Self-control mediates the relationship between time perspective and mobile phone addiction in Chinese college students

Study design

In this study, a cross-sectional design was used to investigate the relationship among time perspective, self-control and MPA in Chinese college students through three self-rated questionnaires.

Procedures

First, the research proposal was submitted to the Academic Ethics Committee of Chongqing University of Arts and Sciences for review, and the research was carried out after approval (CQWL-2022-10-8). Then, advertisements were distributed on university campuses to recruit participants, and surveys were conducted using an online questionnaire tool (Questionnaire Star, https://www.wjx.cn/). Before the subjects formally completed the questionnaires, they were informed of the purpose and specific methods of the survey. The questionnaires were conducted anonymously. All subjects needed to provide informed consent online. After the participants completed the questionnaires, they received a gift for their participation.

Participants

Participants in this study were from three universities in Chongqing, Hubei and Guizhou provinces. All subjects completed the questionnaires within 20 min. A total of 526 college students participated in the questionnaire survey, including 401 female students (76.2%) and 125 male students (23.8%). There were 148 freshmen (28.1%), 154 sophomores (29.3%), 136 juniors (25.9%) and 88 seniors (16.7%). The average age was 19.73 ± 1.37 years old.

Measures

The Zimbardo Time Perspective Inventory (ZTPI) (Zimbardo & Boyd, 1999) was used to investigate the time perspectives of college students. Originally comprising 56 items, the ZTPI has been revised in several countries, resulting in multiple versions for different cultural backgrounds and groups, each with a different number of questions and content (Orkibi, 2015). In this study, the Chinese version revised by Wang et al. (2015a) was used; it can be applied to measure the time perspectives of Chinese college students. This 20-item scale contains five dimensions: the past positive, past negative, present fatalistic, present hedonistic, and future orientation dimensions. A five-point Likert score is adopted, with responses ranging from 1 (complete disagreement) to 5 (complete agreement). Higher scores indicate a higher time perspective level. The Cronbach’s alpha coefficients of the five dimensions are 0.65 (past positive), 0.71 (past negative), 0.65 (present hedonistic), 0.65 (present fatalistic), and 0.61 (future). In this study, the Cronbach’s alpha coefficients of the five subscales were 0.699, 0.804, 0.554, 0.704 and 0.646, respectively. The reliability of the ZTPI-56 version was found to be better than that of the short version (Przepiorka, Sobol-Kwapinska & Jankowski, 2016). Cortina (1993) pointed out that the smaller the number of items in the scale, the smaller the α coefficient will be. Therefore, the α coefficient of the ZTPI-56 is generally larger than that of the short scale. However, a scale with too many items will take a long time to complete and place a certain psychological burden on the subjects. Considering that this study required the subjects to complete three scales at the same time, to reduce the workload of the subjects, the short version of the ZTPI was adopted, as done by other researchers (e.g., Orkibi, 2015; Zhang, Howell & Bowerman, 2013). Confirmatory factor analysis of the structural equation model was used to test the factor structure of this version, and the fit indices were obtained (χ²/df = 2.509, p < 0.001, Tucker Lewis index (TLI) = 0.886, comparative fit index (CFI) = 0.904, root mean square error of approximation (RMSEA) = 0.054), indicating that the fitting of the five-factor model of this scale was acceptable.

The self-control scale developed by Tangney, Baumeister & Boone (2004) was used to measure trait self-control. The revised Chinese version (Tan & Guo, 2008) can effectively measure the self-control levels of Chinese college students and is scored on a 5-point Likert scale, ranging from complete disagreement (1) to complete agreement (5). Higher scores indicate higher levels of self-control. In the present study, the Cronbach’s α coefficient of the total scale was 0.868.

The Mobile Phone Addiction Tendency Scale for college students developed by Xiong et al. (2012) was adopted; it includes four dimensions: withdrawal symptoms (negative physical or psychological effects of not using mobile phones), salience (the use of mobile phones occupies the center of thinking and behavior activities), psychological comfort (the effect of mobile phone use in interpersonal interactions) and mood changes (mood changes caused by mobile phone use). It consists of 16 items scored on a 5-point scale, with scores of 1–5 ranging from complete disagreement to complete agreement. The scores of the four dimensions are summed to obtain the total score of the scale, and the higher the total score is, the more serious the MPA tendency. The scale was shown to provide a good assessment of MPA among Chinese university students (Chen et al., 2016a). In this study, the Cronbach’s α coefficient of the total scale was 0.909.

Data analysis

The data were analyzed with SPSS 22.0 and AMOS 26.0. Pearson correlation analysis was performed to explore the relationships among time perspective, self-control and MPA. The latent variable mediating effect analysis of the structural equation model was used to investigate the mediating role of self-control in the relationship between time perspective and MPA. Since the scales used in this study contain multiple subscales, item parceling was used to simplify the structural equation model. Specifically, the internal-consistency approach (also known as isolated parceling) was used to package items in the same dimension of the scale (Little et al., 2002). χ²/df, the RMSEA, the CFI and the TLI were used to evaluate the fit of the model in this study. It is generally believed that if χ²/df is less than 2, the RMSEA is less than 0.08, and the CFI and TLI are more than 0.9, then the model fits well (Hu & Bentler, 1999; Steiger, 1990). Bootstrap tests with 5,000 random samples and 95% confidence intervals were used to analyze the significance of the mediating variable. If the 95% confidence interval did not include 0, the mediation effect was significant (Hayes, 2013).

Common method bias

This study used a self-report method to collect data for the three scales, so there may be a common method bias effect. The Harman single-factor test (Fuller et al., 2016) was adopted to test this effect. The results showed that a total of 13 factors had eigenvalues greater than 1, and the variance explained by the first factor was 21.9%, which was less than the critical criterion of 40%. According to the usual criteria (Podsakoff, MacKenzie & Podsakoff, 2012), there was no significant common method bias in this study.

Descriptive statistics and Pearson correlation analysis

Pearson correlation analysis was used to examine the relationships among time perspective, self-control and MPA. As shown in Table 1, past negative (r = 0.397, p < 0.001), present hedonistic (r = 0.207, p < 0.001) and present fatalistic (r = 0.444, p < 0.001) perspectives were significantly positively correlated with MPA. A future time perspective (r =  − 0.200, p < 0.001) and self-control (r =  − 0.599, p < 0.001) were negatively correlated with MPA. In addition, past negative (r =  − 0.409, p < 0.001), present hedonistic (r =  − 0.214, p < 0.001) and present fatalistic (r =  − 0.465, p < 0.001) perspectives were significantly negatively correlated with self-control. In contrast, a future time perspective (r = 0.459, p < 0.001) was significantly and positively correlated with self-control. Notably, a past positive time perspective was not significantly associated with either self-control (r = 0.053, p = 0.225) or MPA (r =  − 0.066, p = 0.13).

The mediating effect of trait self-control on the relationship between time perspective and MPA

The latent variable mediating effect analysis was used to examine the mediating effect of self-control on the relationship between time perspective and MPA. The five time perspective dimensions were put into the model at the same time to examine the unique influence of each dimension on MPA. The results show that the overall fit of the model was good (χ²/ df = 4.317 (χ² = 259.009, df = 60), TLI = 0.895, CFI = 0.931, RMSEA = 0.079). The standardization coefficients are shown in Fig. 2. Specifically, the past negative, present hedonistic, present fatalistic and future orientation perspective dimensions were significant predictors of self-control (β = −0.240, p < 0.001, 95% CI [−0.043 to −0.018]; β = −0.192, p < 0.001, 95% CI [−0.050 to −0.018]; β = −0.277, p < 0.001, 95% CI [−0.055 to −0.028]; β = 0.431, p < 0.001, and 95% CI [0.065–0.099], respectively). Both self-control (β = −0.760, p < 0.001, 95% CI [−1.425 to −0.899]) and a future time perspective (β = 0.210, p < 0.001, 95% CI [0.030–0.098]) were significant predictors of MPA. Before the inclusion of self-control as a mediating variable, a future time perspective negatively predicted MPA (β = −0.105, t = −2.648, p < 0.01). Notably, when self-control was included, the relationship between a future time perspective and MPA reversed.

The mediating effect was tested by the nonparametric percentile bootstrap method with bias correction. The 95% confidence intervals were calculated by repeated random sampling (5000 times). The results (see Table 2) showed that the direct effects of the past negative, present hedonistic and present fatalistic perspectives on MPA were not significant, but the indirect effects of these perspectives on MPA through self-control (β = 0.034, p < 0.001, 95% CI [0.02–0.051]; β = 0.038, p < 0.001, 95% CI [0.02–0.06]; β = 0.047, p < 0.001, 95% CI [0.031–0.065], respectively) and the total effects (β = 0.042, p < 0.001, 95% CI [0.024–0.06]; β = 0.039, p < 0.001, 95% CI [0.013–0.063]; β = 0.062, p < 0.001, 95% CI [0.041–0.083], respectively) were significant. These results suggest that self-control fully mediated the relationships between past negative, present hedonistic, and present fatalistic perspectives and MPA. The direct effect of a future time perspective on MPA (β = 0.062, p < 0.001, 95% CI [0.03–0.098]) and the indirect effect on MPA through self-control (β = −0.093, p < 0.001, 95% CI [−0.123 to −0.069]) were both significant, but the total effect was not significant (β = −0.031, p = 0.053, 95% CI [−0.059–0.001]). Notably, the indirect effect size was approximately 1.5 times that of the direct effect size, but the total effect was smaller than the direct effect, indicating that there was a “suppression effect” (MacKinnon & Lamp, 2021) between a future time perspective and MPA.

The relationship between time perspective and MPA

In this study, we found that the a past positive time perspective did not predict MPA, which is consistent with previous findings (Chittaro & Vianello, 2013). This suggests that there is no significant correlation between positive emotions about past events and MPA. In contrast, holding negative attitudes about the past affects mobile phone use. Some studies have found that individuals with a past negative time perspective have worse self-regulation ability and are more likely to experience negative emotions (Davies & Filippopoulos, 2015; Zhang, Ding & Wang, 2021). A past negative time perspective can predict ruminative thinking (Aström et al., 2019), and rumination weakens individuals’ abilities to suppress and transform negative emotions, leading to a decline in their executive function (Yang et al., 2017). Impaired executive function and the inability to resist temptation and control impulses are key causes of addiction (Lechner et al., 2019). Chittaro & Vianello (2013) surveyed 149 Facebook users and found that a past negative perspective could significantly predict pathological internet use.

A present fatalistic view of time involves a fatalistic, helpless, and hopeless attitude toward the future and life (Chen et al., 2016b). A present hedonistic perspective, on the other hand, aims at the near future, focuses only on immediate benefits, and seeks the thrill and pleasure of the moment (Zimbardo, Keough & Boyd, 1997). The tendency to be powerless over the present can lead to the development of negative mental states and ineffective coping strategies, which may promote the emergence of addictive online behavioral dependence (Settanni et al., 2018). Individuals with a present hedonistic time perspective are more prone to risk-taking (Xu et al., 2021). Individuals with present time perspectives (including present fatalistic and present hedonistic perspectives) tend to make quick decisions without planning for the future, achieving immediate gratification without considering the possible consequences of the future (Settanni et al., 2018). According to compensatory internet use theory (Kardefelt-Winther, 2014), individuals with a present time perspective hold a negative attitude toward the future, pursue immediate gratification, and lack the self-control ability to achieve goal-oriented behaviors. Therefore, they are prone to MPA to meet various psychological needs.

This study found a significant negative correlation between a future time perspective and MPA. A future time perspective implies that the pursuit of goals dominated by the behavior is oriented toward the future (Zimbardo & Boyd, 1999). Two aspects of a future time perspective, planning for the future and the realization of future projects, were found to be effective indicators for predicting internet addiction in adults (Przepiorka, Blachnio & Cudo, 2019). Specifically, individuals with a future time perspective have stronger achievement motivations and self-regulation abilities (Baird et al., 2021) and higher incentive values for perceived delayed goals (Teuscher & Mitchell, 2011). Therefore, their corresponding delayed gratification abilities (Kim, Kim & Kim, 2020) and self-control are stronger (Barber et al., 2009), which can reduce the risk of MPA (Han et al., 2017).

The mediating role of self-control

This study found that the impact of time perspective on MPA is mediated through self-control, which verified Hypothesis 2. Specifically, self-control completely mediated the relationships between past negative, present hedonistic, and present fatalistic perspectives and MPA. Individuals with a past negative time perspective are prone to experience negative emotions (Cao et al., 2019), and negative emotions can impair self-control (Chester et al., 2016). Individuals with a present fatalistic view of time believe that life is determined by fate and is difficult to change; they lack a sense of internal control over the future (Orkibi & Ronen, 2019). Individuals with a present hedonistic view of time display highly impulsive sensation seeking (Muro et al., 2015); they pursue the experience of present stimuli and have defects in delay gratification (Chavarria et al., 2015; Cosenza et al., 2017), which leads to a low level of self-control. In contrast, individuals with a future time perspective tend to think about the possible consequences of the future and engage in rational decision behaviors to obtain long-term benefits (Agarwal & Srivastava, 1981). As a result, they increase the incentive value of distal motivation and decrease the incentive value of proximal motivation, thus increasing their self-control (Dreves & Blackhart, 2019). Many studies have confirmed that the stronger an individual’s self-control is, the lower their risk of MPA (Han et al., 2017; Li et al., 2021). A recent meta-analysis revealed that the positive effects of a future time perspective on individuals are mainly mediated by self-regulation (Baird et al., 2021). The dual system theory of addiction holds that addiction is the result of an imbalance between the reflective system in the prefrontal cortex and the impulsive system in the amygdala, and the reflective system controls the impulsive system through multiple mechanisms after an individual learns social rules (Bechara, 2005). A functional brain imaging study found that core clinical symptoms of addiction include impaired self-control(manifested by impulsivity and compulsion), and symptoms associated with impaired self-control include reduced activity in control networks, including the anterior cingulate, adjacent prefrontal cortex and striatum (Tang et al., 2015). These studies confirm the key role of self-control in the development of MPA.

Interestingly, this study found that a future time perspective negatively predicted MPA before including self-control as the mediating variable, that a suppression effect was generated after introducing the mediating variable, and that the relationship between a future time perspective and MPA was reversed. The indirect effect size of self-control was much larger than the direct effect size of a future time perspective, indicating that a future time perspective and self-control together have a greater impact on MPA. A recent study further confirmed that a high future time perspective combined with high self-control predicted a low level of MPA (Peng et al., 2022). A future time perspective and self-control, while effective in reducing the risk of MPA, may not be beneficial at higher levels. The pursuit of long-term goals in the future while limiting the satisfaction of one’s current needs can have the opposite effect. On the one hand, setting the pursuit goal too far into the future will make individuals perceive that the goal is out of reach, which will have a time discount effect on the value of the future goal (McClure et al., 2007). On the other hand, giving too much thought to the realization of long-term goals at the expense of the satisfaction of current needs can reduce happiness and adversely affect an individual (Wiese et al., 2018). The mediating effect analysis revealed that a future time perspective was inversely associated with MPA, which may be related to these reasons. Clearly, more research is needed in the future to delve deeper into the underlying mechanisms of this phenomenon. These results also suggest the importance of maintaining a reasonable future time perspective level and self-control.

The present study has a few limitations. On the one hand, it was conducted by self-reporting, and subjective reports may have certain measurement biases. For example, the results may have been influenced by the survey environment and the effects of social approval. Future studies should eliminate the influence of these interfering factors by means of objective measurements such as behavioral experiments. It is also possible to use empirical sampling (Conner et al., 2009) to continuously measure the behaviors of the subjects and evaluate their dynamic change processes to understand the real situation of mobile phone use in college students’ daily lives. On the other hand, this study adopted a cross-sectional design and could not examine the causal relationship between variables. Future research should use experimental studies to explore the effects of different levels of time perspective and self-control on MPA. In particular, intervention studies on college students’ time perspectives and self-control should be carried out to investigate the effects of different intervention schemes, providing empirical evidence for follow-up mental health work.