Mobile phones have become a significant component of our daily life due to their multifunctionality. The current statistics show that the number of mobile phone users in China has reached 1.59 billion (Ministry of Industry and Information Technology of the People’s Republic of China 2019). Mobile phones provide convenient methods for payment, shopping, entertainment, and learning that greatly simplify people’s lives. However, the increasing negative impacts of mobile phones cannot be ignored. College students are also under the influence of the impact of mobile phones, leading some students to have a lower GPA, suffer from anxiety, and experience lower well-being and life satisfaction (Lepp et al. 2014). Many studies revealed that college students who are mobile phone overusers are more likely to suffer from poor-quality sleep (Sahin et al. 2013; Thomée et al. 2011). After a long period, their situation could worsen due to physical and psychological issues. Furthermore, mobile phone dependence is related to increased impulsiveness in college students (Mei et al. 2018). Mobile phone addiction is a related issue that has also become widespread among college students (Qing et al. 2019), especially long-term addiction and uncontrolled behavior.

Mobile Phone Dependence among College Students

College students who depend on mobile phones and develop social and emotional disorders (Yen et al. 2009) exhibit a series of behavioral and technological addictions (Billieux 2012). High-frequency mobile phone use occupies college students’ time, leading to lower academic performance. Academic performance is an important indicator of students’ professional competence, and prior studies have shown that academic anxiety and procrastination among college students are positively correlated with their mobile phone use (Yang et al. 2019). Mobile phone dependence is positively related to depressive symptoms (Sara 2011) and negatively correlated with interpersonal relationships, self-esteem, self-regulation, and life satisfaction among young people (Lepp et al. 2014). Such poor and unsound social relationships could further increase individuals’ dependence on their mobile phones, and thus, they reduce their number of social activities to alleviate their anxiety regarding social interactions. This situation presents a vicious circle (Demirci et al. 2015). Studies have shown that overreliance on mobile phones or mobile phone addiction can lead to physical and psychological discomfort that is detrimental to the social functioning of individuals (Walsh et al. 2011). Thus, it is vital to prevent and improve the excessive use of mobile phones among college students.

Physical Activity

Physical activity is a progressive, nondisposable type of activity that is planned, developed, and repeatedly implemented; it can improve or maintain good physical fitness, including improved cardiopulmonary function, muscle strength, refined body composition, flexibility, etc. (Puetz et al. 2006). A meta-analysis found that there is no threshold of physical activity for health benefits (Warburton and Bredin 2017). Thus, sedentary people can easily gain observable benefits from minimal physical activity. In recent years, people have started to recognize the multifaceted positive effects of physical activity on the body and mind and have become convinced of the benefits of physical activity, even activity involving relatively low exertion.

Self-Control

Self-control is the ability of an individual to regulate reactions to overcome impulses, which requires individuals to change not only their inherent or spontaneous behavior and their way of thinking but also their levels of effort and consciousness to help them independently adjust their behavior to match their values and social expectations. Controlling impulses enables people to make better behavioral choices and maintain better long-term behaviors (Tangney et al. 2004). In many fields, self-control is considered a characteristic or the ability to influence behavior, and can be divided into state self-control and trait self-control. State self-control is a short-term, temporary resource-based characteristic, while trait self-control is a long-term and more stable characteristic. A higher level of trait self-control indicates that more self-control resources can be used (Baumeister et al. 2006), which also suggests that while trait self-control is a long-term, stable cognitive ability, it is also a trainable behavior.

The self-control strength model (Baumeister et al. 1998) considers that the resources available for self-control are limited. Once used, an individual’s limited mental energy temporarily declines, which is also known as self-loss and has an impact on subsequent behaviors that require self-control (Finne et al. 2019). On the basis of previous studies, Zhang and Zhang (2017) proposed that self-control can be divided into inhibition self-control and persistence self-control. Inhibition self-control mainly refers to refraining from engaging in a behavior, while persistence self-control refers to maintaining a behavior.

Self-Control and Mobile Phone Dependence

Self-control is among the psychological factors that effectively predict mobile phone addiction (e.g., Bianchi and Phillips 2005; Khang et al. 2012; Jiang and Zhao 2016; Mei et al. 2018). People with weak self-control are less able to resist automated habitual actions, such as the use of mobile phones, and tend to pursue immediate enjoyment, current desires, novel experiences, etc. Poor planning skills, low reaction inhibition, and the tendency to feel lost without their mobile phones and become bored easily are the characteristics of college students who rely on their mobile phones. Self-control can effectively restrain the unhealthy use of mobile phones. Impulse is a dimension of self-control, and many studies have shown that impulse is among the critical dimensions determining mobile phone overuse (Roberts and Pirog 2013). The more impulsive the student is, the more desperate they are to use their phones to satisfy their desires, and the more time they lose that could be used for studying; thus, they cannot control their behavior well and have a greater urge to rely on their phones, suggesting that their overuse of mobile phones is positively related to their impulses (Mei et al. 2018). Self-control involves not only healthy physical and mental behaviors, habits and concentration both at work and while studying, but also impulse control, controlled desire for entertainment, and resistance to temptation (Tan and Guo 2008). These are important qualities that help college students find a balance between their studies and mobile phone usage. Self-control implies that one engages in the appropriate behaviors needed to achieve one’s goals.

Physical Activity and Self-Control

An increasing number of studies have confirmed that physical activity not only shapes the body, improves weight loss, and strengthens the musculature but also has psychological benefits in terms of emotions, thinking, etc., and studies have shown that people with low self-control can benefit from physical activity (Schöndube et al. 2017). To date, limited research has investigated whether physical activity can improve self-control. Zou et al. (2016) found that five weeks of aerobic exercise could be a potentially effective intervention to improve self-control in college students. The authors assert that through low-intensity but regular exercise, individuals can train and enhance their self-control. Other studies have demonstrated that moderate-intensity physical activity can effectively promote self-control among exercise participants (Kamijo et al. 2007).

Hypothesis and Aims

As mobile phone dependence shares characteristics with Internet addiction, physical activity could improve behaviors that are similar to those observed with Internet addiction (Gao et al. 2012). On the one hand, physical activity objectively occupies a certain amount of time. On the other hand, physical activity can improve individuals’ capacity for self-control and ameliorate their mobile phone dependence. Physical activity and the use of mobile phones are considered dual tasks performed within a short period, warranting the introduction of the self-control strength model (Baumeister et al. 1998). Based on this model, it is easier for people to lose their restraint regarding unconscious behaviors at night (such as playing on their mobile phones for a long time before going to bed), possibly due to depletion of their self-control resources after completing their daytime activities (Baumeister 2002). When self-control is regarded as a long-term stable state, accomplishing various tasks that consume one’s self-control resources can actually improve one’s level of self-control (Muraven 2010). Thus, a more stable level of self-control ability ensures that an individual has greater capacity to refrain from mobile phone use.

A meta-analysis (de Ridder et al. 2012) indicated that trait self-control has a small to moderate effect on behavior in many areas of life; i.e., self-control is positively correlated with adaptive, idealized behaviors (e.g., habitual, automated behavior) and negatively correlated with nonadaptive, nonidealized behaviors. The comprehensive model proposed by Hagger (2013) regarding trait self-control and its diverse influence on behaviors is similar: Hagger argued that self-control exhibits individual differences, reflecting the ability and availability of one’s resources to resist impulses and habitual responses to participate in goal-oriented activities. Is there a mediation effect on the influence of self-control on mobile phone dependence?

The purpose of this study is to explore how physical activity can decrease an individual’s dependence on mobile phones and how self-control plays a mediating role, i.e., how it decreases mobile phone dependency. According to prior research, self-control has a limiting effect on mobile phone dependence. Thus, it is proposed that if physical activity has a positive effect on self-control, then it is reasonable to speculate that self-control can affect the relationship between physical activity and mobile phone dependence as a mediating variable. Therefore, this paper establishes a structural equation model with physical activity as the independent variable, mobile phone dependence as the dependent variable, and self-control as the mediating variable, and the stability of the model is verified.

Methods

Participants

In total, 418 undergraduate students were recruited via convenience sampling from a university in Beijing (151 freshmen; 42 sophomores; 64 juniors; and 137 seniors). The sample consisted of 115 males (29.2%) and 279 females (70.8%). In total, 220 (55.8%) participants were from rural areas, while 174 (44.2%) participants were from urban areas. Additionally, 137 (34.8%) participants were from a one-child family, while 256 (65.2%) participants were not.

Measures

Physical Activity Rating Scale (PARS-3: Liang 1994)

The PARS-3 is used to measure the amount of physical activity that participants engaged in during the previous month. In total, there are 3 items measuring exercise intensity, time and frequency (e.g., What is the intensity of your physical activity? If the frequency of several activities is similar, please select the activity with the maximum exercise intensity.). Each dimension is divided into 5 levels. Levels 1–5 of the exercise intensity correspond to 0–4 points. The exercise time and exercise frequency correspond to 1–5 points. The following formula is used: amount of physical activity = exercise intensity × exercise time × exercise frequency; thus, the minimum score of physical activity is 0 points, and the highest possible score is 100 points. The evaluation standard for the amount of physical activity is as follows: a small exercise amount is ≤19 points, a medium exercise amount is 20 to 42 points, and a large exercise amount is ≥43 points. The scale test-retest reliability is 0.82, and the internal consistency coefficient is 0.75.

College Students’ Mobile Phone Dependence Questionnaire (CSMDQ: Wang 2013)

The CSMDQ was developed by Wang (2013). In total, there are 20 items, and the items are rated on a 5-point Likert scale. The items cover withdrawal, i.e., the negative emotions that occur when an individual is involuntarily separated from his/her phone or is unable to use the applications normally; conflict, i.e., the negative influence on an individual’s work or study; prominence, i.e., the meaning of mobile phones to the individual; continuousness, i.e., the number of uninterrupted hours of phone use; and technicality, i.e., an individual’s habits during the operation of mobile media functions. The following is a sample item: “Cell phones are more important than clothes and food.” “Complete disagreement” is scored as 1 point, “basic disagreement” is scored as 2 points, “uncertain” is scored as 3 points, “basic agreement” is scored as 4 points, and “complete agreement” is scored as 5 points. Thus, the higher the score is, the stronger the dependency trend of that individual. The consistency coefficient of the scale is 0.832, and the internal consistency coefficient of each factor ranges from 0.517 to 0.717. The correlation between each factor of the questionnaire ranges between 0.133 and 0.521, and the correlation between each factor and the total questionnaire ranges between 0.523 and 0.833.

College Student Self-Control Scale (SCS: Tan and Guo 2008)

This scale consists of 19 items, and the Chinese version of the SCS has 5 dimensions (Tangney et al. 2004). The five dimensions are impulse control, resisting temptation, focusing on work or study, healthy habits and moderation. Statements such as “I am good at resisting temptation” are rated on a 5-point Likert scale. “Completely disagree” is scored as 1 point, “basically disagree” is scored as 2 points, “unsure” is scored as 3 points, “basically agree” is scored as 4 points, and “completely agree” is scored as 5 points. The scale includes 15 reverse-scored items. Thus, the higher the score is, the stronger the student’s self-control ability. The internal consistency coefficient of the scale is 0.86.

Procedures

We published the electronic questionnaire online for 5 days, and each participant completed the three scales. The questionnaire included the instructions, requirements and precautions and was completed anonymously. In total, 418 questionnaires were collected; 24 questionnaires were removed because the completion time was too short, and 394 valid questionnaires were obtained, yielding an effective recovery rate of 94.26%.

Data Analysis

The results of the study were discussed and analyzed using SSPS20.0. A hierarchical linear regression analysis was used to investigate the effects of physical activity on self-control and mobile phone dependence after controlling for the demographic variables. An independent samples t-test was utilized to compare the high-level and low-level physical activity study participants in terms of their self-control and mobile phone dependence. AMOS 17.0 was used to establish a structural equation model and test whether the mediating effect of self-control was significant via the bootstrap method.

Results

Common Method Deviation Test

The Harman single factor test was used to conduct a common method deviation test using exploratory factor analysis. Without rotation, nine factors with a trait root greater than 1 were obtained. The first common factor interpretation percentage was 23.905%, which was less than 40%, suggesting that there was no serious common method bias.

Regression Analysis and Difference Tests of Mobile Phone Dependence

After controlling for gender, place of origin and only-child status, the impact of physical activity on college students’ mobile phone dependence was analyzed. Table 1 shows that model 1 was statistically significant before including physical activity in the model (R2 = 0.03, F (2, 391) = 5.78, p < .01). Notably, gender had a predictive effect on mobile phone dependence, as females reported more dependence on their mobile phones than males. Additionally, the student’s place of origin could predict mobile phone dependence, as the college students from cities relied on their mobile phones more than the college students from rural areas. After including the amount of physical activity, model 2 was found to be statistically significant (R2 = 0.06, F (2, 391) = 8.00 (p < .01), ΔR2 = 0.03), and physical activity was significantly associated with mobile phone dependence (ΔF = 12.10, p < .01). The amount of physical activity was negatively associated with mobile phone dependence, whereas the place of origin remained significantly associated with mobile phone dependence.

Table 1 The effect of college students’ physical activity on mobile phone dependence using different demographic characteristics
Full size table

Because the numbers of college students who performed moderate and high amounts of exercise were not large enough, the two groups were combined, and the subjects were divided into the high- and low-exercise groups as follows: those receiving ≤19 points were assigned to the low-exercise group, and those receiving ≥20 points were assigned to the high-exercise group. The results of the independent samples t-test showed (Table 2) that the persistence of mobile phone dependence significantly differed between the two physical activity groups (t = 2.88, p < .01), as the low-exercise group used their mobile phones for longer periods. Technicality also significantly differed between the physical activity groups (t = 2.13, p < .05), as the high-exercise group used their mobile phones to pursue technical learning more frequently than the low-exercise group. The difference in the level of withdrawal between the genders according to the amount of physical activity was also significant (t = 3.08, p < .01); the low-exercise students were more inseparable from their mobile phones. Overall, the results revealed that mobile phone dependence is significantly different between the low- and high-exercise groups (t = 3.34, p < .01); i.e., college students with high exercise levels are less dependent on their mobile phones.

Table 2 Analysis of the difference in physical activity amounts, mobile phone dependence dimensions and total scores
Full size table

Regression Analysis and Difference Tests of Self-Control

We analyzed the impact of physical activity on the self-control of college students after controlling for their gender, place of origin and only-child status (Table 3). Model 1′ is statistically significant (R2 = 0.04, F (2, 391) = 7.07, p < .01) before physical activity is entered into the model. The students’ place of origin influences their self-control as students from urban areas were found to have a higher self-control capacity. Gender also exerted an influence on students’ self-control capacity, as females showed worse self-control than males. After adding physical activity to the model, model 2′ was statistically significant (R2 = 0.05, F (2, 391) = 6.12, p < .01), and the effect of physical activity was significant (ΔF = 4.11, p < .05). Students with higher levels of physical activity had stronger self-control. Moreover, the effect of place of origin on students’ self-control remained significant.

Table 3 The effect of physical education on the self-control of college students with different demographic characteristics
Full size table

The independent samples t-tests (Table 4) revealed a significant difference in impulse control between the groups of different levels of physical activity (t = −2.39, p < .05). Healthy habits were positively associated with physical activity (t = −3.31, p < .01), and students with low exercise levels were found to have poor health habits. Additionally, the temptation dimension significantly differed between the low-exercise and high-exercise groups (t = − 3.15, p < .01); compared with the high-exercise group, the low-exercise group had less capacity to resist temptation. Focus on work also significantly differed based on the level of physical activity (t = −2.37, p < .05), as students with high activity levels were more likely to concentrate on their work or studies than those with low levels. Overall, the results revealed that self-control is significantly different between the low- and high-exercise groups (t = −3.31, p < .01); i.e., a higher level of physical activity leads to greater self-control.

Table 4 Analysis of the differences in physical activity amount and self-control dimensions
Full size table

Mediation Effect Test of the Hypothesized Model

Mediating Effect Test of the Overall Model

According to the results above, physical activity has an impact on college students’ mobile phone dependence, and there is a relationship between physical activity and self-control. We sought to establish the mechanism linking physical activity, self-control and mobile phone dependence. A structural equation model of physical activity and mobile phone dependence with self-control as a mediating variable was established. The model was verified as follows.

First, the mediating effect model (see Fig. 1) was established, including the self-control dimensions (impulse control, healthy habits, resisting temptation, and focus on work), physical activity, and the dimensions of mobile phone dependence (persistence, technicality and withdrawal). Self-control was used as a mediator. The results indicate that the model fit is not ideal (x2/df = 3.55, GFI = 0.96, AGFI = 0.93, RMSEA = 0.08).

Fig. 1
figure 1

The impact of physical activity on mobile phone dependence: a mediation model of self-control

Full size image

Based on the modification index and the practical significance of the research, the two pairs of error terms (resisting temptation and healthy living) are covariates that can reduce the chi-square values by 9.24 and 10.35, respectively. The fit of the modified final structural equation model is good (x2 = 38.58, df = 16, x2/df = 2.41, GFI = 0.98, AGFI = 0.95, CFI = 0.97, RMSEA = 0.06, SRMR = 0.04).

The bootstrap test was used to repeat the sampling 2000 times (Table 5), and the deviation-corrected nonparametric percentile bootstrap method was used as the test basis (Hayes and Scharkow 2013). The indirect effect value is −0.15, and the 90% deviation correction confidence interval is [−0.22, −0.08], suggesting the presence of a mediating effect. The direct effect value is −0.09, and the 90% confidence interval is [−0.17, 0.01] and contains 0, indicating that the direct impact of physical activity on mobile phone dependence in the mediation model is not significant (Zhao et al. 2010). In practical research, cases of complete intermediation are rare (Iacobucci 2008). If self-control fully intervenes in the relationship between physical activity and mobile phone dependence, the possibility of other intermediaries is excluded (Pituch et al. 2005); thus, complete mediation should not be considered (Preacher and Hayes 2008). After adding self-control, the direct effect of physical activity on mobile phone dependence is not significant. Bootstrapping verifies that in this model, self-control has a mediating effect on the relationship between physical activity and mobile phone dependence.

Table 5 Mediation effect test
Full size table

Cross-Group Model Stability Test

To verify whether the mediation model has universal applicability, this study conducted a cross-group analysis of the model based on gender and place of origin (see Table 6). The model fit degree is good (x2 = 87.55, RMSEA = 0.04, GFI = 0.95). The measurement weight model showed that there is no significant difference in the factor loads; i.e., the loads of the male and female models are equivalent (p > .05). The measurement weight model and the structural covariance model were compared and analyzed, and △x2 reaches the critical value of x2 (0.05, △df). This finding indicates that there are significant differences in the covariates of the mediation models among college students of different genders. Thus, there are gender differences within the model. Further analysis shows that gender affects only the path from physical activity to self-control. The path coefficients of the male and female students are 0.12 and 0.16, respectively. The critical ratio of the differences between the parameters is 1.05 < 2.13 (Wu 2010). Thus, gender has a moderating effect on the path from physical activity to self-control. Physical activity is more effective in improving self-control in females than males.

Table 6 Model nested comparison table of the gender and place of origin cross-group analysis
Full size table

The model fit results (x2 = 58.06, RMSEA = 0.04, GFI = 0.97) revealed no difference with respect to the students’ place of origin. Thus, the model is widely applicable to different groups from rural and urban areas. The model nested comparison table shows the results of the comparison and analysis of the measurement weight model and the structural covariance model. △x2 does not reach the critical value of x2 (0.05, △df); thus, there are no significant differences in the factor load values in the mediation models of college students from different places of origin. Additionally, while comparing the measurement residual model with the structural covariance model, △x2 does not reach the critical value of x2 (0.05, △df), indicating that the residuals are also equivalent. Therefore, this model is equivalent for students from different places of origin.

Discussion

The Influence of Demographic Variables and Physical Activity on Mobile Phone Dependence

Physical activity can help decrease college students’ mobile phone dependence. Gender affects college students’ mobile phone dependence, and females’ mobile phone dependence tends to be greater than that of males. This result is consistent with previous findings. Females spend more time and energy on mobile social networking and online shopping than males (Li et al. 2017). In addition, a student’s place of origin has an impact on mobile phone dependence, as college students from cities are more dependent on mobile phones than those from rural areas. After controlling for the demographic variables, physical activity still has an overt negative effect on mobile phone dependence, indicating that physical activity can reduce college students’ mobile phone dependence; moreover, the gender differences in mobile phone dependence are mainly a consequence of different levels of exercise. Physical activity can effectively reduce the dependence of college students on mobile phones, especially the persistence, technicality and withdrawal aspects of mobile phone dependence. Here, persistence includes both the duration of the use of mobile phones and the continuing psychological impact of the overuse of mobile phones (Wang 2013). Technicality is reflected in the various applications of mobile phones, including ringtones, wallpaper, and camera functions, which are attractive to college students. When overusers are involuntarily separated from or unable to use their mobile phones, they exhibit negative emotions, low self-efficacy, low self-esteem and other withdrawal reactions; physical activity can decrease these symptoms.

The Influence of Demographic Variables and Physical Activity on Self-Control

Physical activity has a positive effect on self-control. College students of different genders and places of origin have different self-control capacities. Females have weaker self-control than males, and college students from rural areas have stronger self-control than those from urban areas. Daily city life is colorful, convenient and fast-paced but also full of all types of temptations, including overeating, crime, violence and addictive behavior, which tend to occur later in the day. This finding may be attributable to a depletion of self-control resources after the completion of an individual’s daytime activities (Baumeister 2002). For curious youths, resisting the temptation of the mobile phone world is challenging. A student’s place of origin has a greater influence than gender.

Physical activity positively influences the self-control dimensions of impulse control, healthy habits, resistance to temptation and the ability to focus on work. Engaging in physical activity can deplete an individual’s self-control resources. This physical activity could lead to some discomfort due to the accumulation of the exercise load, thereby motivating the impulse to cease the activity. At this time, an individual’s self-control system can activate and force the person to persist in training; in this way, an individual’s self-control can actually be improved (Baumeister and Tice 2007). According to the multidimensional self-control perspective proposed by Berkman et al. (2012), self-control can be divided into three parts, namely, emotional control, behavioral control and cognitive control, and a certain amount of exercise can alleviate negative emotions, restrain unhealthy/bad behaviors, and improve memory. Studies have shown that the mechanisms underlying the benefits of physical activity may be related to specific changes in the brain. For example, an increase in aerobic fitness in childhood is linked to a larger dorsal striatum region, which is the region of the brain that promotes cognitive control (Chaddock et al. 2010).

The Impact of Physical Activity on Mobile Phone Dependence: The Mediating Effect of Self-Control

The process by which physical activity decreases mobile phone dependence through self-control can be explained by the self-control resource model. The self-control strength model (e.g., Baumeister et al. 1998; Baumeister and Vohs 2016) argues that individuals who undertake tasks that require self-control resources temporarily consume some limited resources that are not immediately recovered (Baumeister et al. 1998). Trainable self-control is divided into two forms, namely, strength (baseline ability) and endurance (ability to endure exhaustion), and self-control can only improve self-control (Muraven 2010). When both physical activity and mobile phone dependence are regarded as long-term and stable behaviors, after the limited resources are consumed, self-control recovery can break through the ceiling of the original self-control resources. This process leads to continuous improvement in an individual’s trait self-control capacity. Due to the improvement in the self-control capacity, an individual’s dependence on mobile phones decreases.

Self-control can help regulate one’s behavior to meet specific criteria, such as ideals, values, ethics, and social expectations, and support the pursuit of long-term goals. The effects of physical activity and self-control are mutual. Many studies have explored the role of self-control in healthy behaviors, especially physical activity (Kinnunen et al. 2012; Galla and Duckworth 2015; Gerdtham et al. 2019; Finne et al. 2019). When engaging in exercise of a certain frequency, duration and intensity, it is necessary to use self-control resources. Lower self-control can lead the individual to fail to create an exercise plan or even meet low-level exercise intentions (Englert and Rummel 2016). Sport psychologists have divided self-control into inhibition and persistence self-control (Zhang and Zhang 2017), and physical activity is a long-term behavior that is not stable. As a dynamic process, physical activity can be divided into the following five stages: the pre-expected phase, the expected phase, the preparation phase, the action phase and the maintenance phase (Zong 2019). People with lower exercise levels are more likely to be in a stage prior to the difficult action phase, while individuals with higher exercise levels tend to be in the maintenance stage. Individuals with lower exercise levels need to use self-control resources more than individuals who have a higher exercise capacity when performing a certain amount of physical activity (Schöndube et al. 2017). Dependence on mobile phones has become a habitual and spontaneous behavior for some people. Individuals need to use cognitive inhibition, i.e., inhibitive self-control, to reduce their use of mobile phones (Zhang and Zhang 2017). In the short term after exercise, individuals with a low exercise level have fewer self-control resources than individuals with a high exercise capacity, reflecting a lower self-control capacity to restrain from using mobile phones. However, in the long term, the more self-control resources are used in the exercise process, the more beneficial exercise is in restoring the original self-control resources; i.e., people with low self-control have more room for improvement. Individuals with higher exercise levels also consume self-control resources but to a lesser extent. This finding indicates that even if the exercise level is high, the individual has formed a habit with respect to exercise intensity, duration and form, and then, self-control is used. Less exercise is not conducive to improvement in self-control. High exercise levels are the result of long-term adherence to physical activity, and improvement in self-control ability can plateau. Therefore, people who have reached a certain level of physical activity should not continue to exercise in the same way. Researchers recommend more diverse physical activities, including different types, more diverse forms of participation, different rules, and different environments.

The multigroup path analysis indicated that gender has a moderating effect only in terms of the impact of physical activity on self-control. Women can improve their self-control through physical activity more easily than men. Generally, women undertake less physical activity than men; the results of this work and others have shown that females are more apt to benefit from physical activity (Jun and Park 2019) and improve their self-control. To improve the mental health of college students, increasing the amount of their exercise is a suitable method for cultivating self-control ability in the early stage of exercise. After a certain level of exercise has been reached and certain habits have been formed, it is recommended that exercise activities be diversified. The key is to use certain self-control resources in the exercise process, thereby improving one’s self-control ability and decreasing one’s dependence on mobile phones. Eliminate a series of physical and mental problems caused by the tendency of addiction in the long term.

Limitations

In this study, physical activity level was assessed via a questionnaire. Thus, the level of physical activity subjectively reported by the participating respondents may be imprecise. In the future, the amount of physical activity could be tracked to achieve more accurate measurements. Most respondents who participated in the study were women, and most participants did not have a high level of physical activity; thus, the sample was unevenly distributed and lacked universality. Currently, mobile phones are rapidly developing, and the functions of mobile phones are not limited to simply making calls and sending text messages; mobile phones have become multifunctional. The environment on which the mobile phone dependence questionnaire relied was mostly based on the era of non-smartphones. Given the development of mobile phones, the sociological valence needs to be improved. In the future, a more scientific and effective questionnaire should be developed to address this issue. The advent of the era of Big Data helps researchers collect much larger sample sizes for analysis, thereby increasing universality and reducing bias in various factors.

Conclusion

Physical activity has a predictive effect on college students’ mobile phone dependence, and the tendency is lower among students with higher exercise levels. Physical activity also has an influence on the self-control of college students, and those with high levels of physical activity have stronger self-control. Self-control plays a mediating role in the relationship between the physical activity level and mobile phone dependence. Gender influences how physical activity affects self-control, as women can more easily enhance their self-control through physical activity than men. This study suggests that people should engage in physical activity to improve their self-control to achieve the ultimate goal of decreasing mobile phone dependence.