Introduction

Perfectionism is a personality construct characterized by having unrealistic performance standards and excessively critical self-evaluations [1]. Although perfectionism has been measured in different ways, it is widely recognized that perfectionism is a multi-dimensional construct with adaptive and maladaptive dimensions [1–3]. The adaptive form of perfectionism consists of intrinsic and self-directed beliefs for striving for perfection [1], and it is associated with positive domains of well-being such as personal growth [4], adaptive coping [2,5], and high self-efficacy [6]. The maladaptive form of perfectionism, however, consists of extrinsic beliefs for holding perfectionist standards—believing that others have unrealistic and unattainable expectations for them [1]—and it is associated with depression and anxiety [7], hopelessness [8], low self-efficacy [9], and avoidant coping styles [2,5].

In addition to emotional outcomes, maladaptive and adaptive forms of perfectionism relate differently to performance outcomes. People high on adaptive perfectionism tend to achieve higher grades and perform better on academic exams [10–12] and music competitions [12]. In contrast, maladaptive perfectionism doesn’t consistently relate to performance. Some studies have found that maladaptive perfectionism has a negative relationship with achievement [11,13], but other studies have found no relationship [6,14]. Overall, the adaptive component appears consistently related to high achievement, but the maladaptive form is negatively or unrelated to performance.

One reason adaptive perfectionism could affect achievement is that perfectionists may put more effort into achieving their goals. Few studies have directly examined the relation between perfectionism and effort. Two studies found that musicians high on adaptive perfectionism had higher effort, which was measured by time spent practicing a musical instrument [12,15]. In another study, people high on adaptive perfectionism rated higher subjective effort during a lab task [14]. Self-reported time and effort, however, have limitations as indicators of effort. First, time spent on a task and effort are not necessarily related: people might work intensely for brief periods or listlessly for long periods [16,17]. Second, self-report items, while widely used in motivation research, are sensitive to demand characteristics and self-presentational issues [18].

Because of the limitations of behavioral and self-report indicators of effort, a large tradition in biological psychology has developed that assesses physiological activity linked to effort. Research rooted in Brehm’s motivational intensity theory [19,20], for example, emphasizes effort-related cardiovascular changes that occur when people confront both physical and mental challenges [21,22]. A contemporary measure of beta-adrenergic sympathetic activation, the cardiac pre-ejection period (PEP), is often used as a physiological indicator of mental effort. As the sympathetic branch of the autonomic nervous system become more active on the heart, it contracts more forcefully, which is apparent in faster PEP scores [23]. Although sympathetic measures of effort are more common [24], recent research has emphasized examining both branches of the autonomic nervous system. Grit, a personality construct that captures the ability to sustain effort towards long-term goals, was related to co-activation of the sympathetic and parasympathetic branches [25]. Specifically, Perseverance, a subscale from the Short Grit Scale [26], predicted higher levels of effort engagement—reflected in faster PEP and higher RSA—during a mental effort task. This finding draws attention to the importance of looking at both branches of the autonomic nervous system during mental tasks.

To date, only one study has examined perfectionism and cardiac reactivity. Besser, Flett, Hewitt, and Guez [27] looked at how maladaptive and adaptive perfectionism interacted with task performance, confidence, and manipulated feedback in predicting systolic blood pressure (SBP), a measure of sympathetic activation [24]. Although this project showed a relation between perfectionism and sympathetic activation, it did not measure physiological activity during the task. Therefore, its findings reflect effects of perfectionism on reactions to false feedback rather than to the physiological engagement of effort when faced with a challenge.

In addition to examining motivational aspects of perfectionism, using physiological measures in research on perfectionism can help bridge basic scientific research on perfectionism with health outcomes. Specifically, recent research indicates that high levels of perfectionism may increase mortality risk [28]. Moreover, research on perfectionism and coping suggests that perfectionists may have dysfunctional stress responses, which may explain some of the increased risk for maladaptive health outcomes [29]. The use of physiological measures, such as cardiac reactivity, in the perfectionism literature may provide additional information on the biology of perfectionism and help connect the behavioral research on perfectionism to health psychology research.

Methods

Results

Descriptive Statistics and Analytic Approach

Descriptive statistics are presented in Table 1. For each physiological outcome, scores for the 5 baseline minutes were averaged to form a baseline value, and scores for the 3 parity task minutes were averaged to form a task value. We analyzed the data using regression models, as in our recent work [43,44]. Reactivity scores—change from baseline to task—were computed by subtracting the baseline value from the task value. These reactivity scores served as the dependent variables. The predictors were the perfectionism factors (SOP and SPP) and the physiological baseline value, which was included to control for potential initial-value effects. The predictors were centered at the sample’s grand mean, giving them a mean of 0. The intercept thus represents the average within-person change from baseline to task, and the regression coefficients represent how the predictors influence reactivity.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Descriptive statistics.

https://doi.org/10.1371/journal.pone.0160340.t001

The models were estimated using Mplus 7.4, using maximum likelihood with robust standard errors. Unless noted otherwise, all regression coefficients are standardized, and 95% confidence intervals are in square brackets. The standardized regression weights are can be viewed as effect sizes in the r metric based on conventional cut-offs: .10 (small), .30 (medium), and .50 (large). The raw data and Mplus input are available at Open Science Framework (https://osf.io/8sjhr/).

Task Performance and Self-Report Outcomes

Table 2 displays the descriptive statistics and standardized effects for task performance and the subjective self-report items. A multivariate regression model estimated the effects of SOP and SPP on the number of correct items, response times for correct responses, self-reported task difficulty, and self-reported performance. The primary performance outcome was the number of correct items because the task’s goal and incentive structure emphasized getting as many correct as possible.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Standardized Effects of SOP and SPP on parity task performance and self-report outcomes.

https://doi.org/10.1371/journal.pone.0160340.t002

Neither SOP nor SPP significantly predicted the number of correct responses on the parity task; the effect for SPP was marginal (β = -0.16 [-.34, .02], p = 0.008; see Table 2), suggesting that people higher in socially prescribed perfectionism tended to do worse. For response times, which were necessarily highly correlated with the number of correct responses (r = -.72), neither SOP (β = -0.05 [-.26, .15], p = 0.606) nor SPP (β = .07 [-.12, .27], p = 0.470) had a significant effect. For self-reported task difficulty, SPP significantly predicted ratings of how hard the task was (β = 0.27 [.08, .46], p = 0.005), reflecting higher rated difficulty at higher levels of SPP. SOP’s effect on self-reported task difficulty was marginal but in the other direction (β = -0.19 [-.39, .02], p = 0.072): self-oriented perfectionists tended to rate the task as easier. Neither SOP nor SPP significantly predicted participants’ self-report ratings of how well they did on the task (see Table 2).

Discussion

The current study examined how individual differences in perfectionism are related to effort, measured by autonomic cardiac reactivity during a self-paced task with monetary incentives. The parity task did engage participants as expected. The sample as whole showed faster PEP values during the task, indicating greater sympathetic reactivity, as well as higher RSA values during the task, indicating greater parasympathetic reactivity. The task and its incentives thus successfully motivated the participants in ways that past research would predict. These reactivity effects, however, were not moderated by perfectionism. Neither SOP nor SPP significantly predicted the degree of physiological change from baseline to task for either PEP or RSA. The sample as a whole thus showed increased effort engagement, but differences in how much people engaged were not associated with perfectionism.

The findings from the current study provide additional support for motivational intensity theory and for co-activation of both branches of the autonomic nervous system. Overall, people tried harder during the task, which was evidenced by increased sympathetic and parasympathetic activation. This fits past research, which finds that tasks with unfixed difficulty provide a paradigm where importance of success can be easily measured with effort intensity [19,30]. Few studies have examined both branches of the autonomic nervous system during appetitive tasks requiring effort. However, these findings, along with Silvia and colleagues’ [25] findings, show evidence of a co-activated profile. A review by Segerstrom and colleagues [34] also highlights evidence of a co-activated profile, which may occur during a task that is both challenging and requires self-regulation. Moreover, research indicates that co-activation may be an optimal profile for short-term engagement [33] because it increases cardiac efficiency via higher contractility and a slower heart rate, which affords longer ventricular filling times [49].

The lack of interactions involving perfectionism suggests that perfectionism may not influence effort measured by physiological reactivity. Past research has shown that individual differences in perfectionism predict effort measured by self-reported effort and time spent practicing [12,14,15]. However, within the current study, perfectionism did not predict effort in a paradigm that has been effective in examining whether individual-differences predict effort (i.e., grit, self-focused attention, and anhedonic depression; [25,36,43]). Moving forward with research on perfectionism and motivation, it will be important to use physiological measures of effort in addition to self-report and performance measures because these indicators of effort commonly diverge.

Additionally, perfectionism predicted task performance and appraisal of task difficulty. SPP predicted higher ratings on the self-report item assessing the perception of task difficulty on the parity task. This finding fits with past research, which found that people high on SPP are more likely to rate tasks as more difficult [50]. Furthermore, SPP had a marginally significant, small effect (β = -.16) on the number of correct responses, suggesting they were somewhat less effective in working toward the task’s goal of getting as many correct responses as possible. Maladaptive perfectionism is either negatively or inconsistently associated with performance [6,11,13,14]. The findings from the current study suggest that people high on SPP perceived the task to be more difficult and did not perform as well on it, which contributes to the literature suggesting that SPP may have more negative motivational and performance outcomes.

It is possible that the parity task used in the current study somehow failed to activate perfectionistic goals that would draw out individual differences in effort. Although the parity task isn’t appreciably different from other tasks that have been used to examine perfectionism [14,51,52], it is fundamentally an appetitive task that seeks to activate approach-oriented motivation by offering incentives that are tied to performance and allowing people to work as quickly or slowly as they like. Future studies could use priming techniques to increase the importance of tasks by priming self-relevance or social evaluation [51]. Additionally, SOP was expected to predict effort intensity based on previous findings measuring effort with self-report and performance outcomes; however, the current study used an extrinsic incentive (i.e., 3 cents per correct answer) to increase motivation during the task. SOP has been associated with the pursuit of intrinsically-motivated goals. Therefore, future research should examine whether there is an effect of SOP on effort intensity using an intrinsic motivation task.

In the current study, perfectionists’ appraisal of their performance was not associated with levels of perfectionism. However, performance appraisal was collected with only a single item. Past research on perfectionism and task performance suggests that the different domains of perfectionism may have differing physiological reactions to perceived versus actual task performance [27]. Furthermore, research suggests that adaptive forms of perfectionism are associated with higher levels of self-efficacy [3,6], which may impact one’s motivation during tasks. Therefore, future studies should include more detailed measures of perceived appraisal of performance to examine whether this explains increased or decreased effort in perfectionists. More extensive measurement of perfectionists’ evaluation of tasks, their performance, and their physiological effort may allow for further understanding of the increased likelihood of that perfectionists engage in maladaptive motivational behavior.

Another limitation of the current study is the inclusion of only female participants. This was due to the participant pool being comprised of mostly female students and the limited availability of male research assistants. Based on previous research, gender differences have occasionally been found for physiological reactivity as measured by diastolic blood pressure [53,54] and heart rate [53], although the motivational intensity literature as a whole finds few gender differences.

The current study highlights the importance of assessing physiological measurements of effort, which can provide additional information about task engagement beyond self-report and behavioral measures. Although research suggests individual differences in perfectionism predict performance outcomes, the current study indicates this may not be due to differences in mental effort measured by autonomic reactivity. In the future, identifying discrepancies between performance, evaluation of performance, and effort may be beneficial for understanding perfectionsts’ tendency towards self-defeating behaviors, such as procrastination and self-criticism.

Acknowledgments

We thank Ashley McHone and Zuzana Mironovová for their assistance with this project.