 8.

Table 8.

Independent Samples t test for Learning Motivation.

Group	N	Mean (pre-test)	t-value	df	p
EG	35	5.17 ± 0.58	14.300***	67	<.001
CG	34	3.14 ± 0.59	14.300***	67	<.001

Table 9.

Independent Samples t test for Subdimensions of Learning Motivation.

Motivation subdimensions	Group	N	Mean	t-value	df	p
EGO	EG	35	5.44 ± 0.98	5.463***	67	<0.001
EGO	CG	34	4.10 ± 1.06
IGO	EG	35	4.89 ± 0.87	15.359***	67	<0.001
IGO	CG	34	2.18 ± 0.56

Discussion

The Role and Mechanism of 3DLP-Based MetaClassroom in Programming Education (RQ1)

The study began by exploring the construction of MetaClassroom, a virtual programming learning platform grounded in the 3DLP concept. This platform integrated the three core dimensions of 3DLP-DCI, SEP, and CCC-into the design of eight learning scenes and functional modules. To enhance the educational value, MetaClassroom also incorporated multiple educational theories, bridging knowledge acquisition, application, and transfer, and creating an immersive ecosystem where theory guided practice and practice informed theory (Kaldaras et al., 2021).

The DCI concept was seamlessly integrated into MetaClassroom through the creation of multidimensional and multilayered knowledge systems, such as Knowledge Corridor and Textbook Library. This design transcended simple presentation of knowledge; it systematically and structurally organized the core knowledge concepts of Python programming, helping students to develop a deep understanding of the fundamentals of this discipline. The concept of SEP was fully realized through Virtual Classroom and Game Corner. These two learning scenes moved beyond traditional teaching methods by offering practice environments that simulated real-world programming scenarios. Through real-time interactions in the virtual classroom and gamified problem-solving tasks, students could apply their acquired knowledge in near-authentic contexts, developing both practical programming skills and problem-solving abilities. The concept of CCC was primarily achieved through Website Bookshelf and Memory Tree. These two learning scenes not only offered a wealth of interdisciplinary resources but, more importantly, demonstrated the intrinsic connections between Python programming and other disciplines through knowledge mapping. This approach helped foster students’ interdisciplinary thinking and innovation skills.

The innovation of MetaClassroom also lay in the formative assessment mechanism. Data Dashboard implemented a multidimensional, formative assessment mechanism based on the 3DLP concept. Unlike traditional summative assessments, this mechanism enabled a more comprehensive formative evaluation of students’ knowledge comprehension, practical application, and interdisciplinary thinking. It also offered data-driven support for personalized and adaptive learning, which were consistent with the perspectives widely held by current programming education researchers (Anindyaputri et al., 2020; Troussas, Krouska, & Sgouropoulou, 2020). Furthermore, MetaClassroom integrated various educational theories, including self-regulated learning theory, and the concept of personalized adaptive learning, into the design. By offering a variety of learning resources and allowing for flexible, self-directed choices, the platform empowered students to craft personalized learning strategies that aligned with their individual needs and preferences. This approach could help promote students’ metacognitive skills and foster their stronger self-regulated learning habits.

Undoubtedly, MetaClassroom represented an innovative attempt to bring advanced educational theories into concrete practice. It was more than just an implementation of the 3DLP concept within programming education; it served as a holistic platform that unified teaching, learning, and assessment. By creating an immersive and interactive programming learning environment, MetaClassroom provided a virtual space that supported students’ deep learning, facilitated their knowledge transfer, and fostered their higher-order thinking skills. This groundbreaking platform opened up new possibilities for the future development of programming education, offering insights into the integration of theory and practice in educational technology.

Analysis of the Impact of MetaClassroom on Students’ PLA (RQ2)

The positive impact of MetaClassroom on students’ PLA indicated that programming instruction in a virtual learning environment could effectively enhance students’ cognitive development, which aligned with previous research showing that VR and AR technologies could enhance students’ programming performance (Yolcu & Demirer, 2023; Çetin & Demircan, 2020). This insight also highlighted the potential value of integrating educational technology with both educational psychology and cognitive psychology to enhance students’ learning achievements. Rather than focusing solely on technical aspects of programming education, MetaClassroom provided a holistic support system for both knowledge acquisition and practical application. By applying the 3DLP concept, MetaClassroom created an environment that not only facilitated learning but also nurtured higher-order thinking skills, contributing to improved PLA. In summary, the design and implementation of MetaClassroom emphasized improving students’ visible PLA while simultaneously optimizing their long-term learning experiences, promoting deeper cognitive restructuring, and strengthening their problem-solving and innovation skills in the process of tackling complex problems.

Despite the stable and statistically positive results, it was crucial to maintain an objective and rational perspective regarding the role of technology in programming education. While the improvements in PLA after technological intervention appeared encouraging, it was necessary to carefully evaluate whether other factors might have contributed to this outcome (Liu et al., 2024). During the teaching practice, our research team observed that students in the experimental group were noticeably more focused and engaged, potentially due to the novelty and curiosity sparked by MetaClassroom, which might have increased their motivation to learn. However, this could have also led to the possibility of a Hawthorne effect, where participants modified their behavior because they were aware of being observed, causing a deviation in the experimental results from actual conditions (Diaper, 1990; Jing et al., 2024a; Sedgwick & Greenwood, 2015). Such effects have been a concern in many quasi-experimental studies (Adair et al., 1989; Mo et al., 2022), and regrettably, this study did not effectively control for its influence.

Our research team also observed that, as the learning activities progressed, although there was already a significant difference between the experimental and control groups during the first mid-term test, this gap did not substantially widen in the second mid-term test or the final post-test. As illustrated in Figure 6, this trend could potentially indicate the presence of a Hawthorne effect, where initial improvements tapered off as the novelty of the intervention faded. However, to systematically explore the underlying reasons behind this occurrence, the experimental period needs to be extended in future studies.

Analysis of the Impact of MetaClassroom on Students’ SRL Skills (RQ3)

The evaluation of MetaClassroom’s impact on students’ SRL skills revealed significant overall improvements, with varied effects across different subdimensions. The experimental results showed notable enhancements in subdimensions, including Metacognitive Skills, Persistence, and Seeking Help, highlighting the platform’s effectiveness in supporting students’ SRL. The improvements in these subdimensions demonstrated the platform’s capacity for supporting students’ cognitive development and revealed its potential for fostering their higher-order thinking skills and promoting their self-regulated learning habits.

MetaClassroom, by creating an immersive learning environment and incorporating a real-time formative assessment mechanism, effectively enhanced students’ internalization and application of self-monitoring, reflection, and regulation strategies. This deep-level cognitive process optimization supported by MetaClassroom extended beyond surface-level skill enhancements, leading to significant growth in students’ metacognitive skills, which extended the findings of previous research on the application of VR technology in programming education (Loksa et al., 2022). However, this study identified a lack of significant progress in time management skills, highlighting a potential issue with the task structure design within MetaClassroom. This suggests that improving time management skills within highly flexible learning environments is still a critical area needing further research.

From a broader perspective, the design and implementation of MetaClassroom, grounded in the 3DLP concept, marked a significant shift in modern programming education paradigms. The benefits of this shift were evident not only in the improvements in students’ PLA but also, more significantly, in the substantial advancements in their SRL skills. The enhancement of SRL skills through MetaClassroom highlighted the platform’s success in transcending traditional knowledge delivery methods by emphasizing the holistic development of students’ core competencies. This progress offered valuable insights for the future development of educational technology, suggesting that such technology has the potential to not only optimize knowledge acquisition but also enrich learning experiences, promote cognitive development, and cultivate essential 21st-century skills.

Undoubtedly, future educational developments will increasingly emphasize the cultivation of advanced cognitive abilities and SRL skills to meet the growing complexity of societal needs and the demands of lifelong learning (Jing et al., 2024c). This study demonstrated that MetaClassroom effectively enhanced certain subdimensions of students’ SRL skills in programming education. However, the specific mechanisms through which MetaClassroom influenced these skills were not systematically explored. Also, future research should investigate how to strategically apply MetaClassroom to develop more comprehensive and effective educational technology interventions, with the goal of fully optimizing students’ SRL performance.

Analysis of the Impact of MetaClassroom on Students’ Learning Beliefs and Motivation (RQ4)

This study not only examined the impact of MetaClassroom on students’ cognitive development (such as PLA) and core competency development (such as SRL skills) but also investigated its effect on their learning experience, specifically in terms of learning beliefs and motivation. The findings revealed that students in the experimental group using MetaClassroom exhibited significantly higher learning beliefs and motivation compared to the control group in the traditional classroom settings. This can be attributed to the immersive and interactive nature of MetaClassroom, which promoted greater student engagement and intrinsic motivation, thereby boosting their confidence in programming learning progression. Moreover, the real-time feedback provided by Data Dashboard promoted more coherent learning behaviors, significantly enhancing students’ overall learning experience (Kazemitabaar et al., 2024).

Previous research demonstrated that engaging and simulated learning environments, combined with appealing instructional methods, could significantly boost students’ motivation to engage in learning activities (Moazami et al., 2014). MetaClassroom effectively incorporated these principles by creating an immersive and interactive learning experience, which is especially crucial for high school students, who are in the key developmental stage for abstract logical thinking (Al-Ajmi & Ambusaidi, 2022). Programming learning, which demands high-level logical thinking, often requires external support to facilitate effective learning (Jing et al., 2024c). MetaClassroom provided such support by offering an immersive environment that enhanced student agency, encouraged diverse learning interactions, and offered a variety of learning resources. This environment helped lower the cognitive barriers associated with programming by providing scaffolding for the development of logical thinking skills. These findings echoed prior research on virtual learning environments, which showed that such environments could significantly improve students’ overall learning experience (Yasmin et al., 2019).

In designing MetaClassroom, our research team prioritized students’ learning experiences and perceptions, while remaining mindful of the risks associated with technology misuse. The application of any technology must align with specific educational goals (Wang, Chen, et al., 2024). However, as mentioned in the introduction section, the rapid advancement of technology and the rise of technocentrism have shifted the focus towards the convenience and efficiency brought by intelligent technologies. This shift has led education stakeholders to gradually overlook the core objective of “sustainable human development and cultivation” (Birdman et al., 2022). Attention has increasingly been drawn to immediately measurable outcomes, such as performance and test scores, which has led to the misuse of intelligent technologies in educational contexts (Kulesza et al., 2011). To mitigate this phenomenon, the design of intelligent technologies for educational purposes must also take into account learners’ holistic learning experiences. If the primary focus remains on improving efficiency through adaptive learning techniques and personalized recommendations without considering the complexities of human cognitive development, there is a risk of straying from core educational goals. This misalignment can lead to conflicts between educational values and technological applications, ultimately diminishing the inherent educational purpose and undermining the core mission of fostering humanistic care and holistic development (Chan & Hu, 2023; Tang & Tang, 2021).

Therefore, the core value orientation of MetaClassroom emphasized that technology functioned as a tool to achieve educational objectives, not as the objective itself. Throughout the development, the research team was careful to avoid the potential risk of technological determinism, maintaining a rational and critical stance to ensure that technology supported, rather than detracted from, the primary educational goals. MetaClassroom prioritized three key functions: facilitating students’ knowledge acquisition, promoting skill development, and enhancing their learning experience. In this context, educational practice was evaluated not only based on students’ achievement outcomes but also through incorporating educational values, as well as considering students’ learning perceptions, and experiences. By creating simulated learning scenes and facilitating immersive programming experiences (Hamutoglu et al., 2020), MetaClassroom enhanced students’ sense of presence (Zhang & Lin, 2021), making programming education more engaging and authentic. This, in turn, made it easier to enhance students’ programming learning self-efficacy, strengthen their learning beliefs and boost their learning motivation, while minimizing their aversion towards programming. Ultimately, MetaClassroom fostered students’ intrinsic motivation, encouraging more SRL behaviors and ensuring sustained, positive engagement throughout students’ programming learning progression.

Conclusions and Prospects

Conclusions

This study designed, implemented, and evaluated MetaClassroom based on the 3DLP concept, thoroughly investigated the impact of this virtual programming learning environment on the students’ multidimensional development. The findings indicated that MetaClassroom successfully integrated the 3DLP into eight specific learning scenes and functional modules, creating an innovative immersive learning ecosystem that bridged theory and practice (RQ1). In terms of students’ PLA, MetaClassroom significantly improved students’ programming skills, demonstrating the potential of virtual learning environments to optimize the processes of knowledge acquisition and application (RQ2). Regarding SRL skills, the study found that MetaClassroom had a significant positive impact on key subdimensions such as Metacognitive Skills, Persistence, and Seeking Help, highlighting the platform’s unique value in fostering students’ higher-order thinking skills and SRL habits (RQ3). Additionally, MetaClassroom’s positive effects on students’ learning beliefs and motivation further validated its effectiveness in enhancing learning experiences, as well as intrinsic and extrinsic motivation (RQ4). Overall, this study not only confirmed MetaClassroom’s effectiveness in enhancing students’ PLA, but also introduced an innovative model that integrates teaching, learning, and assessment, offering insights and directions for the development of educational technology in programming education and broader educational practices.

Implications

This study integrated the 3DLP concept into the design of a virtual programming learning environment, offering a new perspective for the development of programming education theories. The design and implementation of MetaClassroom not only validated the applicability of 3DLP in programming education but also demonstrated how abstract educational concepts could be transformed into concrete learning scenes. This process of bridging theory and practice enriched the intersection of educational technology and learning sciences. The findings revealed the multifaceted impact of virtual learning environments on students’ cognitive, metacognitive, and affective performance, providing empirical evidence for understanding the mechanisms behind technology-enhanced programming learning. Moreover, this study introduced SRL theory into the field of programming education, exploring how virtual environments foster students’ SRL skills, thereby expanding the application of SRL theory within specific disciplinary contexts.

From a practical perspective, this study provided an actionable model and guidance for innovative implementation in programming education. The design framework of MetaClassroom innovated a paradigm for integrating disciplinary knowledge, practical skills, and interdisciplinary thinking in programming education, serving as a guide for developing more effective programming teaching strategies and learning environments. The findings underscored the importance of fostering students’ metacognitive skills and other SRL skills in programming education, proposing new instructional goals and assessment dimensions for educators. Furthermore, the study’s focus on learning beliefs and motivation reminded educators to balance learning achievements with learning experiences when designing educational environments. The successful implementation of MetaClassroom provided valuable insights into the future direction of programming education and served as a reference for designing virtual learning environments across other disciplines. It demonstrated how to effectively balance technological integration with educational theories and leverage technology to enhance students’ core competencies.

Limitations and Future Research

This study, while offering valuable insights, had several limitations. First, it relied on self-reported questionnaires, which might have introduced subjectivity, as participants might have struggled to accurately recall and describe their past learning behaviors. This reliance on memory could have led to cognitive bias, potentially affecting the accuracy of the evaluation results to some extent (Li et al., 2020).

Secondly, the MetaClassroom platform was constrained by its web-based 3D architecture and did not incorporate advanced immersive 3D technologies that could have offered a more comprehensive and engaging learning environment. Future iterations of the platform could consider exploring the integration of advanced technologies to further enhance user experience and interactivity.

Furthermore, as MetaClassroom was still under development by the research team, the design and refinement were ongoing. One significant limitation was the absence of generative AI in the platform, despite its widely recognized potential as a virtual learning companion (Chen, Hu, & Wang, 2024; Lin & Yu, 2024; Wang, Wang, et al., 2024). Future efforts could consider incorporating generative AI into MetaClassroom, potentially transforming it into an interactive and adaptive component of the educational ecosystem. Such integration might offer opportunities to dynamically adjust learning pathways and enhance personalized learning experiences, contributing to the ongoing development and effectiveness of the platform.

Supplemental Material

Supplemental Material - MetaClassroom: A New Paradigm and Experience for Programming Education

Supplemental Material for MetaClassroom: A New Paradigm and Experience for Programming Education by Chengliang Wang, Xiaojiao Chen, Yifei Li, Pengju Wang, Haoming Wang, and Yuanyuan Li in Journal of Educational Computing Research

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Chengliang Wang

Yuanyuan Li

Supplemental Material

Supplemental material for this article is available online.

Appendix

Author Biographies

Chengliang Wang is a postgraduate in Department of Education Information Technology, Faculty of Education, East China Normal University. His research interests include technology-supported programming instruction, social media learning, AI for social science, educational technology philosophy and learning environment supported by technology.

Xiaojiao Chen is an undergraduate in College of Educational Science and Technology, Zhejiang University of Technology. Her research interests include technology-enabled programming learning, human-centered educational technology design.

Yifei Li is an undergraduate in Department of Education Information Technology, Faculty of Education, East China Normal University. Her research interests include intelligent learning environment construction, computer-supported collaborative learning.

Pengju Wang is an undergraduate in Department of Education Information Technology, Faculty of Education, East China Normal University. Her research interests include online learning analysis, metaverse learning environment design and application.

Haoming Wang is a postgraduate in Department of Education Information Technology, Faculty of Education, East China Normal University. His research interests include programming education, educational technology philosophy and artificial intelligence in education.

Yuanyuan Li is a researcher at Shanghai Normal University. Her research interest include simulation-based learning, dialogic teaching and extended reality technology-supported teacher education.

References

Abdüsselam

M. S.

Turan-Güntepe

Durukan

Ü. G.

(2022). Programming education in the frameworks of reverse engineering and theory of didactical situations. Education and Information Technologies, 27(5), 6513–6532. https://doi.org/10.1007/s10639-021-10883-8

Adair

J. G.

Sharpe

Huynh

C. L.

(1989). Hawthorne control procedures in educational experiments: A reconsideration of their use and effectiveness. Review of Educational Research, 59(2), 215–228. https://doi.org/10.3102/00346543059002215

Al-Ajmi

Ambusaidi

(2022). Scientific argumentation the level of scientific argumentation skills in chemistry subject among grade 11th students: The role of logical thinking. Science Education International, 33(1), 66–74. https://doi.org/10.33828/sei.v33.i1.7

Alonzo

A. C.

Gotwals

A. W.

(Eds.), (2012). Learning progressions in science: Current challenges and future directions. Springer Science & Business Media.

Anindyaputri

N. A.

Yuana

R. A.

Hatta

(2020). Enhancing students’ ability in learning process of programming language using adaptive learning systems: A literature review. Open Engineering, 10(1), 820–829. https://doi.org/10.1515/eng-2020-0092

Bandura

(1986). The explanatory and predictive scope of self-efficacy theory. Journal of Social and Clinical Psychology, 4(3), 359–373. https://doi.org/10.1521/jscp.1986.4.3.359

Birdman

Wiek

Lang

D. J.

(2022). Developing key competencies in sustainability through project-based learning in graduate sustainability programs. International Journal of Sustainability in Higher Education, 23(5), 1139–1157. https://doi.org/10.1108/IJSHE-12-2020-0506

Capel

M. I.

Tomeu

A. J.

Salguero

A. G.

(2017). Teaching concurrent and parallel programming by patterns: An interactive ICT approach. Journal of Parallel and Distributed Computing, 105(1), 42–52. https://doi.org/10.1016/j.jpdc.2017.01.010

Çetin

Demircan

H. Ö.

(2020). Empowering technology and engineering for STEM education through programming robots: A systematic literature review. Early Child Development and Care, 190(9), 1323–1335. https://doi.org/10.1080/03004430.2018.1534844

10.

Chan

C. K. Y.

(2023). Students’ voices on generative AI: Perceptions, benefits, and challenges in higher education. International Journal of Educational Technology in Higher Education, 20(1), 43. https://doi.org/10.1186/s41239-023-00411-8

11.

Chau

Wong

Wang

Lai

Chan

K. W. Y.

T. M. H.

Chu

Chan

I. K. W.

Sung

(2013). Using 3D virtual environments to facilitate students in constructivist learning. Decision Support Systems, 56(1), 115–121. https://doi.org/10.1016/j.dss.2013.05.009

12.

Chen

Xiao

Chen

Song

Sun

(2024a). ChatScratch: An AI-augmented system toward autonomous visual programming learning for children aged 6-12. In Proceedings of the CHI conference on human factors in computing systems (pp. 1–19). Retrieved from: https://doi.org/10.1145/3613904.3642229

13.

Chen

X. J.

Z. B.

Wang

C. L.

(2024b). Empowering education development through aigc: A systematic literature review. Education and Information Technologies, 29(13), 17485–17537. https://doi.org/10.1007/s10639-024-12549-7

14.

Chen

Herskovitz

Matute

April

Yi W.

Sang

Y. L.

Lasecki

W. S.

Oney

(2020). EdCode: Towards personalized support at scale for remote assistance in CS education. In 2020 IEEE symposium on visual languages and human-centric computing. Retrieved from: https://doi.org/10.1109/vl/hcc50065.2020.9127260

15.

Cheng

Poon

L. K.

Lau

W. W.

Zhou

R. C.

(2019). Exploring the relationship between self-regulated learning strategies and computer programming achievement in higher education. Proceedings of the International Conference on Education, 5(1), 67–74. https://doi.org/10.17501/24246700.2019.5108

16.

Cloete

A. L.

(2017). Technology and education: Challenges and opportunities. HTS Teologiese Studies / Theological Studies, 73(4), 1. https://doi.org/10.4102/hts.v73i4.4589

17.

Coşkunserçe

(2023). Comparing the use of block based and robot programming in introductory programming education: Effects on perceptions of programming self efficacy. Computer Applications in Engineering Education, 31(5), 1234–1255. https://doi.org/10.1002/cae.22637

18.

Deci

E. L.

Ryan

R. M.

(2013). Intrinsic motivation and self-determination in human behavior. Springer Science & Business Media.

19.

Deci

E. L.

Vallerand

R. J.

Pelletier

L. G.

Ryan

R. M.

(1991). Motivation and education: The self-determination perspective. Educational Psychologist, 26(3-4), 325–346. https://doi.org/10.1080/00461520.1991.9653137

20.

Department for Education . (2013). National curriculum in England: Framework document. Retrieved from: https://www.gov.uk/government/publications/national-curriculum-in-england-framework-for-key-stages-1-to-4

21.

Diaper

(1990). The Hawthorne effect: A fresh examination. Educational Studies, 16(3), 261–267. https://doi.org/10.1080/0305569900160305

22.

Duncan

T. G.

McKeachie

W. J.

(2005). The making of the motivated strategies for learning questionnaire. Educational Psychologist, 40(2), 117–128. https://doi.org/10.1207/s15326985ep4002_6

23.

EDUCAUSE . (2023). 2023 EDUCAUSE horizon report: Teaching and learning edition. Retrieved from: https://library.educause.edu/resources/2023/5/2023-educause-horizon-report-teaching-and-learning-edition

24.

Engineers Australia . (2018). National STEM strategy 2019-2023: Create tomorrows engineers’ strategic priority. Retrieved from: https://www.engineersaustralia.org.au/publications/national-stem-strategy-2019-2023

25.

Gottfried

A. E.

Marcoulides

G. A.

Gottfried

A. W.

Oliver

P. H.

Guerin

D. W.

(2007). Multivariate latent change modeling of developmental decline in academic intrinsic math motivation and achievement: Childhood through adolescence. International Journal of Behavioral Development, 31(4), 317–327. https://doi.org/10.1177/0165025407077752

26.

Hamutoglu

N. B.

Gemikonakli

Duman

Kirksekiz

Kiyici

(2020). Evaluating students experiences using a virtual learning environment: Satisfaction and preferences. Educational Technology Research & Development, 68(1), 437–462. https://doi.org/10.1007/s11423-019-09705-z

27.

Haynes

(2013). Benjamini–hochberg method. In Dubitzky

Wolkenhauer

Cho

K. H.

Yokota

(Eds.), Encyclopedia of systems biology. Springer. Retrieved from: https://doi.org/10.1007/978-1-4419-9863-7_1215

28.

Shin

Kaldaras

Krajcik

(2024). Integrating artificial intelligence into learning progression-based learning systems to support student knowledge-in-use: Opportunities and challenges. In Jin

Yan

Krajcik

(Eds.), Handbook of research on science learning progressions. Routledge.

29.

Hidayatullah

Csíkos

(2024). The role of students’ beliefs, parents' educational level, and the mediating role of attitude and motivation in students’ mathematics achievement. The Asia-Pacific Education Researcher, 33(2), 253–262. https://doi.org/10.1007/s40299-023-00724-2

30.

Hwang

W.-Y.

Wang

C.-Y.

Hwang

G.-J.

Huang

Y.-M.

Huang

(2008). A web-based programming learning environment to support cognitive development. Interacting with Computers, 20(6), 524–534. https://doi.org/10.1016/j.intcom.2008.07.002

31.

Jansen

R. S.

Van Leeuwen

Janssen

Kester

Kalz

(2017). Validation of the self-regulated online learning questionnaire. Journal of Computing in Higher Education, 29(1), 6–27. https://doi.org/10.1007/s12528-016-9125-x

32.

Jin

Sun

Pan

Lin

C. H.

(2025). High heels, compass, spider-man, or drug? Metaphor analysis of generative artificial intelligence in academic writing. Computers & Education, 228(1), 105248. https://doi.org/10.1016/j.compedu.2025.105248

33.

Jing

Y. H.

Dai

Wang

C. L.

Shen

Shadiev

(2024a). Unleashing the power of virtual learning environment: Exploring the impact on learning outcomes through a meta-analysis. Interactive Learning Environments, 2(1), 1–18. https://doi.org/10.1080/10494820.2024.2335481

34.

Jing

Y. H.

Wang

C. L.

Chen

Z. Y.

Shen

S. S.

Shadiev

(2024b). A bibliometric analysis of studies on technology-supported learning environments: Hot topics and frontier evolution. Journal of Computer Assisted Learning, 40(3), 1185–1200. https://doi.org/10.1111/jcal.12934

35.

Jing

Y. H.

Wang

H. M.

Chen

X. J.

Wang

C. L.

(2024c). What factors will affect the effectiveness of using ChatGPT to solve programming problems? A quasi-experimental study. Humanities & Social Sciences Communications, 11(1), 319. https://doi.org/10.1057/s41599-024-02751-w

36.

Kaldaras

Akaeze

Krajcik

(2021). Developing and validating Next Generation Science Standards‐aligned learning progression to track three‐dimensional learning of electrical interactions in high school physical science. Journal of Research in Science Teaching, 58(4), 589–618. https://doi.org/10.1002/tea.21672

37.

Kaldaras

Akaeze

H. O.

Krajcik

(2023). Developing and validating an Next Generation Science Standards aligned construct map for chemical bonding from the energy and force perspective. Journal of Research in Science Teaching, 61(7), 1689–1726. https://doi.org/10.1002/tea.21906

38.

Kao

Matlen

Weintrop

(2022). From one language to the next: Applications of analogical transfer for programming education. ACM Transactions on Computing Education, 22(4), 1–21. https://doi.org/10.1145/3487051

39.

Kazemitabaar

Wang

Henley

A. Z.

Denny

Craig

Grossman

(2024). Codeaid: Evaluating a classroom deployment of an llm-based programming assistant that balances student and educator needs. In Proceedings of the CHI conference on human factors in computing systems (pp. 1–20). Retrieved from: https://doi.org/10.1145/3613904.3642773

40.

Kong

S. C.

Chiu

M. M.

Lai

(2018). A study of primary school students' interest, collaboration attitude, and programming empowerment in computational thinking education. Computers & Education, 127(1), 178–189. https://doi.org/10.1016/j.compedu.2018.08.026

41.

Kryeziu

S. A.

Avdiu

T. A.

Avdiu

(2021). Examining the teachers, administrators and parents’ view on drawbacks of technology use in education. Ilköğretim Online, 20(2), 1.

42.

Kulesza

DeHondt II

Nezlek

(2011). More technology, less learning? Information Systems Electronic Journal, 9(7), 4. https://isedj.org/2011-9/N7/ISEDJv9n7p4.html

43.

Lee

H. Y.

Lin

C. J.

Wang

W. S.

Chang

W. C.

Huang

Y. M.

(2023b). Precision education via timely intervention in K-12 computer programming course to enhance programming skill and affective-domain learning objectives. International Journal of STEM Education, 10(1), 52. https://doi.org/10.1186/s40594-023-00444-5

44.

Lee

S. W. Y.

Liang

J. C.

Hsu

C. Y.

Tsai

M. J.

(2023a). Students’ beliefs about computer programming predict their computational thinking and computer programming self-efficacy. Interactive Learning Environments, 32(8), 4088–4108. https://doi.org/10.1080/10494820.2023.2194929

45.

Baker

Warschauer

(2020). Using clickstream data to measure, understand, and support self-regulated learning in online courses. Internet and Higher Education, 45(1), 100727. https://doi.org/10.1016/j.iheduc.2020.100727

46.

Y.-F.

Guan

J.-Q.

Wang

X.-F.

Chen

Hwang

G.-J.

(2024). Examining students' self-regulated learning processes and performance in an immersive virtual environment. Journal of Computer Assisted Learning, 40(6), 2948–2963. https://doi.org/10.1111/jcal.13047

47.

Lin

Niu

(2023). Roles of ambiguity tolerance and learning effectiveness: Structural equation modeling evidence from EFL students’ perceptions of factors influencing peer collaboration. Language Teaching Research, 13621688231216201. https://doi.org/10.1177/13621688231216201

48.

Lin

(2024). Learner perceptions of artificial intelligence-generated pedagogical agents in language learning videos: Embodiment effects on technology acceptance. International Journal of Human-Computer Interaction, 41(2), 1606–1627. https://doi.org/10.1080/10447318.2024.2359222

49.

Liu

(2024). Toward artificial intelligence-human paired programming: A review of the educational applications and research on artificial intelligence code-generation tools. Journal of Educational Computing Research, 62(5), 1385–1415. https://doi.org/10.1177/07356331241240460

50.

Liu

Y. C.

G. R.

Yuan

Zhou

Wang

C. L.

(2024). Assessment as learning: Evidence based on meta-analysis and quantitative ethnography research. Studies In Educational Evaluation, 83(1), 101423. https://doi.org/10.1016/j.stueduc.2024.101423

51.

Loksa

Margulieux

Becker

B. A.

Craig

Denny

Pettit

Prather

(2022). Metacognition and self-regulation in programming education: Theories and exemplars of use. ACM Transactions on Computing Education, 22(4), 1–31. https://doi.org/10.1145/3487050

52.

Lukkarinen

Malmi

Haaranen

(2021). Event-driven programming in programming education: A mapping review. ACM Transactions on Computing Education, 21(1), 1–31. https://doi.org/10.1145/3423956

53.

Lye

S. Y.

Koh

J. H. L.

(2014). Review on teaching and learning of computational thinking through programming: What is next for K-12? Computers in Human Behavior, 41(1), 51–61. https://doi.org/10.1016/j.chb.2014.09.012

54.

Qian

Gong

(2023). Promoting programming education of novice programmers in elementary schools: A contrasting cases approach for learning programming. Education and Information Technologies, 28(7), 9211–9234. https://doi.org/10.1007/s10639-022-11565-9

55.

Magen-Nagar

Cohen

(2017). Learning strategies as a mediator for motivation and a sense of achievement among students who study in MOOCs. Education and Information Technologies, 22(3), 1271–1290. https://doi.org/10.1007/s10639-016-9492-y

56.

Messer

Brown

N. C. C.

Kölling

Shi

(2023). Automated grading and feedback tools for programming education: A systematic review. ACM Transactions on Computing Education, 24(1), 1–43. https://doi.org/10.1145/3636515

57.

C. Y.

Wang

C. L.

Dai

Jin

P. Q.

(2022). Video playback speed influence on learning effect from the perspective of personalized adaptive learning: A study based on cognitive load theory. Frontiers in Psychology, 13(1), 839982. https://doi.org/10.3389/fpsyg.2022.839982

58.

Moazami

Bahrampour

Azar

M. R.

Jahedi

Moattari

(2014). Comparing two methods of education (virtual versus traditional) on learning of Iranian dental students: A post-test only design study. BMC Medical Education, 14(1), 45–55. https://doi.org/10.1186/1472-6920-14-45

59.

Morales-Navarro

Giang

M. T.

Fields

D. A.

Kafai

Y. B.

(2023). Connecting beliefs, mindsets, anxiety and self-efficacy in computer science learning: An instrument for capturing secondary school students’ self-beliefs. Computer Science Education, 34(3), 387–413. https://doi.org/10.1080/08993408.2023.2201548

60.

National Research Council . (2012). A framework for K-12 science education: Practices, crosscutting concepts, and core ideas. National Academies Press.

61.

Panadero

Jonsson

Pinedo

Fernández-Castilla

(2023). Effects of rubrics on academic performance, self-regulated learning, and self-efficacy: A meta-analytic review. Educational Psychology Review, 35(4), 113. https://doi.org/10.1007/s10648-023-09823-4

62.

Reisoğlu

Topu

Yılmaz

Karakuş Yılmaz

Göktaş

(2017). 3D virtual learning environments in education: A meta-review. Asia Pacific Education Review, 18(1), 81–100. https://doi.org/10.1007/s12564-016-9467-0

63.

Schunk

D. H.

(2008). Metacognition, self-regulation, and self-regulated learning: Research recommendations. Educational Psychology Review, 20(4), 463–467. https://doi.org/10.1007/s10648-008-9086-3

64.

Sedgwick

Greenwood

(2015). Understanding the Hawthorne effect. BMJ, 351(1), h4672. https://doi.org/10.1136/bmj.h4672

65.

Seralidou

Douligeris

(2021). Learning programming by creating games through the use of structured activities in secondary education in Greece. Education and Information Technologies, 26(1), 859–898. https://doi.org/10.1007/s10639-020-10255-8

66.

Sobral

S. R.

(2021). Teaching and learning to program: Umbrella review of introductory programming in higher education. Mathematics, 9(15), 1737. https://doi.org/10.3390/math9151737

67.

Sorva

Karavirta

Malmi

(2013). A review of generic program visualization systems for introductory programming education. ACM Transactions on Computing Education, 13(4), 1–64. https://doi.org/10.1145/2490822

68.

Y. S.

Shao

Zhao

(2022). Effect of mind mapping on creative thinking of children in scratch visual programming education. Journal of Educational Computing Research, 60(4), 906–929. https://doi.org/10.1177/07356331211053383

69.

Sung

C. Y.

Cherng

F. Y.

Chiu

Y. L.

Chen

P. H.

Chen

B. Y.

(2024). 3CPEs: Concrete computational concepts programming environments for elementary computer science education. In Extended abstracts of the CHI conference on human factors in computing systems (pp. 1–7). Retrieved from: https://doi.org/10.1145/3613905.3650833

70.

Tang

W. Y. R.

Tang

W. Y. R.

(2021). Education, whole person, and the sustainability of a society. Transforming Education in Practice: In Search of a Community of Phronimos, 1(1), 119–140. https://doi.org/10.1007/978-981-16-6871-5_7

71.

Tellhed

Björklund

Kallio Strand

(2022). Sure I can code (but do I want to?). Why boys' and girls’ programming beliefs differ and the effects of mandatory programming education. Computers in Human Behavior, 135(1), 107370. https://doi.org/10.1016/j.chb.2022.107370

72.

Tellhed

Björklund

Kallio Strand

Schöttelndreier

(2023). “Programming is not that hard!” When a science center visit increases young women’s programming ability beliefs. Journal for STEM Education Research, 6(2), 1–23. https://doi.org/10.1007/s41979-023-00094-w

73.

The White House . (2017). Expanding access to high quality STEM and computer science education provides more pathways to good jobs. Retrieved from: https://trumpwhitehouse.archives.gov/briefings-statements/expanding-access-high-quality-stem-computer-science-education-provides-pathways-good-jobs/

74.

The White House . (2018). Charting a course for success:America’s strategy for STEM education. Retrieved from: https://www.whitehouse.gov/wp-content/uploads/2018/12/STEM-Education-Strategic-Plan-2018.pdf

75.

TIOBE . (2023). TIOBE index | TIOBE - the software quality company. Retrieved from: https://www.tiobe.com/tiobe-index/

76.

Trontl

(2010). Problem of computer programming education for technical engineers. Computer Applications in Engineering Education, 18(1), 200–202. https://doi.org/10.1002/cae.20187

77.

Troussas

Krouska

Sgouropoulou

(2020). A novel teaching strategy through adaptive learning activities for computer programming. IEEE Transactions on Education, 64(2), 103–109. https://doi.org/10.1109/TE.2020.3012744

78.

UNESCO . (2023). Generative AI and the future of education. Retrieved from: https://unesdoc.unesco.org/ark:/48223/pf0000385877

79.

Vinnervik

(2023). Programming in school technology education: The shaping of a new subject content. International Journal of Technology and Design Education, 33(4), 1449–1470. https://doi.org/10.1007/s10798-022-09773-y

80.

Magis-Weinberg

Jansen

B. R.

van Atteveldt

Janssen

T. W.

Lee

N. C.

van der Maas

H. L. J.

Raijmakers

M. E. J.

Sachisthal

M. S. M.

Meeter

(2022). Motivation-achievement cycles in learning: A literature review and research agenda. Educational Psychology Review, 34(1), 39–71. https://doi.org/10.1007/s10648-021-09616-7

81.

Wang

A. Y.

Head

Zhang

A. G.

Oney

Brooks

(2023). Colaroid: A literate programming approach for authoring explorable multi-stage tutorials. In Proceedings of the 2023 CHI conference on human factors in computing systems (pp. 1–22). Retrieved from: https://doi.org/10.1145/3544548.3581525

82.

Wang

C. L.

Chen

X. J.

Z. B.

Jin

X. Q.

(2025). Deconstructing university learners’ adoption intention towards aigc technology: A mixed-methods study using ChatGPT as an example. Journal of Computer Assisted Learning, 41(1), e13117. https://doi.org/10.1111/jcal.13117

83.

Wang

C. L.

Chen

X. J.

Liu

Y. D.

Jing

Y. H.

(2024a). Education reform and change driven by digital technology: A bibliometric study from a global perspective. Humanities & Social Sciences Communications, 11(1), 256. https://doi.org/10.1057/s41599-024-02717-y

84.

Wang

C. L.

Wang

H. M.

Y. Y.

Dai

X. Q.

(2024b). Factors influencing university students’ behavioral intention to use generative artificial intelligence: Integrating the theory of planned behavior and AI literacy. International Journal of Human-Computer Interaction, 1(1), 1–23. https://doi.org/10.1080/10447318.2024.2383033

85.

Wei

Wang

Tang

Lei

Peng

(2024). Influence of programming education modalities on the computational thinking in young children: A comprehensive review and meta analysis. Journal of Computer Assisted Learning, 40(5), 2385–2397. https://doi.org/10.1111/jcal.13038

86.

Zhao

Liew

Zhou

Kogut

(2023). Synthesizing research evidence on self-regulated learning and academic achievement in online and blended learning environments: A scoping review. Educational Research Review, 39(1), 100510. https://doi.org/10.1016/j.edurev.2023.100510

87.

Yasmin

Naseem

Masso

I. C.

(2019). Teacher-directed learning to self-directed learning transition barriers in Pakistan. Studies In Educational Evaluation, 61(1), 34–40. https://doi.org/10.1016/j.stueduc.2019.02.003

88.

Yolcu

Demirer

(2023). The effects of educational robotics in programming education on students' programming success, computational thinking, and transfer of learning. Computer Applications in Engineering Education, 31(6), 1633–1647. https://doi.org/10.1002/cae.22664

89.

Zdawczyk

Varma

(2022). Engaging girls in computer science: Gender differences in attitudes and beliefs about learning scratch and python. Computer Science Education, 33(4), 600–620. https://doi.org/10.1080/08993408.2022.2095593

90.

Zhang

Jia

Cheng

Son

J. Y.

Stigler

J. W.

(2025). Empowering novice programmers: Redesigning documentation for effective informal learning. Journal of Educational Computing Research, 1(1), 1–33. https://doi.org/10.1177/07356331241311505

91.

Zhang

Lin

C. H.

(2021). Effects of community of inquiry, learning presence and mentor presence on K-12 online learning outcomes. Journal of Computer Assisted Learning, 37(3), 782–796. https://doi.org/10.1111/jcal.12523

92.

Zimmerman

B. J.

(2002). Becoming a self-regulated learner: An overview. Theory Into Practice, 41(2), 64–70. https://doi.org/10.1207/s15430421tip4102_2

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.07 MB