- Email: [email protected]
Multidimensional perfectionism and perceived stress: Group differences and test of a coping mediation model
Personality and Individual Differences 119 (2017) 106–111
Contents lists available at ScienceDirect
Personality and Individual Differences journal homepage: www.elsevier.com/locate/paid
Multidimensional perfectionism and perceived stress: Group differences and test of a coping mediation model Jeffrey S. Ashby a,⁎, Philip B. Gnilka b a b
Georgia State University, United States Virginia Commonwealth University, United States
a r t i c l e
i n f o
Article history: Received 1 May 2017 Received in revised form 10 July 2017 Accepted 11 July 2017 Available online xxxx Keywords: Perfectionism Coping Stress
a b s t r a c t This study examined the relationship between multidimensional perfectionism, four types of coping (task-oriented, emotion-oriented, social diversion, distraction), and perceived stress in a sample of 323 undergraduate students. Specifically, results of the study offered support for the tripartite model of perfectionism with three classes (adaptive, maladaptive, and non-perfectionists) who differed from one another on levels of stress and coping. The relationship between maladaptive perfectionism and perceived stress was mediated by task-oriented, emotional-oriented, and distraction coping in support of the general vulnerability model of perfectionism. In regard to adaptive perfectionism, only task-oriented and emotion-oriented coping were significant mediators. © 2017 Published by Elsevier Ltd.
Perfectionism is a multidimensional personality construct that has garnered increased attention in recent years (e.g., Stoeber & Otto, 2006). Of interest has been the relationship of perfectionism to a variety of measures of mental health and well-being (e.g., Gnilka, Ashby, & Noble, 2013). In addition, numerous authors (e.g., Dunkley, Mandel, & Ma, 2014) have highlighted the important role of stress in the relationship between perfectionism and mental health issues. Dunkley et al. (2014) argue for a “general vulnerability model” of perfectionism that suggests that individuals with higher levels of perfectionism and higher levels of stress are particularly vulnerable to emotional distress (e.g., Enns, Cox, & Clara, 2005). The purpose of this study was to investigate the differential role of coping in the relationship of multidimensional perfectionism to perceived stress. While there are differing definitions, conceptualizations, and measures of perfectionism (e.g., Broman-Fulks, Hill, & Green, 2008; Flett et al., 2016; Hewitt & Flett, 1991; Slaney, Rice, Mobley, Trippi, & Ashby, 2001; Smith, Saklofske, Stoeber, & Sherry, 2016), a number of factor analytic studies have offered evidence for two dimensions of perfectionism (e.g., Dunkley, Zuroff, & Blankstein, 2003). Researchers have shown that the two dimensions of perfectionism (i.e., perfectionistic strivings and perfectionistic concerns) can be used to identify discrete groups of adaptive, maladaptive, and non-perfectionists (e.g., Suh, Gnilka, & Rice, 2017) otherwise known as Parker's (1997) tripartite model of perfectionism. While both types of perfectionism are marked by the pursuit of high personal standards, maladaptive perfectionism includes an intense self-depreciation when personal standards are not reached (Hamachek, 1978) and is associated with various negative ⁎ Corresponding author. E-mail address: [email protected] (J.S. Ashby).
http://dx.doi.org/10.1016/j.paid.2017.07.012 0191-8869/© 2017 Published by Elsevier Ltd.
outcomes. For instance, maladaptive perfectionism has been consistently linked to depression (Ashby, Noble, & Gnilka, 2012), anxiety (Gnilka, Ashby, & Noble, 2012), and increased levels of perceived stress (Rice & Van Arsdale, 2010). In contrast, adaptive perfectionism is consistently associated with lower levels of depression (Rice, Ashby, & Slaney, 1998) and lower levels of stress and anxiety (Corry et al., 2013). While research consistently find differences in the stress levels of adaptive, maladaptive, and non-perfectionists (e.g., Ashby et al., 2012), there is a paucity of research investigating paths between adaptive and maladaptive perfectionism and different levels of stress. Several studies indicate that maladaptive and adaptive perfectionists use different coping patterns and styles. For instance, individuals with higher levels of maladaptive perfectionism are more likely than those with adaptive perfectionism to use ineffective methods of coping including avoidance (O'Connor & O'Connor, 2003) and emotion-based coping (Rice & Lapsley, 2001). Conversely, adaptive perfectionists more frequently use task-oriented (O'Connor & O'Connor, 2003) and problemfocused (Rice & Lapsley, 2001) coping resulting in better mental health outcomes. In summary, studies suggest higher levels of maladaptive perfectionism is associated with higher levels of emotional distress. There is consistent evidence for the general vulnerability model of maladaptive perfectionism in which stress creates the vulnerability that maladaptive perfectionists may have for negative emotional outcomes. The research regarding the relationship of stress to adaptive perfectionism in predicting outcomes is more varied. However, there is evidence for the mediating role of stress in the relationship of adaptive perfectionism to satisfaction with life and depression (Ashby et al., 2012). Despite this evidence supporting the important role of stress, few studies have been conducted to investigate the relationship between perfectionism and
J.S. Ashby, P.B. Gnilka / Personality and Individual Differences 119 (2017) 106–111
stress. The purpose of this study was to investigate the relationship between multidimensional perfectionism, coping, and stress. Specifically, we hypothesized that: (a) The results of latent profile analysis would support a three-class model of adaptive, maladaptive, and non-perfectionists. (b) The three identified perfectionist classes would significantly differ from one another in stress levels and coping, with maladaptive perfectionists showing the highest levels of stress, social diversion, distraction, and emotion-oriented coping and the lowest levels of task-oriented coping and adaptive perfectionists having the lowest levels of stress, social diversion, distraction, and emotion-oriented coping and the highest levels of task-oriented coping. (c) All four forms of coping would mediate the relationship between adaptive perfectionism and perceived stress (i.e., higher adaptive perfectionism would be positively associated with task-oriented coping and negatively associated with social diversion, distraction, and emotion-oriented coping, which in turn would lead to lower stress levels). (d) All four forms of coping would mediate the relationship between maladaptive perfectionism and perceived stress (i.e., higher maladaptive perfectionism would be negatively associated with task-oriented coping and positively associated with social diversion, distraction, and emotion-oriented coping, which in turn would lead to higher stress levels).
1. Methods 1.1. Participants & procedure Three hundred twenty three participants were randomly selected from a larger sample of 1229 undergraduate students from a large urban southeastern university. Due to a clerical error, specific demographic data for the participants in this study were not available, though overall demographics were available for the larger sample. In the larger sample, 66% of participants identified as female, 33% of participants identified as male, and 1% declined to identify their gender. Participant ethnicity for the sample was the following: 37.3% White, 33.5% African American, 10.1% Multiracial, 7.2% Hispanic, 6.5% Asian American, 3.1% Other ethnicity, and 2.2% declined to answer. 1.2. Instruments 1.2.1. Almost Perfect Scale — Revised (APS-R; Slaney et al., 2001) The APS-R is a 23-item inventory designed to measure the multidimensional construct of perfectionism through three subscales: Standards, Order, and Discrepancy. The Standards subscale is designed to measure personal standards, the Order subscale measures a participant's organization and need for order, and the discrepancy subscale is designed to measure distress caused by the discrepancy between performance and standards. The Order subscale was administered but was not needed for classifying participants (e.g., Rice & Ashby, 2007). Factor analyses, convergent and divergent validity of the APS-R has been demonstrated by previous studies (e.g., Slaney et al., 2001). Rice and Ashby (2007) reported high Cronbach's coefficients alphas with a college student sample. 1.2.2. Coping Inventory for Stressful Situations (CISS; Endler & Parker, 1994) The CISS is a 48-item self-report measure for assessing four types of coping orientations: task, emotion, social diversion, and distraction. The Task subscale measures coping strategies that actively mitigate the stressful event. The Emotion subscale measures coping strategies that assist in managing the emotions due to the stressor. The Social Diversion
107
subscale measures coping strategies where individuals seek out other people for support. Lastly, the Distraction subscale taps into individuals who initiate activities and behaviors to distract from a potential stressor. Several factor analyses have been conducted that have supported the structure of the scales (Endler & Parker, 1994). Acceptable Cronbach's coefficients alphas with college student samples has been previous reported by various authors for the subscales ranging between 0.76 and 0.91. 1.2.3. Perceived Stress Scale (PSS; Cohen, Kamarck, & Mermelstein, 1983) The PSS is a 14 item measure designed to appraise perceived stress. Cohen et al. (1983) report that internal consistency reliability estimates ranged from 0.84 to 0.86 across two groups of college-aged students and on a group of participants in a community-based smoking-cessation program. The authors also offered evidence for the concurrent validity of the measure and Martin, Kasarian, and Brieiter (1995) found support for the factor structure of the PSS. 1.3. Statistical methods Latent profile analysis (LPA) was conducted with Mplus (Version 7.4; Muthén & Muthén, 1998–2017) using a robust maximum likelihood estimator (MLR). LPA offers a sophisticated statistical approach to creating categorical profiles that offers advantages over cluster analysis such as the ability to compare competing models with various fit indices. Covariance coverage was 1.00, which allowed for reliable model convergence (Muthén & Muthén, 1998–2017). Analyses were based on two measured variables from the scales (APS-R Standards and Discrepancy) as continuous indicators of a latent class variable. Five thousand random sets of starting values were used and after 100 iterations, 500 optimizations were used in the final stage. The final stage log likelihood values were replicated across the optimizations in the final stage, and all models converged on proper solutions. A model comparison approach was used to determine the number of classes. First, a model specifying a single class was calculated. Next, a two-class model was calculated and compared to the single class model, and so forth, until reaching a five-class model. As an indicator of fit, we used several methods that are widely accepted (Nylund, Asparouhov, & Muthén, 2007). We compared models with different numbers of classes using the Bayesian Information Criterion (BIC) scores and the sample-size-adjusted BIC (aBIC) which were the primary indicators of model fit in these analyses. The Lo-Mendell-Rubin (LMR) likelihood ratio test and the Bootstrap Likelihood Ratio Test (BLRT) were used to determine the fit between two nested models that differ by a single. Lastly, models were also evaluated based on practical and theoretical considerations that may limit interpretability of results. Next, we conducted a series of analyses to determine whether there were differences in the mean levels of perceived stress and four forms of coping (i.e., task, emotion, and social diversion, and distraction) across the different latent profile classes. The Mplus BCH method (Bakk & Vermunt, 2016) was used which provides an omnibus test in addition to individual comparisons between the various latent classes on each outcome variable. In order to test the hypothesis that the four coping styles (task, emotion, social diversion, and distraction) mediated the relationships between the two dimensions of perfectionism and perceived stress, a measured variable path analyses was conducted using Mplus (Version 7.4; Muthén and Muthén, 1998–2017). As suggested by Cole and Preacher (2014), all measured variables were adjusted by their respective reliability coefficients. A series of model comparisons were conducted using the maximum likelihood method. Multiple fit indices were used to compare fit of competing models including chi-square, the root mean square error of approximation (RMSEA), comparative fit index (CFI), and standardized root mean square residual (SRMR). Next, we estimated the fit of the overall model followed by testing a
108
J.S. Ashby, P.B. Gnilka / Personality and Individual Differences 119 (2017) 106–111
Table 1 Descriptive statistics and correlations for study subscales (N = 323).
1. APSR standards 2.APSR discrepancy 3. Perceived stress 4. Task oriented 5. Emotion oriented 6. Social diversion 7. Distraction
M
SD
α
1
2
3
4
5
6
7
38.88 46.45 26.70 56.02 45.03 17.43 23.91
7.52 18.50 8.00 11.44 12.54 5.00 6.82
0.92 0.77 0.80 0.91 0.90 0.83 0.79
– −0.05 −0.16⁎ 0.42⁎ −0.13⁎ 0.20⁎ 0.03
– 0.53⁎ −0.24⁎ 0.62⁎ −0.09 0.16⁎
– −0.37⁎ 0.64⁎ −0.03 0.15⁎
– −0.13⁎ 0.29⁎ 0.05
– 0.12⁎ 0.38⁎
– 0.45⁎
–
Note. Standards = Almost Perfect Scale Revised (APSR) standards subscale; Discrepancy = APSR discrepancy subscale; Perceived stress = perceived stress scale; Task oriented, Emotion oriented, Social diversion, and Distraction are subscales of the Coping in Stressful Situations Scale. ⁎ p b 0.05.
fully mediated model where the direct paths between both perfectionism dimensions and perceived stress were restrained to be zero. To evaluate the strength of the four coping strategies as mediators, a bias-corrected bootstrapping analysis was used to test the indirect effects as products of coefficients (Preacher & Hayes, 2008). The null hypothesis was that the product of the indirect paths (i.e., from the predictors [APS-R Discrepancy and Standards] to the mediators [CISS Task, Emotional, Social Diversion, and Distraction] and from the mediators to the outcome [Perceived Stress]) — is equal to zero. Each unique mediator pathway can be tested for significance. A total of 10,000 random samples were created from the sample.
2. Results Descriptive statistics, Cronbach's coefficients alphas, and bivariate correlations among study scales are reported in Table 1. Cronbach's coefficient alphas were consistent with previous studies. Fit indices for the LPA are presented in Table 2. Results indicated that BIC values decreased for the one- to three-class models then began to rise for the four-class model. The BIC values decreased for the one-to-three class models then began to level off for the four-class model. Entropy value was higher for the two-, three-, and five-class models than for the other models. The LMR suggested a three-class model fit the data best; however, the BLRT suggested a four-class model best fit the data. We examined the four-class model to see if it would fit the 2 × 2 Model of Perfectionism (Gaudreau & Thompson, 2010). The fourth class was similar to the third class (i.e., similar Standards levels and slightly increased Discrepancy levels) and was not significantly different on emotionbased coping, avoidant coping, social diversion coping, and perceived stress. Given the similar outcomes between scores in the two classes, the inconsistent discrepancy score with the 2 × 2 model for the fourth class, and the other fit indices suggesting a 3-class model, a 3-class solution was viewed as more parsimonious and consistent with the tripartite model. (see Fig. 1). Next, several sets of analyses using the BCH method in Mplus were conducted to determine whether the means of perceived stress and four forms of coping (task, emotion, and social diversion, and distraction) differed across the latent classes. Four of the five omnibus tests were significant (p b 0.05), with the exception of distraction coping (p N 0.05). Results of these tests are reported in Table 3.
To investigate the potential mediating role of the four styles of coping between adaptive and maladaptive perfectionism and perceived stress, mediation analyses were conducted. Results of the analyses indicated that Model 1 showed poor fit, χ2 (6, N = 323) = 144.51, p b 0.001, CFI = 0.78, RMSEA = 0.27, 90% CI [0.23, 0.31], SRMR = 0.10, and modification indices (MIs) suggested allowing the CISS Social Diversion subscale to covary with the CISS Distraction subscale. After completing the modification and rerunning the model, the path between the subscales was significant (p b 0.05). The model fit statistics indicated that the model with the modifications was a better fit than the original model, Δχ2 (1) = 80.38, p b 0.001. The CFI increased to 0.91, the RMSEA decreased to 0.19, 90% CI [0.15, 0.23], and SRMR decreased to 0.07. In conclusion, the modification improved the fit of the model; however, the overall model was still a poor fit. Consistent with the earlier modification in allowing various coping styles to covary, the MIs suggested allowing the CISS Distraction subscale to covary with the CISS Emotion-oriented subscale. After rerunning the model with this modification, the path between the subscales was significant (p b 0.05) and resulted in a better fitting model, Δχ2 (1) = 25.82, p b 0.001, the CFI increased to 0.95, the RMSEA decreased to 0.16, 90% CI [0.12, 0.21], and SRMR decreased to 0.05. The overall model, however, was still a poor fit. The MIs suggested allowing the CISS Social Diversion subscale to covary with the CISS Emotion-oriented subscale. The model was rerun with this modification resulting in a significant path between the subscales (p b 0.05) and a significant change in chi-square, Δχ2 (1) = 23.74, p b 0.001. The other fit indices also improved (CFI = 0.98; RMSEA = 0.11, 90% CI [0.06, 0.17]; SRMR = 0.03). The MIs were consulted one last time and suggested allowing the CISS Social Diversion subscale to covary with the CISS Task-based subscale. A rerun model with this modification resulted in a significant path between the subscales (p b 0.05) and a significant change in chisquare Δχ2 (1) = 8.81, p b 0.001. The other fit indices also improved: the CFI increased to 0.99, the RMSEA decreased to 0.07, 90% CI [0.00, 0.15], and SRMR decreased to 0.02 suggesting the overall model was a good fit for the data. No further MIs that were suggested could be theoretically justified. Next, to test for full mediation between both perfectionism dimensions and Perceived Stress, we restrained direct paths as described earlier. This model resulted in a non-significant increase in chi-square Δχ2 (2) = 3.26, p N 0.05. In addition, RMSEA decreased to 0.06, 90% CI [0.000,
Table 2 Fit indices and entropy for one- to six-class models. Model
BIC
aBIC
LMR value
LMR p-value
BLRT p-value
Entropy
1-Class model 2-Class model 3-Class model 4-Class model 5-Class model 6-Class model
5042.195 4995.511 4956.521 4953.819 4958.821 4973.247
5029.508 4973.307 4924.802 4912.585 4908.071 4912.982
– 60.525 53.250 18.942 11.659 2.748
– b0.001 b0.001 N0.05 N0.05 N0.05
– b0.001 b0.001 b 0.001 b0.05 N0.05
– 0.801 0.776 0.774 0.790 0.760
Note. BIC = Bayesian Information Criterion; aBIC = sample size adjusted BIC; LMR = Lo-Mendell-Rubin test; BLRT = Bootstrap Likelihood Ratio Test
J.S. Ashby, P.B. Gnilka / Personality and Individual Differences 119 (2017) 106–111
109
Fig. 1. Characteristics of classes.
0.12], SRMR and CFI remained constant. Thus, the four coping strategies fully mediated the relationships between both dimensions of perfectionism and stress (see Fig. 2). Next, we conducted additional analyses to determine the relative strength of the various mediation pathways by bootstrapping the pathways. For maladaptive perfectionism, taskoriented, emotional-oriented, and distraction coping were significant mediators. In regard to adaptive perfectionism, only task-oriented and emotion-oriented coping were significant mediators. Social diversion was found not to be a significant mediator for either dimension of perfectionism. 3. Discussion The results of this study are consistent with previous research that identified clear groups of adaptive, maladaptive, and non-perfectionists (e.g., Suh et al., 2017). Consistent with other studies (e.g., Rice & Ashby, 2007), the non-perfectionist group had low standards and moderate levels of discrepancy while the identified adaptive perfectionist group had high standards and low discrepancy and the maladaptive perfectionists had high standards and high discrepancy. These groups are also consistent with the theoretical literature (e.g., Hamachek, 1978) that indicated that both adaptive and maladaptive perfectionists were distinct from non-perfectionists in that that they hold high standards and consistently strive for perfection. The distinction between the groups of perfectionists is their relative ability to productively move forward when their standards are not met. Adaptive perfectionists, while disappointed at not having met their perfectionistic standards, are able to relatively quickly regroup and take on the next task. In contrast, maladaptive perfectionists are more likely to be highly self-critical.
In addition to clearly identifying groups of perfectionists, the results of this study offered evidence for the perfectionistic groups' differing levels of coping and stress. In this study, the adaptive perfectionists had significantly lower levels of perceived stress than either the maladaptive perfectionists or non-perfectionists. The maladaptive perfectionist group had significantly higher levels of perceived stress than the non-perfectionists. These results are consistent with numerous other studies (e.g., Ashby et al., 2012; Chang, Watkins, & Banks, 2004) that have examined the relationship between perfectionism and stress and found evidence for differences across perfectionism group. Relatedly, the results of this study also indicated that the three groups of perfectionists differed significantly in coping scores. Adaptive perfectionists, maladaptive perfectionists, and non-perfectionists all differed significantly in levels of task-oriented and emotional-oriented coping. Adaptive perfectionists in the study reported significantly higher use of task-oriented coping than either of the other two groups while maladaptive perfectionists reported the lowest use of this coping style. In contrast, maladaptive perfectionists reported the highest use of emotion-oriented coping while adaptive perfectionists reported the lowest use of this coping style. Adaptive perfectionists reported significantly higher use of social diversion coping than the other two groups who did not differ significantly from one another. The results of the study revealed no significant differences between adaptive perfectionists, maladaptive perfectionists, and non-perfectionists in their use of the distraction copying style. While a number of studies have offered evidence for the role of stress in the relationship of perfectionism to various outcomes, there has been limited investigation of the role of coping in the relationship between perfectionism and stress. Consistent with the Transactional
Table 3 Latent profile means, standard deviations, and Wald chi-square tests of mean equality using the BCH procedure. Auxiliary variable
Perceived stress Task-oriented Emotion-oriented Social diversion Distraction
Non-perfectionists (N = 90)
Maladaptive perfectionists (N = 100)
Adaptive perfectionists (N = 127)
M
SD
M
SD
M
SD
27.02 49.52 46.50 15.95 23.95
7.15 10.56 12.14 5.08 6.87
31.80 55.03 53.41 17.09 24.30
8.64 15.29 14.20 5.79 7.75
22.51 61.43 37.50 18.77 23.57
8.68 9.97 11.83 5.57 8.08
Note. A = adaptive perfectionists, M = maladaptive perfectionists, N = non-perfectionists. ⁎ p b 0.05. ⁎⁎ p b 0.01. ⁎⁎⁎ p b 0.001.
Global χ2
N vs. M
N vs. A
M vs. A
58.13⁎⁎⁎ 70.42⁎⁎⁎ 79.16⁎⁎⁎
16.06⁎⁎⁎ 7.72⁎⁎ 12.05⁎⁎ 0.10 1.92
17.06⁎⁎⁎ 69.44⁎⁎⁎ 29.16⁎⁎⁎
57.76⁎⁎⁎ 11.74⁎⁎ 73.24⁎⁎⁎ 0.43 4.31⁎
0.43 14.84⁎⁎
0.14 14.63⁎⁎⁎
110
J.S. Ashby, P.B. Gnilka / Personality and Individual Differences 119 (2017) 106–111
Fig. 2. Path model of the relations between adaptive and maladaptive perfectionism, coping, and perceived stress. The path coefficients are standardized regression coefficients. *p b 0.05.
Model of Stress (TMS; Lazarus & Folkman, 1984) suggesting that coping can buffer or augment the stress response, the results of this study found support for the role of coping in the prediction of stress. Specifically, the results of this study found evidence for the mediating role of task-oriented and emotion-oriented coping in the relationship between both adaptive and maladaptive perfectionism and stress. These results suggest that the higher levels of stress reported by maladaptive perfectionists may be related to these individuals' inability to address stressor directly (e.g., task-oriented coping) and “thereby increases the severity and duration of stress, leading to a greater susceptibility to experience additional stressors” (Dunkley, Solomon-Krakus, & Moroz, 2016, p. 160). In addition, maladaptive perfectionists may be more likely to engage in emotion-oriented coping consistent with the harsh self-evaluations generally associated with this cognitive style that may lead them to “magnify the negative aspects of events, thereby interpreting even mundane stressors as major threatening stressors” (Dunkley et al., 2016, p. 160). The results of this study also suggest that, in contrast to their maladaptive counter parts, adaptive perfectionists had higher levels of task-oriented coping and lower levels of emotion-oriented coping. Their tendency to engage in more task-oriented coping and less emotion-oriented coping may potentially buffer the impact of potential stressors, resulting in lower levels of perceived stress. This combination of coping is consistent with previous studies that have found adaptive perfectionists are as likely to utilize similar levels of active coping processes as maladaptive perfectionists in combination with significantly fewer unhealthy coping processes. For example, both Gnilka et al. (2012) and Noble, Ashby, and Gnilka (2014) found that while maladaptive perfectionists and adaptive perfectionists utilized similar levels of healthier coping processes, adaptive perfectionists utilized significantly fewer unhealthy coping processes. When looking at multiple coping processes simultaneously, it is important to note that these processes may vary by circumstance (Folkman, Lazarus, Dunkel-Schetter, DeLongis, and Gruen, 1986). One possible explanation for this may be that both adaptive and maladaptive perfectionists are active in facing stressors when first encountered but, over time, adaptive perfectionists may be less likely to shift to patterns of avoidance that would contribute to more chronic stress. This lack of a chronic stress response suggests maladaptive perfectionists may have a blunted cortisol response as suggested by Richardson, Rice, and Devine (2014).
This study had several limitations. First, we used an undergraduate student sample from a large university in the southeast. Future researchers should consider community and clinical samples to see if similar results are found. Second, this study was cross-sectional. While this type of design can add some understanding about how these different constructs may work together, we still need more longitudinal studies to be conducted that use three or more waves to determine causality. Third, more ethnically diverse populations should be utilized to see if the results of this study can be replicated. Fourth, this study used self-report measures and did not account for social desirability. Future researchers may want to consider controlling for social desirability. References Ashby, J. S., Noble, C. L., & Gnilka, P. B. (2012). Multidimensional perfectionism, depression, and satisfaction with life: Differences among perfectionists and tests of a stress-mediation model. Journal of College Counseling, 15, 130–143. Bakk, Z., & Vermunt, J. K. (2016). Robustness of stepwise latent class modeling with continuous distal outcomes. Structural Equation Modeling, 23, 20–31. Broman-Fulks, J. J., Hill, R. W., & Green, B. A. (2008). Is perfectionism categorical or dimensional? A taxometric analysis. Journal of Personality Assessment, 90, 481–490. Chang, E. C., Watkins, A., & Banks, K. H. (2004). How adaptive and maladaptive perfectionism relate to positive and negative psychological functioning: Testing a stress-mediation model in Black and White female college students. Journal of Counseling Psychology, 51, 93–102. Cohen, S., Kamarck, T., & Mermelstein, R. (1983). A global measure of perceived stress. Journal of Health & Social Behavior, 24, 385–396. Cole, D. A., & Preacher, K. J. (2014). Manifest variable path analysis: Potentially serious and misleading consequences due to uncorrected measurement error. Psychological Methods, 19, 300–315. Corry, J., Green, M., Roberts, G., Frankland, A., Wright, A., Lau, P., ... Mitchell, P. (2013). Anxiety, stress and perfectionism in bipolar disorder. Journal of Affective Disorders, 151, 1016–1024. Dunkley, D. M., Mandel, T., & Ma, D. (2014). Perfectionism, neuroticism, and daily stress reactivity and coping effectiveness 6 months and 3 years later. Journal of Counseling Psychology, 61, 616–633. Dunkley, D. M., Solomon-Krakus, S., & Moroz, M. (2016). Personal standards and self-critical perfectionism and distress: Stress, coping, and perceived social support as mediators and moderators. In F. Sirois, & D. Molnar (Eds.), Perfectionism, health, and wellbeing (pp. 157–176). New York, NY: Springer. Dunkley, D. M., Zuroff, D. C., & Blankstein, K. R. (2003). Self-critical perfectionism and daily affect: Dispositional and situational influences on stress and coping. Journal of Personality & Social Psychology, 84, 234–252. Endler, N. S., & Parker, J. D. (1994). Assessment of multidimensional coping: Task, emotion, and avoidance strategies. Psychological Assessment, 6, 50–60. Enns, M. W., Cox, B. J., & Clara, I. P. (2005). Perfectionism and neuroticism: A longitudinal study of specific vulnerability and diathesis-stress models. Cognitive Therapy & Research, 29, 463–478.
J.S. Ashby, P.B. Gnilka / Personality and Individual Differences 119 (2017) 106–111 Flett, G. L., Hewitt, P. L., Besser, A., Su, C., Vaillancourt, T., Boucher, D., ... Gale, O. (2016). The child-adolescent perfectionism scale. Journal of Psychoeducational Assessment, 34, 634–652. Folkman, S., Lazarus, R., Dunkel-Schetter, C., DeLongis, A., & Gruen, R. (1986). Dynamics of a stressful encounter: Cognitive appraisal, coping, and encounter outcomes. Journal of Personality & Social Psychology, 50, 992–1003. Gaudreau, P., & Thompson, A. (2010). Testing a 2 × 2 model of dispositional perfectionism. Personality & Individual Differences, 48, 532–537. Gnilka, P. G., Ashby, J. S., & Noble, C. M. (2012). Multidimensional perfectionism and anxiety: Differences among individuals with perfectionism and tests of a coping-mediation model. Journal of Counseling & Development, 90, 427–436. Gnilka, P. B., Ashby, J. S., & Noble, C. M. (2013). Adult attachment style differences among adaptive, maladaptive, and non-perfectionists. Journal of Counseling & Development, 91, 78–86. Hamachek, D. (1978). Psychodynamics of normal and neurotic perfectionism. Psychology, 15, 27–33. Hewitt, P. L., & Flett, G. L. (1991). Perfectionism in the self and social contexts: Conceptualization, assessment, and association with psychopathology. Journal of Personality & Social Psychology, 60, 456–470. Lazarus, R. S., & Folkman, S. (1984). Coping and adaptation. In W. D. Gentry (Ed.), The handbook of behavioral medicine (pp. 282–325). New York: Guilford. Martin, R. A., Kasarian, S. S., & Brieiter, H. J. (1995). Perceived stress, life events, dysfunctional attitudes, and depression in adolescent psychiatric inpatients. Journal of Psychopathology & Behavioral Assessment, 17, 81–95. Muthén, L. K., & Muthén, B. O. (1998–2017). Mplus user's guide (Eighth edition ). Los Angeles, CA: Muthén & Muthén. Noble, C. L., Ashby, J. S., & Gnilka, P. B. (2014). Multidimensional perfectionism, coping, and depression: Differential prediction of depression symptoms by perfectionism type. Journal of College Counseling, 17, 80–94. Nylund, K. L., Asparouhov, T., & Muthén, B. O. (2007). Deciding on the number of classes in latent class analysis and growth mixture modeling: A Monte Carlo simulation study. Structural Equation Modeling, 14, 535–569.
111
O'Connor, R., & O'Connor, D. (2003). Predicting hopelessness and psychological distress: The role of perfectionism and coping. Journal of Counseling Psychology, 50, 362–372. Parker, W. D. (1997). An empirical typology of perfectionism in academically talented children. American Educational Research Journal, 34, 545–562. Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879–891. Rice, K. G., & Ashby, J. S. (2007). An efficient method for classifying perfectionists. Journal of Counseling Psychology, 54, 72–85. Rice, K. G., Ashby, J. S., & Slaney, R. B. (1998). Self-esteem as a mediator between perfectionism and depression: A structural equation analysis. Journal of Counseling Psychology, 45, 304–314. Rice, K. G., & Lapsley, D. (2001). Perfectionism, coping, and emotional adjustment. Journal of College Student Development, 42, 157–168. Rice, K. G., & Van Arsdale, A. (2010). Perfectionism, perceived stress, drinking to cope, and alcohol-related problems among college students. Journal of Counseling Psychology, 57, 439–450. Richardson, C. M., Rice, K. G., & Devine, D. P. (2014). Perfectionism, emotion regulation, and the cortisol stress response. Journal of Counseling Psychology, 61, 110–118. Slaney, R. B., Rice, K., Mobley, M., Trippi, J., & Ashby, J. (2001). The revised almost perfect scale. Measurement & Evaluation in Counseling & Development, 34, 130–145. Smith, M. M., Saklofske, D. H., Stoeber, J., & Sherry, S. B. (2016). The big three perfectionism scale: A new measure of perfectionism. Journal of Psychoeducational Assessment, 34, 670–687. Stoeber, J., & Otto, K. (2006). Positive conceptions of perfectionism: Approaches, evidence, challenges. Personality and Social Psychology Review, 10, 295–319. Suh, H., Gnilka, P. B., & Rice, K. G. (2017). Perfectionism and well-being: A positive psychology framework. Personality & Individual Differences, 111, 25–30.
Contents lists available at ScienceDirect
Personality and Individual Differences journal homepage: www.elsevier.com/locate/paid
Multidimensional perfectionism and perceived stress: Group differences and test of a coping mediation model Jeffrey S. Ashby a,⁎, Philip B. Gnilka b a b
Georgia State University, United States Virginia Commonwealth University, United States
a r t i c l e
i n f o
Article history: Received 1 May 2017 Received in revised form 10 July 2017 Accepted 11 July 2017 Available online xxxx Keywords: Perfectionism Coping Stress
a b s t r a c t This study examined the relationship between multidimensional perfectionism, four types of coping (task-oriented, emotion-oriented, social diversion, distraction), and perceived stress in a sample of 323 undergraduate students. Specifically, results of the study offered support for the tripartite model of perfectionism with three classes (adaptive, maladaptive, and non-perfectionists) who differed from one another on levels of stress and coping. The relationship between maladaptive perfectionism and perceived stress was mediated by task-oriented, emotional-oriented, and distraction coping in support of the general vulnerability model of perfectionism. In regard to adaptive perfectionism, only task-oriented and emotion-oriented coping were significant mediators. © 2017 Published by Elsevier Ltd.
Perfectionism is a multidimensional personality construct that has garnered increased attention in recent years (e.g., Stoeber & Otto, 2006). Of interest has been the relationship of perfectionism to a variety of measures of mental health and well-being (e.g., Gnilka, Ashby, & Noble, 2013). In addition, numerous authors (e.g., Dunkley, Mandel, & Ma, 2014) have highlighted the important role of stress in the relationship between perfectionism and mental health issues. Dunkley et al. (2014) argue for a “general vulnerability model” of perfectionism that suggests that individuals with higher levels of perfectionism and higher levels of stress are particularly vulnerable to emotional distress (e.g., Enns, Cox, & Clara, 2005). The purpose of this study was to investigate the differential role of coping in the relationship of multidimensional perfectionism to perceived stress. While there are differing definitions, conceptualizations, and measures of perfectionism (e.g., Broman-Fulks, Hill, & Green, 2008; Flett et al., 2016; Hewitt & Flett, 1991; Slaney, Rice, Mobley, Trippi, & Ashby, 2001; Smith, Saklofske, Stoeber, & Sherry, 2016), a number of factor analytic studies have offered evidence for two dimensions of perfectionism (e.g., Dunkley, Zuroff, & Blankstein, 2003). Researchers have shown that the two dimensions of perfectionism (i.e., perfectionistic strivings and perfectionistic concerns) can be used to identify discrete groups of adaptive, maladaptive, and non-perfectionists (e.g., Suh, Gnilka, & Rice, 2017) otherwise known as Parker's (1997) tripartite model of perfectionism. While both types of perfectionism are marked by the pursuit of high personal standards, maladaptive perfectionism includes an intense self-depreciation when personal standards are not reached (Hamachek, 1978) and is associated with various negative ⁎ Corresponding author. E-mail address: [email protected] (J.S. Ashby).
http://dx.doi.org/10.1016/j.paid.2017.07.012 0191-8869/© 2017 Published by Elsevier Ltd.
outcomes. For instance, maladaptive perfectionism has been consistently linked to depression (Ashby, Noble, & Gnilka, 2012), anxiety (Gnilka, Ashby, & Noble, 2012), and increased levels of perceived stress (Rice & Van Arsdale, 2010). In contrast, adaptive perfectionism is consistently associated with lower levels of depression (Rice, Ashby, & Slaney, 1998) and lower levels of stress and anxiety (Corry et al., 2013). While research consistently find differences in the stress levels of adaptive, maladaptive, and non-perfectionists (e.g., Ashby et al., 2012), there is a paucity of research investigating paths between adaptive and maladaptive perfectionism and different levels of stress. Several studies indicate that maladaptive and adaptive perfectionists use different coping patterns and styles. For instance, individuals with higher levels of maladaptive perfectionism are more likely than those with adaptive perfectionism to use ineffective methods of coping including avoidance (O'Connor & O'Connor, 2003) and emotion-based coping (Rice & Lapsley, 2001). Conversely, adaptive perfectionists more frequently use task-oriented (O'Connor & O'Connor, 2003) and problemfocused (Rice & Lapsley, 2001) coping resulting in better mental health outcomes. In summary, studies suggest higher levels of maladaptive perfectionism is associated with higher levels of emotional distress. There is consistent evidence for the general vulnerability model of maladaptive perfectionism in which stress creates the vulnerability that maladaptive perfectionists may have for negative emotional outcomes. The research regarding the relationship of stress to adaptive perfectionism in predicting outcomes is more varied. However, there is evidence for the mediating role of stress in the relationship of adaptive perfectionism to satisfaction with life and depression (Ashby et al., 2012). Despite this evidence supporting the important role of stress, few studies have been conducted to investigate the relationship between perfectionism and
J.S. Ashby, P.B. Gnilka / Personality and Individual Differences 119 (2017) 106–111
stress. The purpose of this study was to investigate the relationship between multidimensional perfectionism, coping, and stress. Specifically, we hypothesized that: (a) The results of latent profile analysis would support a three-class model of adaptive, maladaptive, and non-perfectionists. (b) The three identified perfectionist classes would significantly differ from one another in stress levels and coping, with maladaptive perfectionists showing the highest levels of stress, social diversion, distraction, and emotion-oriented coping and the lowest levels of task-oriented coping and adaptive perfectionists having the lowest levels of stress, social diversion, distraction, and emotion-oriented coping and the highest levels of task-oriented coping. (c) All four forms of coping would mediate the relationship between adaptive perfectionism and perceived stress (i.e., higher adaptive perfectionism would be positively associated with task-oriented coping and negatively associated with social diversion, distraction, and emotion-oriented coping, which in turn would lead to lower stress levels). (d) All four forms of coping would mediate the relationship between maladaptive perfectionism and perceived stress (i.e., higher maladaptive perfectionism would be negatively associated with task-oriented coping and positively associated with social diversion, distraction, and emotion-oriented coping, which in turn would lead to higher stress levels).
1. Methods 1.1. Participants & procedure Three hundred twenty three participants were randomly selected from a larger sample of 1229 undergraduate students from a large urban southeastern university. Due to a clerical error, specific demographic data for the participants in this study were not available, though overall demographics were available for the larger sample. In the larger sample, 66% of participants identified as female, 33% of participants identified as male, and 1% declined to identify their gender. Participant ethnicity for the sample was the following: 37.3% White, 33.5% African American, 10.1% Multiracial, 7.2% Hispanic, 6.5% Asian American, 3.1% Other ethnicity, and 2.2% declined to answer. 1.2. Instruments 1.2.1. Almost Perfect Scale — Revised (APS-R; Slaney et al., 2001) The APS-R is a 23-item inventory designed to measure the multidimensional construct of perfectionism through three subscales: Standards, Order, and Discrepancy. The Standards subscale is designed to measure personal standards, the Order subscale measures a participant's organization and need for order, and the discrepancy subscale is designed to measure distress caused by the discrepancy between performance and standards. The Order subscale was administered but was not needed for classifying participants (e.g., Rice & Ashby, 2007). Factor analyses, convergent and divergent validity of the APS-R has been demonstrated by previous studies (e.g., Slaney et al., 2001). Rice and Ashby (2007) reported high Cronbach's coefficients alphas with a college student sample. 1.2.2. Coping Inventory for Stressful Situations (CISS; Endler & Parker, 1994) The CISS is a 48-item self-report measure for assessing four types of coping orientations: task, emotion, social diversion, and distraction. The Task subscale measures coping strategies that actively mitigate the stressful event. The Emotion subscale measures coping strategies that assist in managing the emotions due to the stressor. The Social Diversion
107
subscale measures coping strategies where individuals seek out other people for support. Lastly, the Distraction subscale taps into individuals who initiate activities and behaviors to distract from a potential stressor. Several factor analyses have been conducted that have supported the structure of the scales (Endler & Parker, 1994). Acceptable Cronbach's coefficients alphas with college student samples has been previous reported by various authors for the subscales ranging between 0.76 and 0.91. 1.2.3. Perceived Stress Scale (PSS; Cohen, Kamarck, & Mermelstein, 1983) The PSS is a 14 item measure designed to appraise perceived stress. Cohen et al. (1983) report that internal consistency reliability estimates ranged from 0.84 to 0.86 across two groups of college-aged students and on a group of participants in a community-based smoking-cessation program. The authors also offered evidence for the concurrent validity of the measure and Martin, Kasarian, and Brieiter (1995) found support for the factor structure of the PSS. 1.3. Statistical methods Latent profile analysis (LPA) was conducted with Mplus (Version 7.4; Muthén & Muthén, 1998–2017) using a robust maximum likelihood estimator (MLR). LPA offers a sophisticated statistical approach to creating categorical profiles that offers advantages over cluster analysis such as the ability to compare competing models with various fit indices. Covariance coverage was 1.00, which allowed for reliable model convergence (Muthén & Muthén, 1998–2017). Analyses were based on two measured variables from the scales (APS-R Standards and Discrepancy) as continuous indicators of a latent class variable. Five thousand random sets of starting values were used and after 100 iterations, 500 optimizations were used in the final stage. The final stage log likelihood values were replicated across the optimizations in the final stage, and all models converged on proper solutions. A model comparison approach was used to determine the number of classes. First, a model specifying a single class was calculated. Next, a two-class model was calculated and compared to the single class model, and so forth, until reaching a five-class model. As an indicator of fit, we used several methods that are widely accepted (Nylund, Asparouhov, & Muthén, 2007). We compared models with different numbers of classes using the Bayesian Information Criterion (BIC) scores and the sample-size-adjusted BIC (aBIC) which were the primary indicators of model fit in these analyses. The Lo-Mendell-Rubin (LMR) likelihood ratio test and the Bootstrap Likelihood Ratio Test (BLRT) were used to determine the fit between two nested models that differ by a single. Lastly, models were also evaluated based on practical and theoretical considerations that may limit interpretability of results. Next, we conducted a series of analyses to determine whether there were differences in the mean levels of perceived stress and four forms of coping (i.e., task, emotion, and social diversion, and distraction) across the different latent profile classes. The Mplus BCH method (Bakk & Vermunt, 2016) was used which provides an omnibus test in addition to individual comparisons between the various latent classes on each outcome variable. In order to test the hypothesis that the four coping styles (task, emotion, social diversion, and distraction) mediated the relationships between the two dimensions of perfectionism and perceived stress, a measured variable path analyses was conducted using Mplus (Version 7.4; Muthén and Muthén, 1998–2017). As suggested by Cole and Preacher (2014), all measured variables were adjusted by their respective reliability coefficients. A series of model comparisons were conducted using the maximum likelihood method. Multiple fit indices were used to compare fit of competing models including chi-square, the root mean square error of approximation (RMSEA), comparative fit index (CFI), and standardized root mean square residual (SRMR). Next, we estimated the fit of the overall model followed by testing a
108
J.S. Ashby, P.B. Gnilka / Personality and Individual Differences 119 (2017) 106–111
Table 1 Descriptive statistics and correlations for study subscales (N = 323).
1. APSR standards 2.APSR discrepancy 3. Perceived stress 4. Task oriented 5. Emotion oriented 6. Social diversion 7. Distraction
M
SD
α
1
2
3
4
5
6
7
38.88 46.45 26.70 56.02 45.03 17.43 23.91
7.52 18.50 8.00 11.44 12.54 5.00 6.82
0.92 0.77 0.80 0.91 0.90 0.83 0.79
– −0.05 −0.16⁎ 0.42⁎ −0.13⁎ 0.20⁎ 0.03
– 0.53⁎ −0.24⁎ 0.62⁎ −0.09 0.16⁎
– −0.37⁎ 0.64⁎ −0.03 0.15⁎
– −0.13⁎ 0.29⁎ 0.05
– 0.12⁎ 0.38⁎
– 0.45⁎
–
Note. Standards = Almost Perfect Scale Revised (APSR) standards subscale; Discrepancy = APSR discrepancy subscale; Perceived stress = perceived stress scale; Task oriented, Emotion oriented, Social diversion, and Distraction are subscales of the Coping in Stressful Situations Scale. ⁎ p b 0.05.
fully mediated model where the direct paths between both perfectionism dimensions and perceived stress were restrained to be zero. To evaluate the strength of the four coping strategies as mediators, a bias-corrected bootstrapping analysis was used to test the indirect effects as products of coefficients (Preacher & Hayes, 2008). The null hypothesis was that the product of the indirect paths (i.e., from the predictors [APS-R Discrepancy and Standards] to the mediators [CISS Task, Emotional, Social Diversion, and Distraction] and from the mediators to the outcome [Perceived Stress]) — is equal to zero. Each unique mediator pathway can be tested for significance. A total of 10,000 random samples were created from the sample.
2. Results Descriptive statistics, Cronbach's coefficients alphas, and bivariate correlations among study scales are reported in Table 1. Cronbach's coefficient alphas were consistent with previous studies. Fit indices for the LPA are presented in Table 2. Results indicated that BIC values decreased for the one- to three-class models then began to rise for the four-class model. The BIC values decreased for the one-to-three class models then began to level off for the four-class model. Entropy value was higher for the two-, three-, and five-class models than for the other models. The LMR suggested a three-class model fit the data best; however, the BLRT suggested a four-class model best fit the data. We examined the four-class model to see if it would fit the 2 × 2 Model of Perfectionism (Gaudreau & Thompson, 2010). The fourth class was similar to the third class (i.e., similar Standards levels and slightly increased Discrepancy levels) and was not significantly different on emotionbased coping, avoidant coping, social diversion coping, and perceived stress. Given the similar outcomes between scores in the two classes, the inconsistent discrepancy score with the 2 × 2 model for the fourth class, and the other fit indices suggesting a 3-class model, a 3-class solution was viewed as more parsimonious and consistent with the tripartite model. (see Fig. 1). Next, several sets of analyses using the BCH method in Mplus were conducted to determine whether the means of perceived stress and four forms of coping (task, emotion, and social diversion, and distraction) differed across the latent classes. Four of the five omnibus tests were significant (p b 0.05), with the exception of distraction coping (p N 0.05). Results of these tests are reported in Table 3.
To investigate the potential mediating role of the four styles of coping between adaptive and maladaptive perfectionism and perceived stress, mediation analyses were conducted. Results of the analyses indicated that Model 1 showed poor fit, χ2 (6, N = 323) = 144.51, p b 0.001, CFI = 0.78, RMSEA = 0.27, 90% CI [0.23, 0.31], SRMR = 0.10, and modification indices (MIs) suggested allowing the CISS Social Diversion subscale to covary with the CISS Distraction subscale. After completing the modification and rerunning the model, the path between the subscales was significant (p b 0.05). The model fit statistics indicated that the model with the modifications was a better fit than the original model, Δχ2 (1) = 80.38, p b 0.001. The CFI increased to 0.91, the RMSEA decreased to 0.19, 90% CI [0.15, 0.23], and SRMR decreased to 0.07. In conclusion, the modification improved the fit of the model; however, the overall model was still a poor fit. Consistent with the earlier modification in allowing various coping styles to covary, the MIs suggested allowing the CISS Distraction subscale to covary with the CISS Emotion-oriented subscale. After rerunning the model with this modification, the path between the subscales was significant (p b 0.05) and resulted in a better fitting model, Δχ2 (1) = 25.82, p b 0.001, the CFI increased to 0.95, the RMSEA decreased to 0.16, 90% CI [0.12, 0.21], and SRMR decreased to 0.05. The overall model, however, was still a poor fit. The MIs suggested allowing the CISS Social Diversion subscale to covary with the CISS Emotion-oriented subscale. The model was rerun with this modification resulting in a significant path between the subscales (p b 0.05) and a significant change in chi-square, Δχ2 (1) = 23.74, p b 0.001. The other fit indices also improved (CFI = 0.98; RMSEA = 0.11, 90% CI [0.06, 0.17]; SRMR = 0.03). The MIs were consulted one last time and suggested allowing the CISS Social Diversion subscale to covary with the CISS Task-based subscale. A rerun model with this modification resulted in a significant path between the subscales (p b 0.05) and a significant change in chisquare Δχ2 (1) = 8.81, p b 0.001. The other fit indices also improved: the CFI increased to 0.99, the RMSEA decreased to 0.07, 90% CI [0.00, 0.15], and SRMR decreased to 0.02 suggesting the overall model was a good fit for the data. No further MIs that were suggested could be theoretically justified. Next, to test for full mediation between both perfectionism dimensions and Perceived Stress, we restrained direct paths as described earlier. This model resulted in a non-significant increase in chi-square Δχ2 (2) = 3.26, p N 0.05. In addition, RMSEA decreased to 0.06, 90% CI [0.000,
Table 2 Fit indices and entropy for one- to six-class models. Model
BIC
aBIC
LMR value
LMR p-value
BLRT p-value
Entropy
1-Class model 2-Class model 3-Class model 4-Class model 5-Class model 6-Class model
5042.195 4995.511 4956.521 4953.819 4958.821 4973.247
5029.508 4973.307 4924.802 4912.585 4908.071 4912.982
– 60.525 53.250 18.942 11.659 2.748
– b0.001 b0.001 N0.05 N0.05 N0.05
– b0.001 b0.001 b 0.001 b0.05 N0.05
– 0.801 0.776 0.774 0.790 0.760
Note. BIC = Bayesian Information Criterion; aBIC = sample size adjusted BIC; LMR = Lo-Mendell-Rubin test; BLRT = Bootstrap Likelihood Ratio Test
J.S. Ashby, P.B. Gnilka / Personality and Individual Differences 119 (2017) 106–111
109
Fig. 1. Characteristics of classes.
0.12], SRMR and CFI remained constant. Thus, the four coping strategies fully mediated the relationships between both dimensions of perfectionism and stress (see Fig. 2). Next, we conducted additional analyses to determine the relative strength of the various mediation pathways by bootstrapping the pathways. For maladaptive perfectionism, taskoriented, emotional-oriented, and distraction coping were significant mediators. In regard to adaptive perfectionism, only task-oriented and emotion-oriented coping were significant mediators. Social diversion was found not to be a significant mediator for either dimension of perfectionism. 3. Discussion The results of this study are consistent with previous research that identified clear groups of adaptive, maladaptive, and non-perfectionists (e.g., Suh et al., 2017). Consistent with other studies (e.g., Rice & Ashby, 2007), the non-perfectionist group had low standards and moderate levels of discrepancy while the identified adaptive perfectionist group had high standards and low discrepancy and the maladaptive perfectionists had high standards and high discrepancy. These groups are also consistent with the theoretical literature (e.g., Hamachek, 1978) that indicated that both adaptive and maladaptive perfectionists were distinct from non-perfectionists in that that they hold high standards and consistently strive for perfection. The distinction between the groups of perfectionists is their relative ability to productively move forward when their standards are not met. Adaptive perfectionists, while disappointed at not having met their perfectionistic standards, are able to relatively quickly regroup and take on the next task. In contrast, maladaptive perfectionists are more likely to be highly self-critical.
In addition to clearly identifying groups of perfectionists, the results of this study offered evidence for the perfectionistic groups' differing levels of coping and stress. In this study, the adaptive perfectionists had significantly lower levels of perceived stress than either the maladaptive perfectionists or non-perfectionists. The maladaptive perfectionist group had significantly higher levels of perceived stress than the non-perfectionists. These results are consistent with numerous other studies (e.g., Ashby et al., 2012; Chang, Watkins, & Banks, 2004) that have examined the relationship between perfectionism and stress and found evidence for differences across perfectionism group. Relatedly, the results of this study also indicated that the three groups of perfectionists differed significantly in coping scores. Adaptive perfectionists, maladaptive perfectionists, and non-perfectionists all differed significantly in levels of task-oriented and emotional-oriented coping. Adaptive perfectionists in the study reported significantly higher use of task-oriented coping than either of the other two groups while maladaptive perfectionists reported the lowest use of this coping style. In contrast, maladaptive perfectionists reported the highest use of emotion-oriented coping while adaptive perfectionists reported the lowest use of this coping style. Adaptive perfectionists reported significantly higher use of social diversion coping than the other two groups who did not differ significantly from one another. The results of the study revealed no significant differences between adaptive perfectionists, maladaptive perfectionists, and non-perfectionists in their use of the distraction copying style. While a number of studies have offered evidence for the role of stress in the relationship of perfectionism to various outcomes, there has been limited investigation of the role of coping in the relationship between perfectionism and stress. Consistent with the Transactional
Table 3 Latent profile means, standard deviations, and Wald chi-square tests of mean equality using the BCH procedure. Auxiliary variable
Perceived stress Task-oriented Emotion-oriented Social diversion Distraction
Non-perfectionists (N = 90)
Maladaptive perfectionists (N = 100)
Adaptive perfectionists (N = 127)
M
SD
M
SD
M
SD
27.02 49.52 46.50 15.95 23.95
7.15 10.56 12.14 5.08 6.87
31.80 55.03 53.41 17.09 24.30
8.64 15.29 14.20 5.79 7.75
22.51 61.43 37.50 18.77 23.57
8.68 9.97 11.83 5.57 8.08
Note. A = adaptive perfectionists, M = maladaptive perfectionists, N = non-perfectionists. ⁎ p b 0.05. ⁎⁎ p b 0.01. ⁎⁎⁎ p b 0.001.
Global χ2
N vs. M
N vs. A
M vs. A
58.13⁎⁎⁎ 70.42⁎⁎⁎ 79.16⁎⁎⁎
16.06⁎⁎⁎ 7.72⁎⁎ 12.05⁎⁎ 0.10 1.92
17.06⁎⁎⁎ 69.44⁎⁎⁎ 29.16⁎⁎⁎
57.76⁎⁎⁎ 11.74⁎⁎ 73.24⁎⁎⁎ 0.43 4.31⁎
0.43 14.84⁎⁎
0.14 14.63⁎⁎⁎
110
J.S. Ashby, P.B. Gnilka / Personality and Individual Differences 119 (2017) 106–111
Fig. 2. Path model of the relations between adaptive and maladaptive perfectionism, coping, and perceived stress. The path coefficients are standardized regression coefficients. *p b 0.05.
Model of Stress (TMS; Lazarus & Folkman, 1984) suggesting that coping can buffer or augment the stress response, the results of this study found support for the role of coping in the prediction of stress. Specifically, the results of this study found evidence for the mediating role of task-oriented and emotion-oriented coping in the relationship between both adaptive and maladaptive perfectionism and stress. These results suggest that the higher levels of stress reported by maladaptive perfectionists may be related to these individuals' inability to address stressor directly (e.g., task-oriented coping) and “thereby increases the severity and duration of stress, leading to a greater susceptibility to experience additional stressors” (Dunkley, Solomon-Krakus, & Moroz, 2016, p. 160). In addition, maladaptive perfectionists may be more likely to engage in emotion-oriented coping consistent with the harsh self-evaluations generally associated with this cognitive style that may lead them to “magnify the negative aspects of events, thereby interpreting even mundane stressors as major threatening stressors” (Dunkley et al., 2016, p. 160). The results of this study also suggest that, in contrast to their maladaptive counter parts, adaptive perfectionists had higher levels of task-oriented coping and lower levels of emotion-oriented coping. Their tendency to engage in more task-oriented coping and less emotion-oriented coping may potentially buffer the impact of potential stressors, resulting in lower levels of perceived stress. This combination of coping is consistent with previous studies that have found adaptive perfectionists are as likely to utilize similar levels of active coping processes as maladaptive perfectionists in combination with significantly fewer unhealthy coping processes. For example, both Gnilka et al. (2012) and Noble, Ashby, and Gnilka (2014) found that while maladaptive perfectionists and adaptive perfectionists utilized similar levels of healthier coping processes, adaptive perfectionists utilized significantly fewer unhealthy coping processes. When looking at multiple coping processes simultaneously, it is important to note that these processes may vary by circumstance (Folkman, Lazarus, Dunkel-Schetter, DeLongis, and Gruen, 1986). One possible explanation for this may be that both adaptive and maladaptive perfectionists are active in facing stressors when first encountered but, over time, adaptive perfectionists may be less likely to shift to patterns of avoidance that would contribute to more chronic stress. This lack of a chronic stress response suggests maladaptive perfectionists may have a blunted cortisol response as suggested by Richardson, Rice, and Devine (2014).
This study had several limitations. First, we used an undergraduate student sample from a large university in the southeast. Future researchers should consider community and clinical samples to see if similar results are found. Second, this study was cross-sectional. While this type of design can add some understanding about how these different constructs may work together, we still need more longitudinal studies to be conducted that use three or more waves to determine causality. Third, more ethnically diverse populations should be utilized to see if the results of this study can be replicated. Fourth, this study used self-report measures and did not account for social desirability. Future researchers may want to consider controlling for social desirability. References Ashby, J. S., Noble, C. L., & Gnilka, P. B. (2012). Multidimensional perfectionism, depression, and satisfaction with life: Differences among perfectionists and tests of a stress-mediation model. Journal of College Counseling, 15, 130–143. Bakk, Z., & Vermunt, J. K. (2016). Robustness of stepwise latent class modeling with continuous distal outcomes. Structural Equation Modeling, 23, 20–31. Broman-Fulks, J. J., Hill, R. W., & Green, B. A. (2008). Is perfectionism categorical or dimensional? A taxometric analysis. Journal of Personality Assessment, 90, 481–490. Chang, E. C., Watkins, A., & Banks, K. H. (2004). How adaptive and maladaptive perfectionism relate to positive and negative psychological functioning: Testing a stress-mediation model in Black and White female college students. Journal of Counseling Psychology, 51, 93–102. Cohen, S., Kamarck, T., & Mermelstein, R. (1983). A global measure of perceived stress. Journal of Health & Social Behavior, 24, 385–396. Cole, D. A., & Preacher, K. J. (2014). Manifest variable path analysis: Potentially serious and misleading consequences due to uncorrected measurement error. Psychological Methods, 19, 300–315. Corry, J., Green, M., Roberts, G., Frankland, A., Wright, A., Lau, P., ... Mitchell, P. (2013). Anxiety, stress and perfectionism in bipolar disorder. Journal of Affective Disorders, 151, 1016–1024. Dunkley, D. M., Mandel, T., & Ma, D. (2014). Perfectionism, neuroticism, and daily stress reactivity and coping effectiveness 6 months and 3 years later. Journal of Counseling Psychology, 61, 616–633. Dunkley, D. M., Solomon-Krakus, S., & Moroz, M. (2016). Personal standards and self-critical perfectionism and distress: Stress, coping, and perceived social support as mediators and moderators. In F. Sirois, & D. Molnar (Eds.), Perfectionism, health, and wellbeing (pp. 157–176). New York, NY: Springer. Dunkley, D. M., Zuroff, D. C., & Blankstein, K. R. (2003). Self-critical perfectionism and daily affect: Dispositional and situational influences on stress and coping. Journal of Personality & Social Psychology, 84, 234–252. Endler, N. S., & Parker, J. D. (1994). Assessment of multidimensional coping: Task, emotion, and avoidance strategies. Psychological Assessment, 6, 50–60. Enns, M. W., Cox, B. J., & Clara, I. P. (2005). Perfectionism and neuroticism: A longitudinal study of specific vulnerability and diathesis-stress models. Cognitive Therapy & Research, 29, 463–478.
J.S. Ashby, P.B. Gnilka / Personality and Individual Differences 119 (2017) 106–111 Flett, G. L., Hewitt, P. L., Besser, A., Su, C., Vaillancourt, T., Boucher, D., ... Gale, O. (2016). The child-adolescent perfectionism scale. Journal of Psychoeducational Assessment, 34, 634–652. Folkman, S., Lazarus, R., Dunkel-Schetter, C., DeLongis, A., & Gruen, R. (1986). Dynamics of a stressful encounter: Cognitive appraisal, coping, and encounter outcomes. Journal of Personality & Social Psychology, 50, 992–1003. Gaudreau, P., & Thompson, A. (2010). Testing a 2 × 2 model of dispositional perfectionism. Personality & Individual Differences, 48, 532–537. Gnilka, P. G., Ashby, J. S., & Noble, C. M. (2012). Multidimensional perfectionism and anxiety: Differences among individuals with perfectionism and tests of a coping-mediation model. Journal of Counseling & Development, 90, 427–436. Gnilka, P. B., Ashby, J. S., & Noble, C. M. (2013). Adult attachment style differences among adaptive, maladaptive, and non-perfectionists. Journal of Counseling & Development, 91, 78–86. Hamachek, D. (1978). Psychodynamics of normal and neurotic perfectionism. Psychology, 15, 27–33. Hewitt, P. L., & Flett, G. L. (1991). Perfectionism in the self and social contexts: Conceptualization, assessment, and association with psychopathology. Journal of Personality & Social Psychology, 60, 456–470. Lazarus, R. S., & Folkman, S. (1984). Coping and adaptation. In W. D. Gentry (Ed.), The handbook of behavioral medicine (pp. 282–325). New York: Guilford. Martin, R. A., Kasarian, S. S., & Brieiter, H. J. (1995). Perceived stress, life events, dysfunctional attitudes, and depression in adolescent psychiatric inpatients. Journal of Psychopathology & Behavioral Assessment, 17, 81–95. Muthén, L. K., & Muthén, B. O. (1998–2017). Mplus user's guide (Eighth edition ). Los Angeles, CA: Muthén & Muthén. Noble, C. L., Ashby, J. S., & Gnilka, P. B. (2014). Multidimensional perfectionism, coping, and depression: Differential prediction of depression symptoms by perfectionism type. Journal of College Counseling, 17, 80–94. Nylund, K. L., Asparouhov, T., & Muthén, B. O. (2007). Deciding on the number of classes in latent class analysis and growth mixture modeling: A Monte Carlo simulation study. Structural Equation Modeling, 14, 535–569.
111
O'Connor, R., & O'Connor, D. (2003). Predicting hopelessness and psychological distress: The role of perfectionism and coping. Journal of Counseling Psychology, 50, 362–372. Parker, W. D. (1997). An empirical typology of perfectionism in academically talented children. American Educational Research Journal, 34, 545–562. Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879–891. Rice, K. G., & Ashby, J. S. (2007). An efficient method for classifying perfectionists. Journal of Counseling Psychology, 54, 72–85. Rice, K. G., Ashby, J. S., & Slaney, R. B. (1998). Self-esteem as a mediator between perfectionism and depression: A structural equation analysis. Journal of Counseling Psychology, 45, 304–314. Rice, K. G., & Lapsley, D. (2001). Perfectionism, coping, and emotional adjustment. Journal of College Student Development, 42, 157–168. Rice, K. G., & Van Arsdale, A. (2010). Perfectionism, perceived stress, drinking to cope, and alcohol-related problems among college students. Journal of Counseling Psychology, 57, 439–450. Richardson, C. M., Rice, K. G., & Devine, D. P. (2014). Perfectionism, emotion regulation, and the cortisol stress response. Journal of Counseling Psychology, 61, 110–118. Slaney, R. B., Rice, K., Mobley, M., Trippi, J., & Ashby, J. (2001). The revised almost perfect scale. Measurement & Evaluation in Counseling & Development, 34, 130–145. Smith, M. M., Saklofske, D. H., Stoeber, J., & Sherry, S. B. (2016). The big three perfectionism scale: A new measure of perfectionism. Journal of Psychoeducational Assessment, 34, 670–687. Stoeber, J., & Otto, K. (2006). Positive conceptions of perfectionism: Approaches, evidence, challenges. Personality and Social Psychology Review, 10, 295–319. Suh, H., Gnilka, P. B., & Rice, K. G. (2017). Perfectionism and well-being: A positive psychology framework. Personality & Individual Differences, 111, 25–30.