The Impact of AI-integrated Sport Blended Learning on Primary School Students' Sports Skills and Attitudes

2.1.1. Artificial Intelligence

Artificial Intelligence (AI) is broadly defined as the simulation of human intelligence by a machine, enabling it to perform tasks such as learning, reasoning, and decision-making [25, 26]. In Primary school PE, AI has shown significant benefits, including increasing student engagement through gamification and contextualization, providing personalized feedback for skill development, and creating an immersive PE learning environment that meets individual needs [27]. For example, wearables, virtual technology, exercise apps, and AI-powered apps enable teachers to monitor students' physical activity in real-time, providing data-based adjustments that can improve cognition and body composition [15, 28]. The process of technology development in PE typically consists of several phases: a needs analysis to determine student and instructor requirements, an algorithm design to create a personalized learning model, integration of the tool into the curriculum, and an iterative evaluation based on feedback [12]. These steps ensure that AI technology can be effectively embedded in educational environments to maximize its transformative potential in Primary school PE learning [29, 30]. Similar definitions emphasize that the approach creates a more engaging and personalized learning experience by incorporating virtual tools such as wearables, gamified apps, and interactive feedback systems into actual physical activity [31]. In Primary school PE, blended learning in SBL offers a number of advantages, including increased student engagement, enhanced skill acquisition, and meeting diverse learning needs. For example, accurate monitoring of exercise loads through digital platforms and immersive gamification design can enhance students' motivation to learn. At the same time, AI-based feedback systems can provide personalized instruction to improve motor skills [32]. The process of developing blended learning in SBL typically involves several stages: identifying educational goals, integrating digital tools into the curriculum, designing interactive activities, and continually evaluating the program's effectiveness and engagement [33]. These steps ensure the effective implementation of blended learning in SBL, providing both cognitive and physical benefits to Primary school students [34].

2.1.2. Sport Skills

Sports skills are the abilities acquired by an individual through learning to perform physical activities effectively and efficiently, and are typically divided into basic sports skills and specialized skills required for specific sports (e.g., dribbling and passing in soccer) [35]. Related definitions emphasize the need for sport skills to integrate physical, cognitive, and behavioral components for success in sport and physical activity [36]. Developing sports skills in Primary school education has many benefits, including improving physical health, enhancing cognitive functioning, and promoting social skills such as teamwork and communication [37]. In addition, sports skills help increase students' self-confidence and self-efficacy, encouraging lifelong participation in sports [38]. Combined with technological tools, such as AI-based feedback systems and interactive platforms, this process can be further enhanced by providing personalized feedback, tracking progress, and increasing student engagement [39]. By integrating these steps, educators can ensure that students effectively acquire and apply sports skills for their overall development.

2.1.3. Sports Attitude

The sports learning attitude scale is composed of four aspects: (1) students' sports attitude, (2) sports learning purpose, (3) sports learning time, and (4) sports learning evaluation. The cognitive dimension encompasses beliefs and perceptions about the benefits of physical activity; the affective component involves preferences or aversions to physical activity, and the behavioral dimension reflects willingness to participate in or resistance to physical activity [40]. Yavuz (2019) stated that attitudes toward sport are key factors influencing sport participation and performance, and are influenced by a combination of personal experiences, social influences, and learning environments. For Primary school students, developing positive attitudes toward sport has many benefits, including increased participation, improved physical fitness, enhanced social interaction skills, and enhanced academic performance [41]. Positive attitudes toward sports also help students form lifelong habits of physical activity, which promotes overall health [40]. The development of attitudes toward sports typically involves several stages: exposure to fun and inclusive physical activity, reinforcement through positive experiences, and the development of intrinsic motivation through personalized and meaningful learning experiences [42]. The use of AI in PE further supports this process by providing personalized feedback, gamification elements, and adaptive learning environments to enhance student engagement and foster positive attitudes [43]. Through these steps, educators can effectively develop positive attitudes toward PE among Primary school students, thereby encouraging sustained physical activity participation.

2.1.4. Other Relevant Research

AI-integrated SBL can enhance students' sport skills and attitudes toward sport. The intervention first unfolds through a personalized feedback mechanism that uses AI-powered wearables and motion-tracking systems to analyze student movements and provide targeted suggestions for improvement [44]. Next, gamification using AI-based platforms creates engaging learning environments with points, rewards, and competitive elements that incentivize engagement and skill acquisition [45]. Collaborative activities are then introduced, and the AI system facilitates teamwork, peer learning, and cloud-based competitions by assigning roles and tracking team dynamics [46]. The SBL study demonstrated that these interventions significantly enhanced sport skills, cognitive understanding, and social interactions, while fostering positive attitudes toward PE [47]. In addition, AI integration enables continuous progress tracking to ensure a personalized and effective learning experience, as well as an after-school exercise platform that provides exercise prescriptions tailored to students' individual needs [48]. By focusing on both the physical and emotional domains, these interventions create a holistic learning environment that maximizes student development and engagement in PE.

3.1.1. Exercise Load Detection System (ELMS)

ELMS is a sports load monitoring system developed by Shanghai Xianshi Education Technology Co., based on the National PE and Health Curriculum Standard 2022. The system is divided into two parts: hardware and core functions. The hardware consists of a student watch charger (50 pieces) and a teacher box. There are six core functions (Fig. 1).

In the classroom (Figs. 2 and 3), the teacher and students use the ELMS, and the students wear load watches on their wrists, which do not interfere with the PE movement. During class, the sports watches worn by the students are monitoring their physical exertion levels. Through the color blocks on the large screen, both students and teachers can see at any time the individual's physical exertion status. The primary purpose is to monitor students' heart rates in real-time during exercise. Taking a 9-year-old student as an example, the blue color block represents a low-intensity exercise state. At this time, the student's physical exertion is relatively small, and the heart rate is below 106 b/m; the green color block corresponds to a moderate-intensity exercise, and the student can better withstand the current exercise load, with the heart rate ranging from 106 to 148 b/m; the yellow color block indicates a moderately high-intensity exercise, and the student's body can receive good exercise, but it has not reached the limit state, with the heart rate ranging from 148 - 180 b/m; the red color block represents a high-intensity exercise. This suggests that teachers and students should monitor exercise intensity to prevent excessive fatigue or injury, especially when the heart rate exceeds 180 beats per minute. Students can view their exercise data on a large screen, which fosters self-motivation and enables them to track their progress. Teachers can use ELMS to obtain real-time data on students' exercise intensity, heart rate, pace, and other metrics. Before the lesson (Fig. 4), teachers use the exercise data collected by ELMS to review and analyze the exercise load in the classroom. The data from Fig. (4) reveals several key metrics. The classroom load achievement rate is 28.2%. The average heart rate is 117 beats per minute, which is far below the standard (140-160 beats per minute). Although the maximum heart rate is 146 beats per minute, it does not exceed the standard. The pass rate is 100%, and the duration of heart rate ≥ 120 beats per minute is 17′35″. Regarding exercise density, the group exercise density is 78%, the practice density is 58%, and the effective exercise density is 42%. All of these meet the standards, indicating that most students are qualified. However, despite the average calories (146) and the average heart rate index (156) being close to the standard, the number of qualified students is relatively small. The fluctuation of the heart rate curve of the class in the chart shows that the heart rate did not reach the ideal range for most of the time, the average heart rate in each stage was the highest but did not meet the standard either, and the distribution of exercise intensity was dominated by the 0 - 60% intensity range. Overall, the students' exercise load in this course is insufficient. Further adjustments to the teaching are needed to increase the intensity (This is a throwing distance course, and the students' heart rate is not easy to raise, which is the limitation of sports projects). Teachers also use the data (Fig. 5) provided by ELMS to discuss the relationship between exercise and health with their students.

3.1.2. Daily Jump Rope App (DJR)

The Daily Jump Rope App was used to support sports learning before and after class, while the control group continued with traditional teaching methods. The DJR is an intelligent sports platform (Fig. 6), based on big data and algorithmic models, utilizing intelligent technology to calibrate human skeletal points, with a portrait recognition function combined with motion capture technology. In addition, the platform features AI interactive technology, including sports homework assignments, cloud competitions, and cloud physical testing, as well as a variety of combination sports and physically engaging activities. Furthermore, the DJR has an anti-cheating system and intelligent checking and correction of homework. The platform's video supervision effectively ensures that students participate and complete several exercises on a regular basis. The platform's intelligent homework correction function automatically generates a learning report for each exercise and creates visual icons of the overall class data, allowing teachers to see, at a glance, the completion status of students' homework. The intelligent supervision and data feedback function can effectively reduce the burden on teachers and parents.

3.2.1. Traditional Face-to-face Sport Learning (TFSL) Models

The TFSL model (Fig. 7) is centered on being teacher-led, with student participation, and typically includes teacher preparation, classroom instruction and demonstration, student practice, and on-site instruction by the teacher [29]. Teachers help students master motor skills through direct instruction and demonstration, and students improve their abilities through practice in the classroom and receive feedback and assessment from the teacher [44]. The model emphasizes the effective use of classroom time, with student learning primarily achieved through face-to-face interactions, and minimal post-class review is involved [45]. The core feature of TFSL is that teachers play a leading role, directly controlling classroom content and student activities.

Teaching usually focuses on the transfer of skills and practical operation.

3.2.2. AI-integrated Sports Blended Learning (SBL) Model

SBL teaching model after the experiment, including the process of teachers' and students' movement, AI technology-enabled PE learning in the three periods of time: before, during, and after the lesson (Fig. 8).

Before the lesson (Fig. 9), the teacher shares resources, collects information on students' learning conditions, presents problems that need to be solved during the lesson, assigns a PE teaching design, sets the course load for the lesson, and checks whether the students are wearing their watches. The students watch micro-videos related to the content being taught and complete the pre-study sheet. In class, teachers adjust the intensity and density of exercise in the classroom according to the intuitive interface statistical indicators and charts, and accurately provide personalized guidance (real-time monitoring and tracking of students' heart rate, intensity, density, participation rate, etc., in the class); students: according to the use of the ELMS in the PE classroom, can see their own real-time high and low levels of intensity of exercise (especially important to note that, when the exercise load exceeds the maximum value, the movement bracelet and the exercise bracelet system are able to detect students’ intensity of exercise and automatically send out an alarm to remind the instructor and the students of the need to adjust the amount and duration of exercise).

After class (Fig. 10), the teacher processes and analyzes the data collected by the ELMS in the classroom (the system allows customized metrics to be used for various types of cross-sectional, longitudinal, and multidimensional analyses across classes, grades, school districts, and cities), conducts assessment of teaching effectiveness, researches teaching methods, identifies problems, and improves PE teaching, uses the DJR to post classroom-related exercise tasks for students to practice in a targeted manner after class. Students use the DJR to complete the exercise tasks issued by teachers. After completing these, they can then conduct independent exercise practice on the daily jump rope platform, based on their own exercise interests. DJR can provide real-time guidance and evaluation, which is superior to traditional face-to-face teaching. Additionally, there is no way to guide students' PE learning and practice outside of class. Such situational, personalized, precise, and efficient sports platforms can increase students' interest in sports, improve their attitudes towards sports, and enhance their sports skills.

4.3.1. Student Sport Skills Test

To ensure authority and validity, all the pre-and post-test sport skills tests refer to the test requirements and standards of the 2014 Revised Edition of the Physical Fitness Standard for Chinese Students. The tests were administered simultaneously to 94 students at the same location, in the same test group, with the same test organization, encompassing a total of four test items: vital capacity, Sit-up-and-bend, 50-meter run, and 1-minute Jump rope. The results of the test items, according to the standards, were assessed as “excellent,” “good,” “qualified,” or “failed” (Appendix B).

4.3.2. Student Sports Learning Attitude Scale

The sports learning attitude scale was adapted from a study conducted by Wang et al. and consists of four dimensions: (1) students' Sports Attitude (SA), (2) Sports Learning Evaluation (SLE), (3) Sports Learning Time (SLT), and (4) Sports Learning Outcome (SLO) [51]. The scale consisted of more than 30 items. The reliability of this scale was demonstrated by a high Cronbach's alpha value of 0.869. The validity of the scale was further verified by factor analysis. This scale was also subjected to the same validity and reliability tests as in the first phase of the study. After collecting data on students' acquisition of sports skills, the students in the experimental group were given a questionnaire and asked open-ended questions about their perceptions of the learning activity. This covered a wide range of aspects relating to the participants' perceptions of the learning activities they received.

6.1.1. Before Class

First, Pre-class send micro-video-enabled students to preview sports content (e.g., movement demonstrations, warm-up guidelines) via mobile apps, fostering engagement and autonomous learning through self-challenge opportunities [1, 15, 59]. Research has shown that multimedia learning, such as videos, can significantly enhance students' understanding of skills and concepts, thereby increasing the efficiency of actual practice [60]. In contrast, traditional teaching methods typically lack such pre-course preparation components, resulting in less efficient course execution [61].

6.1.2. In Class

Second, the in-class ELMS, as a core component of the SBL model of AI technology [26], is able to monitor students' physical loads in real-time, provide personalized feedback to students, and adjust the intensity of exercise according to their performance [60]. This system can avoid generalized feedback and make the learning process more refined, compared to traditional teaching methods [61, 62]. Research has shown that this real-time data-based feedback can improve learning outcomes by optimizing exercise intensity and duration [31, 58].

6.1.3. After Class

Finally, Post-class DJR implementation bridges classroom learning with extracurricular practice by recommending progress-aligned exercises and assigning targeted home workouts for skill consolidation [59, 63, 64]. Additionally, students can track their progress, receive personalized advice, and continue learning after school [62]. This fun and contextualized learning experience not only contributes to the continuous improvement of students' skills but also avoids the learning disconnect and lack of continuity that often occur at the end of class in TFSL instruction [65].

6.2.1. Vital Capacity

In terms of Vital capacity, the AI model adjusts the content of PE learning by monitoring students' exercise loads, allowing for real-time adjustments of exercise intensity during practice to enhance students' cardiorespiratory fitness [63]. It also collects data on exercise intensity over time to assign exercise prescriptions to each student. Research has shown that AI can optimize the training load and enhance students' physical health [31].

6.2.2. 50-meter Run

In the 50-meter sprint, the use of ELMS, combined with reviewing classroom videos, helped students monitor their exercise loads and adjust their running posture and stride, resulting in improved speed and agility [33]. Real-time performance tracking and feedback have been shown to improve students' sprinting performance [30].

6.2.3. Sit-up-and-bend

Students' flexibility also improved significantly during the Sit-up-and-bend test, indicating that, with the use of the DJR system under AI, and the monitoring of exercise data after class, the AI teacher was able to provide students with specific stretching exercises to help improve flexibility [59, 62].

6.2.4. 1-minute Jump Rope

In addition, in the 1-minute Jump rope test, the DJR system with AI can provide a fun, situational and immersive PE learning experience, provide instant guidance from the AI teacher and a rich library of exercise resources, drive students to actively learn and practice after class, buttress the classroom learning content, and consolidate learning and practice effects. AI intelligent technology keeps records, so that the movements used are collected, and real-time feedback can help students to improve their coordination and endurance [57, 60]. These results are consistent with previous research suggesting that personalized feedback and continuous practice play an important role in improving students' motor skills [1, 65].

In conclusion, the data suggest that the AI-integrated SBL instructional model significantly improves Primary school students' performance in various physical fitness tests, supporting the hypothesis that AI-driven personalized learning is more effective in enhancing students' sport skills than traditional teaching methods.

6.3.1. Improvement in Primary School Students' Interest in PE Learning Attitudes

AI-integrated SBL enhances primary students' PE engagement through personalized learning and real-time feedback across pre-/in-/post-class phases [67]. By tailoring activities to individual capabilities, AI fosters autonomous motivation, transforming passive task completion into enjoyable challenges that reinforce the value perception of activities [37, 47, 68]. This approach enhances both skill acquisition and curricular identity, shifting student attitudes from compulsory participation to proactive involvement with sustained interest in sports [11, 69]. This change demonstrates that AI not only has a positive impact on skill enhancement but also plays a crucial role in increasing students' interest in sports and their long-term willingness to participate. This suggests that the AI-integrated SBL model can effectively stimulate students' interest in PE learning and change the lack of student interest in traditional PE [40, 70].

6.3.2. Primary School Students' Perception of the Purpose of PE Learning Changed

The perception of primary school students regarding the purpose of PE learning has also changed, and AI-supported PE learning emphasizes the development of lifelong sport skills and sustainable development [12]. The AI platform has enabled PE learning to be accessible anytime and anywhere, making it more scientific, intelligent, and modern. Primary school students participate in PE learning intuitively and with interest, transforming the experience from the tedious TFSL sport learning method. There is an emphasis on attending to each student's individualized development, as well as on diversified evaluation [71]. This shift helps Primary school students to have a broader and more positive perception of PE learning [37].

6.3.3. Primary School Students became more Committed to PE Learning Time

The integration of AI-integrated SBL instructional models helped students better understand the relevance and value of PE, motivating them to invest more time in both the classroom and extracurricular activities [58]. AI instruments, such as the before- and after-school DJR apps, made it easier for students to monitor their own progress, which further facilitated sustained attention to physical activity [62]. This finding is consistent with existing research, which shows that technology can enhance student engagement and sustain their interest in physical activity [40]. The quality of time spent in PE learning increased significantly from passive to active physical activity.

6.3.4. Primary School Students' Perceptions of PE Learning Assessment changed for the Better

The AI-integrated SBL instructional model also changed Primary school students' perceptions of PE learning assessment. Through continuous feedback and progress tracking, students' attitudes toward assessment became more positive and less anxious [57]. The ability of AI to provide objective, real-time assessment helped students to focus on improvement, rather than just results, thus fostering a growth mindset in PE [17, 59]. This is consistent with research suggesting that technology-based feedback systems can reduce the stress and practitioner workload associated with traditional assessments and encourage students to focus on personal development [18, 19]. In summary, this study demonstrates the transformative potential of AI-integrated SBL instructional models in primary school physical education. Through its tripartite framework (pre-class micro-learning, in-class ELMS monitoring, and post-class DJR reinforcement), the SBL model outperforms traditional TFSL methods in three critical dimensions: