Gritty People Try Harder:Grit and Effort-Related Cardiac Autonomic Activity During an Active Coping Challenge

Abstract

Grit, a recently proposed personality trait associated with persistence for long-range goals, predicts achievement in a wide range of important life outcomes. Using motivational intensity theory, the present research examined the physiological underpinnings of grit during an active coping task. Forty young adults completed the Short Grit Scale and worked on a self-paced mental effort task. Effort-related autonomic nervous system (ANS) activity was assessed using impedance cardiography, which yielded measures of sympathetic activity (pre-ejection period; PEP) and parasympathetic activity (respiratory sinus arrhythmia; RSA). Multilevel models revealed that people high on the Perseverance of Effort subscale showed autonomic coactivation: both PEP and RSA became stronger during the task, reflecting higher activity of both ANS divisions. The Consistency of Interest subscale, in contrast, predicted only weaker sympathetic activity (slower PEP). Taken together, the findings illuminate autonomic processes associated with how “gritty” people pursue goals, and they suggest that more attention should be paid to the facets’ distinct effects.

1. Introduction

Why do some people try so much harder than others? The role of individual differences in effort is a classic problem in motivation science (Atkinson, 1964). Recently, researchers have proposed a trait known simply as grit, which represents “perseverance and passion for long-term goals” (Duckworth, Peterson, Matthews, & Kelly, 2007, p. 1087). The growing literature on grit shows that it predicts an impressive set of real-world markers of motivation and perseverance. Gritty adults have higher educational attainment and higher college GPAs (Duckworth et al., 2007), gritty children perform better in the National Spelling Bee (Duckworth et al., 2007), gritty military cadets are more likely to graduate from an elite military academy (Duckworth et al., 2007; Maddi, Matthews, Kelly, Villarreal, & White, 2012), and gritty teachers foster better academic performance in their students (Duckworth, Quinn, & Seligman, 2009). Grit predicts success in part by promoting self-control, thus allowing people to persist in repetitive, tedious, or frustrating behaviors that are necessary for success (Duckworth, Kirby, Tsukayama, Berstein, & Ericsson, 2011).

To date, no psychophysiological studies have examined how individual differences in grit manifest themselves in biological mechanisms of effort and motivation. Given the role of grit in fostering real-world success, it is important to understand the biological processes that allow gritty people to achieve long-range goals. Motivational intensity theory (Brehm & Self, 1989), a general model of how people regulate effort (Gendolla, Wright, & Richter, 2012; Wright, 1996; Wright & Kirby, 2001), offers a natural platform for developing predictions about how grit influences effort-related physiology. Motivational intensity theory proposes that effort is a function of two factors: the importance of success and the perceived difficulty of attaining the goal. The importance of success defines how much effort people are willing to expend; the difficulty of attaining it defines actual effort.

How might grit affect effort? According to motivational intensity theory, a trait can affect effort by making goals more or less important or by making achieving the goal seem more or less difficult. We expect, given theorizing about grit, that grit influences effort primarily via the importance pathway. Many traits have been shown to affect effort by affecting the importance of success (e.g., Capa & Audiffren, 2009; Capa, Audiffren, & Ragot, 2008; Silvia, Jones, Kelly, & Zibaie, 2011b). When a goal is more valuable, meaningful, or relevant to the self-concept, people are willing to expend more effort when necessary (Gendolla & Richter, 2010). Research on grit indicates that people high in grit are more passionate about their goals and more dedicated to accomplishing them (Duckworth et al., 2007), so it seems reasonable that the importance of success—and hence the level of potential effort—should be higher for gritty people.

Fortunately, motivational intensity theory offers a robust set of paradigms for illuminating the factors underlying effort, so it is easy to evaluate whether grit affects effort by making goals more important. One way to determine if higher success importance is involved is to analyze effort for unfixed-difficulty tasks (Wright, Killebrew, & Pimpalapure, 2002). Also known as self-paced, piece-rate, and do-your-best tasks, unfixed tasks allow people to work at their own pace and hence achieve as much or as little as they wish. For unfixed tasks, motivational intensity theory predicts that effort is solely a function of importance (Wright, 2008), a pattern that many experiments have found (e.g., Gendolla, Richter, & Silvia, 2008; Silvia, 2012; Silvia, Jones, Kelly, & Zibaie, 2011a).

In the present research, we examined how grit affected cardiac autonomic activity during an active coping challenge. Grit has two facets—Perseverance of Effort and Consistency of Interest, hereafter Perseverance and Consistency—and we were interested in whether these facets differentially affected the physiological mobilization of effort. Perseverance reflects commitment and effort toward one’s goals; it is measured with items such as “I finish whatever I begin.” Consistency reflects focus and dedication to a small set of important goals; it is measured with items such as “New ideas and projects sometimes distract me from previous ones” (reversed).

Despite the important conceptual differences between these facets, most published work on grit has used the total score and has not reported subscale-specific effects. The subscales correlate highly (r = .59; Duckworth & Quinn, 2009) but not so highly to preclude distinct effects. For example, Perseverance more strongly predicted middle- and high-school students’ GPAs and extracurricular activities, and Consistency more strongly predicted making fewer career changes during adulthood (Duckworth & Quinn, 2009). Regarding effort on a novel task, one could foresee potential differences. Perseverance should predict higher effort due to a tendency to take goals seriously and to try hard, whereas Consistency may not—novel goals that fall outside the small sphere of central goals might be seen as unimportant.

Effort was quantified in terms of cardiac autonomic activity. Our primary outcome was the cardiac pre-ejection period (PEP; the systolic interval between ventricular depolarization and the opening of the aortic valve), which is commonly used to measure beta-adrenergic sympathetic impact on the heart (Kelsey, 2012; Richter & Gendolla, 2009). Research on motivational intensity theory typically measures how the sympathetic division of the autonomic nervous system (ANS) influences cardiovascular activity (Wright & Gendolla, 2012). Sympathetic activity is a reliable indicator of motivational engagement in contexts requiring “active coping” (Obrist, 1976, 1981), such as when people must expend effort to attain an attractive goal or reward. Most past research has used systolic blood pressure (SBP) to quantify sympathetic impact during effortful engagement (Gendolla et al., 2012; Wright, 1996). In recent research and the present study, impedance cardiography is used to assess sympathetic influences on the heart more directly (Kelsey, 2012).

Sympathetic cardiac processes are more central to effort and motivational intensity theory, so PEP is our primary measure of interest. We also measured, as a secondary and exploratory outcome, parasympathetic influences on the heart. Respiratory sinus arrhythmia (RSA; high-frequency variability in heart rate as a function of respiration) was used to measure parasympathetic activity. A large literature supports PEP and RSA as markers of the activity of the sympathetic and parasympathetic ANS divisions (e.g., Berntson et al., 1994; Berntson, Cacioppo, & Quigley, 1993; Berntson, Norman, Hawkley, & Cacioppo, 2008; Cacioppo et al., 1994; Schächinger, Weinbacher, Kiss, Ritz, & Langewitz, 2001).

For our purposes, RSA was a secondary but interesting dependent variable: it isn’t a standard measure of effort during appetitive tasks, but it can illuminate emotional control and “cooling” in response to stress (Segerstrom et al., 2012). Measuring both sympathetic and parasympathetic influences is worthwhile because the divisions of the ANS can operate independently, jointly, or reciprocally (Berntson, Cacioppo, & Quigley, 1991; Koizumi & Kollai, 1992; Paton, Boscan, Pickering, & Nalivaiko, 2005). Moreover, some recent applications of motivational intensity theory have suggested measuring RSA, if only to provide additional information about autonomic processes during goal engagement (Richter, 2010). One example comes from recent work on how appraisals of goal relevance and goal conduciveness influence the relative activation of the sympathetic and parasympathetic branches (Kreibig, Gendolla, & Scherer, 2012). When an event was appraised as both relevant and conducive to a goal, people showed higher sympathetic and parasympathetic activation, which suggests a profile of heightened motivational engagement in response to an opportunity for reward and achievement (cf. Kreibig, Gendolla, & Scherer, 2010). Thus, one can see bridges between how goal-related appraisals affect autonomic activity (Kreibig et al., 2012) and motivational intensity theory’s historical interest in active coping processes (Wright, 1996). If both sympathetic and parasympathetic autonomic divisions reflect the cognitive appraisals involved in how people evaluate goals, then it is worth measuring both when evaluating how people engage with goals.

2. Method

2.1. Participants and Design

A total of 40 adults—24 men and 16 women—participated in the experiment. Participants were students in psychology courses who volunteered and received credit toward a research option or adults who responded to flyers advertising the study and received $10. Four participants were excluded, leaving a final sample of 36 people (21 men, 15 women). The sample was young (M = 19.83, SD = 2.52) and mostly European American (58%) or African American (36%). All participants provided informed consent. The experiment was approved and monitored by UNCG’s Institutional Review Board.

2.2. Materials and Procedure

Each participant was tested individually. The sessions were conducted by a gender-matched experimenter, who explained that the study was about how mental effort was reflected in the body, particularly in the activity of the cardiovascular system. The participants completed a series of personality questionnaires and then a computer-based cognitive task requiring mental effort. The questionnaires and tasks were completed using MediaLab and DirectRT (Empirisoft, NY). After the experimenter prepped the skin with alcohol wipes and placed the electrodes, the participants sat quietly and completed demographic and personality questionnaires on the computer for approximately 10 minutes. During the middle of this period, five minutes of baseline readings were taken.

2.2.1. Grit scale.

Grit was assessed using the Short Grit Scale (Grit-S; Duckworth & Quinn, 2009). This scale has eight items—four for Perseverance, four for Consistency—rated on five-point scales (1 = not like me at all, 5 = very much like me). Based on Cronbach’s alpha, internal consistency was α = .52 for Perseverance, α = .69 for Consistency, and α = .74 for the full scale. The Perseverance and Consistency subscales correlated r = .54 (p < .001), which is similar to past work (r = .59; Duckworth & Quinn, 2009). The scores were relatively normal: neither skewness nor kurtosis values were significant for the total scale or the subscales.

2.2.2. Parity task.

After the baseline period, participants were introduced to the mental effort task. The parity task, developed by Wolford and Morrison (1980), involves presenting a word flanked by two numbers, such as “3 BENCH 8.” Participants must ignore the word and decide if the two numbers have the same parity (both even or both odd) or a different parity (one even, one odd). The parity task is challenging because the centrally presented word is hard to ignore and because digit parity judgments are unfamiliar (Aquino & Arnell, 2007; Harris & Pashler, 2004), and it has been used successfully in past research on behavioral effort (Silvia & Phillips, 2013).

The numbers 2, 3, 5, and 8 were used to flank 16 neutral nouns (e.g., BOAT, CHAIR, MARKET). Each trial began with an intertrial interval (750 ms) and a fixation cross (350 ms) followed by a parity item. The black text was displayed in 28 pt Tahoma on a white background. The parity item remained on screen until people gave a response. They pressed a green key for “same parity” and a red key for “different parity,” using a DirectIN high-speed keyboard (Empirisoft, NY), which has a timing accuracy of under 1 ms.

The parity task was “unfixed” in difficulty because people could work at their own pace (Wright et al., 2002). The trial didn’t terminate until people responded, so responding more quickly allowed people to potentially get more correct. The task terminated after five minutes. Unfixed do-your-best tasks can become “quasi-fixed” if people are asked about their impressions of the task’s difficulty beforehand or if a fixed standard for performance is implied (Richter, 2010). To avoid this, we did not ask for ratings of the task’s difficulty or of people’s impressions of the task, and the task instructions focused on people doing their best and getting as many correct as they could.

2.3. Measurement of Cardiac Autonomic Activity

Cardiac activity was measured using a Bionex hardware system (Mindware, Gahanna, OH). An electrocardiogram (ECG) was assessed using a modified Lead-II configuration, in which electrodes were placed on the right clavicle and the left and right lowest ribs. An impedance cardiogram (ICG) was assessed using disposable spot electrodes placed in a standard tetrapolar configuration: two receiving electrodes were placed on the front of the body (an upper one 4 cm above the suprasternal notch, and a lower one at the base of the sternum on the xiphoid process), and two were placed on the back (one 4 cm above the horizontal plane of the first and one 4 cm below the horizontal plane of the xiphoid, roughly at the 3^rd and 9^th vertebrae). The signals were sampled at 1000 Hz and processed offline using bandpass filters (ECG and dZ/dt, .5 to 45 Hz; Z₀, 10 to 45 Hz).

The five-minute baseline period and five-minute task period were divided into 60-second epochs, and PEP and RSA values were obtained for each epoch. For PEP, the ECG and ICG dZ/dt waveforms were ensemble-averaged (Kelsey et al., 1998) for each 60-second epoch, and PEP was computed as the difference between the ECG Q-point (denoting the onset of ventricular depolarization; Berntson, Lozano, Chen, & Cacioppo, 2004) and the dZ/dt B-point (denoting the opening of the aortic valve and hence left ventricular ejection onset; Lozano et al., 2007). These points were automatically identified by the software (Mindware, Gahanna, OH). When corrected manually in a small number of cases, B-points were identified as the notch or inflection prior the dz/dt wave’s final upstroke (see Sherwood et al., 1990). For RSA, spectral methods were used to compute high-frequency heart rate variability in the respiratory frequency range (0.12–0.40 Hz) for each 60-second epoch. Respiration rate was estimated from the ICG Z₀ thoracic impedance signal (see Ernst, Litvack, Lozano, Cacioppo, & Berntson, 1999). Variation in Z₀ due to respiration can be identified using spectral analysis (see Ernst et al., 1999, for details). Only respiration rate (in cycles per minute), not peak volume or inspiration/exhalation intervals, was analyzed. Several studies have shown that thoracic impedance measures of respiratory rates are very close to spirometric measures (de Geus, Willemsen, Klaver, & van Doornen, 1995; Ernst et al., 2009; Houtveen, Groot, & de Geus, 2006).

3. Results

3.1. Analytic Model and Specification

We used multilevel models to accommodate the repeated-measures data structure (Kristjansson, Kircher, & Webb, 2007; Llabre, Spitzer, Siegel, Saab, & Schneiderman, 2004). The analyses were conducted with Mplus 7 using maximum likelihood with robust standard errors. Task period—baseline (scored 0) vs. task (scored 1)—was a within-person predictor. Grit scores—the Perseverance and Consistency subscales—were between-person predictors. In each model, we estimated the intercepts and slopes as random effects. The random intercepts and slopes were allowed to covary to reflect possible initial value effects. Because RSA varies with respiration rate (Berntson et al., 1993; Grossman & Taylor, 2007), the multilevel model for RSA included respiration rate as a time-varying (within-level) predictor. All within-person predictors were centered at each person’s own mean; all between-person predictors were centered at the sample’s grand mean (Enders & Tofighi, 2007). Table 1 displays the descriptive statistics.

Table 1.

Descriptive Statistics

	M	SE
Grit: Perseverance of Effort	3.78	.09
Grit: Consistency of Interests	2.71	.14
Number of Correct Responses	136.05	1.92
Response Time	964.32	32.11
PEP (Baseline)	125.06	1.87
PEP (Task)	124.84	2.18
RSA (Baseline)	6.06	.21
RSA (Task)	6.06	.18
Respiration Rate (Baseline)	18.09	.48
Respiration Rate (Task)	18.85	.48
IBI (Baseline)	760.20	18.17
IBI (Task)	773.26	17.03

Note. n = 36. PEP = pre-ejection period (in ms); RSA = respiratory sinus arrhythmia (in ms²); IBI = interbeat interval (in ms); respiration rate is in cycles per minute.

3.2. PEP

Did grittier people show higher sympathetic activity during the parity task? Our first multilevel model estimated the effects of task period and grit scores on PEP. The within-person main effect of task period was significant, b = −1.26, SE = .59, p = .032. As expected, PEP became faster from baseline to task, reflecting higher contractility due to greater sympathetic activity. The between-person main effects were not significant: neither Perseverance (b = 2.29, SE = 4.94, p = .643) nor Consistency (b = −.47, SE = 2.78, p = .865) significantly predicted PEP overall.

The interactions between task period and both subscales, however, were significant, albeit in different directions. The Perseverance subscale had a significant negative effect, b = −2.46, SE = .72, p = .001: as Perseverance increased, PEP declined more from baseline to task, reflecting greater effort. The Consistency subscale, in contrast, had a significant positive effect, b = 1.73, SE = .51, p = .001: as Consistency increased, PEP declined less from baseline to task. Stated differently, people high in Perseverance showed higher sympathetic activity during the task (faster PEP values), whereas people high in Consistency showed weaker sympathetic activity (slower PEP values).

3.3. RSA and Respiration Rate

Did grittier people show higher parasympathetic activity during the parity task? The within-person main effect of task period was not significant—overall, RSA values did not change from baseline to task, b = .09, SE = .09, p = .284. Moreover, there were no between-person main effects of the Grit subscales on RSA (Perseverance: −.31, SE = .39, p = .418; Consistency: b = −.03, SE = .26, p = .912). There was, however, a marginally significant interaction for Perseverance, b = .32, SE = .17, p = .056. As Perseverance increased, RSA went up from baseline to task, reflecting higher parasympathetic activity. The interaction for Consistency was not significant, b = −.06, SE = .13, p = .654.

For respiration rate, the within-person main effect of task period was significant: respiration rates were higher in the task period than the baseline period (b = .72, SE = .36, p = .044). There were no between-person main effects of the Grit subscales on respiration rate (Perseverance: b = 1.30, SE = 1.25, p = .299; Consistency: b = −.85, SE = .76, p = .264), and there were no interactions (Perseverance: b = 1.03, SE = .73, p = .162; Consistency: b = −.42, SE = .53, p = .430).

3.4. Task Performance

The goal for the task was to get as many correct as possible within the 5-minute task period. The task yielded two scores: the total number of correct responses achieved and response time for correct responses, which were correlated, r = −.81, p < .001. To explore performance effects, we ran a multivariate regression model with Perseverance and Consistency as predictors and the number of correct responses and response time as outcomes. Perseverance had a weak effect on the number of correct responses (β = .26, p = .149) and a marginally significant effect on response time (β = −.35, p = .080). Consistency was essentially unrelated to the number of correct responses (β = .05, p = .805) and to response time (β = .11, p = .629). Overall, then, grit scores had larger effects on the mobilization of effort during the task than on actual achievement, which is common given the imperfect relationships between effort and performance for most cognitive tasks.

Regarding task performance and cardiac autonomic activity, for the sample as a whole, as PEP declined the number of correct trials increased weakly (r = −.22, p = .214) and response time declined (r = .23, p = .162), but not significantly so. As RSA changed, neither the number of correct responses (r = −.03, p = .869) nor response time (r = −.01, p = .942) changed.

4. Discussion

4.1. Grit and Effort

Given the role of grit in real-world goal achievement, the present experiment examined how people who vary in facets of grit responded to a mental effort challenge. The study found support for an influence of individual differences in grit on effort-related cardiac activity. Based on theorizing about grit, which emphasizes dedication to goals (Duckworth et al., 2007), we expected that grit would influence effort via the potential motivation pathway: if grit makes goals more valuable, then people high in grit would be willing to expend more effort to achieve a goal. People high in Perseverance showed faster PEP values and marginally higher RSA values during the active coping task. Because the degree of effort during an unfixed, do-your-best task reflects the importance of success (Wright et al., 2002), this pattern suggests that people high in Perseverance appraised the task as more important, as one would expect.

From an autonomic space perspective (Berntson et al., 1991), people high in Perseverance showed a pattern of autonomic coactivation: sympathetic and parasympathetic activity both increased during the task. These ANS branches are often coactivated, and their joint influence can confer functional benefits (Koizumi & Kollai, 1992; Paton et al., 2005). When both branches are active, for example, the sympathetic branch can increase myocardial contractility without simultaneously increasing heart rate, which is reduced by the parasympathetic branch. This allows for more efficient cardiac output by affording longer ventricular filling times paired with higher contractility (Koizumi, Terui, Kollai, & Brooks, 1982).

People high in Consistency, in contrast, showed slower PEP values and no change in RSA. The PEP finding suggests that Consistency reduced the importance of succeeding at the parity task. Although not intuitive, this finding is consistent with the psychological basis of having consistent interests. The Consistency subscale measures focusing on and sticking with a small cluster of valuable goals. In past work, research examined how Consistency predicted the achievement of long-term goals that participants selected and found personally meaningful (Duckworth et al., 2007, 2009). In the present experiment, in contrast, people worked on a short-term goal with a standard that was assigned by the experimenter. Because the parity task is short-term and surely not in people’s set of valued goals, it’s possible that people high in Consistency found it less important. This interpretation is speculative and awaits further research, but it does suggest that there are cases in which grit, by fostering the selective prioritization of effort, can occasionally lead to lower effort. One way to test this possibility in future work would be to manipulate self-relevance, which should moderate the effect of Consistency on effort.

The effects were stronger for effort-related physiology than for behavioral task performance, a common finding in our research (e.g., Silvia, 2012; Silvia et al., 2011a, 2011b; Silvia, Kelly, Zibaie, Nardello, & Moore, in press; Silvia, McCord, & Gendolla, 2010) and in motivational intensity research generally. Effort is one of several contributions to task performance. For cognitive tasks, cognitive abilities (e.g., fluid intelligence, mental speed) and task strategies contribute substantial variance as well. Indeed, in instances of “compensatory effort”—such as when people try to overcome the effects of fatigue or sleep deprivation—high effort will be associated with poorer performance (e.g., Hockey, 1997; Schmidt, Richter, Gendolla, & Van der Linden, 2010; Wright & Stewart, 2012). It is notable, however, that the effect sizes were in the expected directions: both higher Perseverance and faster PEP were associated with getting more correct responses, albeit not significantly so.

One limitation of the proposed work was the inability to control for possible preload and afterload effects. PEP values can be influenced both by sympathetically-mediated contractility and by changes in ventricular preload and afterload (Sherwood et al., 1990). It is difficult to evaluate the potential contributions of preload and afterload effects without assessments of blood pressure, which weren’t available in the present analyses. An additional issue to be considered for future work is the manipulation of task difficulty. The present work clearly indicated that grit was associated with differences in potential motivation, one pathway in motivational intensity theory, but it might also influence perceptions of task difficulty, the other pathway. Because task difficulty wasn’t varied in the present study, this issue is worth examining in future research. Finally, we recognize that the assessment of heart rate variability is complex and contentious, with a great many metrics and disagreement among researchers (e.g., Allen, Chambers, & Towers, 2007). Our study used the spectral approach while controlling for (at the within-person level) variation in respiration rate, which is consistent with best practices. One virtue of our assessment approach is that it follows past work on autonomic coactivation, which typically used measures of PEP and frequency-based RSA as indicators of sympathetic and parasympathetic activity (e.g., Berntson et al., 2008). Nevertheless, we recognize the complexities involved in assessing RSA (e.g., Grossman & Taylor, 2007), which await resolution from future research.

4.2. Implications for Motivational Intensity Theory

Motivational intensity theory has historically been concerned with sympathetic activity. Most studies have assessed measures of sympathetic cardiovascular activity, particularly systolic blood pressure and PEP (see Gendolla et al., 2012; Wright, 1996). This is largely due to the influence of Obrist’s (1981) active coping approach, which emphasized sympathetic adjustments to mental and behavioral challenges, on early motivational intensity research (e.g., Wright, Contrada, & Patane, 1986). Nevertheless, recent research has begun to consider parasympathetic parameters, such as time- and frequency-domain measures of heart-rate variability (Richter, 2010).

Expanding motivational intensity research to include both ANS branches has several virtues. First, there is a large literature on the role of parasympathetic activity in both the functional and dysfunctional regulation of emotion and action (e.g., Segerstrom & Nes, 2007; Rottenberg, 2007). Motivational intensity theory, as a general model of effort regulation, could offer insight into such processes. Second, and more generally, assessing both branches affords a look at their joint activation. Some contexts promote coactivation of the sympathetic and parasympathetic branches. Kreibig et al. (2012), for example, found that appraising an event as relevant and congruent with a goal caused coactivation, which they interpreted as an engagement profile. In the present research, people high in Perseverance showed coactivation, which implies a more efficient cardiovascular profile when coping with the challenge (Koizumi et al., 1982). But relatively little is known about how these branches interact during active coping or what the consequences might be for effort, performance, recovery, and health. Understanding coactivation may thus further illuminate how the body mobilizes effort in the service of confronting challenges and achieving goals.

We thank Christina Chai Chang, Lance Moore, Joseph Nardello, and Rachel Sopko for their assistance with this research.