Objectives. This review examines how artificial intelligence (AI) can support university physical education (PE) pedagogy and curriculum development while preserving motor learning, meaningful movement time, teacher judgment, safety, and equity, and proposes a context-sensitive framework for Vietnamese higher education. Materials and Methods. A structured narrative review and conceptual synthesis were conducted using peer-reviewed research and authoritative policy and guidance sources identified through PubMed and targeted scholarly searches of journal and publisher platforms. The search was updated on 23 September 2026. Priority was given to systematic and scoping reviews, recent empirical studies, and PE-focused sources addressing motor learning, movement analysis, feedback, assessment, wearables, teacher competence, ethics, and governance. Thirty-four sources were retained: 29 peer-reviewed journal articles and reviews and five policy or guidance documents. Coding and theory-driven consolidation were used to map recurring findings to framework elements, with greater interpretive weight given to reviews and authoritative guidance. Results. Five application domains were identified: curriculum needs analysis and learning design, personalized motor-learning support, movement recognition and feedback, multimodal movement and participation monitoring, and teacher and program support. Educational value depended on context-specific validity, teacher competence, data governance, equitable access, and preservation of active learning time. The proposed AI-PE framework links six components—learning outcomes, learner data, learning design, AI support, multi-source assessment, and curriculum improvement—under three safeguards: human decision-making, data safety and ethics, and equitable access. Conclusions. AI should function as a pedagogical and decision-support layer rather than an autonomous instructor or assessor. The framework places embodied learning, motor competence, physical activity, and teacher responsibility at the center and provides a basis for institutional pilots and empirical validation in Vietnamese universities.
Background. The literature on school outdoor environments spans research on physical activity, health, and education. Adventure education and the concepts of competence and physical literacy are subthemes that need to be mapped without treating them as requirements for all documents. Objectives. To map publication output, citations, contributors, the co-occurrence structure of author keywords, and differences in keyword representation across periods in the literature retrieved using a search strategy on school environments and movement activities. Materials and Methods. The study employed a descriptive bibliometric design. The Scopus export dated September 6, 2026, contained 289 English-language records published between 2017 and 2026. Two records labeled Retracted, two Erratum records, and one record with a duplicate DOI were excluded, yielding an analytical corpus of 284 documents. The main network used normalized author keywords, full counting, and a threshold of five documents. The temporal comparison used document counts and proportions for 2017–2021 and 2022–2026. Results. The documents were distributed across 182 sources, with a mean of 15.01 citations per document. Publication output peaked in 2022 and 2025, with 38 documents each; the 2019 publication cohort had the highest mean citation count, at 33.64. The United States had the greatest representation based on affiliation countries. The author keyword network comprised 33 nodes and 187 links. Physical activity was recorded in 78 documents (27.46%). The proportion of documents using the keyword physical education increased from 1.69% to 7.83%, outdoor education from 1.69% to 4.22%, and natural environment from 5.08% to 8.43%. Motor competence and physical literacy were recorded in three and two documents, respectively. Conclusions. The retrieved corpus represents physical activity more frequently than pedagogical and competence-related keywords. Several pedagogical terms had higher proportions in the second period. These findings describe terminological representation within the corpus rather than intervention effectiveness, the origins of themes, or the absence of relevant research.
Background. Full factorial experiments provide opportunities not only to estimate the effects and interactions of controlled factors but also to construct mathematical models, optimize pedagogical interventions, and support decision-making. However, the extent to which these methodological functions are implemented in physical education and sport research remains insufficiently established. Objectives. To determine the functional evolution of full and other multifactorial experimental designs in physical education and sport, distinguishing their use for mathematical modeling, optimization, state recognition, programming, control, and management support. Methods. A systematic methodological review with evidence mapping was conducted using Scopus publications from 2000 to 2026. The specialized corpus was expanded from 244 to 291 publications. Twenty-five full-text primary empirical publications were subjected to functional analysis at three levels: publication, experimental dataset, and research trajectory. The analysis examined factorial design structures, regression and discriminant models, optimization procedures, programming, pedagogical control, and management support. Potential reuse of experimental datasets was assessed to distinguish publications from independent experimental units. Results. The confirmed core of publications demonstrating constructive use of experimental design increased from 11 to 29. The 25 primary publications represented approximately 18-21 independent experimental units. Factorial designs were confirmed in 18 publications (72%), regression models in 10 (40%), discriminant or classification models in 14 (56%), and optimization in 14 (56%). Programming was demonstrated in four publications (16%), pedagogical control in 10 (40%), and management support in four (16%). No study demonstrated experimentally validated closed-loop adaptive management. Two complementary analytical trajectories were identified: regression modeling for estimating factor effects and optimizing intervention parameters, and discriminant modeling for recognizing experimentally formed states, modes, and training effects. The principal functional discontinuity occurred between local optimization and temporally organized programming. Conclusions. The full factorial experiment functions as a methodological link between controlled experimentation, mathematical modeling, and model-based pedagogical decision-making. Its constructive potential is realized most consistently through models of influence, state recognition, and optimization, whereas programming and management support remain less developed. The distinction between publications, independent datasets, and research trajectories is essential for evaluating the evidence base. Further methodological development requires prospective validation of adaptive management systems integrating state assessment, pedagogical decisions, intervention, reassessment, and subsequent adjustment.
Background. Experimental design is the principal methodological framework for establishing causal relationships in sports science. Traditional classifications describe experimental designs according to structural characteristics, such as the number of factors, randomization procedures, measurement organization, or statistical methods. However, these approaches do not explain how the methodological functions of experimental design evolve from hypothesis testing to mathematical modelling, programming, and model-based decision making. Objectives. To investigate the functional evolution of experimental design in sports science, determine the methodological role of full factorial experiments (2^k), and develop a functional concept of experimental design based on bibliometric and systematic methodological analyses. Material and Methods. The study was conducted in two stages. First, a bibliometric analysis of publications indexed in the Scopus database (2000–2026) was performed to identify the structure of research on experimental designs and to construct an analytical corpus. After duplicate removal, the bibliometric corpus comprised 1653 publications. Second, a systematic methodological analysis of a specialized corpus of 244 publications was carried out to evaluate the realization of the cognitive and constructive potential of experimental designs. Particular attention was paid to studies integrating full factorial experiments with mathematical modelling, regression analysis, and discriminant analysis. Results. The bibliometric analysis demonstrated that contemporary sports science is dominated by subject-oriented terminology, whereas methodological concepts related to experimental design are rarely represented in author keywords or publication titles. The systematic methodological analysis showed that experimental designs with similar structural characteristics perform substantially different methodological functions. A functional concept of experimental design was developed, according to which its methodological significance is determined by the level of realization of its cognitive and constructive potential. Full factorial experiments were identified as the methodological link connecting experimental analysis with mathematical modelling. Three levels of realization of the constructive potential of experimental design were identified: local model-based control, programming of the pedagogical process, and system-level management. Conclusions. The study proposes a functional concept of experimental design that complements traditional structural classifications. The methodological contribution of the study consists in introducing a new principle for classifying experimental designs according to the level of realization of their cognitive and constructive potential. Full factorial experiments occupy a central methodological position in this framework by enabling the transition from hypothesis testing to mathematical modelling, programming, and model-based management of pedagogical and training processes.
Background. Small-sided games (SSGs) are widely used in soccer training to simultaneously develop physical and physiological performance. However, transition-specific SSG formats combined with work-to-rest ratio manipulation may further influence training load, although their interactive effects remain unclear. Objectives. This study aimed to examine the effects of transition small-sided game format and work-to-rest ratio on internal and external training load in university soccer players. Materials and Methods. Eighteen male university soccer players participated in this study. Two transition game formats (3v3v3+2 and 3v3v3+2 Change) and four work-to-rest ratios (4:4, 4:2, 2:2, and 2:1) were applied. Internal load was assessed using heart rate responses, and external load was measured using Global Navigation Satellite System (GNSS) devices. A two-way repeated-measures ANOVA was used to analyze the data. Results. Both the work-to-rest ratio and the playing format significantly affected internal load, with higher mean heart rate and greater time spent above 80% of HR under shorter recovery (4:2), as well as higher mean heart rate under the No-Change format overall. External load variables showed significant main and interaction effects, with the 2:2 ratio producing the highest total distance, acceleration, and deceleration, while and the 3v3v3+2 Change format combined with the 2:2 ratio eliciting the greatest high-intensity running and peak speed. Maximum heart rate indicated a significant Format × Work-to-Rest interaction, although no individual condition-pair comparison remained significant after Bonferroni correction. Conclusions. The findings suggest that combining transition game format with an appropriate work-to-rest ratio can be used as an effective strategy for manipulating training loads in soccer players.
