A qualitative study of physical education teachers’ perceptions of artificial intelligence and influencing factors based on social cognitive theory - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice BMC Psychol . 2026 Mar 6;14:530. doi: 10.1186/s40359-026-04295-x Search in PMC Search in PubMed View in NLM Catalog Add to search A qualitative study of physical education teachers’ perceptions of artificial intelligence and influencing factors based on social cognitive theory Weilei Yang Weilei Yang 1 School of Physical Education, Guangdong Polytechnic Normal University, Tianhe, Guangzhou, China Find articles by Weilei Yang 1 , Haidong Chen Haidong Chen 2 School of Physical Education and Sports Science, South China Normal University, Panyu, Guangzhou, China Find articles by Haidong Chen 2, ✉ , Feixue Rao Feixue Rao 3 Guangzhou Huanan Business College, Baiyun, Guangzhou, China Find articles by Feixue Rao 3 Author information Article notes Copyright and License information 1 School of Physical Education, Guangdong Polytechnic Normal University, Tianhe, Guangzhou, China 2 School of Physical Education and Sports Science, South China Normal University, Panyu, Guangzhou, China 3 Guangzhou Huanan Business College, Baiyun, Guangzhou, China ✉ Corresponding author. Received 2025 Jul 12; Accepted 2026 Mar 2; Collection date 2026. © The Author(s) 2026 Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ . PMC Copyright notice PMCID: PMC13081590 PMID: 41792799 Abstract With the application of artificial intelligence in physical education, how physical education teachers understand and embrace AI technology in their actual teaching has become a key issue in the reform of physical education. However, existing research has paid insufficient attention to physical education teachers’ cognitive states and their influencing factors, making it difficult to provide a basis for effectively integrating AI technology into physical education classrooms. This study aims to reveal physical education teachers’ perceptions of AI and clarify the core factors influencing their acceptance of AI technology. Based on Social Cognitive Theory, the research primarily sets three objectives: First, to analyse physical education teachers’ basic perceptions and attitudes toward AI; second, to examine the personal, environmental, and behavioural factors affecting their acceptance of AI technology; and third, to construct a model of the mechanism underlying physical education teachers’ acceptance of AI technology. This study employed qualitative research methods, utilising purposive sampling to conduct semi-structured interviews with 16 physical education teachers. Data coding and thematic analysis were performed using NVivo software. Findings revealed that a combination of multidimensional factors, including personal, environmental, situational, and resource factors, influences physical education teachers’ acceptance of AI. Among these, self-efficacy serves as the core factor driving physical education teachers’ acceptance of AI. Outcome expectancy significantly influences behavioural decisions, while subjective norms mediate the relationship between self-efficacy and these decisions. By integrating SCT with SDT, this study provides a more context-sensitive explanation of PE teachers’ AI acceptance in embodied teaching settings. Keywords: Social cognitive theory, Physical education, Physical education teachers, Artificial intelligence, Self-efficacy Introduction Artificial Intelligence, as the core engine of the digital technology revolution, is typically defined as “a technological system that simulates, extends, or even surpasses human intelligence through algorithms, data, and computational power” [ 1 ]. The implementation of this technology is not an isolated event, but rather deeply dependent on the foundational digital transformation of society and industry [ 2 ]. In education, AI has demonstrated the potential to reshape traditional educational paradigms. By integrating learning analytics, cognitive simulation, and self-correcting mechanisms, it drives personalised learning, automated assessment, and the optimisation of distance education [ 3 ]. Notably, AI has gradually expanded into the sports sector, which is characterised by physical activity and open-field settings. From recognising athletes’ training movements to supporting sports competition systems [ 4 , 5 ], AI is comprehensively reshaping the operational logic of the sports industry. However, compared to other disciplines, the application of AI technology in popular sports settings faces unique challenges. Focusing specifically on school physical education reveals that implementing AI technology is not merely a matter of tool adaptation. As the pivotal intermediaries bridging technology and the classroom, physical education teachers or coaches’ level of understanding, willingness to adopt, and practical operational skills regarding AI technology directly determine whether the technology can transition from a “laboratory tool” to a “teaching or training assistant.” This study addresses this core contradiction by revealing physical education teachers’ perceptions of AI and clarifying the key factors influencing their acceptance of AI technology. The application of AI in physical education The integration of AI into education is reshaping learning processes and teaching practices, and this broader trend has also extended to physical education (PE) [ 6 ]. In PE contexts, AI is increasingly used to improve instructional efficiency and support students’ learning outcomes through functions such as personalised guidance, real-time feedback, and performance assessment [ 7 , 8 ]. Existing studies further show that AI-based approaches can contribute to physical fitness development and provide practical support for improving the quality of PE programs, particularly in institutional settings [ 9 ]. In addition, AI-enabled interactive training systems can process training data and deliver immediate feedback, helping learners adjust strategies in real time and strengthening the continuity of teaching and training [ 10 ]. At the university level, AI is also viewed as a driver of broader transformation, including data management, governance, teaching services, research support, and athletic evaluation [ 11 ]. Related technologies such as virtual reality, cloud computing, and big data analytics have further expanded the possibilities of PE by enabling immersive learning environments, data visualisation, and more integrated communication and support systems [ 12 , 13 ]. In K-12 physical education, K-12 refers to the entire basic education stage from preschool through high school in countries such as the United States and the United Kingdom, where teaching demands are often shaped by large class sizes and diverse student needs, AI has likewise attracted growing attention. Trendowski examined the use of generative artificial intelligence (GAI) in K-12 PE curriculum planning and proposed a checklist to support the systematic and critical evaluation of GAI-generated content [ 14 ]. Ahn further suggested that AI can function in multiple instructional roles—such as operational assistant, personal trainer, group coach, and evaluator—thereby helping teachers address classroom management, feedback delivery, team organisation, and assessment tasks more effectively [ 15 ]. Overall, current research consistently highlights AI’s value in optimising sports education and related sports practices. However, much of this work focuses on technological functions and application scenarios, while paying less attention to how PE teachers perceive AI and to the factors that shape those perceptions. This gap directly motivates the present study. Opportunities and challenges facing physical education teachers in the wave of AI Research indicates that the teaching community generally holds a positive attitude toward AI education and is willing to promote the integration of AI-related content into school curricula, providing an important foundation for AI adoption in education [ 16 ]. At the same time, evidence from other disciplines shows that teachers often encounter substantial implementation burdens when using AI systems, including increased instructional effort and stress, suggesting that these difficulties are not subject-specific and are likewise relevant to physical education teachers [ 17 ]. Against this background, the rapid development of AI is raising new expectations for physical education teachers’ professional competence. Teachers are increasingly required to adjust their pedagogical skills, technical knowledge, and instructional thinking to adapt to technology-rich teaching environments [ 18 ]. Existing studies further emphasise that physical education teacher training should align with the development of smart and digital sports by updating training concepts, content, and methods, so that teachers can understand AI technologies, operate AI tools, and apply them to sports-related instructional problems [ 19 ]. This need has become even more urgent with the emergence of AI-driven tools such as ChatGPT, which have introduced new challenges and decision-making demands for teacher education in health and physical education [ 20 ]. Although AI offers substantial opportunities for sports education, its implementation in practice remains constrained by multiple barriers. Prior research identifies persistent challenges, including limited teacher preparedness and familiarity with AI, concerns about data privacy and security, and insufficient infrastructure support [ 21 ]. In physical education specifically, inadequate digital competence remains a prominent issue, shaped by both individual and institutional constraints on technology integration [ 22 ]. Related studies in China further show that problems in the professional development system of physical education teachers—such as outdated training content, limited development opportunities, and narrow evaluation mechanisms—continue to affect teachers’ ability to respond to new educational demands [ 23 ]. In addition, while teachers have increasingly engaged with new curriculum standards, some studies suggest that their understanding remains relatively superficial, which may limit content selection and pedagogical innovation [ 24 ]. Taken together, the influence of AI on physical education teachers extends beyond technical adoption; it also encompasses professional development, pedagogical beliefs, and broader instructional transformation. More broadly, AI is reshaping human-technology interaction and has significant implications for teacher education [ 25 ]. While previous studies have examined physical education teachers’ readiness to use wearable technology in PE settings [ 26 ], AI applications in physical education extend far beyond wearables. Accordingly, this study focuses on PE teachers’ perceptions of AI and the factors influencing those perceptions through a psychological lens. By doing so, it contributes to interdisciplinary research at the intersection of physical education, AI, and educational psychology. It extends the explanatory value of technology-related acceptance perspectives in PE contexts. Furthermore, by drawing on frameworks such as Social Cognitive Theory (SCT), this study provides a deeper account of how technological perceptions are linked to teachers’ pedagogical beliefs, self-efficacy, and technological anxiety. Theoretical foundations Bandura’s Social Cognitive Theory (SCT) serves as the core framework for this study, informing the research design, variable selection, and explanatory logic. SCT emphasises the triadic reciprocal interaction among individual cognition, behavioural performance, and the social environment [ 27 , 28 ]. Its key constructs—such as self-efficacy, observational learning, and self-regulation—closely align with the processes by which physical education (PE) teachers perceive, evaluate, and adopt AI technologies. In this process, teachers’ beliefs and attitudes, their actual technology-use behaviours, and contextual factors such as school support and peer influence are not independent variables; rather, they form a dynamic and mutually reinforcing system. This perspective is highly consistent with the real-world conditions under which PE teachers engage with AI in educational settings. Although this study adopts SCT as its primary framework, prior research on technology acceptance also provides important background support. The Technology Acceptance Model (TAM) has shown strong explanatory value in educational technology adoption, particularly in clarifying how self-efficacy shapes perceived usefulness and perceived ease of use [ 29 ]. Related research has further confirmed that self-efficacy directly influences individuals’ acceptance of digital tools in educational and sport-related contexts [ 30 ]. Cross-domain studies likewise suggest that background factors may affect behavioural tendencies through self-efficacy [ 31 ], and that environmental influences often need to be translated into behavioural intention through mediating mechanisms such as self-efficacy and trust [ 32 ]. Taken together, these findings support the use of an SCT-based pathway to analyse how PE teachers’ AI-related cognition, confidence, and school environment jointly shape adoption behaviour. However, SCT alone may be less effective in explaining the motivational quality underlying teachers’ AI adoption—particularly why some teachers engage proactively. In contrast, others remain passive, even under similar external conditions. To address this issue, this study introduces Self-Determination Theory (SDT) as a complementary framework. SDT, proposed by Deci and Ryan, explains how high-quality, sustainable motivation is generated [ 33 ]. It conceptualises motivation as a continuum from external regulation to intrinsic motivation and argues that more autonomous forms of motivation are associated with greater persistence and better performance. According to SDT, such motivation depends on the satisfaction of three basic psychological needs: autonomy, competence, and relatedness [ 34 ]. In educational research, SDT has been widely used to explain teacher motivation and professional behaviour. Existing studies show that when teachers experience greater autonomy and fulfilment of basic psychological needs, they are more likely to demonstrate higher self-efficacy, job satisfaction, professional commitment, and positive teaching engagement [ 35 ]. Conversely, unmet psychological needs may weaken motivation and contribute to burnout or withdrawal. For the present study, SDT therefore provides a useful lens for understanding the internal motivational differences behind PE teachers’ responses to AI technologies(Table 1 ). Table 1. Core Elements of SCT and SDT and Their Specific Manifestations in This Study Theoretical Framework Core Elements The specific manifestation in this study SCT Triadic Interaction (Individual/Behaviour/Environment) Teachers’ Perceptions of AI, Usage Behaviours, Environmental Feedback, and Cognitive Restructuring Self-efficacy Teachers’ Beliefs in Their AI Usage Competence Observation learning Demonstration Effect of Companion AI Usage Behaviour SD T Autonomy Teachers’ right to independently choose the type of AI tools and their usage scenarios Sense of competence Confidence in Using AI-Assisted Teaching and Perceived Effectiveness Sense of belonging Support for the School AI Technology Exchange Community Open in a new tab SCT and SDT are thus theoretically complementary. SCT explains the dynamic interaction among cognition, behaviour, and environment in teachers’ AI adoption, whereas SDT clarifies the motivational mechanisms that shape the quality and sustainability of that adoption. Based on this complementarity, the present study uses SCT as the primary analytical framework and SDT as a supporting lens. Through qualitative inquiry, it examines how PE teachers’ AI-related perceptions are formed and translated into practice, and how self-efficacy, environmental support, and psychological need satisfaction jointly influence their technology adoption. This combined framework not only strengthens the theoretical basis for interpreting PE teachers’ AI acceptance but also provides clearer guidance for future teacher training and intervention design. Research methods Research design This study employed a qualitative research approach grounded in an interpretivist paradigm to explore physical education teachers’ cognitive experiences, subjective attitudes, and behavioural intentions toward AI. Given that AI remained in its nascent stages within physical education, qualitative research offered richer contextual details than quantitative data, thereby facilitating the revelation of underlying motivations and complex human-machine interactions. Participants and sampling strategy This study employed purposive sampling to recruit participants. The inclusion criteria were as follows: (1) currently employed physical education teachers; (2) having some understanding of, or exposure to, digital technologies or AI tools; and (3) representing diversity in gender and teaching experience to enhance sample variation. Given that AI adoption in physical education is currently more evident in higher education than in primary and secondary schools, this study focused on university physical education teachers as interview participants. A total of 16 university physical education teachers participated in the study, including 8 males and 8 females, with teaching experience ranging from 1 to 21 years. All participants signed informed consent forms before the interviews, and the study strictly followed academic ethics and privacy protection principles (Table 2 ). Table 2. Interviewed Teachers Information Form ID Gender Age Educational Level Years of teaching experience Do you have experience using AI 1 Female 30 Master’s degree 5 Yes 2 Male 44 Bachelor’s degree 19 Yes 3 Female 46 Bachelor’s degree 21 Yes 4 Male 34 Master’s degree 7 Yes 5 Male 31 Master’s degree 4 Yes 6 Male 30 Master’s degree 2 Yes 7 Female 31 Master’s degree 2 Yes 8 Female 28 Master’s degree 2 Yes 9 Male 44 Bachelor’s degree 21 Yes 10 Female 25 Master’s degree 2 Yes 11 Male 29 Master’s degree 1 Yes 12 Male 32 Master’s degree 6 Yes 13 Female 30 Master’s degree 2 Yes 14 Male 31 Master’s degree 4 Yes 15 Female 32 Master’s degree 3 Yes 16 Female 35 Master’s degree 5 Yes Open in a new tab Data collection This study primarily used semi-structured interviews for data collection. An SCT-informed interview guide was developed to align the interview content with the study’s analytical framework. The guide included 10 core questions targeting relationships among individual cognition, behaviour, and environment, and was supplemented with flexible follow-up probes to capture participants’ personal experiences and emergent insights (Table 3 ). Given participants’ geographical distribution and teaching schedules, interviews were conducted in a hybrid format (face-to-face or online voice call). Each interview lasted approximately 30 min. With participants’ consent, all interviews were audio-recorded and transcribed verbatim for subsequent analysis. Table 3. Interview Guide for Teachers’ Perceptions of AI Technology and Influencing Factors Interview Purpose Questions Teaching Background and Initial Experience 1. Please briefly describe your years of teaching experience and primary teaching subjects. 2. When did you first encounter AI technology (such as smart devices or data analysis tools)? 3. Could you describe that experience? Confidence and Experience in Using New Technology 1. How confident were you in your ability to utilise AI technology to support teaching? What factors might have influenced your learning of new technologies (such as training, age, or experience)? 2. Could you share an experience where you tried using a new technology? What challenges did you encounter during that experience, and what kind of support did you receive? Observing Others’ Experiences and Influences 1. Have you ever seen other teachers using AI in the classroom? Please describe the most memorable instance you recalled. 2. Did that teacher’s approach make you think, “I could try that too”? Did this observation influence your attitude toward new technologies? What changes did you experience? Expectations and Concerns About AI Technology 1. What practical benefits did you believe AI technology could bring to physical education instruction? Please provide examples. 2. Did you have any concerns about potential risks or issues associated with this technology? For instance, were you worried it might alter teachers’ roles in the classroom? School Environment and External Support 1. Was your school promoting and supporting AI technology? If so, what specific measures had been implemented? How had these measures impacted you? 2. Could you share how you had adjusted your perspective on and approach to using this new technology with the training or policy support provided by your school? Interaction Between Personal Effort and External Support 1. In your experience using AI technology, what kind of interaction had you observed between yourself and your school or colleagues? For example, had your efforts elicited positive feedback from your surroundings, and what kind of feedback had emerged? 2. How did this mutual influence between the individual and the environment affect your willingness to adopt new technologies in the future? Overall Experience and Suggestions 1. Overall, how did you view the prospects for applying AI in sports education? 2. For schools or education departments, what recommendations did you have to help teachers better adapt to and utilise this technology? Open in a new tab 1. Explain the purpose and use of the interview. 2. The interviewer should not deliberately guide the conversation, maintaining neutrality and encouraging the interviewee to express their own feelings as much as possible. 3. When the interviewee’s response to a question is vague or unclear, appropriate open-ended follow-up questions should be used. 4. Record the entire interview and transcribe it Data analysis methods Qualitative data were analysed using a hybrid thematic analysis that combined deductive and inductive coding. This approach followed the hybrid coding logic proposed by Fereday and Muir-Cochrane [ 36 ] and the thematic development procedures outlined by Braun and Clarke [ 37 ]. Deductive coding was guided by Social Cognitive Theory (SCT), particularly its triadic framework of individual, behaviour, and environment, which served as the a priori analytical structure [ 38 ]. Inductive open coding was then applied to data segments not fully captured by the SCT framework, thereby expanding and refining the coding system based on participants’ accounts. Through iterative comparison and revision, codes were organised into themes and subthemes. NVivo 12 was used for data management, coding, and node organisation [ 39 ]. To ensure research rigour, the analysis was conducted with attention to credibility, transferability, dependability, and confirmability [ 40 ], and the reporting process was aligned, where applicable, with the COREQ checklist [ 41 ]. The analysis proceeded in three steps. First, after obtaining participants’ consent, all interviews were transcribed verbatim and reviewed repeatedly to identify salient points and preliminary analytic impressions. Second, an initial deductive coding structure was established in NVivo based on the SCT triadic model (“individual,” “behaviour,” and “environment”), followed by inductive open coding for data that extended beyond the predefined framework. Third, deductive and inductive codes were compared, merged, and hierarchically organised into themes and subthemes through iterative refinement. During this process, analytic memos, coder discussion, and cross-checking of node relationships were used to strengthen consistency and maintain a clear evidence chain for the final thematic interpretation (see Fig. 1 ). Fig. 1. Open in a new tab Qualitative data analysis procedure Result analysis Results of deductive coding analysis under SCT SCT emphasises the reciprocal interactions among personal cognition, environmental factors, and behaviour. The interview data show that physical education (PE) teachers’ AI-related perceptions and practices vary across contexts. Based on the deductive coding framework, the findings are organised into three dimensions (Table 4 ): personal factors (cognition, attitudes, and beliefs toward AI), environmental factors (school support, peer influence, and broader context), and behavioural factors (willingness to use, actual use patterns, and expected outcomes). Table 4. Deductive Coding Framework (Based on SCT) Encoding Category Encoding Subcategory Code Definition Interview Transcript Example (Must be authentic) Personal factors Cognition Respondents’ level of understanding of AI technology “In my teaching work, I haven’t had much exposure to AI yet. I believe AI might find more applications in specialised sports instruction or athlete training. ” Attitude Respondents’ emotional attitudes toward AI “It feels like this era is truly the age of AI. Before, my understanding of AI was still quite superficial… I feel that new technologies have brought many conveniences to teaching, and I’m very eager to learn and experiment with them. ” Belief Teachers’ comprehensive understanding of their own AI proficiency (self-efficacy) and their value judgments regarding the consequences of AI application (outcome expectations). “I believe AI can still offer valuable assistance in teaching… When it comes to the most cutting-edge knowledge in sports dance, if I haven’t gained it through extensive reading of academic papers or news articles, I can ask the AI. It can provide me with a summary of the latest information—essentially delivering the answer to me in the fastest way possible.” Environmental factors School Support The school’s resources and policy initiatives to promote and support AI technology. “The school is gradually implementing AI technologies, such as planning to provide relevant training for teachers and procuring intelligent physical fitness testing and sports training equipment. ” Peer influence Did respondents experience changes in their perceptions, attitudes, or behaviours toward AI technology due to the influence of their colleagues? “Regarding AI, we maintain excellent communication among colleagues and foster positive interactions… Overall, our colleagues are highly enthusiastic about new technologies, creating a very supportive environment. ” Social Environmental Impact The influence of society’s overall promotion and acceptance of AI technology on respondents “The more hype there is around environmental issues and the greater the exposure, the more my awareness of using AI will increase. ” Behavioral factors willingness to use Respondents’ propensity to use AI technology in future teaching “I believe this experience will make me even more eager to adopt new technologies in the future, as they have genuinely enhanced teaching effectiveness and motivated me to explore further.” Actual usage The frequency, duration, and scope of teachers’ current use of AI in teaching. “Primarily used for writing lesson plans, news releases, and similar materials, but due to the unique nature of physical education classes, these technologies have not yet been applied in teaching settings—mainly because of high costs.” Usage Instructions Specific strategies and forms of AI implementation by teachers in practical work, such as assisting with content creation, human-machine collaborative lesson preparation, or classroom teaching support. “Use it to help polish my thesis, as well as gather and compile materials.” Open in a new tab Personal factors At the personal level, teachers’ self-efficacy showed clear task-specific variation. Confidence was generally higher for text-based or preparatory tasks, where AI was perceived as directly improving efficiency: “My first encounter with AI was during my graduate studies… Using resources from our studio, I employed ChatGPT to help polish my thesis… That experience opened my eyes to a new understanding of AI… I remain highly confident in applying this technology to teaching.” (Interviewee 11). By contrast, self-efficacy declined when AI use involved embodied skill instruction in PE classes. Teachers reported that general AI tools often could not directly address discipline-specific instructional needs and still required substantial manual adjustment: “I asked him to draft an 18-week lesson plan… He’s not incapable of producing what I want, but I constantly have to fill in the gaps for him… Ultimately, what he delivers still requires me to make numerous adjustments based on the rough outline he provides… By the time the final product emerges, I have to process it all through my own brain.” (Interviewee 6). Environmental factors At the environmental level, school-level support and peer-level influence functioned differently. School promotion increased awareness of AI, but many teachers described institutional support as general rather than PE-specific, with limited equipment, funding, or systematic training: “The school doesn’t offer systematic training; it’s all self-taught… There’s no AI-related equipment at the school, and funding is inadequate. The school invites experts to give lectures, which are shared in group chats for everyone to watch at their own pace… Without training or policy support in this area, it mainly relies on self-study.” (Interviewee 13). In contrast, peer demonstration had a stronger influence on teachers’ willingness to try AI. Observing colleagues use AI in actual teaching situations provided concrete reference points and prompted further exploration: “I did see a colleague use it… He captured a student’s Tai Chi movements using the video recording feature, then applied AI to analyse the student’s form based on those movements… This had a significant impact on me… It really affected me at the time, and I ended up having a lengthy discussion with the teacher about it.” (Interviewee 3). Behavioural factors At the behavioural level, AI use was concentrated in lesson preparation and administrative work, while classroom integration remained limited. Most teachers described AI as a tool for drafting, planning, and reducing paperwork: “I used Doubao AI to draft an outline, then refined it myself. I primarily use AI to write lesson plans and teaching schedules… Besides teaching, there’s a lot of administrative work, and sometimes I’m overwhelmed. So I use AI to improve my work efficiency.” (Interviewee 16). In contrast, AI use in live PE teaching and teacher–student interaction was rare. Teachers reported low adoption in classroom or field-based instruction, especially in curriculum delivery and skill teaching: “Regarding physical education instruction, the adoption rate remains relatively low. Most inquiries focus on how to deliver effective lessons. However, when it comes to integrating AI into curriculum delivery—such as through multimedia devices—these technologies are still largely absent from physical education classes.” (Interviewee 2). “It’s mainly used for writing lesson plans, news releases, and such. But due to the unique nature of physical education classes, these technologies haven’t been applied during actual instruction yet.” (Interviewee 15). Overall, the deductive coding results indicate a patterned gap across the three SCT dimensions: PE teachers reported relatively strong confidence and frequent AI use in preparatory tasks, but limited classroom application under conditions of weak subject-specific support and insufficient peer-validated practice models. Thematic emergence results under inductive coding Beyond the categories captured by the deductive framework, the inductive analysis identified three recurring themes in PE teachers’ accounts of AI adoption: embodied rejection, resource stratification, and performative adoption (Table 5 ). These themes reflect discipline-specific concerns that were repeatedly mentioned across interviews. Table 5. Thematic Coding Framework (Themes Emerged from Interviews) Core Theme Sub-theme Code Definition Sample Interview Transcript The Repellent Effect of Embodied Disciplinarity The irreplaceable presence of the body Emphasising that physical education relies on bodily perception, on-the-spot correction, and interpersonal interaction, it is argued that AI lacks physical experience and cannot impart tacit knowledge through hands-on instruction. “Sports will always be physical activity. No matter how advanced your smart devices are, they still require physical movement to function. ”(T6) Non-standardisation of teaching scenarios It highlights the disconnect between the physical education teaching environment (outdoor, dynamic) and AI’s current strengths in standardised text generation, leading to limited technological applicability. “Content generated by AI can be useful, but it differs from actual teaching practices and should not be applied wholesale.” (T8) Pragmatic Resource Stratification Competition-Oriented Privileged Applications AI equipment is prioritised for high-level athletic teams capable of producing results (gold medals/honours) rather than for inclusive public physical education classes. “These can be used in the school’s training teams… to bring honour to the school; but for public physical education classes… there’s no need to use such high-end equipment.” (T2) Resource Shortfall in Routine Teaching Standard physical education classes face severe hardware deficiencies (no large screens, no internet access), preventing the implementation of technology. “Downloading DeepSeek is free, but training and education cost money; purchasing equipment costs even more… Ultimately, it all comes down to money. ”(T15) The Crisis of Professional Identity and Performance The Anxiety of De-Skilling Concerns that AI may reduce teachers to auxiliary roles (device administrators), undermining their professional authority and emotional value. “This may reduce some physical education teachers to mere equipment managers, stripping them of their ability to provide personalised instruction.” (T1) Symbolic performative adoption Not driven by genuine educational needs, but rather to meet external demands for “educational modernisation,” AI is being used as a mere ‘showpiece’ or “decoration.” “We will no longer allow society to harbour a misconception about schools… that physical education is outdated.” (T6) Open in a new tab Embodied rejection The embodied nature of PE was repeatedly described as a barrier to AI integration [ 42 ]. Teachers emphasised that PE teaching relies on in-person demonstration, immediate correction, and interpersonal interaction, which they viewed as difficult for current AI tools to replace. “Because physical education is still about studying human beings… The AI you mentioned can explain things and provide examples. But when it comes to verbal expression and the personal experiences inherent in human interaction, AI currently cannot replace that… It cannot replace us, physical education teachers, or the elements of our physical education classes. However, I believe it can fully replace the theoretical aspects of teaching.” (Interviewee 7). Teachers also highlighted the real-time and organisational demands of PE classes, especially in skill correction and classroom control, and questioned whether AI could provide comparable instructional guidance. “Since physical education involves teaching technical skills, instructors must not only explain and demonstrate movements but also correct students’ form… If AI were to provide guidance, it would likely depend on students’ self-discipline, unlike the structured organisation provided by teachers.” (Interviewee 5). Resource stratification A second theme was unequal access to AI-related resources. Interview data showed a clear distinction between competitive sports settings and general PE classes, with AI equipment and funding concentrated in performance-oriented teams. “It can be used within the school’s training teams, because these teams exist to help the school achieve good results and bring honour to the institution… So my suggestion is that general education, elective, and physical education courses should focus on improving students’ physical fitness—the goal is for them to get exercise during this process… It’s not necessarily required to use these high-end instruments.” (Interviewee 2). “The ski team receives significant attention from the school due to its strong competition results, with dedicated funding allocated annually—primarily for purchasing equipment like heart rate monitors… Ultimately, it boils down to funding: more investment leads to better development; less investment means falling short.” (Interviewee 15). By contrast, teachers in general PE classes reported limited access to funding and equipment, which restricted the practical use of AI in everyday teaching. “Funding is inadequate… and new technologies are costly.” (Interviewee 13). Performative adoption The third theme was performative adoption. Some teachers expressed concern that AI use might weaken the teacher’s instructional role or reduce teacher–student interaction, especially if technology became over-relied upon. “What concerns me is that technology carries the risk of marginalising the role of teachers… I fear this could reduce some physical education teachers to mere equipment managers. That is, they would completely lose their ability to provide personalised guidance, reduced solely to managing equipment and reviewing data.” (Interviewee 1). “I’m concerned that AI might alter the role of teachers in the classroom. For instance, if we become overly reliant on technology, we risk gradually losing that stage of direct interaction with students, which would undoubtedly impact the emotional connection between teachers and students.” (Interviewee 3). At the same time, some teachers described AI use as a way to signal modernisation or technological advancement, especially under institutional or social expectations. “If we create an AI-powered environment for students… it gives them a sense of, well, technological sophistication and intelligence.” (Interviewee 9). “We must also avoid perpetuating societal misconceptions about our school. Currently, the prevailing view is that our school’s learning theories lag behind societal developments. However, our institution maintains an open approach. It continuously adapts to societal trends, driving teachers to engage in ongoing professional development.” (Interviewee 6). Overall, the inductive coding results indicate that PE teachers’ AI acceptance was shaped not only by personal efficacy and immediate teaching needs, but also by disciplinary norms, resource distribution, and external expectations. Reconstructing cognition and context: an analysis of integrative mechanisms based on SCT SCT conceptualises behaviour as the product of reciprocal interactions among individual factors, environmental conditions, and behavioural outcomes [ 43 ]. Based on the inductive themes identified in this study, an integrative mapping was conducted by aligning emergent themes with SCT constructs according to their explanatory salience (Table 6 ; Fig. 2 ). This process yielded three dominant linkages: from embodied rejection to self-efficacy, from resource stratification to subjective norms/environmental support, and from performative adoption to outcome expectancy. Table 6. Integrated Analysis of Factors Influencing Physical Education Teachers’ Acceptance of AI Theoretical Dimensions Theme Emerging Themes Attribution Judgment Logic Theme of Integration Mechanisms Characteristics Personal factors The Repellent Effect of Embodied Disciplinarity Why do personal beliefs face resistance? Because the “physical presence” in physical education directly challenges the validity of “data algorithms.” This represents the most direct challenge to the perception of ability. Self-efficacy The disconnect between desk-bound confidence and on-site apprehension. Teachers exhibit high efficacy in handling generic text tasks yet low efficacy in core action instruction, resulting in acceptance rates plummeting dramatically across different scenarios. Environmental factors Pragmatic Resource Stratification Why is environmental support ineffective? Because material resources are the hard threshold for implementing teaching practices, and they hold greater sway than policy advocacy. Subjective Norms The “regulatory failure” of soft support and hard resources. The subjective normative pressure imposed by school systems (soft support) is blocked by the resource deficit in regular physical education classes (hard gap), resulting in norms failing to regulate behavior. Behavioral factors The Crisis of Professional Identity and Performance Why does behavior become alienated? Because teachers must strike a balance between job security and external evaluations—a trade-off that shapes outcomes. Expected Outcomes The dual-game of instrumental value and occupational risk. To balance these two diametrically opposed outcome expectations, teachers ultimately adopted a compromise behavior of “performative/exhibitionist use.” Open in a new tab Fig. 2. Open in a new tab Mechanisms Influencing Physical Education Teachers’ Acceptance of AI Embodied rejection and contextualized self-efficacy The first linkage shows that the theme of embodied rejection corresponds most directly to the SCT construct of self-efficacy. Interview data indicate that teachers’ confidence in AI is context-dependent: they report relatively high efficacy in text-based and preparatory tasks, but lower efficacy in embodied skill instruction. One participant described AI as useful for theoretical and informational support in teaching: “AI in teaching… Previously, information in edited books would become outdated, but now with AI… we gain a more comprehensive understanding. During instruction, it becomes easier to grasp concepts and integrate knowledge across the entire spectrum.” (Interviewee 14). In contrast, another participant reported difficulty using AI for discipline-specific instructional problems in movement teaching: “He couldn’t accurately answer my questions, so I had to adjust my phrasing… It couldn’t provide the answers I needed, and ultimately, when it came to presenting the final output, I still had to process it using my own human brain.” (Interviewee 6). These accounts indicate that self-efficacy varies across task types and is reduced when AI use enters the embodied core of PE instruction. Resource stratification and constrained environmental support The second linkage connects resource stratification to the SCT environmental dimension, particularly the practical force of subjective norms and support conditions. Interview data show that teachers perceived school-level encouragement, but repeatedly emphasized the gap between policy rhetoric and actual resource provision. As one participant noted: “The school is promoting… but I know that teaching under these circumstances would definitely require the school to invest funds to purchase this equipment… Given our school’s current situation, physical education isn’t considered a top priority subject. It’s essential but not highly emphasized.” (Interviewee 2). Another participant similarly highlighted the lack of training and equipment: “Without training or policy support in this area, we mainly rely on self-study… The school lacks AI-related equipment, and funding is insufficient.” (Interviewee 13). These findings show that institutional encouragement alone did not translate into usable support when equipment, funding, and PE-specific training were limited. Performative adoption and split outcome expectancy The third linkage maps performative adoption onto the behavioral dimension of SCT, especially teachers’ outcome expectancy. Interview accounts showed both negative and positive expectations regarding AI use, which appeared to coexist. On one hand, teachers expressed concern that technology-centered teaching might weaken their professional role and reduce teacher–student interaction: “What concerns me is that technology carries the risk of marginalizing the role of teachers… I fear this could reduce some physical education teachers to mere equipment managers. That is, they would completely lose their ability to provide personalized guidance, reduced solely to managing equipment and reviewing data.” (Interviewee 1). On the other hand, some teachers described AI-related practices as a way to present the subject or school as modern and responsive to broader educational trends: “The more this topic gains traction and exposure… the more the school will undoubtedly actively support us… to dispel the misconception society has about our institution—that its theoretical curriculum lags behind societal demands.” (Interviewee 6). Taken together, these accounts indicate a split pattern of expected outcomes, in which concerns about professional risk coexist with expectations of institutional recognition and symbolic modernization. Discussion Individual level: the tension between embodied discipline and “situated self-efficacy” Self-efficacy serves as the pivotal link between cognition and behaviour [ 44 ]. However, this study reveals that physical education teachers’ AI self-efficacy exhibits strong “context dependence” rather than stable personal traits. In “disembodied” tasks like lesson planning and documentation, teachers demonstrate high levels of self-efficacy and outcome expectations. Yet in core “embodied” teaching activities such as movement demonstration and error correction, their sense of efficacy rapidly collapses. This collapse stems from the discipline’s inherent “reprocessing” of self-efficacy. That is, physical education teachers’ assessment of AI extends beyond evaluating operational skills; it involves ontological scepticism about whether technology can truly address the core of physical education. From the SCT perspective, when teachers repeatedly experience success in desk-bound tasks, the “mastery experiences” emphasised by SCT do indeed strengthen their technical self-efficacy. However, in classroom segments involving tacit knowledge, experiential judgments, and physical presence, teachers have almost no demonstrable experiences to emulate and struggle to receive positive feedback. Self-efficacy remains persistently low due to the lack of both “successful experiences” and “vicarious experiences.” This aligns with domestic discourse on the “context-dependent” nature of teacher self-efficacy within specific instructional settings. From the perspective of SDT, this “scenario fragmentation” in self-efficacy directly shapes teachers’ sense of competence. SDT posits that whether an individual’s sense of competence, autonomy, and relational needs are met during an activity directly influences the quality and persistence of their motivation [ 45 ]. AI can enhance lesson preparation efficiency and boost teachers’ sense of competence and intrinsic motivation. However, in core subject teaching, teachers repeatedly experience AI as “unhelpful” and “adding to their workload,” leading to feelings of incompetence and helplessness. This indicates that understanding self-efficacy in embodied disciplines requires examining the “sense of competence” in tandem with the “discipline’s ontology,” rather than simply applying the relatively abstract, decontextualised concept of efficacy from the field of technology adoption. Environmental level: the “suspension” of subjective norms and structural constraints of resource stratification In SCT, environmental factors influence individual decision-making by providing resources and setting behavioural constraints, as well as by shaping subjective norms and social expectations. This study reveals that physical education teachers operate within a distinct “dual-layer structure”: the upper layer comprises macro-level policy directives, campus discourse, and the narrative of “educational modernization,” collectively exerting strong subjective normative pressure; the lower layer consists of micro-level constraints such as inadequate facilities, funding prioritization for elite teams, and the marginalization of regular physical education classes. This tension between “discourse abundance and resource scarcity” leads to a pronounced “regulatory failure” in the application of subjective norms in physical education contexts. This finding aligns with recent research on the “discourse-resource gap” in educational technology policies [ 46 ]. From an SCT perspective, if school training, equipment provision, and institutional arrangements can provide sufficient contextual support to physical education teachers, then the chain of “school emphasis—peer modelling—teacher adoption” should operate smoothly [ 47 ]. However, the resource stratification mechanism revealed in this study—prioritising AI equipment allocation to performance-oriented competitive teams while leaving public physical education classes in a prolonged “equipment vacuum”—effectively severed the middle segment of this chain. Even when teachers acknowledged the school’s overall direction in promoting AI, they acutely perceived the hard constraints of “no equipment, no training,” thereby dismissing these normative requirements as “irrelevant background noise.” From the perspective of SDT, this phenomenon’s impact on motivational quality can be further explained. Policy directives and assessment metrics from higher levels are essentially external demands. Without supporting resources and opportunities for participation, they resemble “controlling norms” that readily induce teachers to engage in external regulation driven by obedience and pressure, potentially even triggering psychological resistance [ 48 ]. The long-term allocation of resources toward competitive teams has fostered a sense of “structural injustice” among teachers: ordinary physical education instructors witness the efficient application of AI in elite training settings while remaining trapped in a dilemma of “wanting to use it but lacking the means,” eroding their fundamental sense of fairness and belonging. From an SDT perspective, this simultaneous squeeze on autonomy, competence, and relatedness significantly diminishes teachers’ intrinsic motivation to adopt technology. It leads them to prioritise fulfilling “the required minimum performance” with minimal effort and risk, rather than proactively seeking meaningful integration pathways [ 49 ]. Behavioural level: the dual game of performance adoption, professional identity, and outcome expectations At the behavioural level, SCT emphasises the predictive role of outcome expectancy in behavioural choice [ 50 ]. This study reveals that physical education teachers’ decisions regarding AI adoption are not based solely on weighing whether teaching becomes more effective. Instead, they navigate a dual-outcome expectation: on one hand, “instrumental gains”—such as enhanced lesson preparation efficiency, optimised data analysis, and improved teaching presentation effects; on the other hand, “professional risks”—including potential teacher marginalisation by technology, diminished professional skills, and diluted emotional connections between teachers and students. This dual expectation of “benefits versus risks” is also evident in studies on teacher technology adoption and disciplinary identity: on one hand, technology integration is viewed as a crucial pathway to boost employability and improve teaching quality; on the other hand, teachers worry that their roles may be replaced by “technology platforms.” This prolonged tug-of-war between dual expectations ultimately gave rise to a compromise strategy: During open classes, demonstration lessons, or any setting requiring the presentation of a “modernised image,” teachers proactively incorporate AI to create a classroom landscape that projects “technological sophistication” and “cutting-edge innovation,” thereby addressing external evaluations and organisational expectations. In contrast, during routine instruction, teachers tend to confine AI to desk-based activities, maintaining a traditional classroom model centred on physical demonstrations and direct interaction. Similar “performative technology integration” has been discussed in studies of health and physical education teachers’ technology use, where educators showcase tech applications in high-visibility settings while maintaining low levels of integration in routine classes. On the surface, physical education teachers appear to be “using AI.” Still, in reality, this adoption is more of a “performative use” and “symbolic gesture” aimed at demonstrating to schools and society that “PE isn’t falling behind,” rather than a deep integration driven by intrinsic teaching needs. According to SDT, this behaviour pattern reflects a clear dominance of controlling motives: the primary drivers for teachers using AI stem from external evaluation systems (school assessments, societal perceptions) and avoidance of negative labels (“outdated,” “unambitious”), rather than an intrinsic recognition of the technology’s inherent value. Therefore, understanding outcome expectations in physical education must extend beyond explicit benefits, such as “teaching effectiveness” or “work efficiency,” to include implicit consequences, such as “professional identity preservation,” “quality of teacher-student relationships,” and “educational humanistic values.” Otherwise, the outcome expectation construct in SCT risks being monopolised by “instrumental rationality,” overlooking the profound self-debate unfolding behind physical education teachers’ technology adoption: “Who am I?” and “What is physical education?” Conclusion Through SCT and SDT, this study provides a more contextually grounded explanatory framework for physical education teachers’ AI acceptance mechanisms. On one hand, SCT’s triadic interaction model retains explanatory power in physical education contexts: self-efficacy determines teachers’ confidence and willingness to attempt in different situations; subjective norms shape the “ideal picture” of technology use through policy discourse and peer modelling; and outcome expectancy guides teachers to weigh tool benefits against professional risks. On the other hand, only when these constructs are situated within the context of physical education’s embodied ontology, hierarchical resource structure, and defense of professional identity—supplemented by SDT’s analysis of motivational quality—can we truly understand why physical education teachers exhibit complex behavioral patterns such as “high desk-bound frequency—low classroom frequency,” “deep application in competitive sports—shallow application in general physical education,” and “proactive in demonstrations—conservative in daily practice.” At the practical level, this integrated analysis implies that promoting the effective integration of AI in physical education cannot remain confined to a linear approach of “technology introduction” or “teacher training.” Instead, it must begin with “contextual reconstruction”: First, systematically rebuild physical education teachers’ technological self-efficacy within authentic classroom settings—not merely teaching them “how to use software,” but providing replicable success experiences in specific projects (such as fitness testing, motion analysis, and rehabilitation training) to cultivate a sense of competence through embodied teaching. Second, bridge the resource divide between competitive teams and public physical education classes. Through equitable equipment allocation and training programs, ensure that subjective norms are underpinned by tangible hardware support, thereby transforming them into actionable protocols. Third, within policy and evaluation frameworks, we must recognise the humanistic value of physical education and the irreplaceable role of physical training. AI should be positioned as a tool to “enhance teacher agency” rather than a yardstick for measuring “who is more modern.” This approach preserves sufficient professional autonomy for educators, supporting their transition from external compliance to internal integration. Overall, in the field of physical education where embodiment is highly emphasised, the key to truly “accepting” AI technology lies not in how advanced the technology itself is, but in whether it can establish a relationship with teachers that is not premised on replacement but rather aimed at empowerment—one that bridges their self-efficacy, basic psychological needs, and professional identity. This also provides crucial empirical foundations and theoretical clues for constructing a specialised theoretical model of “physical education-technology adoption” in the future. The theoretical contribution of this study lies in integrating SCT with SDT, thereby addressing the explanatory limitations of traditional Technology Acceptance Models within the specific context of physical education instruction. Grounded in the embodied ontology, hierarchical resource structure, and professional identity defence of the physical education discipline, the research constructs a more discipline-specific contextual explanatory framework. This provides a logical starting point for establishing a dedicated ‘Theory of Technology Adoption in Physical Education’ in the future. The study not only uncovers the “psychological needs-self-efficacy-professional identity” coupling mechanism underlying physical education teachers’ “complex behavioural combinations,” but also clarifies that AI technology should be positioned as an enabling tool to “enhance teacher agency” in physical education. Furthermore, through systematic analysis of teachers’ adoption process from external compliance to internal integration, it lays crucial theoretical foundations and empirical clues for constructing a specialised “physical education-technology adoption” theoretical model. However, as this study is based on a limited sample size, primarily drawn from qualitative interviews with physical education teachers at universities in the same region, the conclusions are more applicable to interpreting similar contexts. Future research should conduct large-scale sample surveys across different regions and educational levels, employing multiple methodologies, such as questionnaires and experiments, to validate the mechanism hypotheses proposed in this study. Additionally, subsequent research could incorporate multiple perspectives, including those of students and administrators, to comprehensively deepen the systematic understanding of digital transformation in physical education. Acknowledgments Not applicable. Authors’ contributions Haidong Chen and Weilei Yang conceptualised the study, designed the methodology, and drafted the manuscript. Feixue Rao collected and analysed the data and contributed to the discussion section. All authors have read and agreed to the published version of the manuscript. Funding Not applicable. Data availability No datasets were generated or analysed during the current study. Declarations Ethics approval and consent to participate This study involving human participants adheres to the Declaration of Helsinki . In accordance with Article 32 of the Measures for Ethical Review of Life Science and Medical Research Involving Human Beings (deliberated and approved by the National Science and Technology Ethics Committee of China, which serves as the institutional review board [IRB] overseeing such research in China), research using anonymized data, posing no harm, and involving no sensitive personal information is exempt from ethical review requirements. This study meets these exemption criteria: it utilised anonymised interview records with data strictly limited to academic analysis. All participants involved in the study have provided informed consent for their data to be included. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References 1. Jiang Y, Li X, Luo H, Yin S, Kaynak O. Quo vadis artificial intelligence? Discover Artif Intell. 2022;2(1):4. [ Google Scholar ] 2. Holzinger A, Keiblinger K, Holub P, Zatloukal K, Müller H. AI for life: Trends in artificial intelligence for biotechnology. New Biotechnol. 2023;74:16–24. [ DOI ] [ PubMed ] [ Google Scholar ] 3. Nedungadi P, Tang KY, Raman R. The transformative power of generative artificial intelligence for achieving the Sustainable Development Goal of quality education. Sustainability. 2024;16(22):9779. [ Google Scholar ] 4. Liu Z, Wang X. Action recognition for sports combined training based on wearable sensor technology and SVM prediction. Prev Med. 2023;173:107582. [ DOI ] [ PubMed ] [ Google Scholar ] 5. Ma C, Shou M. Sports competition assistant system based on fuzzy big data and health exercise recognition algorithm. Mob Inform Syst. 2021;2021(1):7687178. [ Google Scholar ] 6. Abulibdeh A, Zaidan E, Abulibdeh R. Navigating the confluence of artificial intelligence and education for sustainable development in the era of industry 4.0: Challenges, opportunities, and ethical dimensions. J Clean Prod. 2024;437:140527. 7. KK K. Revolutionizing Physical Education: The Role of Artificial Intelligence in Enhancing Learning and Performance. Volume 44. Library of Progress-Library Science, Information Technology & Computer; 2024. 3. 8. Zhou T, Wu X, Wang Y, Wang Y, Zhang S. Application of artificial intelligence in physical education: a systematic review. Educ Inform Technol. 2024;29(7):8203–20. [ Google Scholar ] 9. Li Z, Wang H. The effectiveness of physical education teaching in college based on Artificial intelligence methods. J Intell Fuzzy Syst. 2021;40(2):3301–11. [ Google Scholar ] 10. Xu M, Liu D, Zhang Y. Design of interactive teaching system of physical training based on artificial intelligence. J Inform Knowl Manage. 2022;21(Supp02):2240021. [ Google Scholar ] 11. Zhang J. Reform and innovation of artificial intelligence technology for information service in university physical education. J Intell Fuzzy Syst. 2021;40(2):3325–35. [ Google Scholar ] 12. Kuleva M. (2024, June). Exploring the integration of virtual reality in physical education: A comprehensive review. In Environment. Technologies. Resources. Proceedings of the International Scientific and Practical Conference (Vol. 2, pp. 197–201). 13. Li G. (2021, January). Research on AI Technology and the New Era of Physical Education. In 2021 International Conference on Information Technology and Contemporary Sports (TCS) (pp. 619–623). IEEE. 14. Trendowski T. A Checklist for Using Generative Artificial Intelligence to Create Lesson Plans in K–12 Physical Education. J Phys Educ Recreation Dance. 2025;96(2):15–27. [ Google Scholar ] 15. Ahn D, Lim H. (2025, April). Exploring K-12 Physical Education Teachers’ Perspectives on Opportunities and Challenges of AI Integration through Ideation Workshops. In Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems (pp. 1–16). 16. Polak S, Schiavo G, Zancanaro M. (2022, April). Teachers’ perspective on artificial intelligence education: An initial investigation. In CHI Conference on Human Factors in Computing Systems Extended Abstracts (pp. 1–7). 17. Wardat Y, Tashtoush M, AlAli R, Saleh S. Artificial intelligence in education: mathematics teachers’ perspectives, practices and challenges. Iraqi J Comput Sci Math. 2024;5(1):60–77. [ Google Scholar ] 18. Wang Y, Wang X. Artificial intelligence in physical education: comprehensive review and future teacher training strategies. Front Public Health. 2024;12:1484848. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 19. Lee HS, Lee J. Applying artificial intelligence in physical education and future perspectives. Sustainability. 2021;13(1):351. [ Google Scholar ] 20. Killian CM, Marttinen R, Howley D, Sargent J, Jones EM. Knock, Knock… Who’s there? ChatGPT and artificial intelligence-powered large language models: reflections on potential impacts within health and physical education teacher education. J Teach Phys Educ. 2023;42(3):385–9. [ Google Scholar ] 21. Karimi J, Karimi M, Mahmoudi S. Teachers’ Attitudes Toward Modern Physical Education Utilizing the Transformative Potentials of Artificial Intelligence. Res Educational Sport. 2025;12(37):87–110. [ Google Scholar ] 22. Wallace J, Scanlon D, Calderón A. Digital technology and teacher digital competency in physical education: a holistic view of teacher and student perspectives. Curriculum Stud Health Phys Educ. 2023;14(3):271–87. [ Google Scholar ] 23. Duan R, Ma LZ, Wang ST. The professional development of high-quality physical education teachers: The contemporary nature, problem areas, and footholds. J Phys Educ. 2024;31(02):74–81. 10.16237/j.cnki.cn44-1404/g8.2024.02.003. [ Google Scholar ] 24. Cao F, Lei M, Lin S, Xiang M. Application of Artificial Intelligence-Based Big Data AI Technology in Physical Education Reform. Mob Inform Syst. 2022;20221:4017151. [ Google Scholar ] 25. Jamal A. The role of artificial intelligence (AI) in teacher education: Opportunities & challenges. Int J Res Anal Reviews. 2023;10(1):139–46. [ Google Scholar ] 26. Almusawi HA, Durugbo CM, Bugawa AM. Innovation in physical education: Teachers’ perspectives on readiness for wearable technology integration. Comput Educ. 2021;167:104185. [ Google Scholar ] 27. Bandura A. Human agency in social cognitive theory. Am Psychol. 1989;44(9):1175–84. 10.1037/0003-066X.44.9.1175. [ DOI ] [ PubMed ] [ Google Scholar ] 28. Bandura A. Perceived Self-Efficacy in Cognitive Development and Functioning. Educational Psychol. 1993;28(2):117–48. 10.1207/s15326985ep2802_3. [ Google Scholar ] 29. Chahal J. Effect of computer and internet self-efficacy on students’ acceptance of e-learning: analysis of tam model using pls sem approach. 2022 Int Conf Comput Communication Intell Syst (ICCCIS). 2022;257:262. [ Google Scholar ] 30. Zhang X, Wang Y, Leung SO. Technology Acceptance Model (TAM) and sports bracelets usage in physical education for freshmen: the role of gender and self-efficacy. Technol Pedagogy Educ. 2023;32(1):45–63. [ Google Scholar ] 31. Chahal J, Dagar V, Dagher L, Rao A, Udemba EN. The crisis effect in TPB as a moderator for post-pandemic entrepreneurial intentions among higher education students: PLS-SEM and ANN approach. Int J Manage Educ. 2023;21(3):100878. [ Google Scholar ] 32. Wu D, Gu H, Gu S, You H. Individual motivation and social influence: a study of telemedicine adoption in China based on social cognitive theory. Health Policy Technol. 2021;10(3):100525. [ Google Scholar ] 33. Deci EL, Olafsen AH, Ryan RM. Self-determination theory in work organizations: The state of a science. Annual Rev organizational Psychol organizational Behav. 2017;4:19–43. [ Google Scholar ] 34. Deci EL, Ryan RM. Self-determination theory. Handb Theor social Psychol. 2012;1(20):416–36. [ Google Scholar ] 35. Fradkin-Hayslip A. Teacher Autonomy, Motivation, and Job Satisfaction: Perceptions of Elementary School Teachers According to Self-Determination Theory. Ilkogretim online. 2021:20(2):198–205. 36. Fereday J, Muir-Cochrane E. Demonstrating rigor using thematic analysis: A hybrid approach of inductive and deductive coding and theme development. Int J qualitative methods. 2006;5(1):80–92. [ Google Scholar ] 37. Braun V, Clarke V. Using thematic analysis in psychology. Qualitative Res Psychol. 2006;3(2):77–101. [ Google Scholar ] 38. Bandura A. Social foundations of thought and action: A social cognitive theory. Prentice-Hall; 1986. 39. Bazeley P, Jackson K. Qualitative data analysis with NVivo. SAGE; 2013. 40. Lincoln YS, Guba EG. (1985). Naturalistic inquiry. sage. 41. Tong A, Sainsbury P, Craig J. Consolidated criteria for reporting qualitative research (COREQ): a 32-item checklist for interviews and focus groups. Int J Qual Health Care. 2007;19(6):349–57. [ DOI ] [ PubMed ] [ Google Scholar ] 42. Larsson H, Nyberg G. It doesn’t matter how they move really, as long as they move.’Physical education teachers on developing their students’ movement capabilities. Phys Educ Sport Pedagogy. 2017;22(2):137–49. [ Google Scholar ] 43. López-Angulo Y, Sáez-Delgado F, Mella-Norambuena J, Bernardo AB, Díaz-Mujica A. Predictive model of the dropout intention of Chilean university students. Front Psychol. 2023;13:893894. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 44. Anderson ES, Winett RA, Wojcik JR. Self-regulation, self-efficacy, outcome expectations, and social support: social cognitive theory and nutrition behavior. Ann Behav Med. 2007;34(3):304–12. [ DOI ] [ PubMed ] [ Google Scholar ] 45. Ryan RM, Deci EL. Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. Am Psychol. 2000;55(1):68. [ DOI ] [ PubMed ] [ Google Scholar ] 46. Zhang K. Teacher adoption of digital education management systems through combined information systems and social cognitive frameworks during the post-COVID era. Sci Rep. 2025;15(1):16810. 10.1038/s41598-025-01552-8. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 47. Rogers EM, Vaughan PW, Swalehe RM, Rao N, Svenkerud P, Suruchi S. Effects of an entertainment-education radio soap opera on family planning behaviour in Tanzania. Stud Fam Plann. 1999;30:193–211. 10.1111/j.1728-4465.1999.00193.x. [ DOI ] [ PubMed ] [ Google Scholar ] 48. Yang SL, Wu HH. The Application of Self-Determination Theory in Higher Education: Intrinsic Motivation and Academic Achievement. Adv Psychol. 2024;14(5):289–96. 10.12677/ap.2024.145314. [ Google Scholar ] 49. Hu XQ, Tang Y, Huang X, Zhang JL, Fan HY. Research Progress on the Effect of Physical Education Learning from the Perspective of Self-determination Motivation Theory. China Sport Sci. 2020;0467–77. 10.16469/j.css.202004008. 50. Bandura A. Social cognitive theory: An agentic perspective. Ann Rev Psychol. 2001;52(1):1–26. [ DOI ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Data Availability Statement No datasets were generated or analysed during the current study. Articles from BMC Psychology are provided here courtesy of BMC ACTIONS View on publisher site PDF (1.1 MB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top