The Influence Mechanism of College Students' Physical Exercise Persistence Behavior from the Perspective of the SCeiP Model

Guoqiang Song1*, Xingyu Yi1,Tao Li1, Yuandai Chen2*

¹ Physical education Institute, Hanjiang Normal University, Shiyan 442000, China

² Physical education Institute, Hubei University Automotive Technology, Shiyan 442000, China

* Correspondence:

Guoqiang Song 1✉ Emailsongguoqiang@hjnu.edu.cn

Xingyu Yi 1

Tao Li 1

Yuandai Chen 1✉

1 Physical education Institute Hanjiang Normal University 442000 Shiyan China

2 Physical education Institute Hubei University Automotive Technology 442000 Shiyan China

Corresponding Author: Guoqiang Song,Yuandai Chen

songguoqiang@hjnu.edu.cn,20230064@huat,edu,cn.

Abstract

Objective

Promoting college students’ participation in physical exercise and cultivating good exercise habits are key measures to achieve the strategies of a Healthy China and Sports Power. To explore the influence mechanism of college students’ physical exercise adherence behavior, this study constructed a structural equation model based on the theoretical framework of the SCeiP behavioral development model. The model includes Subjective Experience (SE) and Perceived Social Support (SS) as the perceptual input layer, Perceived Usefulness (PU) and Behavioral Control (BC) as the decision and motivation layers, and Exercise Persistence (EP) as the behavioral output layer. This study aimed to reveal the mechanism underlying college students’ physical exercise adherence behavior.

Methods

A cluster random sampling method was used to obtain 684 valid questionnaires from three universities in China for analysis. The assessment tools included demographic information and scales for Perceived Social Support (SSRS), Subjective Exercise Experience (SEES), Perceived Usefulness (EEPS), Perceived Behavioral Control (PBCS), and Exercise Persistence Behavior (EPBS). Data were statistically analyzed using SPSS 26.0 and AMOS 28.0.

Results

Significant differences in physical exercise persistence were found between males and females and between upperclassmen and underclassmen (P < 0.001). Subjective experience was significantly and positively correlated with exercise persistence (r = 0.212, p < 0.001), and perceived external support was significantly and positively correlated with exercise persistence (r = 0.154, p < 0.01). Perceived usefulness and behavioral control played a partial remote mediating role in the relationship between subjective experience and exercise persistence (β = 0.006, 95% CI [0.001, 0.015]), accounting for 3.3% of the total effect of the study. They also played a partial mediating role in the relationship between perceived external support and exercise persistence (β = 0.009, 95% CI [0.001, 0.024]), accounting for 4.0% of the total effect.

Conclusion

Subjective exercise experience and perceived external support can directly influence exercise persistence and enhance college students' physical exercise adherence behavior through cognitive transformation and behavioral control. In physical education courses, teachers should enhance the construction of external resources, create a positive exercise atmosphere, and focus on value identification, transformation, and assessment of the behavioral process.

Keywords:

SCeiP model

Physical exercise persistence

Subjective experience

Perceived social support

Perceived usefulness

Behavioral control

1. Introduction

Regular physical exercise not only provides physiological benefits, such as promoting metabolism, enhancing cardiovascular health, and releasing endorphins [1–3] but also positively impacts psychological aspects, such as alleviating personal stress, reducing anxiety, boosting self-confidence, regulating mood, and enhancing well-being [4]. It has become a crucial means of regulating public health. However, in reality, college students often lack sufficient internal drive and sustainability in their exercise behavior, which severely constrains their holistic physical and mental health. Issues such as "persistently high obesity and myopia rates" and the emergence of the "brittle college student" phenomenon highlight existing health crises [5]. Despite the national emphasis on sports and health strategies, the problem of difficulty in maintaining exercise habits persists. The underlying reason is that scholars have overly focused on isolated factors and single effects on physical exercise, neglecting a systematic analysis of the dynamic decision-making chain in behavioral development, thereby hindering the mechanism of translating rational cognition into behavior.

The SCeiP model is an individual behavior development model that integrates the essence of the Theory of Planned Behavior [6], Social Cognitive Theory [7, 8], and Social Exchange Theory. It divides individual behavioral development into perceptual input, decision-making, motivation, and behavioral output layers, revealing the dynamic coupling mechanism of "cognition-affection-behavior" [9]. This model addresses the shortcomings of the Theory of Planned Behavior, which overlooks environmental factors, while enhancing the rigor and systematicity of the behavioral development process. It adds an effect perception layer—value-risk judgment—emphasizing the analysis of value trade-offs and integration in behavior, and comprehensively considers the impact of individual psychological cognition and environmental interaction on behavioral development. Research indicates that the SCeiP model performs well in terms of its predictive and explanatory power.

Therefore, this study aims to extend the application scope of SCeiP theory and validate its effectiveness by proposing a structural equation model to examine the relationships among subjective experience, perceived social support, perceived usefulness, behavioral control, and exercise persistence. Revealing the mechanism of college students' physical exercise adherence behavior holds practical significance for promoting their physical exercise and improving their physical health levels.

1.1 Relationship between Subjective Experience and Exercise Persistence Behavior

Cognitive development theory posits that all human activities result from continuous interaction with the environment, and that the quality of subjective experience influences the development of individual behavior [10]. Physical exercise experience refers to the subjective feelings and personal emotional tendencies an individual obtains during exercise, comprising cognitive and affective tendencies [11, 12]. A positive exercise experience can influence exercise cognition and enhance decision-making for exercise behavior, as well as induce a pleasant psychological state, fostering the generation of self-efficacy [13]. A positive subjective experience marks the beginning of enhanced exercise cognition, promotes exercise behavior, and stimulates the desire to repeatedly engage in an activity. Some scholars have argued that physical exercise releases endorphins and increases adrenaline levels. These chemicals can improve the exerciser's mood and enhance muscular non-excitability, serving as an important source of stress relief and effectively stimulating the recurrence of exercise behavior [14]. Therefore, this study proposes Hypothesis H1: Subjective experience can effectively predict college students' physical exercise adherence behavior.

1.2 Relationship between Perceived Social Support and Exercise Persistence Behavior

Perceived social support refers to the support, encouragement, and assistance individuals receive when engaging in certain activities, primarily encompassing external resources in spiritual and material forms. In physical exercise, perceived social support typically includes environmental facilities, interpersonal support and informational support [15]. Among these, interpersonal support from family, peers, and teachers provides spiritual encouragement, forming an important basis for creating an exercise atmosphere and establishing emotional bonds [16]. The influence of perceived social support on physical exercise has been confirmed in several studies. For example, Wang Kun (2023) pointed out that extraverted individuals show a stronger association between emotional support and exercise [5]; upperclassmen are more influenced by peer exercise commitment, whereas underclassmen are more affected by family emotional support. Based on this, this study proposes Hypothesis H2: Perceived social support can effectively predict college students' physical exercise adherence behavior.

1.3 The Chain Mediating Role of Perceived Usefulness and Behavioral Control

Perceived usefulness is the subjective basis for individuals’ judgments when engaging in certain activities. When faced with urgent and important tasks, people tend to exert greater effort to ensure smooth progress in completing the task. It serves as the central hub for individual information processing and decision-making, determining the development of events [17]. Social Exchange Theory suggests that individuals weigh the pros and cons, balancing benefits and risks during social interactions to decide on the course of events. For instance, the intensity, potential injury, and emotional experience of a single exercise session can influence exercise adherence; positive exercise experiences can make individuals aware of the value and usefulness of exercise, whereas emotional support from friends and family can foster the development of perceived usefulness [18]. Expectancy Theory posits that when people have sufficient confidence and capability regarding an activity, they are more willing to invest effort, thereby overcoming adverse factors to achieve their goals. Existing research indicates that self-efficacy, an important component of behavioral control, influences the occurrence and persistence of individual behavior [19]. In summary, this study proposes the following hypotheses: H3: Perceived usefulness and behavioral control play a chain mediating role in the effect of subjective experience on exercise persistence; H4: Perceived usefulness and behavioral control play a chain mediating role in the effect of perceived social support on exercise persistence.

2. Methods

2.1 Participants

This study surveyed college students from Hubei Province. First, a random sampling method was used to select three universities from all public institutions in Hubei Province: Hanjiang Normal University, Hubei University of Arts and Science, and Hubei University (these universities are located in both provincial capital and non-capital cities, providing a certain level of representativeness). Second, proportional random sampling was conducted based on the ratio of full-time students at a 100:1 ratio. Online questionnaires were distributed through counselors or teachers at the three universities, ensuring a completion time of no less than 10 min. A total of 758 questionnaires were distributed. After excluding 74 questionnaires due to overly short completion time, contradictory reverse-scored items, or missing values, the valid response rate was 90.2%. Among the valid samples, there were 363 men (53.1%) and 321 women (46.9%).

The questionnaire survey was conducted between November 25, 2024, and April 30, 2025. Prior to completing the questionnaire, participants were informed of its purpose, relevant considerations, and privacy protection measures.

All participants provided informed consent and were advised of their right to withdraw from the survey at any time. The study did not involve minors.

2.2 Measurement Tools

Based on the four-tier elements of the SCeiP model (perceptual input, decision-making, motivation, and behavioral output layers) and considering the actual exercise situation of college students, the study identified subjective exercise experience, perceived social support, perceived usefulness, behavioral control, and exercise persistence as the core content of the questionnaire.Subjective Exercise Experience (SE):Adapted from the scale developed by McAuley and Courneya in 1994 [20], it includes two dimensions—affective attitude and cognitive attitude—comprising 4 items. It primarily assesses the exerciser's cognition of exercise outcomes and subjective feelings during exercise.Perceived Social Support (SS): Utilized the scale developed by Zimet et al., including three aspects: peer support, family support, and other support, representing major aspects of external interpersonal support [9, 15].Perceived Usefulness (PU): Adapted the core part of Wang Lei's "Exercise Effect Perception Questionnaire" [9], which includes benefit perception, risk perception, and objective conditions, mainly measuring the exerciser's decision-making regarding exercise benefits and risks.Behavioral Control (BC): Adapted from the "Perceived Behavioral Control Questionnaire" compiled by Foley et al. [21], reflecting the difficulty, behavioral drive, and degree of behavioral control in college students' extracurricular exercise.Exercise Persistence (EP): Self-developed by Wang Shen et al. in 2016 [22], includes four indicators: behavioral habits, effort investment, emotional experience, and sustained effort, assessing students' physical exercise habits and persistence.

2.3 Data Analysis

Data were analyzed using SPSS 26.0 for descriptive statistics, reliability analysis, correlation analysis, and common method bias testing. Structural Equation Modeling (SEM) was conducted using AMOS 28.0 to test the hypothesized model and mediation effects. The maximum likelihood estimation method was used. Model fit was assessed using indices such as χ²/df, GFI, AGFI, CFI, IFI, RFI, RMR, and the RMSEA. Bootstrap resampling (5000 iterations) was used to test the significance of indirect effects.

3. Results

3.1 Common Method Bias Test

Harman’s single-factor test was conducted by performing a principal component factor analysis on all questionnaire items[23]. The results revealed five factors with eigenvalues greater than 1. The most significant factor accounted for 26.168% of the variance ( Table 1), which was below the 40% threshold [24, 25]. Therefore, no significant common method bias was detected in this study.

Table 1

Variance Explained by Factors
	Total	Variance Percentage	Cumulative %	Total	Variance Percentage	Cumulative %
1	5.234	26.168	26.168	5.234	26.168	26.168
2	2.921	14.603	40.77	2.921	14.603	40.770
3	2.172	10.858	51.628	2.172	10.858	51.628
4	1.853	9.263	60.892	1.853	9.263	60.892
5	1.721	8.607	69.498	1.721	8.607	69.498

3.2 Descriptive Statistics and Correlation Analysis

A total of 684 participants were included in the analysis, with males accounting for 53.1% of the sample. The proportions of freshmen, sophomores, juniors, and seniors were 28.1%, 22.2%, 23.2%, and 26.5%, respectively. The ages ranged from 18 to 23 years.

The correlation results are presented in Table 2. Perceived social support was significantly positively correlated with subjective experience (r = 0.202, p ≤ 0.001), perceived usefulness (r = 0.204, p ≤ 0.001), behavioral control (r = 0.021, p ≤ 0.05), and exercise persistence (r = 0.154, p ≤ 0.01). Subjective experience was significantly positively correlated with perceived usefulness (r = 0.247, p ≤ 0.001), behavioral control (r = 0.026, p ≤ 0.05), and exercise persistence (r = 0.212, p ≤ 0.001). Perceived usefulness was significantly and positively correlated with behavioral control (r = 0.105, p ≤ 0.05) and exercise persistence (r = 0.097, p ≤ 0.001). Behavioral control was significantly and positively correlated with exercise persistence (r = 0.283, p ≤ 0.001). All five variables showed significant positive correlations, providing a necessary premise for subsequent chain mediation effect testing.

Table 2

Descriptive Statistics and Correlation Analysis (N = 684)
	Mean	SD	SS	SE	PU	BC	EP
SS	3.18	0.85	1
SE	3.06	0.85	0.202^༊༊༊	1
PU	3.32	0.87	0.204^༊༊༊	0.247^༊༊༊	1
BC	3.33	0.87	0.021^༊	0.026^༊	0.105^༊	1
EP	3.31	0.94	0.154^༊༊	0.212^༊༊༊	0.097^༊༊༊	0.283^༊༊༊	1
Note: SS = Perceived Social Support; SE = Subjective Experience; PU = Perceived Usefulness; BC = Behavioral Control; EP = Exercise Persistence. SD:Standard Deviation. *p ≤ .001, p ≤ .01, *p ≤ .05.

3.3 Control Variable Difference Analysis

Independent samples t-tests were conducted for gender and grade level (grouped as lower-grade: freshmen and sophomores; upper-grade: juniors and seniors). As shown in Table 3, men scored significantly higher than women on exercise persistence, behavioral control, perceived usefulness, subjective experience, and perceived social support. Regarding grade level, juniors scored highest in exercise persistence, behavioral control, and perceived usefulness, while seniors scored highest in subjective experience and perceived social support. Significant differences existed between the groups, with upperclassmen scoring significantly higher than underclassmen across all five aspects.

Table 3

Difference Analysis of Demographic Variables (Mean ± SD)
Group	Number of Cases	Persistence	Behavior Control	Perceived Usefulness	Subjective Experience	Perceived Social Support
GenderMale	363	3.962 ± 0.718	3.892 ± 0.701	3.683 ± 0.721	4.128 ± 0.599	4.085 ± 0.563
Female	321	3.525 ± 0.850	3.539 ± 0.768	3.462 ± 0.744	3.956 ± 0.612	3.943 ± 0.612
F		7.349	7.757	7.003	9.724	7.089
P		.000	.000	.008	.002	.008
Grade: Low	344	3.497 ± 0.871	3.484 ± 0.850	3.206 ± 0.705	3.885 ± 0.689	3.931 ± 0.591
High	340	3.648 ± 0.803	3.643 ± 0.718	3.552 ± 0.748	4.015 ± 0.586	3.949 ± 0.591
F		8.258	9.151	10.699	8.954	9.920
P		.000	.000	.000	.000	.009

3.4 Structural Equation Model Analysis

3.4.1 Model Fit Indices

After multiple model comparisons and modifications using AMOS 28.0, the optimal model was obtained ( Fig. 1). The maximum likelihood method was used with Bootstrap 5000 iterations, selecting Bias-Corrected and Percentile 95% confidence intervals. As shown in Table 4, all model fit indices were excellent, particularly RMSEA and χ²/df, confirming that the final model met the research requirements and was credible.

Fig. 1

Non-standardized Model Diagram

Table 4

Model Fit Indices
Indicator Name	Model Fit Standard	Test Results	Model Fit Judgment	Indicator Name	Model Fit Standard	Test Results	Model Fit Judgment
Chi2/DF	<3	1.23	Acceptable	RFI	>9	0.94	Acceptable
GFI	>0.9	0.98	Acceptable	IFI	>9	0.96	Acceptable
AGFI	>0.9	0.97	Acceptable	RMR	<0.08	0.06	Acceptable
CFI	>0.9	0.91	Acceptable	RMSEA	<0.08	0.07	Acceptable

3.4.2 Path Analysis Among Factors

The regression analysis results (Table 5) showed that subjective experience had a positive predictive effect on exercise persistence (β = 0.177, P = 0.001), confirming H1. After introducing remote mediators (perceived usefulness and behavioral control), the direct effect of subjective experience on exercise persistence weakened (β = 0.171, P = 0.001). Perceived social support had a positive predictive effect on exercise persistence (β = 0.222, P = 0.019), confirming Hypothesis H2. Similarly, after introducing remote mediators, the direct effect of perceived social support on exercise persistence weakened (β = 0.213, P = 0.025).

The combined results showed that perceived usefulness and behavioral control played a partial chain mediating role in the relationship between subjective experience and exercise persistence (Path: SE → PU → BC → EP; β = 0.006, 95% CI [0.001, 0.015]), confirming Hypothesis H3. They also played a partial chain mediating role in the relationship between perceived social support and exercise persistence (Path: SS → PU → BC → EP; β = 0.009, 95% CI [0.001, 0.024]), confirming Hypothesis H4. The contribution rates of these two mediation paths were 3.3% and 4.0%, respectively. Significant differences existed between the direct and mediation effects in both paths.

Table 5

Effect Test of Variables
Parameter	Estimate	Product of Coef		Bias-corrected
Parameter	Estimate	SE	T	Lower	Upper	P
c1(SE→EP)	0.171	0.044	3.886	0.083	0.256	0.001
c2(SS→EP)	0.213	0.092	2.315	0.031	0.389	0.025
ind1(SE→PU→BC→EP)	0.006	0.003	2.000	0.001	0.015	0.014
ind2(SS→PU→BC→EP)	0.009	0.006	1.500	0.001	0.024	0.015
total1(ind1 + c1)	0.177	0.044	4.023	0.089	0.262	0.001
total2(ind2 + c2)	0.222	0.092	2.413	0.044	0.399	0.019
r1(ind1/total1)	0.033	0.023	1.435	0.005	0.109	0.016
r2(ind2/total2)	0.040	0.182	0.220	0.002	0.215	0.034
diff1(ind1-total1)	-0.171	0.044	-3.886	-0.256	-0.083	0.001
diff2(ind2-total2)	-0.213	0.092	-2.315	-0.389	-0.031	0.025

4. Discussion

Based on the results and hypotheses, this study found that males had higher levels of exercise persistence than females, and upperclassmen scored higher than underclassmen, which is consistent with previous research [26, 27]. Subjective experience can directly predict exercise persistence and can also remotely predict it through perceived usefulness and behavioral control, aligning with Chang's findings [28]. Perceived social support can directly predict exercise persistence and can also remotely predict it through perceived usefulness and behavioral control, consistent with Kong's results [29].

4.1 Demographic Differences Analysis

Males exhibited higher exercise persistence than females. Multiple studies have shown significant sex differences in physical exercise. Physiologically, males generally have higher muscle mass, basal metabolic rate, and testosterone levels than females, contributing to better motor performance and object control, manifesting as more active, aggressive, and motivated behaviors [30]. Psychologically, significant differences exist in exercise motivation between men and women [31, 32]. Females show more pronounced clustering characteristics in exercise participation, influenced by exercise atmosphere, identity, and peer companionship, while males score significantly higher on competence, health, and fun motivation [33]. This suggests that male physical exercise is more influenced by internal motivation than female physical exercise. Self-Determination Theory posits that internal motivation is the primary driver of sustained behavior. Finally, differing social role expectations and image norms lead to significant gender differences in physical exercise, constrained by perceived social support and cultural factors, with women often expected to display gentleness and quietness [34]. Therefore, males score higher on exercise persistence and related influencing factors, which are influenced by regional culture, physiological conditions, and psychological drivers.

Upperclassmen exhibited higher exercise persistence than underclassmen. This phenomenon reveals a positive developmental trajectory in college students' physical exercise over time. Yang Xiaojie, investigating the physical exercise status at Nanjing University of Technology, pointed out that compared to underclassmen, upperclassmen have stronger autonomous exercise awareness and are more inclined to choose sports they are interested in (e.g., badminton, running), with significantly increased diversity in exercise choices [35]. As health awareness gradually strengthens, the autonomous exercise intention of upperclassmen increases. Tang and Wang found that higher levels of psychological resilience were associated with greater stability in exercise habits. Upperclassmen have higher psychological resilience than underclassmen, which is directly related to their accumulated life experience [36]. They also noted that male college students' exercise behavior was better than that of female students. Yang Jianwen's survey supports this view, suggesting that although upperclassmen may score higher on somatization and anxiety factors, they possess higher psychological resilience and often use behavioral control to enhance physical exercise [37]. Therefore, as college students accumulate life experiences and disciplinary knowledge, their exercise cognition, perceived benefits, and risk perception further increase, leading to more stable exercise habits and persistence.

4.2 Subjective Experience and Exercise Persistence

As shown in Table 5, subjective experience positively predicted college students' exercise persistence (β = 0.171, significant), supporting H1. This indicates that subjective experience is an influencing factor of exercise persistence, with a linear relationship between them, consistent with previous research. Rodrigues et al. noted that exercise frequency has a significant positive relationship with positive emotions, and positive affect is positively correlated with life satisfaction, self-esteem, and subjective vitality [38]. Subjective experience reflects an individual's subjective response and emotional value provision to behavior, is closely related to self-efficacy, and serves as the source of motivating individual behavior. Self-Determination Theory suggests that internal behavioral motivation evolves from the transformation of external motivation [6, 39]. Emotional experience, as part of the perceptual input layer, is a direct source of information for building a personal exercise cognition. Research by Xu Jing, Yang Chunhui, and Zhu Leqing indicates that subjective exercise experience can significantly positively predict college students' exercise persistence and can also stimulate cognitive decision-making power [40, 41]. The mechanism lies in positive exercise experiences directly acting on internal motivation, allowing individuals to perceive the value and enjoyment of the activity. Therefore, creating positive subjective exercise experiences is key to cultivating the persistent exercise habits of college students.

4.3 Perceived Social Support and Exercise Persistence

As shown in Table 5, perceived social support positively predicted college students' exercise persistence (β = 0.213, significant), supporting Hypothesis H2. Perceived social support includes interpersonal support (family, friends, and teachers) and informational support. Research by Zhang Han, Qiu Zhaoyi, et al. found that interpersonal support positively predicts table tennis enthusiasts' exercise participation, with self-efficacy playing a partial mediating role; informational support positively predicts their exercise participation, with self-efficacy playing a complete mediating role [42]. Perceived social support directly affects exercise persistence. Its core mechanism lies in providing emotional identification, instrumental assistance, and normative pressure, thereby lowering the threshold for physical exercise [43]. Social Cognitive Theory suggests that encouragement, role models, and commitment from peers can significantly enhance an individual's self-efficacy [44–46], a key factor in predicting exercise behavior. Many studies have confirmed that perceived social support can not only directly predict exercise persistence but also indirectly promote it through self-efficacy [42], psychological resilience [36], family parenting styles [47], and social-emotional competence [48]. Therefore, creating an inclusive and encouraging social exercise environment is a necessary prerequisite for transforming college students' occasional exercise behavior into stable habits.

4.4 The Chain Mediating Role of Perceived Usefulness and Behavioral Control

As shown in Table 5 and Fig. 1, perceived usefulness and behavioral control played a partial chain mediating role between subjective experience and exercise persistence (β = 0.006, significant, accounting for 3.3%), supporting Hypothesis H3. They also played a partial chain-mediating role between perceived external support and exercise persistence (β = 0.009, significant, accounting for 4.0%), supporting Hypothesis H4.

Perceived usefulness, as the decision-making layer, is the core factor for individuals to weigh the pros and cons. Social Exchange Theory suggests that individuals only act when the perceived benefits outweigh the costs, which is influenced by cognition, environment, individual skills, and psychological preparedness [49]. Behavioral control refers to people's control over specific activities, which is influenced by self-efficacy, psychological resilience, and self-control, directly reflecting internal motivation. When individuals perceive external support and have positive exercise experiences, they often recognize the meaning and value of physical exercise, subsequently controlling behavior through goal achievement and self-discipline. The Theory of Planned Behavior confirms that individual behavior is primarily driven by behavioral intention and perceived behavioral control. Perceived usefulness and behavioral control are key hubs that transform initial experience and external resources into personal motivation and action information [50]. Self-Determination Theory further supports this view, suggesting that introjected and identified regulations are key paths in the motivation internalization process [51]. Therefore, in the process of motivation internalization and behavior cultivation, information processing (perceived usefulness) and goal achievement (behavioral control) are two critical links that determine the formation of the systemic chain.

4.5 Limitations

This study, based on the four-tier elements of the SCeiP model, selected key influencing factors for a cross-sectional survey design and constructed an influencing factor model for college students' physical exercise adherence behavior. This reflects the relationships among the variables to a certain extent. However, due to the multitude of factors influencing college students' exercise persistence, a more precise promotion mechanism model requires further enrichment and optimization in future studies. Additionally, this study did not consider the effects of exercise type and intensity. Subsequent research should include these variables as control or moderating variables for further analysis.

5. Conclusion

The development of college students' physical exercise adherence behavior is a systematic and complex process. Key factors range from individual subjective perception and external resource utilization to individual information processing and behavioral control. Creating a positive exercise atmosphere and facilitating the systematic transformation of health knowledge are focal points in physical education. Furthermore, in the current era of comprehensive digital empowerment, physical education teachers should utilize informational support channels to provide exercise assistance and guidance, offer timely feedback on problems encountered during exercise, supervise and manage the exercise process, and emphasize the cultivation of health identity mechanisms and goal design related to physical exercise. This approach helps develop physical exercise habits, gradually enhancing the physical and mental health of college students.

Acknowledgement

We are grateful to the participants and their universities for their cooperation and participation in this study.

Funding

This study was supported by the Hanjiang Normal University Scientific Research Fund Key Project (XJ2025A06) and the Hanjiang Normal University Teaching Reform Research Project (2024B01).

Institutional Review Board Statement

The study protocol was approved by the Ethics Committee of Hanjiang Normal University (approval no. 2025 (38)).

Informed consent was obtained from all participants prior to the formal survey and testing.

Data Availability

The raw data supporting the conclusions of this article will be made available by the authors without undue reservation.

Author Contribution

Conceptualization: Guo-qiang Song, Xing-yu YiData curation:Guo-qiang Song, Xing-yu Yi,Tao LiFormal analysis: Guo-qiang Song, Xing-yu Yi,Tao Li.Funding acquisition: Guo-qiang Song, Xing-yu Yi,Tao Li.Investigation: Guo-qiang Song, Xing-yu Yi,Tao Li.Methodology: Guo-qiang Song, Xing-yu Yi,Tao Li.Project administration: Guo-qiang Song, Xing-yu Yi,Tao Li.Resources: Guo-qiang Song,Yuandai ChenSoftware: Guo-qiang Song,Yuandai ChenSupervision: Guo-qiang Song,Yuandai ChenValidation: Guo-qiang Song,Yuandai ChenVisualization: Guo-qiang Song,Yuandai ChenWriting – original draft: Guo-qiang Song,Xing-yu Yi,Yuandai ChenWriting – review & editing: Guo-qiang Song,Xing-yu Yi,

References

Wolf JK, Cossman JS. Exercise, Sleep Quality, and Physical Health during the COVID-19 Pandemic. South Med J. 2024;117(3):150–8. http://dx.doi.org/10.14423/smj.0000000000001659.

He LH, Li Y, Chen ZH. The Effect of Subjective Exercise Experience on Exercise Behavior and Amount of Exercise in Children and Adolescents: The Mediating Effect of Exercise Commitment. Int J Environ Res Public Health. 2022;19(17). http://dx.doi.org/10.3390/ijerph191710829.

Lv HW, Zhang T, Li B, Wang R. The influence of control beliefs on the cardiovascular fitness of college students: the chain mediating effect of subjective exercise experience and exercise adherence. BMC Public Health. 2024;24(1). http://dx.doi.org/10.1186/s12889-023-17509-3.

Zhang Z, Chen Y, Zhang L. Research on the Relations Among Personality Traits, Sports Commitment, and Exercise Behavior - A Case Study of Chinese College Students. Psychol Res Behav Manag. 2024;17:665–78. http://dx.doi.org/10.2147/PRBM.S440540.

Kun W, Yucong P, Muhuan L, Yuhui S, Ying J. An Analysis of the Moderating Role of Extracurricular Activities among College Students of a University in Beijing in the Relationship between Social Support and Physical Exercise. Med Soc. 2023;36(03):78–82. http://dx.doi.org/10.13723/j.yxysh.2023.03.014.

Hwang S, Lee J, Jang D. Climate Change Awareness and Pro-Environmental Intentions in Sports Fans: Applying the Extended Theory of Planned Behavior Model for Sustainable Spectating. Sustainability. 2024;16(8). http://dx.doi.org/10.3390/su16083246.

Bassette LA, Kujawski M, Donges E. Teaching social support to promote inclusive physical activity for people with autism. Adv Autism. 2024;10(4):259–72. http://dx.doi.org/10.1108/aia-04-2024-0028.

Guo SQ, Guo KL, Fu HY. Exercise adherence and meaning in life: The chain mediating effect of perceived social support and optimism. Social Behav Personality. 2023;51(11). http://dx.doi.org/10.2224/sbp.12679.

Wang L. Research on the Influencing Factors and Promotion Strategies of College Students' Exercise Persistence Behavior. Doctor. Shandong University; 2021.

10.

Delgado JLG, Pagola GR, Barea FJP, Madueño AF, Arteaga AS, Corrales BS. Physical Activity and Exercise Interventions in Abdominal Wall Hernia Patients. Am J Phys Med Rehabil. 2025;104(6):519–26. http://dx.doi.org/10.1097/phm.0000000000002657.

11.

Xiang J, Gao J, Gao Y. The effect of subjective exercise experience on anxiety disorder in university freshmen: the chain-mediated role of self-efficacy and interpersonal relationship. Front Psychol. 2024. http://dx.doi.org/10.3389/fpsyg.2024.1292203. 15.

12.

Sun YT, Ji PY, Wang Y, Fan HY. The Association Between the Subjective Exercise Experience of Chinese Women Participating in Square Dance and Group Cohesion: The Mediating Effect of Income. Front Psychol. 2021. 12.http://dx.doi.org/10.3389/fpsyg.2021.700408.

13.

Gibson LP, Giordano GR, Bidwell LC, Hutchison KE, Bryan AD. Acute Effects of Ad Libitum Use of Commercially Available Cannabis Products on the Subjective Experience of Aerobic Exercise: A Crossover Study. Sports Med. 2024;54(4):1051–66. http://dx.doi.org/10.1007/s40279-023-01980-4.

14.

Hu YC, Chen XX, Fan JX, Huang Y, Ye JY, Gu F, Li YM. The Subjective Will and Psychological Experience of Home-Based Exercise in Lung Cancer Patients During Interval of Chemotherapy: A Qualitative Study. J Multidisciplinary Healthc. 2023;16:663–74. http://dx.doi.org/10.2147/jmdh.S403185.

15.

Han L, Zhaoyi Q, Jingtang H. The relationship between environmental support and participation in table tennis: The mediating role of self-efficacy and the moderating effect of technical knowledge. J Beijing Sport Univ. 2024;47(08):105–18. http://dx.doi.org/10.19582/j.cnki.11-3785/g8.2024.08.010.

16.

Zhang Y, Hasibagen, Zhang C. The influence of social support on the physical exercise behavior of college students: The mediating role of self-efficacy. Front Psychol. 2022. http://dx.doi.org/10.3389/fpsyg.2022.1037518. 13.

17.

Huang WZ, Wei RK, Cheng BJ. The impact of exercising self-efficacy on physical education learning engagement: the mediating role of flow experience. Front Psychol. 2025. http://dx.doi.org/10.3389/fpsyg.2025.1588643. 16.

18.

Ren ZB, Hu LL, Yu JJ, Yu Q, Chen ST, Ma YD, Lin JY, Yang L, Li XY, Zou LY. The Influence of Social Support on Physical Activity in Chinese Adolescents: The Mediating Role of Exercise Self-Efficacy. Children-Basel. 2020;7(3). http://dx.doi.org/10.3390/children7030023.

19.

Qin GY, Han SS, Zhang YS, Ye YP, Xu CY. Effect of physical exercise on negative emotions in Chinese university students: The mediating effect of self-efficacy. Heliyon. 2024;10(17):e37194. http://dx.doi.org/10.1016/j.heliyon.2024.e37194.

20.

Yucheng Y, Wenying H, Chang H, Wen Z, Chao S. Analysis of Mediating and Latent Variables between the Core Competencies of Physical Education in the New Curriculum Standards for Middle School Students and Subjective Exercise Experiences. Chin School Health. 2025;46(07):941–5. http://dx.doi.org/10.16835/j.cnki.1000-9817.2025205.

21.

Salleg-Cabarcas MJ, Robledo-Castro C, Monsalve-Vertel CP. Effect of Hit-Sport-Game Exercise Training on Self-Control and Regulation in Children With Attention Deficit Hyperactivity Disorder (ADHD). Glob Pediatr Health. 2024;11:2333794X241287095. .http://dx.doi.org/10.1177/2333794X241287095.

22.

Shen W, Yiping L, Chunqiang G. The mechanism of the impact of the cohesion of amateur sports teams on the exercise persistence of team members: A moderated two-level mediating model. J Wuhan Inst Phys Educ. 2016;50(03):73–80. http://dx.doi.org/10.15930/j.cnki.wtxb.2016.03.012.

23.

Zhang T, Li BW, He XQ, Jia P, Ye ZC. The Effect of Exercise Atmosphere on College Students' Physical Exercise-A Moderated Chain Mediation Model. Behav Sci. 2025;15(4). http://dx.doi.org/10.3390/bs15040507.

24.

Middleton D, Thompson F, James K. Evaluating the outcomes of Australia's first all-age public hospital Sport and Exercise Medicine Outpatient Clinic: a retrospective cross-sectional study. Aust Health Rev. 2024;49. http://dx.doi.org/10.1071/AH24245.

25.

Liu YY, Feng QK, Tong Y. Physical exercise and college students' mental health: Chain mediating effects of social-emotional competency and peer relationships. Social Behav Personality. 2024;52(7). http://dx.doi.org/10.2224/sbp.13159.

26.

Feng YL, Chen C, Liu Y, Li S. Physical exercise behavior characteristics and influencing factors for participation in police officers. Front Public Health. 2025. http://dx.doi.org/10.3389/fpubh.2024.1507066. 12.

27.

Guan T, Sheng J, Tan L. Physical exercise and subjective well-being of Chinese adults: Relationship and mechanisms based on the Chinese General Social Survey 2021. Social Behav Personality. 2025;53(5). http://dx.doi.org/10.2224/sbp.13344.

28.

Huidong C. The Impact of Emotional Experience in Physical Education Class on College Students' Exercise Persistence: The Mediating Role of Peer Support. Master. Liaoning Normal University; 2024.

29.

Fanliang K. A Study on the Impact of Social Support on Physical Self-esteem and Exercise Persistence among Non-Sports Major College Students. Master. Guangzhou University; 2025.

30.

Li N, Wang D, Zhao X, Li Z, Zhang L. The association between physical exercise behavior and psychological resilience of teenagers: an examination of the chain mediating effect. Sci Rep. 2024;14(1):9372. http://dx.doi.org/10.1038/s41598-024-60038-1.

31.

Truong LK, Mosewich AD, Miciak M, Losciale JM, Li LC, Whittaker JL. Social support and therapeutic relationships intertwine to influence exercise behavior in people with sport-related knee injuries. Physiother Theory Pract. 2025;41(1):139–52. http://dx.doi.org/10.1080/09593985.2024.2315520.

32.

Su Z, Zhang ZQ, Zhou YH. Impact of Exercise Atmosphere on Adolescents' Exercise Behavior: Chain Mediating Effect of Exercise Identity and Exercise Habit. Int J Mental Health Promotion. 2024;26(7):579–88. http://dx.doi.org/10.32604/ijmhp.2024.051842.

33.

Matud MP, Diaz A. Gender, exercise, and health: A life-course cross-sectional study. Nurs Health Sci. 2020;22(3):812–21. http://dx.doi.org/10.1111/nhs.12736.

34.

Meyer T, Cobley S. Addressing female underrepresentation in sport & exercise-related research: JSAMS policy for submitted studies (& subsequent considerations). J Sci Med Sport. 2024;27(7):435–6. http://dx.doi.org/10.1016/j.jsams.2024.05.018.

35.

Xiaojie Y. A Study on the Current Situation and Countermeasures of Physical Exercise among Senior Students of Nanjing University of Technology. J Shandong Inst Phys Educ 2004(05):97–8http://dx.doi.org/10.14104/j.cnki.1006-2076.2004.05.033

36.

Fangwu T, Weiqiang W. The Mechanism of College Students' Psychological Resilience on Their Physical Exercise Behaviors: From the Perspectives of Habit Control and Resistance to Temptation. J Phys Educ. 2024;31(05):53–61. http://dx.doi.org/10.16237/j.cnki.cn44-1404/g8.2024.05.008.

37.

Jianwen Y, Yu T, Shenglin Z, Jianwen Y. Survey on the Mental Health Status of Senior Students in Engineering Universities. Mod Prev Med. 2009;36(16):3077–9.

38.

Rodrigues F, Faustino T, Santos A, Teixeira E, Cid L, Monteiro D. How does exercising make you feel? The associations between positive and negative affect, life satisfaction, self-esteem, and vitality. Int J Sport Exerc Psychol. 2022;20(3):813–27. http://dx.doi.org/10.1080/1612197x.2021.1907766.

39.

Mowaad NA, Elgohary R, ElShebiney S. Effect of Stanozolol and/or Cannabis Abuse on Hypertrophic Mechanism and Oxidative Stress of Male Albino Rat Cardiac Tissue in Relation to Exercise: A Sport Abuse Practice. Cardiovasc Toxicol. 2024;24(6):527–38. http://dx.doi.org/10.1007/s12012-024-09859-0.

40.

Jing X, Chunhui Y. The Impact of Subjective Exercise Experience on the Persistence of College Students' Exercise: The Mediating Role of Internal Motivation. Fujian Sports Technol. 2023;42(01):26–30.

41.

Leqing Z, Baolin D. Subjective experience, commitment and exercise persistence: A case of dualism and undifferentiation. J Nanjing Inst Phys Educ (Social Sci Edition). 2016;30(04):82–90. http://dx.doi.org/10.15877/j.cnki.nsic.2016.04.013.

42.

Han L, Zhaoyi Q, Jingtang H. The relationship between environmental support and participation in table tennis: The mediating role of self-efficacy and the moderating effect of technical knowledge. J beijing Univ Phys Educ. 2024;47(08):105–18. http://dx.doi.org/10.19582/j.cnki.11-3785/g8.2024.08.010.

43.

Peng B, Chen WS, Wang HS, Yu T, Kong MM. A study on the relationship between physical exercise and feelings of inferiority among college students: the chain mediating effect of social support and emotional regulation ability. Front Psychol. 2025;15. http://dx.doi.org/10.3389/fpsyg.2024.1521510.

44.

Davis AJ, Crittenden B, Cohen E. Effects of social support on performance outputs and perceived difficulty during physical exercise. Physiol Behav. 2021. http://dx.doi.org/10.1016/j.physbeh.2021.113490. 239.

45.

Craig BAP, McDonough MH, Culos-Reed SN, Bridel W. Types of Social Support Predicting Physical Activity and Quality of Life in Group Exercise Programs for Adults Living with Cancer. Res Q Exerc Sport. 2025;96(2):279–86. http://dx.doi.org/10.1080/02701367.2024.2402525.

46.

Bhuiyan N, Kang JH, Papalia Z, Bopp CM, Bopp M, Mama SK. Assessing the stress-buffering effects of social support for exercise on physical activity, sitting time, and blood lipid profiles. J Am Coll Health. 2022;70(5):1563–9. http://dx.doi.org/10.1080/07448481.2020.1810055.

47.

Renda Z, Jihao Y, Jiarui G, Xiangying W. The correlation between college students' physical exercise behaviors and family upbringing styles as well as psychological resilience. Chin School Health. 2024;45(08):1152–6. http://dx.doi.org/10.16835/j.cnki.1000-9817.2024240.

48.

Shuming W, Hongbo B. The Impact of Physical Exercise on the Social Emotional Competence of Adolescents: The Chain Mediating Role of Social Support and Psychological Resilience. Res Sports Sci. 2023;37(06):24–33. http://dx.doi.org/10.15877/j.cnki.nsic.20240009.001.

49.

Jinhai S, Jie C, Tuojian L. Research on User Willingness to Use Fitness Apps Based on the Two-Factor Theory. J Phys Educ. 2019;26(05):71–8. http://dx.doi.org/10.16237/j.cnki.cn44-1404/g8.20190724.003.

50.

Tang K, Yao J, Wang R, Ruan WQ, Chen XH, Yang LL, Li YP, Zhang R, Zhou CL. Factors influencing proactive health behaviors in pre-frailty older adults: A qualitative study based on theory of planned behavior. Geriatr Nurs. 2024;60:671–6. http://dx.doi.org/10.1016/j.gerinurse.2024.10.029.

51.

Zhao K, Zhao Y, Xu W. Relationship between middle school students' academic stress and physical exercise behavior from the perspective of Self-Determination Theory: The chained mediation of motivation and intention. PLoS ONE. 2025;20(1):e0316599. http://dx.doi.org/10.1371/journal.pone.0316599.

Table 1. Variance Explained by Factors

	Total	Variance Percentage	Cumulative %	Total	Variance Percentage	Cumulative %
1	5.234	26.168	26.168	5.234	26.168	26.168
2	2.921	14.603	40.77	2.921	14.603	40.770
3	2.172	10.858	51.628	2.172	10.858	51.628
4	1.853	9.263	60.892	1.853	9.263	60.892
5	1.721	8.607	69.498	1.721	8.607	69.498

Table 2. Descriptive Statistics and Correlation Analysis (N = 684)

	Mean	SD	SS	SE	PU	BC	EP
SS	3.18	0.85	1
SE	3.06	0.85	0.202^༊༊༊	1
PU	3.32	0.87	0.204^༊༊༊	0.247^༊༊༊	1
BC	3.33	0.87	0.021^༊	0.026^༊	0.105^༊	1
EP	3.31	0.94	0.154^༊༊	0.212^༊༊༊	0.097^༊༊༊	0.283^༊༊༊	1

Table 3. Difference Analysis of Demographic Variables (Mean ± SD)

Group	Number of Cases	Persistence	Behavior Control	Perceived Usefulness	Subjective Experience	Perceived Social Support
GenderMale	363	3.962 ± 0.718	3.892 ± 0.701	3.683 ± 0.721	4.128 ± 0.599	4.085 ± 0.563
Female	321	3.525 ± 0.850	3.539 ± 0.768	3.462 ± 0.744	3.956 ± 0.612	3.943 ± 0.612
F		7.349	7.757	7.003	9.724	7.089
P		.000	.000	.008	.002	.008
Grade: Low	344	3.497 ± 0.871	3.484 ± 0.850	3.206 ± 0.705	3.885 ± 0.689	3.931 ± 0.591
High	340	3.648 ± 0.803	3.643 ± 0.718	3.552 ± 0.748	4.015 ± 0.586	3.949 ± 0.591
F		8.258	9.151	10.699	8.954	9.920
P		.000	.000	.000	.000	.009

Table 4. Model Fit Indices

Indicator Name	Model Fit Standard	Test Results	Model Fit Judgment	Indicator Name	Model Fit Standard	Test Results	Model Fit Judgment
Chi2/DF	<3	1.23	Acceptable	RFI	>9	0.94	Acceptable
GFI	>0.9	0.98	Acceptable	IFI	>9	0.96	Acceptable
AGFI	>0.9	0.97	Acceptable	RMR	<0.08	0.06	Acceptable
CFI	>0.9	0.91	Acceptable	RMSEA	<0.08	0.07	Acceptable

Table 5. Effect Test of Variables

Parameter	Estimate	Product of Coef		Bias-corrected
Parameter	Estimate	SE	T	Lower	Upper	P
c1(SE→EP)	0.171	0.044	3.886	0.083	0.256	0.001
c2(SS→EP)	0.213	0.092	2.315	0.031	0.389	0.025
ind1(SE→PU→BC→EP)	0.006	0.003	2.000	0.001	0.015	0.014
ind2(SS→PU→BC→EP)	0.009	0.006	1.500	0.001	0.024	0.015
total1(ind1 + c1)	0.177	0.044	4.023	0.089	0.262	0.001
total2(ind2 + c2)	0.222	0.092	2.413	0.044	0.399	0.019
r1(ind1/total1)	0.033	0.023	1.435	0.005	0.109	0.016
r2(ind2/total2)	0.040	0.182	0.220	0.002	0.215	0.034
diff1(ind1-total1)	-0.171	0.044	-3.886	-0.256	-0.083	0.001
diff2(ind2-total2)	-0.213	0.092	-2.315	-0.389	-0.031	0.025

Figure 1 Non-standardized Model Diagram

Yes