Shin and Johnson (1978) regarded life satisfaction (LS) as a person's overall assessment of the relative quality of his or her life, or it is a summative view of the individual’s life overall. In other words, it refers to the extent to which can function positively with respect to physical and psychosocial health (Li et al., 2022; Ryff & Singer, 2008). Research has revealed that life satisfaction plays a crucial role in buffering against health-compromising behaviors during the developmental transition period towards adulthood (Hoyt et al., 2012). For examples, past results indicated that students with higher scores on self-reported well-being reported more physical activities (Piqueras et al., 2011), less drug (e.g., marijuana) consumptions, and less risky, sensation-seeking behaviors (Grant et al., 2009). Conversely, lower life satisfaction is implicative of mental health or behavioral problems such as depression, hopelessness, loneliness, alcoholism, substance abuse, suicidal ideation, and physical inactivity (Grant et al., 2009; Hawton et al., 2013; Lai & Ma, 2016). In Hong Kong, although a past study using latent profile analysis revealed that 90% of university students belonged to the mentally healthy categories (Ma & Lai, 2018), contrary findings from another research published within the same year, however, indicated that the prevalence rates of varying degrees of depressive and anxiety symptoms reported by university students in Hong Kong could rise to 69% and 54% respectively (Lun et al., 2018). Even worse, these statistical figures might have escalated recently because of the mass social unrests against “Extradition Law” in 2019 and the threats of the COVID-19 pandemic and its accompanying mandatory policies of social restrictions (Choi et al., 2020; Li et al., 2021a). Therefore, investigating the mechanisms underpinning students’ life satisfaction may shed light on development of preventive strategies against students’ maladaptive health behaviors and mental health issues.

Roles of affective states on life satisfaction

Positive affect (PA) refers to the extent to which an individual experiences positive affects, not only including the state of being alert, inspired, determined, attentive and active, but also enjoyable engagement, full concentration, and high energy (Thompson, 2007). Negative affect (NA), conversely, which typically reflects distress, unpleasant engagement, and lethargy, is usually regarded as a low energy state (Thompson, 2007). The correlation between positive affect and negative affect are usually negative in sign, nonetheless, it can range from positive to near-zero to extremely substantially negative values across samples (e.g., Schimmack, 2007; Schimmack et al., 2002).

Subjective well-being typically comprises the components of (1) a high frequency of positive affective (Positive affect; PA) experiences, (2) a low frequency of negative affective (Negative affect; NA) experiences, and (3) a sense of satisfaction with one's life (i.e., life satisfaction) (Diener, 1984; Diener et al., 2017). Also, the first two components represent individual’s emotional and affective dimensions of the construct, while the latter one refers to the cognitive-judgmental aspect of the same construct (Diener et al., 1985). Several previous studies have suggested that affect and life satisfaction report are consistently found to be empirically distinct (Lucas et al., 1996; Luhmann et al., 2012). Their relationship has been extensively researched using longitudinal and experimental studies (e.g., Metler & Busseri, 2017), ecological momentary assessments (e.g., Jayawickreme et al., 2017), and cross-national studies (e.g., Kuppens et al., 2008). The linear relationship between affective states and life satisfaction varies across studies. While some studies have suggested there is a similar relationship between life satisfaction and the two affective dimensions (Durak et al., 2010; Lazić et al., 2019), others have revealed that the relationship between positive affect and life satisfaction is stronger than the relationship between negative affect and life satisfaction (Busseri, 2018; Gouveia et al., 2009).

Furthermore, many previous studies even show that positive and negative affect both contribute to life satisfaction (Jovanović & Gavrilov-Jerković, 2016; Kööts-Ausmees et al., 2013). In a sample from 46 countries, the effect of positive affect on life satisfaction was found to be significantly greater than the that of negative affect (Kuppens et al., 2008). A plausible explanation is that individuals refer to their affective experiences as a source of information when being asked to evaluate their lives (Kim-Prieto et al., 2005; Kuppens et al., 2008). According to the feeling-as-information theory, for instance, affective states have impacts on how individuals judge their life satisfaction, implying that their emotional states of happiness matter in evaluative judgment (Schwarz, 2011). Some other researchers suggest that pleasant core affect, which is derived from affective experience, is a key determinant of global life satisfaction (Davern et al., 2007; Veenhoven, 2000). Therefore, it is reasonable to suggest a model in which life satisfaction is an output of accumulated affective experience, with positive affect and negative affect contributing to life satisfaction independently (Busseri & Sadava, 2011; Kim-Prieto et al., 2005; Kuppens et al., 2008). Another study suggests that life satisfaction transcends the summation of different affective experiences. For instance, when being asked about their life satisfaction, individuals consider their overall life circumstances rather than just their affective states (Luhmann et al., 2012). To resolve this ambiguity, we therefore revisit the relationship between these three variables in the present study.

Self-control, affective states, and life satisfaction

Self-control (SC) is the trait capacity to adjust an individual's thoughts, feelings, and behavior in order to conform to social norms or to sustain goal-directed behavior (De Ridder et al., 2012; Schmeichel & Zell, 2007). Numerous studies have consistently shown that individuals with high trait self-control do have a wide variety of adaptive and desired life outcomes, including better physical and psychological health, better salary, greater job satisfaction, and higher academic achievement (Allemand et al., 2019; Converse et al., 2016, 2018; Galla & Duckworth, 2015). Among university students in China, a recent study showed that self-control explained the highest proportion of variance of freshmen's adaptation outcomes over time (Li et al., 2021b). On the other hand, low self-control is linked to a wide range of personal and societal problems, including academic failure and underachievement, procrastination, substance abuses, compulsive buying, maladaptive adjustments, aggressive behaviors, heath-compromising diets, and lack of physical exercises (De Ridder et al., 2012; Duckworth & Seligman, 2005; Friese & Hofmann, 2009; Gottfredson & Hirschi, 1990; Moffitt et al., 2011; Patton et al., 1995; Restubog et al., 2010; Vohs & Faber, 2007).

Not only does self-control explain how individuals cope with adversity and strive for optimal functioning, but it also addresses dysfunctions as well. In other words, previous study has suggested that self-control can act as a buffer against impaired mental health functioning and enable university students to maintain high levels of psychological well-being (i.e., sense of environmental mastery, self-acceptance and personal growth) despite elevated perceived stress (Durand-Bush et al., 2015). Ryff and Keyes (1995)stated that students with high self-regulation capacity were more apt to manage daily, academic tasks and multiple roles they had while maintaining a positive attitude and striving to actualize their full potentials through continued learning. Conversely, Vohs and Baumeister (2004) argued that students with compromised life satisfaction and well-being usually had difficulties in effectively managing their behaviors, thoughts, and feelings that were promotive of achieving desirable goals. In other words, these students might possess poorer self-regulation competence that could maintain stable mental health and well-being in the face of adversity (e.g., academic and life stressors).

Several researchers have discovered the associations between self-control abilities and the ability to reach high levels of subjective well-being (Gavriel-Fried et al., 2018; Ronen et al., 2016; Rosenbaum & Ronen, 2013). For some studies, self-control appears to emerge from positive affect since the pleasant mood results in a stable basis for enabling capacities to achieve goals (Baumeister et al., 1998; Tice et al., 2007). Furthermore, across different age groups and regions, self-control is also found to be negatively correlated with stress and anxiety (Cheung et al., 2014; Li et al., 2016; Nielsen et al., 2019, 2020). Self-control abilities, on the other hand, are found to predict higher positive affect rather than reduced negative affect (Hamama et al., 2013; Orkibi et al., 2018). In this vein, self-control is also believed to be positively associated with one’s overall level of life satisfaction.

Besides positive affect and negative affect, several repeatable research studies have discovered that self-control is also associated with better level of life satisfaction (Cheung et al., 2014; Hofmann et al., 2014; Li et al., 2016; Tu & Yang, 2016). Some researchers, however, believe that excessive exertion of self-control may result in diminished life satisfaction (Grant & Schwartz, 2011). According to this viewpoint, there is an inverted-U link between self-control and life satisfaction. Specifically, it represents that as self-control improves, life satisfaction improves, but as self-control reaches a specific level, further self-control only leads to poorer life satisfaction. Contrary to Grant and Schwartz’s (2022) findings, Situ et al. (2016) revealed that high trait self-control could be negatively associated with emotional and behavioral problems, but no U-shaped patterns were observed as scores on trait self-control further increased. Wiese et al. (2018) have scrutinized this problem and revealed that, after adjusting for various psychological confounds, self-control capacity can positively predict life satisfaction and emotional health in a series of experiments. Nevertheless, Friese et al. (2017) have discovered that direct intervention to improve self-control is not as successful as previously found, implying that any attempts to improve one's self-control in order to increase life satisfaction may not be the best method. Therefore, elucidating the underlying mechanisms of the association between self-control, affective states and the life satisfaction will therefore give insights about ways to improve individuals’ overall well-being.

According to Hofmann et al. (2014), higher self-control is associated with better life happiness. In their study, they found that the effect of self-control on life satisfaction is fully mediated by both positive affect and negative affect, suggesting that self-control can lead to better life satisfaction by mitigating interpersonal conflicts and negative emotions (Hofmann et al., 2014). Their findings to some extent may explain why individuals with heightened levels of self-control tend to excel at controlling negative emotions, resolving disputes, and balancing their performance in daily life. Furthermore, a variety of factors have also been identified to explain the association between self-control and life satisfaction, such as promotion-oriented, prevention-oriented, and positive coping propensities (Cheung et al., 2014; Li et al., 2016). As individuals’ striving/coping behaviors may originate from different inner motivations, in this study, we will focus on the moderating role of fear of negative evaluation on the strength of effect of trait self-control on life satisfaction mediated by states of positive affect and negative affect.

Fear of negative evaluation as a moderator

Watson and Friend (1969) first introduced fear of negative evaluation (FNE), which was defined as the “apprehension about other’s evaluation, distress over their negative evaluations and the expectation that others would evaluate negatively.” To put it another way, fear of negative evaluation is the fear of being negatively judged by the presence of others. Findings in both samples of clinical populations and university students suggest that the fear of negative evaluation, presumably stemmed from poor self-evaluation, is hitherto the best-established cognitive-behavioral component of social anxiety, with which they have a strong positive correlation (Cheng et al., 2015; Iqbal & Ajmal, 2018; Leary & Kowalski, 1997; Weeks & Howell, 2012; Weeks et al., 2010, 2012). Social anxiety and affective states research studies have shed insights on the link between fear of negative evaluation and emotional health.

According to Kashdan et al. (2011), those classified as individuals with high social anxiety require more self-control than their counterparts in many social encounters, and the demands for self-control will be particularly elevated during stressful or challenging social interactions. In other words, those who were awkward and socially withdrawn ruminated over a lot about potential ostracism in society (Blackhart et al., 2015). As a result, they typically engaged in avoidance behaviors, including excessive nodding, or looking for validation, expressing very little, or trying to divert attention by questioning others (Clark & Wells, 1995). Similarly, studies have shown that when people focused their attention on controlling their fear or managing their impressions, they frequently exhibited impairments in other goal-directed actions (Finkel et al., 2006; Kashdan et al., 2010; Vohs et al., 2005). These actions could put more strain on the ability to sustain self-control in the long term.

Furthermore, the attenuation of self-control tends to diminish the favorable outcomes and experiences that socially anxious people might encounter in social contexts, aggravating social anxiety symptoms (Kashdan et al., 2011). That is, when people with social anxiety were attempting at regulating their fear and were engaging in behaviors which were intended to prevent social rejection, they failed to divert their attention to the potential benefits of those fruitful social engagements. Thus, they felt less connected to others and have less positive affect following a satisfying social contact (Kashdan et al., 2011) Such feelings might fuel social anxiety and reinforce it, leading to a vicious spiral of actions. Apart from studies targeting on social anxiety, other researchers observed that individuals with greater degree of fear of negative evaluation also have a higher level of negative affect and a lower level of positive affect (Levinson et al., 2013; Weeks & Howell, 2012; Weeks et al., 2010; Wolniewicz et al., 2018). Therefore, with reference to the mediational pathways introduced, we propose a moderated mediation process in which fear of negative evaluation serves as a moderator moderating the strength of association between self-control and life satisfaction via affective states in the current study.

The present research

Considering the distinction between state and trait psychological construct and the necessity of maintaining consistency in operational definition (Imhoff et al., 2014), self-control in the present study is conceptualized as a trait which tends to remain stable across time. Besides, plenty of studies have established links between self-control and the capacity to achieve high levels of life satisfaction, and Hofmann et al. (2014) demonstrated that the effect of self-control on life satisfaction is mediated by both positive affect and negative affect, we therefore hypothesized that positive affect (H1a) and negative affect (H1b) would respectively mediate the positive relationship between self-control and life satisfaction. Details of these hypotheses are as follows:

  • H1a: Self-control is positively correlated with positive affect, which in turn is positively correlated with life satisfaction.

  • H1b: Self-control is negatively correlated with negative affect, which in turn is negatively correlated with life satisfaction.

Moreover, people who have a higher level of fear of negative evaluation also have a higher level of negative affect and a lower level of positive affect. However, to the author’s best knowledge, no previous study has included fear of negative evaluation into consideration when examining similar associations, the present study contributed to the research literature by further hypothesizing that fear of negative evaluation would moderate the mediational effects of positive affect (H2a) and negative affect (H2b) on the relationship between self-control and life satisfaction. Details of these hypotheses are as follows:

  • H2a: When fear of negative evaluation increases, the mediation effect of positive affect connecting self-control and life satisfaction becomes weaker.

  • H2b: When the score of fear of negative evaluation increases, the mediation effect of negative affect connecting self-control and life satisfaction becomes stronger.

Therefore, the present study could potentially enrich the existing literature by testing and comparing the indirect effects of trait-control on life satisfaction via positive affect or negative affect in Chinese university students, as well as further examining the moderating effect of fear of negative evaluation on the two mediation paths.

Method

Participants and procedure

The present sample consisted of a total of 250 undergraduate students in Hong Kong (182 females; Mage = 20.56, SD = 2.39). Subjects were recruited via email invitations and poster advertisements, using convenience sampling method, and all subjects were asked to complete an online survey tapping various psychological constructs. Prior to the start of the study, informed consent was sought from each participant, and confidentiality was assured.

Measures

Self-control

The 13-item Brief Self-Control Scale (BSCS), which is a shortened version of the original 36-item Self-Control Scale, aims to assess individuals’ general trait self-control, with the specific focus on the behavioral aspect including the abilities to break maladaptive habits and preserve through a task (Tangney et al., 2004). Respondents rated each item on a 5-point Likert scales ranging from 1 (strongly disagree) to 5 (strongly agree). Sample items are, “I am able to work effectively toward long-term goals” and “pleasure and fun sometimes keep me from getting work done (reversed item).” Higher total score reflects greater trait self-control. The Cronbach’s alpha of the BSCS was 0.83 in the present study.

Life satisfaction

The 5-item scale of life satisfaction was extracted from the Satisfaction with Life Scale (SWLS) (Diener et al., 1985). SWLS is an instrument which aims to measure individuals’ global cognitive judgements of one’s life satisfaction through which individuals compare their life circumstances with a self-imposed standard. A sample item is, “So far I have gotten the important things I want in life.” Responders rated each of these statements on a 7-point Likert point ranging from 1 (strongly disagree) to 7 (strongly agree). This scale had good reliability, with an internal consistency reliability Cronbach’s alpha of 0.89 in the present study.

Positive affect and negative affect

The 10-item short form of the Positive and Negative affect schedule (I-PANAS-SF) was used to measure individuals’ positive and negative affect over the previous month (Thompson, 2007). While items in the Positive Affect (PA) subscale reflect the extent to which a person is in the state of high energy, full concentration, and pleasurable engagement, items in Negative Affect (NA) assess subjective distress and unpleasable engagement that subsumes a variety of aversive mood states, including anger, contempt, disgust, guilt, fear, and nervousness. In this scale, respondents rated to what extent they felt with reference to 10 different emotion-relevant adjectives on a scale of 1 (little or none at all) to 7 (a lot). Cronbach’s alphas were 0.60 and 0.83 for positive affect (5 items) and negative affect (5 items) subscale, respectively. By deleting an inconsistent item from the positive affect subscale (i.e., the item measuring “alert”), the alpha estimate of the 4-item positive affect subscale increased to 0.67.

Fear of negative evaluation

The 12-item Brief Fear of Negative Evaluation Scale (BFNE) was adopted to assess the degree to which people experience apprehension at the prospect of being evaluated negatively (Leary, 1983). Although contemporary researchers reveal that fear of negative evaluation precedes the development of social anxiety disorder (or social phobia), fear of negative evaluation is still considered as the cognitive symptoms of this psychopathology since this measurement taps the core features based on the cognitive models of social anxiety. Participants rated to what extent these could statements describe their thinking propensity concerning the possibility of being negatively evaluated by others on a scale of 1 (strongly disagree) to 5 (strongly agree). A sample item is, “I am frequently afraid of other people noticing my shortcomings.” Higher score on BFNE reflects more individual’s heightened propensity to avoid potentially threatening social comparison information to a greater extent and indicate that individual easily feels worse about receiving negative evaluation. The Cronbach’s alpha was 0.90 in the present study.

Analytical strategies

All participants provided complete demographic (i.e., age, gender) and survey data. First, Pearson’s correlations were computed to evaluate the bivariate associations among continuously measured variables including self-control, life satisfaction, positive and negative affect, as well as fear of negative evaluation. Next, structural equation modeling (SEM) was adopted to test the proposed mediation model (ME) and the first stage moderated mediation (moME) model, respectively. Traditionally, these mediation, moderation and conditional path models are tested by the regression-based approach, such as the PROCESS macro developed by (Hayes, 2017). However, one limitation about the regression-based approach is that the method does not provide any information regarding the overall model goodness-of-fit. As a result, it is difficult for applied researchers to evaluate to what extent their hypothesized models fit the sample data. To overcome this difficulty, the present study adopted the Variable System (VS) approach, which is an integrated methodological framework for the analysis of conditional path based on SEM (Kwan & Chan, 2018). After depicting the relationships among variables, we applied the VS 1.3 software program (Chan et al., 2016) to (i) transform the conceptual (i.e., proposed) model into a statistical system of variables (i.e., working model) in SEM, and (ii) fit the working model to the sample data under their SEM framework. Moreover, VS 1.3 software program also provides various model goodness-of-fit indices. The following criteria were computed to test whether the proposed ME and moME models have an acceptable fit, namely the Chi-square goodness of fit statistic, root-mean-square errors of approximation (RMSEA) < 0.08 (Steiger, 1980), standardized root-mean-square residual (SRMR) < 0.05 (Hu & Bentler, 1999; Iacobucci, 2010; Kline, 2015; Schermelleh-Engel et al., 2003), comparative fit index (CFI) > 0.90 (Bentler, 1995), Tucker–Lewis index (TLI) > 0.95 (West et al., 2012). Also, all the direct, indirect effects (i.e., path coefficients) in both proposed models were computed by the 95% bias-corrected bootstrap confidence interval (BCI) based on 1000 bootstrap samples.

Results

Descriptive statistics and bivariate correlations

The means, standard deviations and the intercorrelations among all measures are summarized in Table 1. Correlational analyses revealed that self-control correlated positively with life satisfaction (r = 0.27, p < 0.001) and positive affect (r = 0.37, p < 0.001), whereas it correlated inversely with negative affect (r = -0.33, p < 0.001) and fear of negative evaluation (r = -0.43, p < 0.001). On the other hand, fear of negative evaluation correlated negatively with life satisfaction (r = -0.27, p < 0.001), positive affect (r = -0.21, p < 0.01), and correlated positively with negative affect (r = 0.54, p < 0.001). Given these significant correlations, subsequent ME and moME analyses could be performed.

Table 1 Means, standard deviations and the intercorrelations among variables of the study

Testing the mediation model

Overall, the goodness of fit indices for the proposed model were satisfactory (Fig. 1), with χ2 (1, n = 250) = 0.20, p > 0.05, RMSEA < 0.001, SRMR = 0.01, CFI = 1.00, TLI = 1.04. In this model, self-control was significantly associated with both positive affect (\({\widehat{a}}_{1}\) = 0.19, p < 0.001) and negative affect (\({\widehat{a}}_{2}\) = -0.27, p < 0.001), and both positive affect (\({\widehat{b}}_{1}\) = 0.50, p < 0.001) and negative affect (\({\widehat{b}}_{2}\) = -0.41, p < 0.001) were significantly associated with life satisfaction. Furthermore, as hypothesized, the indirect effect of self-control on life satisfaction via positive affect was significant (\({\widehat{a}}_{1}{\widehat{b}}_{1}=.10\), SE = 0.03, p < 0.001, 95% BCI [0.05, 0.14]), whereas the indirect effect of self-control on life satisfaction via negative affect was also significant (\({\widehat{a}}_{2}{\widehat{b}}_{2}=.11\), SE = 0.03, p < 0.001, 95% BCI [0.06, 0.16]). When these mediators were taken into account statistically, the direct effect of self-control on life satisfaction became non-significant (\(\widehat{c}=.04\), p > 0.05), suggesting that positive and negative affects fully mediated the effect of self-control on life satisfaction. Furthermore, a closer examination revealed that the indirect effect of the self-control on life satisfaction via negative affect was larger than the indirect effect of the self-control on life satisfaction via positive affect by 0.016, although the difference was not significant (p > 0.05).

Fig. 1
figure 1

Path coefficients for the mediation model. *** p < 0.001

Overall, these results were supportive of H1a and H1b, and the occurrence of these significant mediational pathways served as the basis for the follow-up moME analysis.

Testing the moderated mediation model

Similar to the ME model, the goodness of fit indices for the proposed moderated mediation (moME) model were also satisfactory (Fig. 2), with χ2 (3, n = 250) = 1.50, p > 0.05, RMSEA < 0.001, SRMR = 0.01, CFI = 1.00, TLI = 1.03. In this model, self-control was significantly associated with both positive affect (\({\widehat{a}}_{1}\) = 0.18, p < 0.001) and negative affect (\({\widehat{a}}_{2}\) = -0.10, p < 0.05), and both positive affect (\({\widehat{b}}_{1}\) = 0.50, p < 0.001) and negative affect (\({\widehat{b}}_{2}\) = -0.41, p < 0.001) were significantly associated with life satisfaction. Interestingly, the interaction effect between fear of negative evaluation and self-control was negatively associated with positive affect (\({\widehat{d}}_{1}\)= -0.01, p < 0.01), implying that the effect of self-control on positive affect was weaker when the score of fear of negative evaluation was higher, and the further interaction effect was illustrated in Fig. 3. On the other hand, the interaction between fear of negative evaluation and self-control did not produce similar effect on negative affect (\({\widehat{d}}_{2}\)= 0.01, p > 0.05). Furthermore, results from moME analysis revealed that the pathway between self-control and life satisfaction mediated by positive affect was moderated by fear of negative evaluation (\({\widehat{d}}_{1}{\widehat{b}}_{1}\)= -0.01, SE = 0.002, p < 0.01, 95% BCI [-0.010, -0.001]), suggesting that this indirect pathway became stronger when the score of fear of negative evaluation was lower, vice versa. However, fear of negative evaluation failed to moderate the pathway between self-control and life satisfaction mediated by negative affect (\({\widehat{d}}_{2}{\widehat{b}}_{2}\)< -0.001, SE = 0.002, p > 0.05, 95% BCI [-0.006, 0.002]). In short, results of moME analysis supported H2a, but not H2b.

Fig. 2
figure 2

Path coefficients for the moderated mediation model. * p < 0.05, ** p < 0.01, *** p < 0.001

Testing alternative models

Given the nature of cross-sectional study design, there was a possibility that competing models fitting the data could exist. According to the broaden-and-build theory, individuals’ experiences of positive affects can expand the repertoire of their momentary thought-action, resulting in the effect of building up an individual’s physical, intellectual, and social resources (Fredrickson, 2004, 2013). As a result, individuals tend to have broader practical solutions and acting abilities for their problems, leading to a higher sense of self-control. To elucidate that the proposed moderated mediation (moME) model (Fig. 2) fits the data best, two alternative models were computed. The first alternative model was similar to that of the model in Fig. 2, except the independent (Self-Control) and dependent (Life Satisfaction) variables were reversed in their positions. The second alternative model was further constructed based on the first alternative model, and the moderating effects were placed between the relationships of the mediators (Positive and Negative Affects) and dependent variable (Self-Control). Details of the goodness-of-indices of the alternative models are shown in Table 2. With reference to the comparisons of the goodness-of-fit indices derived from both alternative models to the proposed moME, the current moME model fitted the data better and was therefore retained. Details of testing the alternative models were shown in Appendix.

Fig. 3
figure 3

Graphical Illustration of the Moderating Effect of Fear of Negative Evaluation (FNE) on the Relationship Between Self-Control (BSCS) and Positive Affect (PA). Note: Scores on Fear of Negative Evaluation were categorized into high (+ 1 SD), medium and low (-1 SD) groups

Table 2 Summary of the goodness-of-fit indices for the proposed and alternative models

Discussion

First, findings obtained from the mediation analysis in the present study revealed the role of positive affect and negative affect as the mediator between self-control and life satisfaction, supporting the hypothesis H1a that self-control was positively associated with positive affect, which in turn was positively associated with life satisfaction, and hypothesis H1b that self-control was negatively associated with negative affect, which was in turn negatively associated with life satisfaction. This research discovery has sparked interest in studies on self-control, positive affect, negative affect, life satisfaction, and subjective well-being. Several studies have discovered a relationship between self-control and the ability to reach high levels of subjective well-being (Gavriel-Fried et al., 2018; Ronen et al., 2016; Rosenbaum & Ronen, 2013). Furthermore, self-control abilities have been found to predict higher levels of positive affect, as well as to be linked with or even predict higher levels of life satisfaction and emotional health (Cheung et al., 2014; Hofmann et al., 2014; Li et al., 2016; Tu & Yang, 2016; Wiese et al., 2018). Besides, self-control is negatively associated to stress and anxiety (Cheung et al., 2014; Li et al., 2016; Nielsen et al., 2019, 2020). On the other hand, Grant and Schwartz (2011) suggested an inverted-U relationship between self-control and life satisfaction, which suggests that excessive effort of self-control may result in decreased life satisfaction. However, the range of the self-control in the current study may not be wide enough to test the inverted-U relationship.

The follow-up moME revealed the moderating role of fear of negative evaluation in the self-control → positive affect → life satisfaction pathway, supporting the hypothesis H2a that when the score of fear of negative evaluation is higher, the mediation effect of positive affect connecting self-control and life satisfaction becomes weaker, vice versa. This study's findings confirm the previously studied relationship between positive affect and social anxiety, or fear of negative evaluation. The possible explanations for the association between social anxiety and self-control include distribution of attentional resources during and/or after social engagement, incentives to engage in self-control during/after social interaction, and possibly ruminations. Several studies have found that people who were socially anxious were more likely to engage in excessive and persistent ruminations(Chen et al., 2013; Fehm et al., 2007; Jose et al., 2012; Kocovski & Rector, 2007; McEvoy & Kingsep, 2006; Nepon et al., 2011; Rachman et al., 2000; Vassilopoulos, 2004). According to Clark and Wells (1995), those with greater levels of social anxiety tended to demonstrate ruminative cognitive styles, as featured by preoccupation of the negative parts of their interpersonal relationships and their perceived social difficulties. Alike experience avoidance, being overwhelmed by the thoughts of undesirable social encounters might attenuate individuals’ sustainability self-control over time.

Interestingly, contrary to the results obtained from the mediation analysis, findings from the moME did not reveal a significant moderation effect of fear of negative evaluation on the self-control → negative affect → life satisfaction pathway, and hypothesis H2b was therefore not supported. From the evolutionary perspective, Hofmann et al. (2004) has suggested that “fight or flight” are the predominant reactions to anxiety. Despite the difficulty in defining anger, it has been argued that anger involves the interpretation of any insult against oneself or interference with one’s cherished values and pursuit of intended goals (Berkowitz & Harmon-Jones, 2004). Past research has provided some evidence that the perceived exposures to social rejection and shame, the constructs which largely overlap with social anxiety, are associated with elevated anger and aggression (Leary et al., 2006; Tangney et al., 1996; Twenge et al., 2001). In this vein, the higher proneness to fear of negative evaluation may not be accompanied by the emotion of sadness, presumably explaining the non-significant moderated effect of fear of negative evaluation on the relationship between self-control and negative affect. Furthermore, a close examination of correlational analyses revealed that scores on fear of negative evaluation and negative affect were highly overlapped (i.e., r = 0.54), implying the occurrence of multicollinearity which may diminish the significance of statistical findings.

Research implications

In line with several previous studies, research findings from the present study suggest that self-control capacity plays a vital role in university students’ mental health functioning and life satisfaction (Durand-Bush et al., 2015; Park et al., 2012). In details, self-control during the undergraduate studies could possibly facilitate students to proactively manage their intended goals, mental health functioning, and shift from being troubled, at-risk, or frustrated to being well-adjusted, adaptive, and even resilient in the face of academic and life stressors. Furthermore, the originality of the present study lied in the inclusion of fear of negative evaluation as a potential factor that moderated the mediation path of self-control on life satisfaction through affective states. Several past studies from the social cognitive perspectives support the notion that the apprehension of being evaluated negatively from others, which is found to rise steadily from childhood to adulthood, is the core feature of social anxiety (Rapee & Heimberg, 1997; Schreier & Heinrichs, 2010; Wells et al., 1995). Therefore, to optimize the treatment outcomes, current psychosocial intervention modalities should adopt a whole-person, holistic approach which addresses both students’ behavioral (i.e., self-control, time management, prioritizing of tasks) and socio-emotional (i.e., social anxiety, adaptability to other classmates within the university context) dimensions.

Limitations and further research

Despite the research implications, the present study also had its limitations. First, as this study adopted a self-reported cross-sectional survey design, cautions need to be taken when drawing causal relationships between the psychological constructs included in the analyses, especially when all the constructs were based on self-reported measures which were subject to participants’ recall bias. While it is plausible that self-control contributes to affective states and life satisfaction, we cannot conclude that self-control has a casual effect on these variables, the interrelations among variables can be conceptualized by alternative theoretical models and mechanisms. While the proposed moME model was chosen in the present study based on current literature, other competing models may fit the data too, leading to the possibility that alternative theoretical frameworks could also be applied to elucidate the results. To see if our hypothesized model is able to explain the data better, we compared it against two alternative models. Results indicated that both alternative models resulted in inadequate fit, and this may provide some support to the validity of our model. Furthermore, according to Hamaker (2012), cross-sectional data analysis cannot disentangle between-subjects relationships from short-term within-subjects relationships. In greater details, even a positive correlation between trait self-control and life satisfaction were found in the present study, it was unknown whether individuals had higher self-control would lead them to higher life satisfaction (a within-subjects effect) or people who on average had higher self-control also tend to experience greater life satisfaction on average (a between-subjects effect). Therefore, an important avenue for further research is to conduct longitudinal study with more advanced analytical strategies (i.e., latent growth modeling, cross-lagged panel analysis) so that the directional dynamic interplays among variables across time may be examined.

Second, although the Brief Self-Control Scale (BSCS) provides information on students’ overall self-control capacity, it does not explicitly inform us which aspects of self-control can be related to psychological health and life satisfaction (Lindner et al., 2015). A solution to this limitation is to adopt validated scale supporting the multi-faceted, multidimensional nature of self-control (Nilsen et al., 2020). Third, the number of samples collected via convenience sampling method from the university students was not large enough, and there was major unequal distribution of male and female subjects, thus findings from the present study may not be generalizable to other populations.

Conclusion

In a nutshell, findings from the present study conducted among university students in Hong Kong supported the hypotheses that positive affect (H1a) and negative affect (H1b) are the full parallel mediators between self-control and life satisfaction in the mediation analysis. Subsequent moderated mediated analysis revealed fear of negative evaluation as the moderator (H2a) influencing the strength of indirect effect from self-control to life satisfaction mediated by positive affect (i.e., self-control → positive affect → life satisfaction). However, fear of negative evaluation has no moderating role when considering the self-control → negative affect → life satisfaction pathway. The thesis statement of this survey study is that although self-control has been found to be correlated with better life satisfaction, the strength of association between self-control and positive affect would be diminished when the score on fear of negative evaluation is high, informing researchers and mental health practitioners the clinical importance of targeting both self-regulation capacity and social anxiety when working with university students.