Individual differences in experience after a control task: Boredom proneness, curiosity, and grit correlate with emotion

Abstract

Objectives

The aims of the present study are to investigate (a) whether a control task (i.e., the grocery imagery task) classically used in emotion research and designed to emulate emotional neutrality is truly neutral and (b) how boredom proneness, curiosity, and grit influence emotions following this control task.

Method

State feeling of emotions were measured on 178 participants before and after the control task. Boredom proneness, curiosity, and grit was measured to examine the effect of these personality traits on participants' emotion before and after the control task.

Results

The findings demonstrate that the control task does not evoke a neutral state, but rather, it results in fluctuations in emotional experience and these fluctuations are only partially explained by personality traits associated with engagement and disengagement (boredom, curiosity, and grit). External boredom proneness was positively correlated with post‐task boredom, resignation, and sadness. Internal boredom proneness and the feeling‐of‐interest dimension of curiosity were both positively correlated with joy and challenge/determination, whereas the feeling‐of‐deprivation dimension of curiosity was negatively correlated with joy. Finally, grit was positively correlated with interest.

Conclusion

The present study found that even following a control task intended to be emotionally neutral, emotional experience differs and personality can paint the way we perceive and thus react to these control conditions. These findings highlight the importance of considering the neutrality of control tasks and the effects of personality on control tasks in emotion research.

Key words:

• boredom
• curiosity
• emotion
• grit
• neutrality

WHAT IS ALREADY KNOWN ABOUT THIS TOPIC?

• In the field of emotion research, control conditions often strive for “emotional neutrality” by having participants engage with a relatively non‐emotional external stimulus.

• The current literature shows that personality, predominately the Big 5, has a significant influence on emotional experience.

• The current research shows that a commonly used control task (i.e., the grocery imagery task) is not neutral and individual differences beyond the Big 5 has a significant effect on emotional experience toward a control task.

WHAT THIS TOPIC ADDS?

• Our research on this topic suggests that future research in emotion should account for the effects of personality traits on emotional baselines and any emotionally neutral benchmark.

Routinely in experimental endeavours, scientists use control conditions as a comparison against experimental conditions of interest. In the field of emotion research, these control conditions often strive for “emotional neutrality” by having participants engage with a relatively non‐emotional external stimulus (Rottenberg, Ray, & Gross, 2007). Contrasting these control conditions with experimental conditions are the basis from which we draw conclusions about the independent variables of interest. Such conclusions may be confounded or distorted if the control condition is not emotionally neutral. Furthermore, individual differences in how participants engage with these control conditions are not well understood. For instance, is the control task really neutral? Do certain personality traits predict emotional responses to control conditions? The present study aimed: (RO1) to investigate if engaging in an emotionally neutral directed imagery task, which is a commonly used method for emotion induction, is truly neutral; and (RO2) to understand how engagement‐related personality traits influence an individual's emotional experience following a control task. Specifically, we examined how boredom proneness, trait curiosity, and grit predict emotions such as boredom, challenge/determination, resignation, interest, happiness/joy, and sadness in a control task.

Emotions are considered to have multiple components including subjective experience and patterns of cognitive appraisal, motivational tendencies, physiological reactivity, and bodily/facial expression (Izard, 2010; Lazarus, 1991). Interest, joy, and determination are emotions considered to be subjectively pleasant; these positive emotions are theorised to motivate attentional engagement (for a review, see Kirby, Morrow, & Yih, 2014 and Silvia, 2008). In contrast, boredom, resignation, and sadness are negative emotions that are subjectively unpleasant and that motivate task disengagement (for a review, see Bench & Lench, 2013 and Smith & Kirby, 2009a, 2009b).

Three personality traits related to task engagement and disengagement might be particularly associated with the engagement‐related positive emotions of interest, joy, and determination as well as the disengagement‐related negative emotions of boredom, sadness, and resignation: (a) boredom proneness; (b) trait curiosity; and (c) grit. Thus, the current research specially explores how these traits influence emotional responses to a control task: thinking about and imagining going to the grocery store. Such a task is similar to the control conditions used in studies that induce emotion using autobiographical reflection or directed imagery (e.g., Lench, Flores, & Bench, 2011; Miller, Patrick, & Levenston, 2002; Velasco & Bond, 1998). Thus, the present study not only contributes to the growing literature on personality and emotion, but also offers substantial insight into individual differences in responding to control conditions commonly used in emotion research, which carries potential implications for the reliability of prior research. Existing research has focused on how a stimulus and its properties impact emotion (Barrett, Mesquita, Ochsner, & Gross, 2007), showing that appraisals of stimulus attributes and their relations to the goals and motivations can differentiate emotions (Tong, 2015). Other research describes how more distal factors, such as personality, modulate the appraisals that influence emotional experience. In fact, a consideration of how individual differences influence emotion provides a richer understanding of emotion and the individual differences in how it is experienced and regulated (Livingston & Srivastava, 2014). Personality might impact our “appraisal styles,” or the tendency to appraise situations in a certain way and thereby influence emotion (Ellsworth & Scherer, 2003). Impressive evidence has accumulated to support the role of personality in affective experience. Larsen and Ketelaar (1991) demonstrated that more neurotic participants showed greater reactivity to a negative emotion induction, whereas more extraverted participants showed greater reactivity to a positive emotion induction compared to their introverted peers. Subsequent studies have supported the relationships between affect and personality, especially with the Big‐5 traits of extraversion and neuroticism (Livingston & Srivastava, 2014; Shiota, Keltner, & John, 2006).

Previous literature has focused on the Big‐5 personality traits, and this research has clearly supported a link between personality and affect. Yet, it is not well understood how personality influences discrete emotions or how other traits that extend beyond the Big‐5 are associated with emotion. How does personality correspond with discrete emotion, especially specific positive emotional experiences? Answers to this question are of relevance to research on emotion differentiation and positive emotion (Smith, Tong, & Ellsworth, 2014; Tong, 2015). For engagement‐related emotions and disengagement‐related emotions alike, traits such as boredom proneness, curiosity, and grit likely play a role in how these emotions are experienced. Studies such as this are necessary to add to a growing body of empirical research on the predictive utility of narrower personality traits, such as grit (Muenks, Wigfield, Yang, & O'Neal, 2016; Skues, Williams, & Wise, 2017). These more focused traits may be able to explain additional variance in emotions, beyond that which can be explained by the Big‐5 (Rimfeld, Kovas, Dale, & Plomin, 2016). In the following sections, we review the literature to develop specific hypotheses on the influence of these traits on emotion.

RELEVANT LITERATURE AND HYPOTHESES DEVELOPMENT

In previous research there is no clear definition of what emotionally “neutral” actually is. It may be conceptualised as the absence of emotion, but a rather more likely definition is that “neutral” refers to a baseline emotional state or no changes to emotional experience. We suggest that there is no way to be truly devoid of emotion and thus a neutral emotional state may involve slight boredom or mild feelings, with strong emotional deviations from this baseline classified as “not‐neutral.” In fact, the grocery task or similar tasks have been widely used as a control condition in the emotion literature (e.g., Chepenik, Cornew, & Farah, 2007; Cohen, Barker, & White, 2018; Lench et al., 2011; Miller et al., 2002; Speer, Bhanji, & Delgado, 2014; Velasco & Bond, 1998). A recent paper by Cohen et al. (2018) examining affect‐related life events also indicate that “many neutral event items involve engaging in a variety of relatively simple and familiar tasks, such as using a calculator or making a grocery list.” (p. 1809). In this study, the task of coming up with a grocery shopping list was used as the control task and this directed imagery task was designed to emulate other similar control conditions used in emotion research whereby participants are asked to imagine neutral life events. Measuring and comparing the level of emotions before and after such a supposedly control task may provide a means to examine whether control tasks are truly neutral and, if not, how such tasks influence our emotions. To address RO1, we therefore hypothesised that:

Hypothesis

The control task is not truly neutral, whereby there will be significant differences in emotions before and after the control task.

Furthermore, if the task is neutral, then specific personality (i.e., boredom proneness) should predict relevant emotion (i.e., boredom) after the control task. For instance, an individual with high boredom proneness should feel more bored after the control task. However, if the task is not neutral, then personality traits may not completely account for the differences in emotion. For instance, the control task may elicit emotions such as boredom regardless of individual differences in boredom proneness. Thus, if the control task lacks neutrality and results in different in emotions, post‐control‐task emotions should not be predicted by boredom proneness, trait curiosity, and grit, at least not as strongly as features of the task itself.

Boredom proneness

Boredom proneness captures an individual's trait‐level desire to, but inability to, engage in satisfying and stimulating activity (Bench & Lench, 2013; Goldenberg, Eastwood, LaGuardia, & Danckert, 2011; Vodanovich, Wallace, & Kass, 2005). Similarly, state experiences of boredom are characterised as a negative emotion associated with deactivation and low arousal (Pekrun, Goetz, Daniel, Stupnisky, & Perry, 2010). Thus, individuals with high boredom proneness are more susceptible to boredom and thus task disengagement (Bench & Lench, 2013; Pekrun et al., 2010), suggesting that boredom proneness may be an important personality trait that explains individual differences following control conditions.

Along with the intuitive connection between boredom proneness and experiences of boredom, previous findings have also linked boredom proneness to different negative affective states related to aggression, amotivation, apathy, depression, hopelessness, loneliness, and negative self‐awareness (Goldenberg et al., 2011; Vodanovich, Verner, & Gillbride, 1991). These relationships are not surprising given that a lack of challenge and meaning has been found to predict boredom (Bench & Lench, 2013). Boredom has been conceptualised as an inverse to interest (Silvia, Henson, & Templin, 2009) that is characterised by a lack of attention, engagement, and meaning, which might correspond with decreased joy (Bench & Lench, 2013; Vodanovich et al., 1991). Given the relationship between boredom and task disengagement, trait‐level boredom proneness should be positively associated with not only boredom but also resignation and sadness, and negatively linked to interest and joy—especially in the context of a control task that is not intended to provide stimulation or elicit emotion.

Furthermore, research has identified two dimensions of boredom proneness. Internal boredom proneness refers to one's inability to generate internal stimulation. In contrast, external boredom proneness refers to one's perception that there is a lack of external stimulation in the surrounding environment (Vodanovich et al., 2005). Internal boredom proneness is related to the inability to generate stimulating activities, through mechanisms such as creativity, whereas external boredom proneness is related to the need for variety and change (Vodanovich et al., 2005). Thus, in terms of experience following an emotionally neutral control condition, we hypothesised that:

2a

Hypothesis

External boredom proneness will have a significant and positive relationship with negative emotions, especially boredom, resignation, and sadness.

2b

Hypothesis

External boredom proneness will have a significant and negative relationship with positive emotions, especially interest and joy.

Trait curiosity

Trait curiosity is an innate desire for novel knowledge and experience (Litman & Jimerson, 2004). Like boredom proneness, research has shown that trait curiosity consists of two dimensions. “Feeling‐of‐interest” is the enjoyment of seeking and learning new knowledge, whereas “feeling‐of‐deprivation” refers to the displeasure caused by a lack of meaningful and substantive information (Litman, Hutchins, & Russon, 2005; Litman & Jimerson, 2004). Although both dimensions of curiosity are related to novelty seeking and engagement, the feeling‐of‐interest dimension motivates acquisition of new knowledge, whereas the feeling‐of‐deprivation dimension motivates the reduction of informational deprivation.

Research has demonstrated a robust positive relationship between trait curiosity and the positive emotion of interest. Silvia (2008) showed that individuals with high trait curiosity experience more engagement toward novel and complex stimuli because curiosity predicts appraisals of coping potential, which refer to evaluations that one can cope with the uncertainty and complexity of a novel experience. Silvia et al. (2009) replicated these findings and showed that curious individuals tend to experience interest in response to abstract artwork. Taken together, the shared relationship with novelty provides a strong theoretical framework for understanding how trait curiosity might give rise to engagement and experiences of interest; these emotional experiences may be a mechanism through which we seek, attain, and understand novelty.

Notably, existing research on curiosity and interest has not differentiated among the dimensions of curiosity. The present study therefore extends this literature by examining whether feeling‐of‐interest and feeling‐of‐deprivation have unique effects on experience following an emotionally neutral control condition. Previous findings suggest that feeling‐of‐interest should be related to positive emotions that motivate novelty seeking, whereas feeling‐of‐deprivation should be related to negative emotions that motivate novelty seeking, such as boredom. Based on the literature, we hypothesised that:

3a

Hypothesis

The feeling‐of‐interest dimension of trait curiosity will have a significant and positive relationship with positive emotions, especially interest.

3b

Hypothesis

The feeling‐of‐deprivation dimension of trait curiosity will have a significant and positive relationship with negative emotions, especially boredom.

Grit

Grit refers to the tendency to engage with and pursue challenging long‐term goals with passion and perseverance (Duckworth, Peterson, Matthews, & Kelly, 2007). Grit is associated with long‐term success and well‐being (Eskreis‐Winkler, Gross, & Duckworth, 2016), and tends to co‐occur with high scores of extraversion and low scores of neuroticism, thereby giving way to a generally positive personality among gritty individuals (Hill, Burrow, & Bronk, 2014). In line with this profile, previous research suggests a relationship between grit and positive affect (Hill et al., 2014). To extrapolate from the relationship between grit and positive affect, we suspect that grit may be negatively correlated with negative affect. However, the literature has not elaborated on the specific emotions, aside from happiness/joy, associated with grit.

The gritty individual is likely to appraise situations as ones that can be effortfully acted upon such that the situation is in line with personal goals. Thus, being gritty may increase the likelihood of appraising high coping potential in the face of motivationally incongruent situations, and thus, grit should predict determination and decreased resignation (Kirby et al., 2014). The determination theorised to be experienced by gritty individuals is likely one mechanism that sustains the engagement necessary to achieve long‐term goals. Thus, establishing a positive association between grit and determination would provide support for the previously proposed notion that experiences of determination are one mechanism through which grit predicts success and well‐being. Based on the existing research, we hypothesised that:

4a

Hypothesis

Grit will have a significant and positive relationship with positive emotions, especially determination and joy.

4b

Hypothesis

Grit will have a significant and negative relationship with negative emotions, especially resignation.

THE CURRENT STUDY

In the present study, we investigated whether the grocery shopping task, a supposedly non‐emotional control condition that relied on methods classically used in the emotion literature, is truly neutral. To do this, we compare participants' emotional experience before and after the control task. Furthermore, if any differences arises between pre‐ and post‐emotion, the study examines whether the difference(s) is attributed to three personality traits associated with engagement and disengagement—boredom proneness, trait curiosity, and grit. The control task involved thinking about an emotionally neutral situation, an approach that has classically been used as a control condition in research that induces emotion using autobiographical reflection or directed imagery (Lench et al., 2011; Miller et al., 2002; Velasco & Bond, 1998). We focused on the effects of personality on emotion, and thus chose an emotionally “blank slate” that would allow for the “painting” of emotion by personality. We build upon the research on personality and emotion by clarifying how boredom proneness, curiosity, and grit influence emotional experience. The present study also sheds lights on how researchers should consider the effects of personality, especially traits related to task engagement, on control tasks that are often used as control conditions in affective science.

METHOD

Participants

To our knowledge, no evidence on the relationships between our personality traits and emotions of interest exists to guide judgments about reasonable effect sizes to address the central aims of the present study. However, previous meta‐analyses across more than 25,000 published psychology studies reported an average effect size of r = .21 (Richard, Bond, & Stokes‐Zoota, 2003). Based on a statistical power analysis of this expected effect size (α = .05, β = .20), we needed at least 175 participants in the present study to be sufficiently powered, and powered our study to detect effects of that threshold.

Thus, the present study included 178 participants (53% female) between 18 and 68-years old (M = 27.16, SD = 10.26) with 66% identifying as White/Caucasian, 17% as Asian, and the remaining 17% as African American, Hispanic, Pacific Islander, or other ethnicities. Our sample included 86 undergraduates (M_age = 19.92, SD_age = 2.06; 67% female; 65% White/Caucasian) from either a university in Australia (43%) or in the United States (57%), who participated in exchange for course credit or money, as well as 92 workers from Amazon Mechanical Turk (M_age = 33.92, SD_age = 10.25; 40% female; 67% White/Caucasian) who participated for money. Attention checks (e.g., “if you are paying attention, please select disagree”) were used to ensure the validity and quality of the data collection from the undergraduate and mTurk workers samples. We also excluded any invalid cases that did not pass all attention check or responded to the whole survey as straight‐liners (i.e., the same response for 5 different questions in a row). The present study was approved by each university's respective review boards.

Measures

Non‐emotional, neutral control task

Participants either wrote about or imagined going to the grocery store, a situation intended to be neutral. As a part of a broader study, participants were assigned to either the writing or the directed imagery condition. Controlling for the delivery of the neutral grocery task (imagining versus writing) in our regression models did not alter our findings, and so we have collapsed these control conditions together.

Emotion questionnaire

We assessed emotion by listing a set of adjectives that describe a particular emotion. Participants rated how much they were currently experiencing each emotion using a 9‐point rating scale (ranging from 1 = “Not at all” to 5 = “Moderately” to 9 = “Extremely”). Based on our hypotheses, we were particularly interested in boredom, determination, interest, joy, resignation, and sadness. Participants rated their current experience of: (a) boredom by responding to “bored, detached, uninterested”; (b) determination by responding to “determined, persistent, motivated”; (c) interest by responding to “interested, engaged, intrigued”; (d) joy by responding to “joyful, happy, glad”; (e) resignation by responding to “defeated, resigned, beaten”; and (f) sadness by responding to “sad, downhearted, blue.” Emotion ratings were collected both before and after the control task. The test–retest reliability, based on correlations between pre‐ and post‐task ratings for each respective emotion, were acceptable, as listed in Table 1 (r ≥ .53).

Table 1. Correlations among all variables

	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19
1. Age	−.09	−.21	−.21	−.18	.27	−.05	.20	−.11	−.12	−.11	.13	−.07	.13	−.08	.16	−.08	.05	.24
2. Sex		−.13	.05	.01	.00	−.07	−.06	−.15	−.11	−.08	.07	−.01	−.03	−.12	.02	−.09	.07	−.09
3. Pre‐boredom			.38	.43	−.40	−.14	−.21	.53	.38	.36	−.18	−.08	−.13	.40	−.29	−.04	−.20	−.29
4. Pre‐resignation				.77	−.37	−.20	−.22	.29	.64	.63	−.11	−.17	−.15	.24	−.30	.00	−.23	−.38
5. Pre‐sadness					−.33	−.33	−.26	.31	.55	.70	−.21	−.17	−.21	.26	−.31	.01	−.28	−.37
6. Pre‐interest						.47	.67	−.27	−.26	−.32	.60	.42	.52	−.20	.43	.19	.31	.40
7. Pre‐joy							.47	−.12	−.14	−.16	.49	.65	.42	−.17	.37	.20	.32	.33
8. Pre‐determination								−.17	−.17	−.24	.49	.40	.70	−.22	.47	.33	.40	.43
9. Post‐boredom									.38	.32	−.42	−.31	−.24	.35	−.24	.01	−.20	−.26
10. Post‐resignation										.65	−.18	−.16	−.09	.34	−.21	.00	−.15	−.29
11. Post‐sadness											−.30	−.25	−.22	.31	−.31	−.03	−.26	−.30
12. Post‐interest												.58	.61	−.25	.31	.09	.22	.37
13. Post‐joy													.53	−.19	.34	.06	.29	.29
14. Post‐determination														−.23	.44	.19	.40	.37
15. External boredom															−.39	−.02	−.15	−.47
16. Internal boredom																.32	.51	.58
17. Curiosity (deprivation)																	.46	.21
18. Curiosity (interest)																		.27
19. Grit

Note: Bolded text indicate significance at the Bonferroni threshold of p = .00029, rather than including exact p values for each correlation.

Personality questionnaires

We used three questionnaires. First, the 12‐item Boredom Proneness Scale‐Short Form (BPS‐SF) was used to assess the two different types of boredom proneness (Vodanovich et al., 2005). The scale computes separate scores for internal and external stimulation. We refer to these subscales as internal and external boredom proneness, respectively.

Participants used a 7‐point rating scale to complete the BPS‐SF, with higher scores indicating greater boredom proneness. We averaged the items in each respective subscale to create internal (Cronbach's α = .77) and external (Cronbach's α = .64) boredom proneness variables.

The 10‐item Interest/Deprivation Trait Curiosity Scale (I/D) was used to measure two aspects of trait curiosity—feeling‐of‐deprivation and feeling‐of‐interest (Litman & Jimerson, 2004). We denote these variants as curiosity (deprivation) and curiosity (interest) in our tables. Participants completed the I/D using a 4‐point rating scale. We averaged the items in each respective subscale to generate our final curiosity deprivation (Cronbach's α = .84) and interest variables (Cronbach's α = .87), with higher scores indicating greater trait curiosity.

Grit was measured using the 12‐item Grit Scale (Duckworth et al., 2007). Participants completed the Grit Scale using a 5‐point rating scale, and items were scored and averaged (Cronbach's α = .88) to create a composite score with higher scores indicating greater grit.

Procedure

After obtaining informed consent, participants completed the emotion questionnaire followed by the control task. Then, participants completed the emotion questionnaire once again to obtain post‐task ratings. Finally, participants completed the personality questionnaires as well as a demographic survey. We used linear regression models to test for relationships between the personality variables of interest and post‐task experiences of each emotion. The six post‐task emotion variables of interest were dependent variables in separate regression models, with each personality variable as a predictor. The six pre‐task emotions was used as covariates to control for pre‐existing emotional state.

RESULTS

Descriptive statistics for personality and emotion

The descriptive statistics for our five personality variables are displayed in Table 2.

Table 2. Descriptive statistics for personality variables

	Mean (SD)	Minimum	Maximum
External boredom proneness	3.59 (0.98)	1.00	6.17
Internal boredom proneness	5.16 (0.97)	2.00	7.00
Curiosity (deprivation)	2.51 (0.65)	1.00	4.00
Curiosity (interest)	3.08 (0.66)	1.00	4.00
Grit	3.34 (0.70)	1.42	5.00

A paired samples t test was used to compare participant emotion pre‐task (M = 3.67, SD = 0.96) to participant emotion post‐task (M = 3.47, SD = 1.02). On average the participants rated 0.20 points, 95% CI [.13, .28], lower on the Emotion Questionnaire after the control task than they did before the control task; indicating that the control task was, in fact, not neutral. This difference was statistically significant, t(177) = 5.25, p < .001, and small, d = 0.20. Paired samples t tests were also run to compare pre‐task and post‐task emotion for boredom, sadness, resignation, determination, interest and joy. Boredom, sadness, resignation and interest were all rated as significantly lower, by participants, after the control task, and this difference was small (Table 3). These finding support Hypothesis 1 and suggest that the control task elicits differences in self‐reported emotions and therefore are not emotionally neutral. The greatest emotional change was in sadness. These findings suggest that the control task generally reduces the feelings of emotions, but not completely.

Table 3. Mean, mean difference, and effect size (d) of emotions pre‐ and post‐task³

	Mean
Variable	Pre‐task (SD)	Post‐task (SD)	Mean difference [95% CI]	t(DF)	d
Boredom	3.39 (2.09)	3.07 (2.09)	0.32 [0.02, 0.61]	2.08 (177)^**	0.15
Sadness	2.54 (2.06)	1.87 (1.44)	0.67 [0.45, 0.89]	6.04 (177)^**	0.38
Resignation	2.26 (1.85)	1.80 (1.51)	0.46 [0.25, 0.68]	4.24 (177)^**	0.27
Determination	5.48 (2.20)	5.48 (2.20)	0.01 [−0.25, 0.26]	0.04 (177)	0.00
Interest	5.94 (1.94)	5.47 (2.18)	0.48 [0.20, 0.75]	3.42 (177)^**	0.23
Joy	4.84 (2.30)	4.65 (2.24)	0.19 [−0.10, 0.47]	1.30 (177)	0.08

Abbreviations: CI, confidence interval; DF, degrees of freedom; SD, standard deviation.

N = 178.

^*p < .05;

** p < .01.

Post‐task emotion

Boredom

As proposed in Hypothesis 2a, external boredom proneness predicted post‐task boredom after controlling for pre‐task emotion as covariates (Table 4). These findings suggest that external boredom proneness has unique effects on how bored participants feel both before and after a control task, with those prone to experiencing boredom in response to their external environments experiencing greater boredom in response to a control task. In other words, these participants feel more bored after engaging in a non‐emotional task. Contrary to Hypothesis 3b, the feeling‐of‐deprivation dimension of trait curiosity did not predict post‐task experiences of boredom (Table 4). There was also no correlation between grit and boredom, as suggested by Hypothesis ^4b. Hypotheses 2b, 3a, and 4a are not relevant as they refer to positive emotion (i.e., joy). In combination, the sub‐dimensions of personality accounted for a significant 16.1% of variance in post‐task boredom, R² = .16, adjusted R²= .11, F (5, 80) = 3.01, p = .014.

Table 4. Unstandardised (B) and standardised (β) regression coefficients, and squared semi‐partial correlations (sr²) for post‐task boredom⁶

Variable	B [95% CI]	β	sr^2
Post‐boredom
Boredom (external)	.68 [0.20, 1.16]^**	.34	.08
Boredom (internal)	.16 [−0.42, 0.73]	.07	.00
Curiosity (deprivation)	.30 [−0.39, 0.10]	.09	.01
Curiosity (interest)	−.62 [−1.40, 0.15]	−.19	.33
Grit	−.21 [−1.02, 0.60]	−.06	.11

Abbreviation: CI, confidence interval.

N = 85.

^*p < .05;

** p < .01.

Sadness

In support of Hypothesis ^2a, external boredom proneness also predicted post‐task sadness after controlling for pre‐task emotion as covariates (Table 5). As with our interpretation of the relationship between external boredom proneness and boredom, these findings suggest that the tendency to feel bored in response to one's external environment predicts how sad one feels in general (before engaging in any particular task) as well as sadness after engaging in a control task. In combination, the sub‐dimensions of personality accounted for a significant 25.1% of variance in post‐task sadness, R² = .25, adjusted R² = .20, F (5, 80) = 5.35, p < .001. There was no correlation between the feeling‐of‐deprivation and sadness or grit and sadness (Hypothesis ^4b).

Table 5. Unstandardised (B) and standardised (β) regression coefficients, and squared semi‐partial correlations (sr²) for post‐task sadness⁹

Variable	B [95% CI]	β	sr^2
Post‐sadness
Boredom (external)	.53 [0.19, 0.87]^**	.35	.09
Boredom (internal)	.04 [−0.36, 0.45]	.03	.00
Curiosity (deprivation)	.43 [−0.07, 0.92]	.18	.03
Curiosity (interest)	−.48 [−1.03, 0.07]	−.20	.34
Grit	−.39 [−0.97, 0.19]	−.15	.26

Abbreviation: CI, confidence interval.

N = 85.

^*p < .05;

** p < .01.

Resignation

Grit predicted post‐task resignation after controlling for pre‐task emotion as covariates, which is in line with Hypothesis ^4b. External boredom proneness was also a significant predictor after controlling for pre‐task emotion as covariates, in line with Hypothesis ^2a (Table 6). Taken together, external boredom proneness predicted negative emotional experiences after a control task. Furthermore, grit appears to negatively correlate with post‐task resignation, suggesting that the gritty individual is less likely to succumb to feelings of resignation when faced with a particular task; such as the control task. Hypothesis ^3b was not supported as there is no correlation between the feeling‐of‐deprivation dimension of trait curiosity and resignation. In combination, the sub‐dimensions of personality accounted for a significant 21.7% of variance in post‐task resignation, R²= .22, adjusted R²= .17, F (5, 80) = 4.45, p = .001.

Table 6. Unstandardised (B) and standardised (β) regression coefficients, and squared semi‐partial correlations (sr²) for post‐task resignation¹²

Variable	B [95% CI]	β	sr^2
Post‐resignation
Boredom (external)	.49 [0.14, 0.83]^**	.32	.08
Boredom (internal)	.39 [−0.02, 0.80]	.23	.03
Curiosity (deprivation)	.25 [−0.25, 0.74]	.10	.01
Curiosity (interest)	−.14 [−0.70, 0.41]	−.06	.10
Grit	−.72 [−1.30, −0.14]^*	−.28	.49

Abbreviation: CI, confidence interval.

N = 85.

* p < .05;

** p < .01.

Determination

As expected, based on Hypothesis ^4a, grit predicted post‐task determination after controlling for pre‐task emotion as covariates; additionally, the feeling‐of‐ deprivation dimension of trait curiosity predicted post‐task experiences of determination (Table 7). This finding suggests that the grittier individual will approach tasks with more intention and determination than their less gritty peers. Furthermore, these findings suggest that when the control task is considered to lack meaningfulness the individual will feel less determined to complete the task. In combination, the sub‐dimensions of personality accounted for a significant 24.9% of variance in post‐task resignation, R² = .25, adjusted R² = .20, F (5, 80) = 5.30, p < .001. There was no correlation between external boredom proneness (Hypothesis ^2b) or the feeling‐of interest (Hypothesis ^3a) and determination.

Table 7. Unstandardised (B) and standardised (β) regression coefficients, and squared semi‐partial correlations (sr²) for post‐task determination¹⁵

Variable	B [95% CI]	β	sr^2
Post‐determination
Boredom (external)	−.17 [−0.62, 0.28]	−.09	.15
Boredom (internal)	.39 [−0.14, 0.93]	.18	.02
Curiosity (deprivation)	−.66 [−1.31, −0.02]^*	−.21	.40
Curiosity (interest)	.50 [−0.22, 1.23]	.16	.02
Grit	.90 [0.14, 1.65]^*	.26	.05

Abbreviation: CI, confidence interval.

N = 85.

* p < .05;

^**p < .01.

Interest

Contrary to Hypotheses 2b and 3a, neither external boredom proneness nor the feeling‐of‐interest dimension of trait curiosity predicted post‐task experiences of interest after controlling for pre‐task emotion as covariates (Table 8). There was no evidence for Hypothesis ^4a, as grit did not correlate with post‐task interest. In combination, the sub‐dimensions of personality accounted for a significant 13.6% of variance in post‐task interest, R² = .14, adjusted R² = .08, F (5, 80) = 2.51, p = .036.

Table 8. Unstandardised (B) and standardised (β) regression coefficients, and squared semi‐partial correlations (sr²) for post‐task interest¹⁸

Variable	B [95% CI]	β	sr^2
Post‐interest
Boredom (external)	−.50 [−1.02, 0.01]	−.24	.41
Boredom (internal)	−.02 [−0.62, 0.59]	−.01	.01
Curiosity (deprivation)	−.12 [−0.84, 0.63]	−.03	.06
Curiosity (interest)	−.09 [−0.92, 0.72]	−.03	.05
Grit	.75 [−0.11, 1.61]	.20	.03

Abbreviation: CI, confidence interval.

N = 85.

^*p < .05;

^**p < .01.

Joy

After controlling for pre‐emotion as covariates, external boredom proneness did not predict post‐task joy as originally proposed in Hypothesis ^2b, but rather, grit predicted post‐task joy, in line with Hypothesis ^4a (Table 9). This result indicates that participants motivated to seek and pursue challenging long‐term goals tend to experience increased joy. In combination, the sub‐dimensions of personality accounted for a significant 25.2% of variance in post‐task joy, R² = .25, adjusted R² = .21, F (5, 80) = 5.38, p < .001. No positive correlation between the feeling‐of‐interest dimension of trait curiosity and joy was found (Hypothesis ^3a), however the feeling‐of‐deprivation dimension negatively correlated with post‐task joy. This finding suggests that the more individuals see information as unsubstantive the more they are likely to feel displeasure (or a lack of joy).

Table 9. Unstandardised (B) and standardised (β) regression coefficients, and squared semi‐partial correlations (sr²) for post‐task joy²¹

Variable	B [95% CI]	β	sr^2
Post‐joy
Boredom (external)	−.45 [−0.93, 0.03]	−.21	.36
Boredom (internal)	.28 [−0.29, 0.85]	.12	.01
Curiosity (deprivation)	−.71 [−1.40, −0.02]^*	−.21	.40
Curiosity (interest)	.43 [−0.35, 1.20]	.13	.01
Grit	.82 [0.01, 1.62]^*	.22	.04

Abbreviation: CI, confidence interval.

N = 85.

* p < .05;

^**p < .01.

DISCUSSION

In summary, we sought to investigate: (a) whether a classical control task used in emotion research, a directed imagery task involving grocery shopping, is truly neutral; and (b) how personality traits related to engagement—namely, boredom proneness, curiosity, and grit—would map onto experiences of engagement‐related and disengagement‐related emotions in the context of the control task. This directed imagery task was designed to emulate control conditions used in emotion research. In particular, we first hypothesised that external boredom proneness would be associated with boredom, sadness, resignation, interest, and joy. Indeed, we found that participants who scored higher on external boredom proneness reported more negative emotions—namely, feeling more bored, sad, and resigned—following the control task. However, we did not find a negative association between external boredom proneness and positive emotions. Second, we hypothesised that trait curiosity (interest) would be positively correlated with interest, and that curiosity (deprivation) would be positively correlated with boredom following the task. However, we did not find support for these hypotheses. Finally, we correctly hypothesised that grit would be associated with determination, joy, and resignation; we found that grittier individuals endorsed feeling less resigned, more determined and more joyful following the task. Taken together, our findings suggest that personality traits only partially account for the differences in emotional experiences toward the control task. Furthermore, control conditions or tasks that are often used in emotion research and assumed to be neutral may, in fact, not be.

Implications

The current findings provide four major implications. First, if the directed imagery task was truly neutral then we would expect personality to predict post‐control‐task emotions that are highly related and account for the differences in post‐task emotions, which was sometimes the case (i.e., external boredom proneness predicted post‐task boredom). However, personality did not always correlate with post‐task emotions (e.g., the feeling‐of‐interest dimension of trait curiosity and post‐task interest), suggesting that the directed imagery task may possibly have elicited emotions above and beyond what personality could predict. In other words, the directed imagery task, which was intended to be emotionally neutral, was not neutral. The implications of this finding are significant when considering prior studies that have employed a similar control condition on the basis of the assumption that such a task is neutral. We do not presume to propose that said studies are incorrect in their findings, but rather that researchers should acknowledge that some effects found to significantly differ from baseline, or a control task may be partially explained by the non‐neutrality of the supposedly control task itself. Furthermore, future research should be aware of how the control condition fits the definition of emotional neutrality. Indeed, it may be necessary to assess how we, as a field, conceptualise emotional neutrality. Gasper (2018) defines five common conceptualisations of emotionally neutral tasks which create (a) minimal affective states, (b) in‐the‐middle states, (c) deactivated states, (d) typical states, or (e) indifferent states. Each of these states is affective, and therefore not devoid of emotion, because a weak reaction is still some type of reaction to a task. Furthermore, it appears that there is no way to be truly neutral before and after a task‐ in the sense that one is devoid of emotion. Assuming that control tasks are likely to change one's emotional state, it may be useful to define the parameters around what changes are acceptable (to avoid influencing the experimental condition) and what changes are not. It may also be prudent to acknowledge that “neutral” control conditions are worthy of investigating, in their own right (Gasper, 2018). Research defining neutral emotion, is scare and it may be useful for future research to test the neutrality of other commonly utilised neutral or control tasks to investigate changes in emotion.

Second, the present study builds upon the literature on boredom proneness. Prior research has identified an internal dimension of boredom proneness that indicates one's inability to generate stimulating activity, and an external dimension that reflect one's perception of a lack of stimulation in the environment (Vodanovich et al., 2005). We found that external boredom proneness only predicted post‐task negative emotion (boredom, resignation, and sadness), supporting the link between boredom proneness and negative affective states such as depression (Goldenberg et al., 2011) and extending this line of research by demonstrating the effect of external boredom proneness on negative emotion after a control task. We found that external boredom proneness predisposes an individual to feel boredom, resignation, and sadness in response to a control task. Interestingly we found that internal boredom proneness does not influence the experiences of positive or negative emotion. Perhaps this is due to the nature of the task which required participants to complete a task, not generate their own stimulating activity. Future research is needed to investigate how external and internal boredom proneness modulate emotion, as well as the cognitions and behaviours associated with these experiences.

Third, the current findings challenge the conventional assumption that trait curiosity predicts interest and suggest that such a relationship is stimulus dependent. In fact, the present research showed that neither dimension of trait curiosity predicted interest. However, the feeling‐of‐deprivation dimension of trait curiosity negatively correlated with both post‐task joy and post‐task determination. Litman & Jimerson (2004) discussed how curiosity (deprivation) is evoked when a person appraises the target stimulus to be deficient in providing new or novel information. We found that the more the stimulus was perceived to be lacking in novelty, the greater the participant's displeasure. Thus, our findings support the notion that the feeling‐of‐deprivation dimension of trait curiosity brings pleasure by reducing feelings of tension that arise from a lack of new information (Litman & Jimerson, 2004). We also found that curiosity (deprivation) negatively correlated with determination. In other words, individuals presented with apparent unsubstantial or meaningless information felt less determined following the control task. Prior research has discussed a desire for new information as a motivator for behaviour (i.e., knowledge seeking; Litman et al., 2005). We did not find a relationship between the feeling‐of‐deprivation dimension of trait curiosity and negative emotion. Our findings suggest that the dimensions of trait curiosity may not necessarily differ in valence, prompting a reconsideration of the conceptualisation of these dimensions in the extant literature. Future research should continue to examine whether the dimensions of curiosity differ in their affective and motivational correlates.

Fourth, the current research provides further support for the relationship between grit and engagement‐related emotions. We found support for an association between grit and joy, and grit also predicted post‐task experiences of determination, supporting the relationship between grit and positive affect (Hill et al., 2014). Experiences of determination, as previously suggested (Kirby et al., 2014), and interest may be two mechanisms through which grit promotes perseverance. Our findings suggest that grit may prompt an appraisal style of high motivational relevance, motivational congruence, and high coping potential. Thus, grit may sustain attention and perseverance toward long‐term goals, as supported by the respective relationships with interest and determination in the present study. As expected we also found a negative correlation between grit and post‐task resignation. This supports the concept of grit in the literature as grittier individuals are less inclined to give up or feel resigned (Duckworth et al., 2007). Critically, we were not able to directly test relationships between grit and specific appraisals associated with these emotions because the present study did not survey cognitive appraisal. Future studies should continue to investigate the relationships between grit and discrete positive emotions while incorporating measures of appraisal, coping, and emotion regulation. Future research should also test how interest and other emotions influence the well‐established relationships between grit and long‐term outcomes (Eskreis‐Winkler et al., 2016).

Limitations and future research

The present study had at least two limitations. First, the present study was correlational in nature, as it is not feasible to manipulate stable personality traits. Studies using other research designs, such as longitudinal survey methods, are needed to further investigate and replicate our findings on the relationships between these engagement‐related personality traits and emotions. Second, we only tested the effects of boredom proneness, trait curiosity, and grit on pre‐ and post‐task emotion in the context of engaging in a non‐emotional task intended to represent control conditions used in emotion research. It is unclear how our findings would generalise to emotion‐inducing or emotionally intense tasks.

In summary, the present study found that even following a control task intended to be emotionally neutral, emotional experience differ and personality can paint the way we perceive and thus react to these control conditions. Future directions include evaluating the effects of boredom proneness, curiosity, grit, and other dispositional traits on experience during other tasks, such as those that induce emotion. Future research should continue to assess how personality predicts engagement‐related emotion and behaviour during emotional and control tasks.

CONFLICT OF INTERESTS

The authors declare no potential conflict of interest.

AUTHOR CONTRIBUTIONS

B.S. contributed to the paper's study design, data collection, and manuscript writing. J.Y. contributed to the paper's study design, data collection, data analysis, and manuscript writing. N.W. contributed to the paper's manuscript writing.

ACKNOWLEDGEMENTS

This manuscript is not supported by any funding source.