Motivation and Emotion

, Volume 37, Issue 3, pp 413–423

Self-control mediates the link between perfectionism and stress

Authors

    • Chair for Social and Economic PsychologyZeppelin University
  • Ute C. Bayer
    • University of Konstanz
Original Paper

DOI: 10.1007/s11031-012-9321-6

Cite this article as:
Achtziger, A. & Bayer, U.C. Motiv Emot (2013) 37: 413. doi:10.1007/s11031-012-9321-6

Abstract

The relationship between perfectionism and stress is well-established. Recent research has focused on identifying the mediators of this link. Starting from a multidimensional perspective on perfectionism, we investigated the role of self-control and found it to be a mediator between perfectionism and stress in a sample of university freshmen. Further, perfectionistic concerns (i.e., discrepancy; Slaney et al. 2001) were positively correlated with stress, whereas perfectionistic strivings (i.e., high standards; Slaney et al. 2001) were negatively correlated with stress. Practical implications regarding overcoming maladaptive perfectionism are discussed.

Keywords

PerfectionismSelf-controlCopingStressFreshmen

Introduction

The present study combined two strands of the literature to examine the determinants of psychological adjustment, and, in particular, responses to stress. On the one hand, research on perfectionism has examined the influence of this personality disposition on psychological adjustment. On the other hand, a separate strand of the literature has recently pointed out the importance of self-control as a psychological resource influencing psychological adjustment. We combined both approaches by investigating self-control as a possible mediator of the perfectionism-stress link.

Perfectionism is characterized by the setting of high performance standards and striving for flawlessness, accompanied by tendencies to extreme critical evaluations of one’s own behavior (Frost et al. 1990; Hewitt and Flett 1991; Flett and Hewitt 2002). Most researchers have focused on direct links between perfectionism and psychological adjustment (e.g., academic achievement: Accordino et al. 2000; Rice and Slaney 2002) and misadjustment (e.g., psychological distress and clinical disorders; Flett and Hewitt 2002). Recent research has started to identify mediators (e.g., maladaptive coping styles, rumination, cognitive emotion regulation) to explain mechanisms underlying the link between perfectionism and its negative consequences (e.g., O’ Connor et al. 2007; Rudolph et al. 2007; Park et al. 2010). The present study followed this line of research and investigated self-control as a possible mediator of the perfectionism-stress link.

The link between perfectionism and stress

Before turning to the role of self-control, we briefly review the literature on perfectionism and psychological adjustment, with special emphasis on stress. A first distinction is drawn in research on “normal perfectionism” (in contrast to “neurotic perfectionism”) between perfectionistic strivings and perfectionistic concerns (Hamacheck 1978; Stoeber and Otto 2006). Perfectionistic concerns capture negative feelings of discrepancies between one’s expectations and one’s results, concerns over mistakes, doubts about actions, and concerns of socially prescribed perfectionism. In contrast, perfectionistic strivings capture facets of perfectionism that are linked to perfectionistic standards, such as high personal standards, self-oriented striving for perfectionism, and striving for excellence.

Perfectionistic concerns were repeatedly observed to be related to various forms of maladjustment (Flett and Hewitt 2002; Chang et al. 2004; Rudolph et al. 2007). For instance, a study by Flett et al. (2002) showed that individuals with high levels of socially prescribed perfectionism reported a high level of rumination response orientation and experienced more cognitive intrusions in response to stressful events than individuals with low levels of socially prescribed perfectionism. Moreover, discrepancy (see Slaney et al. 2001) was positively related to an external locus of control (Rotter 1954) meaning that individuals with high levels of discrepancy believe that their fate is s determined by chance or outside forces beyond their own control (Periasamy and Ashby 2002). This can lead to feelings of helplessness and loss of control about one’s own life.

Moving closer to our object of study, there is abundant empirical evidence implying a link between perfectionistic concerns and stress (e.g., Dunkley et al. 2003; Stoeber and Rennert 2008; Hill et al. 2010; Rice and Van Arsdale 2010; Tashman et al. 2010). For instance, Dunkley et al. (2003) reported that the relation between self-criticism, which is an important element of perfectionistic concerns, and daily affect can be explained by maladaptive tendencies as for instance avoidant coping and low perceived social support. Findings of this study also indicated that individuals high in self-critical perfectionism were especially emotionally reactive to stressors that implied possible failures, loss of control, and criticism from other people. Moreover, Dunkley et al. (2003) reported that strategies that are usually successful in coping with stress (e.g., problem-focused coping) were ineffective for people high in self-critical perfectionism compared to people low in self-critical perfectionism. In a later study, Dunkley et al. (2006) showed that self-criticism is one of the most robust predictors of daily stress and avoidance coping. Concerning the link between perfectionism and stress, Flett et al. (1995a) provided support for diathesis-stress models. These models maintain that when perfectionists are exposed to stress they become vulnerable to symptoms of depression.

Perfectionistic strivings were also shown to be related to psychological adjustment in general. For instance, it was observed that students with high levels of high standards (measured by the APS-R; Slaney et al. 2001) showed higher academic performance indicated by grade point average (GPA) than students with low levels of high standards (e.g., Accordino et al. 2000; Blankstein et al. 2008). In accordance with these findings, Klibert et al. (2005) reported that self-oriented perfectionism correlated positively with achievement motivation (see also Stoeber and Rambow 2007). Concerning these adaptive aspects of perfectionism it is hypothesized that individuals with high levels of perfectionistic strivings focus on rewards and initiate performance in order to obtain praise and feelings of mastery and control (Burns et al. 2000).

In contrast to perfectionistic concerns, however, the empirical support for a link between perfectionistic strivings and stress is rather mixed. Some authors have not found a link between these concepts (e.g., Rice and Van Arsdale 2010), whereas others even reported that perfectionistic strivings correlate negatively with stress (e.g., Stoeber and Rennert 2008; Hill et al. 2010; Tashman et al. 2010). The latter studies indicate that people with high levels of perfectionistic strivings might better be able to cope with stress than people with low levels (see Dunkley et al. 2000).

We are interested in the mechanisms linking perfectionism and stress. The relation between frequent perfectionistic thoughts and maladaptive cognitive emotion regulation (e.g., rumination, self-blame, lack of positive reappraisals) was observed by Rudolph et al. (2007). This finding implied that deficits in cognitive emotion regulation might be responsible for the positive link between socially prescribed perfectionism and psychological distress. Moreover, there is evidence on the mediating role of coping style in the link between perfectionism and stress (e.g., Dunkley et al. 2000; Hewitt and Flett 2002; Hill et al. 2010). Thus, it is well documented that the various aspects of perfectionism are associated with different coping styles and it is assumed that coping styles might be mechanisms underlying the link between perfectionism and stress. Specifically, perfectionist concerns were frequently shown to be associated with avoidance coping whereas perfectionistic strivings were associated with active coping (e.g., Dunkley et al. 2000, 2003; Stoeber and Rennert 2008).

The link between self-control and stress

Without referring to perfectionism, research on stress has recently started to investigate the role of self-control as a psychological resource that helps to deal with demanding situations as for instance at the workplace (e.g., Schmidt et al. 2012) or in academic contexts (Oaten and Cheng 2005, 2006). As explained above, earlier research on perfectionism provided insights into motivational aspects of perfectionism. We intended to combine approaches on self-control and perfectionism in order to explore whether self-control mediates the link between perfectionism and stress.

Self-control refers to attempts to override or alter one’s dominant response tendencies and to interrupt undesired behaviors (Baumeister et al. 1994; Baumeister and Vohs 2004). Thus, breaking bad habits, resisting temptations, and keeping self-discipline reflect the ability of the self to control itself. It is argued that a lack of self-control is behind the majority of social and personal problems (Baumeister et al. 1994). There is empirical evidence of individual differences concerning self-control capability—some people are better able than others to control their thoughts and emotions, resist temptations, hold their temper and so on. In line with these considerations, self-control was conceptualized as a dispositional, trait-like construct that differs across individuals and that can be measured by questionnaires as for instance the Self-Control Scale (SCS) by Tangney et al. (2004).

Studies investigating individual differences in self-control revealed that high levels of self-control are linked to a broad range of positive outcomes. For instance, it was reported that goal achievement (Tangney et al. 2004; Duckworth and Seligman 2005), impulse control (Friese and Hofmann 2009), emotion regulation (Tangney et al. 2004) and the control of procrastination (Tice and Baumeister 1997) are strongly supported by peoples’ self-control capacity. Finally, high levels of self-control are associated with high interpersonal skills (Tangney et al. 2004).

Important for the present research were studies showing that self-control is negatively linked to stress. For instance, it was demonstrated that exercising self-control is associated with psychological strain and impaired well-being (Muraven et al. 1998; Muraven and Baumeister 2000; Schmidt et al. 2007). In line with this research, Oaten and Cheng (2005) found out that students at stressful times (i.e., during examination periods) often fail at self-control (e.g., breaking up with a diet, drinking more alcohol than usual, smoking more). Hence it can be concluded that managing stress requires self-control.

There is also evidence that self-control as a personality trait might be a protective factor (or psychological resource) that buffers stress. For instance, Bowlin and Baer (2011) reported that self-control as measured by the SCS (Tangney et al. 2004) correlated negatively with stress. Moreover, it was shown that high self-control demands at the workplace are linked to negative outcomes as burnout and absenteeism in a variety of professions (Diestel and Schmidt 2011), but that people high in self-control (measured by the SCS; Tangney et al. 2004) were better able to deal with these demands than people low in self-control (Schmidt et al. 2012). Taken together, there is a lot of empirical evidence showing that self-control is a protective factor against stress. As no study has combined research on self-control and perfectionism in order to test the influence of both personality traits on stress simultaneously before, we run the present study.

The link between self-control and perfectionism

Concerning perfectionism, Tangney et al. (2004) observed that self-control correlates positively with self-oriented perfectionism and negatively with socially-oriented perfectionism (measured by the MPS; Hewitt and Flett 1991). Tangney et al. (2004) did not explore whether self-control mediates or moderates the effects of perfectionism on important negative outcomes such as stress. This should be tested, however, as self-control might serve as a protective factor against stress (see above). Linking self-control and perfectionism, one might refer to Mills and Blankstein (2000) who reported that high personal standards were associated with adaptive work habits, goal striving, and high academic achievement (see also Accordino et al. 2000 concerning the positive link between high standards and work orientation—the desire to work hard and to do a good job). Mills and Blankstein (2000) also assumed that self-oriented perfectionism (and hence also high standards) should be related to strategies that enable perfectionists to meet their self-imposed high standards. Therefore self-oriented perfectionism is hypothesized to include both the tendency to set high standards for oneself and the motivational component of the need to strive to fulfill those standards (Hewitt and Flett 1991).

These assumptions were supported by findings suggesting that self-oriented perfectionism is positively related to achievement motivation (e.g., Flett et al. 1994), resourcefulness, striving, and perceived personal control (Flett et al. 1991, 1995a, b). For instance, Flett et al. (1995b) observed that planning behavior, time and effort management (i.e., examples of self-control behavior) correlated positively with self-oriented perfectionism in a sample of university students (see also Mills and Blankstein 2000). Similarly, Parker and Stumpf (1995) reported a positive correlation between the personal standards subscale of the FMPS (Frost et al. 1990) and conscientiousness (measured by the NEO-PI; Costa and McCrae 1992). The same is true for a recent study by Dunkley et al. (2012) in which personal standards correlated positively with conscientiousness (and achievement motivation).

Klibert et al. (2005) reported that self-oriented perfectionism is related to traits like self-esteem, perceived control (measured by the subscale Control of the Hardiness Scale [HS]; Bartone et al. 1989), and achievement motivation (measured by the Mehrabian Achieving Tendency Scale; Mehrabian 1994) in college students. For the present research it is important to note that the subscale Control of the HS (Bartone et al. 1989) does not measure self-control in the sense of a general human function that helps to override one’s impulses (see Baumeister et al. 1994; Muraven and Baumeister 2000), but rather feelings of autonomy and control over one’s future life.

Flett et al. (1995b) reported that self-oriented perfectionism correlated positively (r = .61, p < .01) with high personal standards of the FMPS (Frost et al. 1990) in a sample of university students. Taken together it can be argued that these motivational aspects (setting high achievement standards and striving for them) of perfectionism are responsible for the adaptive behavior (e.g., high achievement) often reported for self-oriented perfectionists.

In contrast to self-oriented perfectionism and perfectionistic strivings, socially prescribed perfectionism and perfectionistic concerns are thought to be associated with motivational deficits in general, a decreased level of intrinsic motivation, and irrational beliefs (Hewitt and Flett 1991). In accordance with these assumptions, Klibert et al. (2005) observed that socially prescribed perfectionism was negatively associated with perceived control (i.e., autonomy, control over one’s future life; Bartone et al. 1989; see above) and it also correlated negatively with achievement motivation (see also Flett et al. 1991). Moreover, in a study by Parker and Stumpf (1995) negative associations between perfectionistic concerns (measured by the FMPS; Frost et al. 1990) and conscientiousness (measured by the NEO-PI; Costa and McCrae 1992) were observed. Finally, there is evidence that self-criticism (which is positively associated with concerns about mistakes; see Dunkley et al. 2003) is negatively related to goal progress and that this link is mediated by self-efficacy and implementation intentions (Powers et al. 2012). For the negative link between self-criticism and goal progress see also Powers et al. (2011).

These negative motivational aspects of socially prescribed perfectionism and perfectionistic concerns might be responsible for misadaptive behavior that leads to negative outcomes (e.g., failing to succeed in school and university, experiencing high levels of stress, depression, and neuroticism) that are often reported in the literature on perfectionism (review by Stoeber and Otto 2006). The present study aimed at contributing to this line of research by exploring the role of self-control (measured as a personality trait by the SCS; Tangney et al. 2004) as a mediator of the links between high standards and stress and of the link between discrepancy and stress.

The present study

We intended to test whether self-control mediates the influence of perfectionistic strivings and perfectionistic concerns on stress. Therefore we measured high standards as a facet of perfectionistic striving and discrepancy as a facet of perfectionistic concerns by the Almost Perfect Scale-Revised (APS-R; Slaney et al. 2001) as this scale was validated on a sample of college students and frequently used in academic contexts (e.g., Accordino et al. 2000; Wang et al. 2007). In contrast to earlier studies (e.g., Parker and Stumpf 1995; Klibert et al. 2005) we intended to investigate self-control in accordance with research that defines it as an important general psychological resource characterized by the following domains: controlling thoughts, emotions, impulses, and performance (Baumeister et al. 1994). These dimensions are all included in the SCS (Tangney et al. 2004) as an overall index of self-control. Another reason to employ the SCS was that it has been proven to have a good reliability and validity in different contexts and samples (see Tangney et al. 2004, Studies 1 and 2; academic achievement: Duckworth and Seligman 2005; stress at work: Schmidt et al. 2012). Stress was measured by the Perceived Stress Scale (PSS; Cohen et al. 1983) as Oaten and Cheng (2005, 2006) also used this to scale to explore the link between self-control and stress. Moreover, it has also a high reliability and validity (Cohen et al. 1983).

We focused on stress among university freshmen because it is quite often observed that many students extremely suffer from stress in their first months at university (e.g., Van Yperen and Hagedoorn 2008). The reason for this might be that this situation is usually characterized by high self-control demands as it requires adapting to new life situations as for instance moving away from home, looking for a job, finding new friends, being confronted with new academic requirements and so on. Moreover, academic contexts are well suited to investigate correlates of perfectionism because this environment induces stress and is characterized by high performance demands (see Klibert et al. 2005). Another argument for testing our hypotheses on a sample of university students was that the link between self-control and stress has already been demonstrated for university students (e.g., Oaten and Cheng 2005, 2006).

When measuring students’ stress in their first 3 months at the university, we expected significant links between both aspects of perfectionism and stress. Specifically, we predicted a positive relation between discrepancy and stress, and a negative relation between high standards and stress. Moreover, it was expected that students with high levels of self-control (i.e., who have access to high self-control resources) will have less problems with adapting to the new and stressful situation of being a university freshmen and therefore should experience less stress than students with low levels of self-control.

Most importantly, the links between both facets of perfectionism and stress were predicted to be mediated by self-control. We assumed that students with high standards should better be able to deal with stress as this link is mediated by their high self-control capacities. In other words, they should be more able to strive for their self-imposed high academic standards for instance by employing time management strategies (see Klibert et al. 2005), organizing their schedule carefully and thereby reduce stress in their first 3 months than students with low levels of self-control. However, if one controls for self-control, the negative relation between high standards and stress should be reduced as high standards alone should not be helpful in dealing with stressful situations.

Students with high levels of discrepancy should be low in self-control and should therefore have more problems in dealing with stress, as for instance they are occupied with ruminations about possible failures (Rudolph et al. 2007), having a lack of organization and time management skills (see Klibert et al. 2005), having to deal with motivational deficits in general etc. If one controls for self-control in this case, the positive relation between discrepancy and stress should be reduced as we assumed that discrepancy leads to increased levels of stress via decreased levels of self-control. Finally, investigating stress implies controlling the influence of coping behavior. For this reason we measured coping behavior with the brief COPE (Carver 1997) and controlled for its influence on the links between perfectionism, self-control, and stress.

Method

Participants and procedure

Participants were 165 university freshmen (99 female; age: M = 20.61, SD = 2.75. They were enrolled in science (n = 65), humanities (n = 50), and law/economics (n = 50). In exchange for participating, they took part in a lottery in which they could win book vouchers (worth 10 Euros). Participants were given an internet link to the online questionnaires. They were informed that the study was on stress among university freshmen. The questionnaires were presented as follows: First, stress during the first 3 months at university was measured. Then, participants completed the self-control, perfectionism, and coping scales. Finally, participants answered some demographic questions (e.g., gender, age, high school grades).

Measures

Perceived stress

The Perceived Stress Scale (PSS; Cohen et al. 1983) is a 14-item measure of self-appraised life stress. Students were asked about life stress experienced over the last 3 months. The PSS measures stress or the degree to which life is considered as being unpredictable, uncontrollable, and overloaded. There is empirical evidence for the construct validity and reliability of this scale (Cohen 1988). Students were asked to rate the frequency of the items on a 5-point Likert-type scale ranging from 0 (never) to 4 (very often). In the present study, higher scores reflected more perceived stress in the last 3 months.

Self-control

The Self-Control Scale (SCS; Tangney et al. 2004) measures general self-control. It contains 36 items tapping different dimensions of self-control (e.g., achievement and task performance, impulse control, adjustment, and self-control in interpersonal relationships). Items are rated on 5-point Likert-type scales ranging from 1 (not at all like me) to 5 (very much like me). A sum score was used as an index of self-control, with higher ratings representing higher self-appraised self-control. Reliability and internal consistency were good (see Tangney et al. 2004).

Perfectionism

Perfectionism was measured with the brief form of Slaney et al.’s (2001) Almost Perfect Scale-Revised (APS-R) with three subscales: High Standards (7 items), Order (4 items), and Discrepancy (12 items). As in previous studies (see Stoeber and Otto 2006), only high standards and discrepancy scores were used for data analysis in the present study. The high standards subscale reflects the level of standards people set for themselves. The discrepancy subscale reflects the amount of distress felt concerning one’s own standards. Items were to be answered on a 7-point Likert-type scale ranging from 1 (strongly disagree) to 7 (strongly agree). Higher scores represent higher levels of perfectionism. This scale has been shown to exhibit acceptable test–retest reliability and construct validity (e.g., Mobley et al. 2005; Rice et al. 2007). The dimensions of perfectionism yielded good internal consistency in earlier research (e.g., see Rice et al. 2007 reported Cronbach’s α for discrepancy: .91; and for high standards: .86).

Coping

The brief COPE (Carver 1997) is a 28-item instrument designed to assess 14 coping strategies. Participants are asked to think how they generally cope with stressful situations, and to respond to each item on a 4-point scale from 1 (=I usually don’t do this at all) to 4 (=I usually do this a lot). The subscales yielded moderate to good internal consistency in earlier research (Cronbach’s α’s ranging from .50 to .90; see Carver 1997). We excluded the subscale mental disengagement from our data set due to weak internal consistency in earlier studies (Cronbach’s α = .24). A factor analysis by Kapsou et al. (2010) provided mixed results regarding the number of factors underlying the measure. However, these authors reported that the active-coping, planning, positive reframing, and acceptance subscales reliably loaded on the active coping factor in the study by Kapsou et al. (2010). Moreover, the mental disengagement and denial scales loaded on the avoidance factor. Due to this earlier research, we computed an active coping score for each participant by calculating an average score of the subscales described by Kapsou et al. (2010), and an avoidance coping score by calculating an average score of the subscales for denial and behavioral disengagement.

Results

Descriptive statistics

Table 1 presents means, standard deviations, correlations, and reliabilities of all measures. For perfectionism, the mean of the high standards subscale (Slaney et al. 2001) was higher than the mean of the discrepancy subscale (Slaney et al. 2001). For coping styles, the mean for avoidance coping was lower compared to the means of active coping (Carver 1997). For the SCS (Tangney et al. 2004), the mean indicated a quite high level of self-control in the present sample. The mean of the PSS was high compared to previous studies (see Cohen et al. 1983). Cronbach’s α’s reached from acceptable (active coping: α = .69) to excellent (discrepancy: α = .89) with the exception of Cronbach’s α for avoidance coping (α = .54). That means that not all items of the avoidance coping subscale (Carver 1997) measured avoidance coping. As the scale is frequently used in research on stress, we decided to include avoidance coping despite its low reliability. But of course one has to keep in mind that the internal consistency of this subscale was rather low in the present study.
Table 1

Means, standard deviations, correlations, and Cronbach’s alphas for all measures

  

1

2

3

4

5

6

1

High standards (APS-R; Slaney et al. 2001)

(.82)

     

2

Discrepancy (APS-R; Slaney et al. 2001)

.20*

(.89)

    

3

Active coping (COPE; Carver 1997)

.22**

−.08

(.69)

   

4

Avoidance coping (COPE; Carver 1997)

−.06

.41***

−.19*

(.54)

  

5

Self-Control Scale (SCS; Tangney et al. 2004)

.45***

−.35***

.22**

−.27***

(.84)

 

6

Perceived Stress Scale (Cohen et al. 1983)

−.21**

.40***

−.22**

.35**

−.35**

(.82)

 

M

4.84

3.46

2.71

1.74

112.8

40.83

 

SD

1.00

1.08

.42

.49

15.55

7.04

Values in parenthesis are Cronbach’s coefficient alphas; * p < .05, ** p < .01, *** p < .001

Both perfectionism subscales correlated positively with each other. Most important, high standards correlated positively with self-control whereas discrepancy showed a negative correlation with self-control. Concerning stress we observed that both self-control and high standards correlated negatively with this variable, whereas discrepancy correlated positively with stress. High standards correlated positively with active coping whereas discrepancy did not correlate with active coping at all, but it correlated positively with avoidance coping. Self-control correlated positively with active coping and negatively with avoidance coping. Active coping was negatively associated with stress while avoidance coping revealed to be positively associated with stress.

Regression and mediation analysis

First, we ran a mediation test for self-control for the perfectionism-stress link, following a procedure suggested by Baron and Kenny (1986). The results showed that self-control fulfilled the three criteria for a mediation (see Fig. 1). First, stress was significantly predicted by perfectionism (\( R_{adjusted}^{2} \) = .16, F(2,162) = 16.40, p < .001)—high standards were associated with lower levels of stress (β = −.21, t = 2.01, p < .01) and high discrepancy was significantly related to higher levels of stress (β = .40, t = 5.41, p < .001). Moreover, both aspects of perfectionism were significantly linked to self-control (\( R_{adjusted}^{2} \) = .26, F(2,162) = 28.31, p < .001)—high standards were associated with higher levels of self-control, (β = .45, t = 6.51, p < .001) whereas high discrepancy was associated with lower levels of self-control (β = −.35, t = 5.12, p < .001). Second, self-control significantly predicted stress (\( R_{adjusted}^{2} \) = .12, F(1,163) = 22.14, p < .001), indicating that higher self-control was associated with lower levels of stress, (β = −.35, t = 4.71, p < .001). Third, when the influence of perfectionism on stress was controlled for, self-control was still a significant predictor of stress (β = −.22, t = 2.71, p < .01). However, when we controlled for self-control, the effect of high standards on stress was no longer significant (β = −.11, t = 1.41, p = .13) and the effect of discrepancy on stress was reduced (β = .32, t = 4.15, p < .001). Both results were supported by significant Sobel tests (high standards: z = 2.50, p < .01; discrepancy z = 2.38, p < .05) indicating a full mediation for perfectionistic standards and a partial mediation for discrepancy.1
https://static-content.springer.com/image/art%3A10.1007%2Fs11031-012-9321-6/MediaObjects/11031_2012_9321_Fig1_HTML.gif
Fig. 1

Mediation of the link between perfectionism and stress by self-control. Note. Beta weights in parentheses resulted from a regression analysis including the predictors (discrepancy and standards) and the mediator (self-control). *p < .05, **p < .01, ***p < .001

Additional analysis

We tested whether the mediating role of self-control on the perfectionism-stress link was also observed when controlling for active and avoidance coping. For this reason the same mediation analysis as before was computed again, but this time we included active coping and avoidance coping as control variables. We observed that perceived stress was significantly predicted by perfectionism (\( R_{adjusted}^{2} \) = .20, F(4,160) = 11.47, p < .001). High standards were negatively associated with perceived stress (β = −.15, t = 2.08, p < .05) whereas discrepancy was positively associated with perceived stress (β = .29, t = 3.70, p < .001) even when controlling for active coping (β = .20, t = 2.61, p < .01) and avoidance coping (β = −.12, t = 1.70, p < .01). Moreover, self-control was predicted by perfectionism (\( R_{adjusted}^{2} \) = .28, F(4,160) = 4.16, p < .001). High standards were positively associated with self-control (β = .41, t = 5.80, p < .001) whereas discrepancy was negatively associated with self-control (β = - .28, t = 3.74, p < .001) even when we controlled for active coping (β = .08, t < 1) and avoidance coping (β = −.11, t = 1.54, p = .13). Second, as expected, self-control was still negatively associated with perceived stress (β = −.25, t = 3.36, p < .01) even when controlling for active coping (β = −.11, t = 1.57, p = .12) and avoidance coping (β = .26, t = 3.59, p < .001).

Third, the effect of self-control on perceived stress was still significant (β = −.18, t = 2.24, p < .05) but the effect of high standards was no longer significant (β = −.08, t < 1) whereas the effect of discrepancy was still significant (β = .24, t = 2.06, p < .01) when controlling for active coping (β = −.11, t = 1.54, p = .13) and avoidance coping (β = .18, t = 2.35, p < .01). These results were supported by significant Sobel tests (high standards z = 2.07, p < .05, discrepancy z = 1.91, p < .05) indicating a full mediation for high standards and a partial mediation for discrepancy by self-control, even when controlling for coping behavior. In summary, we observed that self-control was a significant mediator of the link between perfectionism and stress even when controlling for individual differences in coping styles.

Discussion

The aim of the present study was to investigate the relation between perfectionism and stress by testing self-control as a possible mediator of this link. A significant perfectionism-stress link was observed in which discrepancy (Slaney et al. 2001) was positively associated with stress, and high standards (Slaney et al. 2001) were negatively related to stress. Moreover, self-control was found to be a significant negative predictor of stress (even when controlling for both high standards and discrepancy). Most importantly, however, self-control was shown to be a full mediator of the link between high standards and stress, and a partial mediator of the link between discrepancy and stress.

Our findings suggested that university freshmen who reported high standards might be able to manage their first months at the university quite well by drawing upon their self-control skills. This argument is in line with research by Flett et al. (1995b) who observed that high personal standards were associated with high goal commitment concerning perfect performance and organization. In the present study, high standards correlated positively with self-control, but not to the extent that one can argue that high standards and self-control are extremely overlapping concepts. Instead the positive correlation between high standards and self-control supported the idea that perfectionism has a strong motivational component and that individuals high in adaptive perfectionism will show both high self-imposed standards and the capability to strive for these standards successfully. This was in line with research as for instance by Burns et al. (2000) showing that “positive perfectionism” (i.e., adaptive perfectionism) is positively associated with feelings of mastery and control.

Self-control in the present study was indicated by students’ scores on the SCS (Tangney et al. 2004) that can be seen as an overall index of self-control as it measures the four main dimensions of self-control. Therefore the present findings implied that self-control that is linked to high standards might be characterized by controlling thoughts, emotions, impulses, and performance (see Baumeister et al. 1994) as these are the dimensions of self-control that are measured by the items of the SCS. When we controlled for the overlap between high standards and self-control to determine the unique impact of both variables on stress by means of regressions, self-control still affected stress negatively. This result provided evidence for the argument that the impact of self-control on stress is unique. Hence it can be concluded that students with high levels of self-control are better able to adapt to new and stressful situations than students with low levels of self-control, even if one controls for high standards (and also for discrepancy—see below). One reason for this might be that time management skills are a significant component of self-oriented perfectionism and these skills are adaptive for university students as they decrease the likelihood of university drop-out and course failure (see Klibert et al. 2005). Hence students with high standards and time management skills (i.e., an aspect of self-control) might well be able to adapt to stressful situations and hence increasing the likelihood that they will be successful in university.

We also observed that the positive correlation between discrepancy and stress was partly mediated by self-control: when controlling for the influence of self-control on the association between discrepancy and stress, this association was significantly reduced, but it remained significant and negative. Students with high levels of discrepancy reported lower levels of self-control than students with low levels of discrepancy. This lack of self-control resources might be at least partially responsible for the higher level of stress students high in discrepancy experience during their first months at university. Tangney et al. (2004) argued that central to their concept of self-control “was the ability to override or change one’s inner responses, as well as to interrupt undesired behavioral tendencies (such as impulses) and refrain from acting on them.” (p. 274). Research on perfectionism demonstrated that perfectionistic concerns and perfectionistic discrepancy are characterized by repeatedly (and unavoidably) occurring thoughts on failures, thoughts of being perfect etc. (see Flett et al. 2002; Rudolph et al. 2007). The results of the present study imply that students high in discrepancy might also tend to have problems with regulating their stream of thought (e.g., by stopping ruminations about their own failures by means of self-control strategies), to alter their negative emotions (e.g., controlling their fear of failure), and to persist in goal striving. Thus, these students might have problems with overriding maladaptive perfectionistic thoughts and to alter their states or behaviors as indicated by the negative relation between discrepancy and self-control. This argument was based on the fact that the SCS was build especially around breaking habits, resisting temptation, and keeping good self-discipline as these behaviors “all reflect the ability of the self to control itself” (Tangney et al. 2004, p. 275). In other words, students high in discrepancy might not be able to reduce the difference between their high expectancies and their performance by exerting self-control for instance by resisting temptations (e.g., preparing for the next seminar instead of having a party), controlling their thoughts (e.g., by not getting involved in negative ruminations about one’s own skills; see Rudolph et al. 2007) or controlling their negative emotions (e.g., fear of failure; Klibert et al. 2005; fear of losing control; Blankstein et al. 1993).

However, the partial mediation of the link between discrepancy and stress by self-control also implied that besides a lack of self-control other processes also determine the positive relation between discrepancy and stress (see also Flett and Hewitt 2002). Maybe students with high levels of discrepancy do not derive pleasure from their labors and efforts (just as socially prescribed perfectionists—see for instance Klibert et al. 2005) and therefore are not able to adapt to university life with its high academic demands. Besides this lack of pleasure from own effort, a helpless orientation could be another possible mediator of the link between discrepancy and stress. For instance, people high in perfectionistic concerns are thought to respond to stressful situations with a helplessness orientation (O’ Connor and O’ Connor 2003), and will experience more fear of failure (Conroy et al. 2007) particularly in new situations (i.e., transition to university life). Future research might address this issue for instance by measuring discrepancy, self-control, and helpless orientation and their impact on stress simultaneously.

The present study also showed a positive, but rather small correlation between self-control and active coping (r = .22). Hence one can argue that self-control measured as a central function of human beings (Baumeister et al. 1994; Muraven and Baumeister 2000) is related to active coping behavior, but it is clearly not the same concept but a core psychological resource that helps in dealing with stress. Moreover, our findings replicated earlier research showing that the two facets of perfectionism are associated with different coping styles (i.e., active coping vs. avoidance coping; Dunkley et al. 2000; Stoeber and Rennert 2008; Hill et al. 2010). However, even when we controlled for coping tendencies, we still observed that self-control fully mediated the link between high standards and stress, and partially the link between discrepancy and stress. These findings demonstrated the impact of self-control on stress management and the mediating role of self-control on the perfectionism-stress link that was rather independent from individual differences in coping styles.

Both aspects of perfectionism were positively and significantly related to each other. This was in accordance with earlier research showing that components of perfectionism are often moderately and positively interrelated (correlations between .3 and .5; e.g., Hewitt and Flett 1991; Mills and Blankstein 2000; Klibert et al. 2005). This result supported the idea that the perfectionism subscales high standards and discrepancy (Slaney et al. 2001) both measure perfectionism and that they therefore have a common component. It is important to notice, however, that the present findings also strongly indicated unique relations between both aspects of perfectionism and stress as they were related to stress in opposite directions– high standards were negatively linked to stress, whereas discrepancy was positively linked to stress. Moreover, high standards and discrepancy were linked to self-control in opposite directions, a result that also supported that argument that unique relations between both facets of perfectionism and self-control existed. This pattern of results also highlighted the importance of controlling for the effects of positively correlated facets of perfectionism on other variables simultaneously (see Hill et al. 2010).

Implications

The present study demonstrated the importance of self-control in adapting to stressful academic situations. Previous research on self-control showed that it can be improved through training (see Baumeister et al. 2006; Muraven et al. 1999; Oaten and Cheng 2005). For instance, Perels et al. (2009) reported that a self-control intervention in regular mathematics lessons of sixth-graders significantly enhances their self-control and their mathematical achievement. Similarly, Oaten and Cheng (2006) found evidence that university students strongly benefitted from repeated practice of self-control, as they were able to improve their regulatory strength what in turn dampened the debilitating effects of exam stress. Based on this research one can suggest to train especially students with high levels of discrepancy how to improve their self-control skills in order to reduce the discrepancy between their high expectations and their actual performance for instance by showing more persistence in goal striving, using time management techniques (see Klibert et al. 2005) etc.

Rudolph et al. (2007) observed that maladaptive cognitive coping tendencies of individuals with high scores in the PCI (Flett et al. 1998) had deficits in positive forms of cognitive coping (e.g., engaging in positive reappraisal). These authors suggested that cognitive interventions designed to remove the distressing automatic thoughts of perfectionists should also focus on changing maladaptive cognitive coping styles and encouraging more positive forms of cognitive emotion regulation in perfectionists. Following this line of argument one could imagine that certain self-regulation strategies as for instance implementation intentions (Gollwitzer 1999) could help to control these distressing automatic thoughts of perfectionism and replace them by positive forms of cognitive coping (see Achtziger et al. 2008).

Limitations

The present findings have some limitations stemming from the study’s cross-sectional design. They cannot shed light on the temporal or causal relationships between perfectionism, self-control, and stress. Future research should therefore employ longitudinal designs in order to clarify these issues. Moreover, participants’ age ranged between 18 and 21 and our sample size was rather small. Our findings therefore should be replicated in bigger and more heterogeneous samples, such as people at the workplace. We focused on self-control as a personality trait, measured with the SCS (Tangney et al. 2004), however, it is also possible to investigate differences in self-control by means of behavioral tasks (e.g., the delay of gratification paradigm) or other self-control scales. Stress can also be measured by other methods than questionnaires (e.g., cortisol level). Notwithstanding these limitations, the present findings make a significant and novel contribution to the understanding of the link between perfectionism and stress. The present research provides the first empirical evidence that self-control is an important mechanism that might explain how perfectionism is linked to stress.

Footnotes
1

As one can argue that the approach to assessing mediation by Baron and Kenny (1986) is rather dated and has some shortcomings (e.g., it underestimates mediation effects, ignores measurement error etc.), we also analyzed our data again by utilizing a bootstrapping method (1,000 samples, confidence interval 95 %). These additional analyses supported our findings as they also revealed a negative and significant link between high standards and stress (β = −.22, t = 2.68, p < .01) and a significant positive link between discrepancy and stress (β = .38, t = 5.13, p < .001). When controlling for self-control, the link between high standards and stress was completely dissolved (β = −.14, t = 1.63, p = .10), whereas the positive link between discrepancy and stress was reduced but still remained significant (β = .32, t = 4.10, p < .001). Moreover, self-control was negatively related to stress (β = −.22, t = 2.73, p < .01) even when we controlled for high standards and discrepancy. As bootstrapping is especially useful for testing mediation effects in small samples, this additional analysis demonstrated that our results reported in the main manuscript were quite robust.

 

Copyright information

© Springer Science+Business Media, LLC 2012