Adolescents in China are divided into two categories: junior high school (Grades 7–9) and senior high school (Grades 10–12). Students in grades seven through 12 were the focus of this investigation. The sample size was determined by applying the rule of 10–15 cases per predictor, with 20 predictors necessitating 200–300 participants. A hybrid sampling strategy that integrates convenience sampling and random sampling was implemented: 10 junior high schools and six senior high schools in Chengdu and Chongqing were selected via convenience sampling. As key drivers of Western China's growth and the national Chengdu-Chongqing Economic Circle, Chengdu and Chongqing exhibit regionally representative school physical education development. Their distinct geographical and cultural characteristics provide an ideal setting for examining how teaching styles relationship student engagement in physical education, while their proximity enables diverse sampling. Random sampling: Two to three classes per grade were randomly selected within each school, and all students in these classes were invited to participate.

Between March and May 2025, Data were collected offline using Questionnaire Star, a Chinese online survey platform. Standardized data collection protocols were implemented to guarantee reliability and validity: (1) the study's purpose, voluntary participation, data utilization scope, and response guidelines were explained to students by researchers during the pre-survey briefing in order to reduce bias and increase engagement. (2) Survey administration: In order to guarantee independent and focused responses, questionnaires were completed in computer labs during physical education courses, with the guidance of a teacher and researcher. (3) Ethical compliance: The study was conducted in accordance with the institutional ethical guidelines and the Declaration of Helsinki. The Ethics Committee of the Shaanxi Vocational and Technical College endorsed the protocol. Informed assent was obtained from minors along with written parental consent: Student assent: Verbal consent from students. Parental consent: Written confirmation from parents/guardians through class-specific communication groups. (4) Data anonymity: No personally identifiable information was gathered. Throughout the storage and analysis processes, all data were anonymized and encrypted. (5) Data processing: To ensure the breadth and representativeness of the sample, a total of 1,857 questionnaires were distributed. During data collation, missing values and invalid answer were deleted from the sample questionnaires (Figure 1), resulting in a final valid sample of 1,629 (87.7% validity rate).

All measurement instruments utilized in this investigation were validated scales with established reliability and validity, which encompassed psychological resilience, classroom satisfaction, learning engagement, and teaching styles. Additionally, demographic queries that were self-designed were implemented to gather data regarding the age, gender, grade, height, and weight of adolescents.

This study employed IBM SPSS Statistics 26.0 for statistical analysis and data processing. In the descriptive statistics, categorical variables were presented as frequencies and percentages, while continuous variables were reported as means and standard deviations. First, the normality of the main variables was confirmed through the Kolmogorov-Smirnov (K-S) test, Q-Q plots, and P-P plots, which supported the use of independent samples t-tests for group comparisons the figures for normality tests are shown in Supplementary material 1. Second, Harman's Single Factor Test was conducted to assess common method bias and evaluate potential systematic errors in the data. Third, Pearson correlation analysis was performed to examine the relationships among psychological resilience, classroom satisfaction, sports learning engagement, and teaching styles. Fourth, multiple linear regression analysis was used to test the predictive association of physical education teaching styles on learning engagement after controlling for confounding variables. During this process, the Variance Inflation Factor (VIF) was calculated to diagnose multicollinearity among independent variables, with thresholds set as follows: VIF = 1 indicated no multicollinearity, 1 < VIF < 5 suggested mild multicollinearity, and VIF ≥ 5 indicated severe multicollinearity. Finally, a parallel mediation model was constructed using the SPSS PROCESS Macro (v4.1) to test the indirect associations of psychological resilience and classroom satisfaction in the relationship between teaching styles and learning engagement. Bootstrap sampling (5,000 iterations with 95% confidence intervals) was used to test significance. The significance level for this study was set at α = 0.05.

In this study, data were collected using self-report methods, which may involve common method bias. Therefore, Harman's single-factor test was employed to examine common method bias. The results showed that there were eight factors with eigenvalues < 1, and the first factor explained only 49.1% (< 50%) of the variance (Podsakoff et al., 2003), indicating that common method bias did not significantly affect the research results.

The study consisted of 1,629 adolescents, with a mean age of 14.31 ± 2.29 years. The cohort was divided between 785 males (48.2%) and 844 females (51.8%). Among these, 1,178 (72.3%) were junior high school students (Grades 7–9) and 451 (27.7%) were senior high school students (Grades 10–12). Within the normal range, the mean BMI was 20.63 ± 4.42 kg/m². Key variables that were assessed were as follows: the perceived PE instructional styles are relatively high (101.36 ± 26.05) (total possible score: 126). Level of engagement in sports learning (total possible score: 105): moderate (64.65 ± 17.93). Relatively high level of classroom satisfaction (total possible score: 175) (141.45 ± 32.99). Level of psychological resilience: moderate (70.73 ± 22.22) (maximum possible score: 100). The demographic and variable characteristics are presented in detail in Table 2.

Male students reported significantly higher levels of physical education learning engagement than for female students (P < 0.05), while there was no statistically significant difference in perceived physical education teachers‘ teaching styles between genders (t = 1.321, P = 0.187). In addition, there were no statistical differences in perceived PE teachers' teaching styles, PE learning engagement, classroom satisfaction, and psychological resilience between grades (p > 0.05). There was no statistical difference in the four types of teaching styles of physical education teachers in HA, RL, IE, and CS as perceived by male and female students between genders and grades (p > 0.05).

In terms of physical education learning engagement, boys were significantly more engaged than girls in vigor engagement, dedication engagement, and concentration engagement (p < 0.05), whereas only concentration engagement differed significantly between grades, and was significantly higher in high school than in middle school (t = −4.054, P = 0.04).

In terms of classroom satisfaction, boys were significantly more satisfied than girls in terms of teaching ability, site equipment, school support, school associationiveness, and classroom climate (p < 0.05), while between grades only site equipment satisfaction was significantly higher in middle school than in high school (t = 2.650, P = 0.008).

In terms of psychological resilience, boys were significantly higher than girls in resilience, self-improvement, and optimism (p < 0.05), however, there was no statistically significant difference in the psychological resilience of students between grades (p > 0.05). The results of the comparative analysis between groups are detailed in Table 3 and Supplementary material 2.

Correlational analyses demonstrated substantial positive correlations between students' perceived PE teaching styles, classroom satisfaction, psychological resilience, and sports learning engagement (refer to Table 4). The following are the primary findings: classroom satisfaction (r = 0.573, p < 0.01), psychological resilience (r = 0.449, p < 0.01), and sports learning engagement (r = 0.543, p < 0.01) exhibited moderately strong positive correlations with perceived PE teaching styles. Classroom satisfaction was moderately positively correlated with psychological resilience (r = 0.445, p < 0.01) and sports learning engagement (r = 0.495, p < 0.01). A moderately strong positive correlation was observed between psychological resilience and sports learning engagement (r = 0.532, p < 0.01). In conclusion, the statistical significance and positive correlation of all pairwise relationships among perceived PE teaching styles, classroom satisfaction, sports learning engagement, and psychological resilience were both significant. In addition, there were significant positive correlations among the sub-dimensions of PE teachers' teaching styles, classroom satisfaction, psychological resilience, and sports learning engagement. The detailed results are shown in Supplementary material 3.

In Table 5, the regression analysis results that investigate the correlation of students' perceived PE teaching styles and other variables on sports learning engagement are presented. Hypothesis H1 was thereby supported by the significant positive association with sports learning engagement that students' perceived PE teaching styles demonstrated after controlling for gender, grade, and BMI (β = 0.371, 95%CI = 0.343–0.398, p < 0.001). Subsequent examination of the four dimensions of teaching style disclosed distinct outcomes: the IE style demonstrated the strongest positive association (β = 0.898, 95%CI = 0.523–1.273, p < 0.001). A moderate positive association was observed for the RL style (β = 0.292, 95%CI = 0.066–0.518, p = 0.011). Engagement was significantly predicted by the CS style (β = 0.489, 95%CI = 0.179–0.800, p = 0.002). The HA style did not have a statistically significant correlation (β = 0.095, 95%CI = −0.092–0.282, p = 0.317). In addition, the VIF values of the independent variables all fell between 1.006 and 1.927, indicating that there was no multicollinearity. These results underscore the fact that IE, RL, and CS teaching styles are the primary factors that relationship sports learning engagement, whereas HA approaches lack significant predictive power.

To investigate the mediating roles of classroom satisfaction and psychological resilience in the relationship between PE teaching styles and sports learning engagement, a parallel mediation model was tested using the SPSS PROCESS Macro. Results (refer to Table 6) illustrated the following: classroom satisfaction: The indirect effect via classroom satisfaction in the association between teaching styles and engagement was significant (β = 0.189, 95%CI = 0.154–0.227, p < 0.05), accounting for 18.9% of the total association. Psychological Resilience: The indirect effect via psychological resilience in the association between teaching styles and engagement was also significant (β = 0.260, 95%CI = 0.215–0.308, p < 0.05), accounting for 26.0% of the total result. The direct association between teaching styles and engagement remained significant after accounting for mediators (β = 0.291, 95%CI = 0.254–0.329, p < 0.05). The partial mediation of both classroom satisfaction and psychological resilience is confirmed by these findings, which corroborate Hypotheses H2 and H3 (refer to Figure 2).

The results confirm a significant positive association between students' perceived PE teaching styles and their engagement (β = 0.371, p < 0.01), thereby validating Hypothesis H1. This finding aligns with self-determination theory (Ryan and Deci, 2017) and prior empirical studies (Leenknecht et al., 2017; Xiong, 2025; Yang and Dong, 2017), which propose that autonomy-supportive, structured, and emotionally nurturing teaching practices are associated with meeting students' psychological and developmental requirements and higher intrinsic motivation and behavioral engagement.

Analysis of Teaching Styles on a Subdimensional Scale: RL Style: By emphasizing structured lesson design and orderly activity organization, teachers mitigate instructional ambiguity, which in turn is associated with enhanced students' absorption engagement by providing predictable and coherent learning experiences (Martin, 2008).

CS Style: This style is associated with improved students' psychological safety and sustained engagement by fostering inclusive classroom climates, fostering emotional bonds between teachers and students, and providing emotional support (He et al., 2011; Shen et al., 2024). IE Style: The incorporation of innovative sports equipment, cutting-edge pedagogical tools, and novel activities contributes to proactive inquiry, creativity, and curiosity, thereby converting students from passive recipients to active learners (Ridwan et al., 2019). HA Style: Although humor can captivate students, Overreliance on entertainment (e.g., frequent quips, exaggerated gestures, or activities that deviate from instructional content) might compromise skill mastery. The positive correlation of these styles may be further diminished by adolescents' rising expectations for classroom professionalism, which may be perceived as unserious or inefficient (Bieg and Dresel, 2014; Bieg et al., 2019; Wanzer et al., 2010).

This investigation verified the partial mediating roles of psychological resilience and classroom contentment in the association between perceived PE teaching styles and sports learning engagement, supporting Hypotheses H2 and H3. To be more precise, classroom satisfaction mediated 18.9% of the total association (indirect association = 0.070) (indirect association = 0.096). Psychological resilience mediated 26.0% of the total association. These findings are consistent with previous research conducted by Betoret (2007) and Fletcher and Sarkar (2016), which suggests that teaching styles indirectly increase engagement by promoting positive classroom environments (via satisfaction) and resilience-building (via psychological adaptability).

This study reveals that physical education (PE) teaching in China needs to adhere to the principle of “teaching through rigor” (Jin and Cortazzi, 2006). PE teachers humor and liveliness should serve instructional objectives, content, and skill acquisition, rather than merely enlivening the classroom atmosphere. Relationship by traditional Chinese culture, classroom instruction emphasizes teacher authority, and classroom discipline (Li and Wegerif, 2014). Consequently, students expect PE teachers to demonstrate professionalism over entertainment. The Confucian concept of being “strict yet caring” helps explain why the “caring-sharing” and “rigorous-logic” teaching styles prove significantly more associationive. The former resonates with the “benevolence” ideal in teacher-student relationships, while the latter aligns with the role responsibility of “imparting knowledge and moral guidance”. Conversely, humor and liveliness disconnected from teaching objectives and content can easily be perceived by students as “inassociationive talk” and “undermining learning seriousness” (Van Praag et al., 2017). This is particularly true in PE, a subject primarily focused on sports skills instruction, where students prioritize the process and experience of learning motor techniques over their teachers' entertaining behaviors. This phenomenon stands in sharp contrast to findings in Western contexts where “humor fosters classroom relaxation” (Bieg and Dresel, 2018; Fredrickson, 2001), further confirming the cultural heterogeneity in teaching style associationiveness (Nisbett, 2010). Therefore, future reforms in Chinese PE should prioritize developing teachers' structured innovation and emotionally immersive care.

Our study also revealed significant gender differences in physical education engagement and psychological resilience, with males scoring significantly higher than females (p < 0.01). This disparity may stem from gender socialization's relationship on sports participation (Chalabaev et al., 2009). Males are often encouraged toward high-intensity competitive sports, while females frequently encounter “negative stereotypes about athletic ability” (Avraam and Anagnostou, 2022) and body image anxiety, reducing their engagement. Therefore, physical education teachers could implement gender-specific activities, and schools might develop differentiated curricula to better safeguard students' rights to enjoy sports.

Regarding grade levels, high school students demonstrated significantly greater focused engagement (p < 0.05) but lower satisfaction with facilities and equipment (p < 0.01). This reflects high schoolers' prioritization of physical education's “stress-relief function” under China's academic pressure (Liu, 2025), coupled with institutional neglect of sports resources. Conversely, interest-driven middle schoolers exhibited greater tolerance for facility limitations (Eccles and Roeser, 2011). We recommend targeted interventions: enhancing psychological resilience training for high school students while ensuring adequate equipment allocation in middle schools to cater to students' interests and needs.

This research contributes to the body of literature in educational psychology and physical education by delineating the correlations between teaching styles and the engagement of students in sports learning. It provides innovative insights for the optimization of pedagogical practices. The results have practical implications for instructional design and teacher development: (1) Improvements to Teacher Training: Emphasize the development of competencies in the areas of IE, RL, and CS. For example, promote the integration of evidence-based pedagogical methods and emerging technologies. Enhance the proficiency of educators in the development of structured lesson plans and the instruction of sport-specific skills. (2) Classroom Satisfaction Optimization: Promote autonomy by providing real-time feedback (e.g., wearable devices for performance monitoring) while ensuring safe and adequate sports facilities. Boost participation by implementing student-centered activities, such as peer-led skill demonstrations. (3) Resilience-Building Interventions: Combine attribution retraining with adversity training modules to reframe setbacks as growth opportunities. (4) Gender-Sensitive Pedagogy: By implementing elective-based modular teaching, the engagement disparity between genders can be bridged, enabling students to select activities that are in accordance with their interests.

The study has limitations, despite the fact that the hypotheses were confirmed: (1) Methodological Constraints: The use of self-report questionnaires and the cross-sectional design limit the ability to infer causal relationships and may neglect contextual factors and nuanced motivations. Qualitative methods, such as classroom observations and interviews, should be employed to triangulate the results of future research. (2) Cross-sectional design: the inability to establish causality necessitates longitudinal or experimental designs (e.g., monitoring engagement changes over a semester under varying teaching styles). (3) The samples in this study were all from the Chinese region, and it is possible that specific cultures may relationship students' perceptions of and responses to teaching styles, and the generalizability of its findings to other cultural contexts needs to be verified. Future research should employ longitudinal or experimental designs to better examine potential causal links, for example, by monitoring engagement changes over time in conjunction with varying teaching styles.