Individual Differences in Emotion Regulation Goals: Does Personality Predict the Reasons Why People Regulate their Emotions?

Abstract

Objective:

We investigated how the Big Five traits predict individual differences in five theoretically important emotion regulation goals that are commonly pursued – pro-hedonic, contra-hedonic, performance, pro-social, and impression management.

Method:

We conducted two studies: (1) a large survey study consisting of undergraduates (N = 394; 18–25 years; 69% female; 56% European-American) and community adults (N = 302; 19–74 years; 50% female; 75% European-American) who completed a newly developed global measure of individual differences in emotion regulation goals and (2) a 9-day daily diary study with community adults (N = 272; 50% female; 84% European-American) who completed daily reports of emotion regulation goals. In both studies, participants completed a measure of the Big Five.

Results:

Across global and daily measures, pro-hedonic goals and pro-social goals were positively associated with agreeableness, performance goals were positively associated with openness, and impression management goals were positively associated with neuroticism. Globally, contra-hedonic goals were also negatively associated with agreeableness and conscientiousness.

Conclusions:

The Big Five systematically predict the emotion regulation goals people typically pursue. These findings have important implications for understanding why people engage in certain forms of regulatory behavior and why personality has consequences for well-being.

Emotion regulation research has largely focused on how people regulate their emotions, that is, the strategies they use. People vary, for instance, in their tendency to keep their emotions to themselves (i.e., expressive suppression) and to reframe how they see a situation (i.e., cognitive reappraisal; Gross & John, 2003; Srivastava, Tamir, McGonigal, John, & Gross, 2009). However, less is known about emotion regulation goals, which we define as the reasons why people manage their emotions. Goals are important because they determine whether (Mauss & Tamir, 2014) and how an individual will regulate their emotions (English, Lee, John, & Gross, 2017), and because pursuing certain emotion regulation goals may be maladaptive for well-being (Millgram, Joormann, Huppert, & Tamir, 2015).

Given that individuals differ in what is important to them (Roccas, Sagiv, Schwartz, & Knafo, 2002), it is unlikely that people all value and pursue the same emotion regulation goals. Indeed, while goals can shift across situations (McCabe & Fleeson, 2016), there are still stable individual differences in goals (e.g., Roberts & Robins, 2000; Reisz, Boudreaux, & Ozer, 2013). Individual differences in emotion regulation goals are important to consider because goals impact how people typically behave (Austin & Vancouver, 1996) and have downstream consequences for well-being if pursued chronically (e.g., Brunstein, Schultheiss, & Grassmann, 1998). The Big Five traits, which represent fundamental dimensions of personality (John, Naumann, & Soto, 2008), systematically predict individual differences in goals across various domains, such as major life goals (e.g., Bleidorn et al., 2010) and personal goals (e.g., Reisz et al., 2013). To the extent that there are common underlying features shared across different goal domains, then the Big Five may also systematically predict individual differences in emotion regulation goals. Furthermore, the Big Five are ideal to consider when predicting individual differences in emotion regulation goals because they are tied to emotional processes, such as individual differences in emotion regulation strategy use (Gresham & Gullone, 2012) and regulatory ability (Ivcevic & Brackett, 2014). For example, introverted people are more likely to hide their emotional expression (Gross & John, 2003).

Emotion Regulation Goals

Despite their critical implications for regulatory behavior and well-being, research on emotion regulation goals is still nascent. In this section, we draw on a recent theoretical taxonomy of emotion regulation motives by Tamir (2016) to lay out the space for emotion regulation goals. Notably, many researchers use the terms motives and goal interchangeably because they consider motives to be a type of goal (Kruglanski et al., 2002). Specifically, they use the term primary goal to describe the reasons why people behave a certain way. Thus, when we use the term emotion regulation goal we are referring to it as a primary goal in the same way that Tamir (2016) uses the term motive. Tamir (2016) proposes two main categories of emotion regulation motives: hedonic and instrumental. Hedonic motives capture how much people want to experience pleasure and pain. Instrumental motives are a broader and larger category, which focus on outcomes besides emotional experience (Tamir, 2009). In the current paper, we focus on two hedonic goal sub-types – pro-hedonic and contra-hedonic – and two instrumental goal sub-types – performance and social. Although Tamir (2016) proposes two additional instrumental goal sub-types (i.e., epistemic, eudaimonic), we focus on performance and social goals specifically because they are commonly pursued in daily life (English et al., 2017; Kalokerinos, Tamir, & Kuppens, 2017). Thus, they are prime candidates for beginning to understand the links between personality and emotion regulation goals.

Pro-hedonic goals describe the desire to feel positively (e.g., wanting to feel happy), whereas contra-hedonic goals describe the desire to feel negatively (e.g., teenager listening to angsty music because they want to feel angry). Performance goals refer to the desire to perform an activity (e.g., wanting to get work done) and social goals refer to the desire to influence social interactions or relationships (e.g., wanting to cheer up a sad friend). Given that social goals are especially prevalent (English et al., 2017; Gross et al., 2006), we find it critical to delineate between social goals even further. Although all social goals involve others, developmental researchers have proposed that social goals fall into different types – wanting an outcome for the self vs. wanting an outcome for others (e.g., Martini, 2011; Zeman & Shipman, 1996). A prime self-focused goal people pursue is impression management, wanting to appear a certain way to others, because it involves influencing relationships for one’s own sake. Meanwhile a prime other-focused goal people pursue is pro-social, wanting to maintain or promote social interactions and relationships, because it involves influencing relationships for the sake of others. Thus, when evaluating social goals, we distinguish between impression management goals and pro-social goals.

Most research on emotion regulation goals has been experimental, instructing people to pursue a particular goal (e.g., collaborate with others; Tamir, Mitchell, & Gross, 2008; Tamir & Ford, 2012). Only recently have studies begun to examine individual differences in emotion regulation goals. This work has informed our understanding of how often people pursue various goals in daily life. For instance, consistent with the idea that people prefer to feel more good than bad (Diener, 2000), they are more likely to pursue pro-hedonic goals than contra-hedonic goals (English et al., 2017; Gross, Richards, & John, 2006; Riediger, Schmiedek, Wagner, & Lindenberger, 2009). Also consistent with the idea that people often regulate in social contexts, they are more likely to regulate for social reasons than for non-social reasons (English et al., 2017). However, little is known about the factors that predict individual differences in emotion regulation goals (Kalokerinos et al., 2017). For instance, who is motivated to frequently regulate their emotions because they want to feel worse as opposed to better? Or who is motivated to frequently regulate their emotions for the sake of others as opposed to for themselves? To begin to answer these questions, we turn to the Big Five as relevant predictors of individual differences in emotion regulation goals.

The Big Five and Goals

The Big Five are a taxonomy of five broad traits that capture people’s stable dispositions: extraversion, agreeableness, conscientiousness, neuroticism, and openness to experience. While they are not the only traits that can describe personality, they are the most widely used taxonomy of traits (Goldberg, 1993; John et al., 2008), are believed to be universal (McCrae & Costa, 1997), and are relatively stable across the lifespan (Roberts & DelVecchio, 2000). Why would the Big Five predict individual differences in emotion regulation goals?

Multiple views have been proposed to understand the relationship between traits and goals. According to one major conceptualization, traits are broad dispositions that manifest themselves in people’s goals, as they do in everyday behavior (e.g., Cantor, 1990; Little, Lecci, & Watkinson, 1992; McCrae & Costa, 2003). That is, traits differentially predict the goals that people typically pursue. We focus on this conceptualization because our aim is to identify the personality profiles of people who pursue various emotion regulation goals. Several studies have also used this conceptualization to investigate the systematic relationships between the Big Five and goals more generally. For instance, agreeableness positively predicts major life social goals (e.g., getting along with others; Bleidorn et al., 2010; Roberts & Robins; 2000), as well as personal social goals (e.g., improving familial relations; Reisz et al., 2013). In addition, changes in the Big Five have even been linked longitudinally to changes in major life goals (Ludtke, Trautwein, & Husemann, 2009; Roberts, O’Donnell, & Robins, 2004), further showing meaningful relations between the Big Five and goals. Given that the Big Five predict goals across various domains, it is plausible for the Big Five to also predict emotion regulation goals.

The Big Five and Emotion Regulation Goals

Below we delineate our predictions about links between the Big Five and individual differences in our subset of five emotion regulation goals: pro-hedonic, contra-hedonic, performance, pro-social, and impression management. In making our predictions, we draw on studies examining broad goal domains (e.g., Roberts & Robins, 2000), as well as the few studies on emotion regulation goals (Kalokerinos et al., 2017). If we do not state hypotheses about links between a specific goal and trait then that is because we do not expect an association.

Hedonic goals.

One key aspect of the Big Five is that it captures consistent patterns of feeling (Costa & McCrae, 1980). For instance, extraverts more strongly experience positive emotions, such as happiness (e.g., Kampfe & Mitte, 2009; Shiota, Keltner, & John, 2006). Meanwhile, people higher in neuroticism tend to experience more negative emotions, such as anger (Watson & Clark, 1992). Most research suggests that people prefer trait-consistent states (Costa & McCrae, 1980; Ford & Tamir, 2014; Lucas, Le, & Dyrenforth, 2008; Tamir, 2005; Tamir, 2009). For example, extraverts report more strongly wanting to experience positive emotions (e.g., Larsen & Ketelaar, 1989; Tsai, Knutson, & Fung, 2006) and individuals higher in neuroticism want to feel more negative emotions, such as worry when performing a demanding task (e.g., taking a test; Kampfe & Mitte, 2009; Tamir, 2005; cf. Augustine, Hemenover, Larsen, & Shulman, 2010; Rusting & Larsen, 1995 on how neurotic individuals sometimes want to feel more positively). Thus, we hypothesize that pro-hedonic goals are positively predicted by extraversion because they may allow extraverts to continue experiencing positive emotions. Meanwhile, contra-hedonic goals may be positively predicted by neuroticism because they could help neurotic people to continue experiencing negative emotions.

One might expect extraversion and neuroticism to be the only Big Five traits relevant for hedonic goals since they are the most strongly affect-based (Larsen & Ketelaar, 1991). However, as with extraversion, agreeable people often experience positive emotions (Shiota et al., 2006; Watson & Clark, 1992). Thus, agreeable individuals might also be highly motivated to pursue pro-hedonic goals. Furthermore, people higher in openness have more intense emotional experiences (Costa & McCrae, 1992) and strongly value their emotional experiences (Terracciano, McCrae, Hagemann, & Costa, 2003) Thus, we hypothesize that highly open individuals are more motivated to pursue both pro-hedonic and contra-hedonic goals to maintain or reach their emotional experiences. Notably, when we consider other goal domains, extraversion and openness are positively associated with hedonistic major life goals (e.g., having fun; Reisz et al., 2013; Roberts & Robins, 2000), which involve enjoying life.

Instrumental goals.

As broad dispositions, the Big Five not only reflect consistent patterns of feeling, but also consistent patterns of thought and behavior (Costa & McCrae, 1980). Thus, people should want to regulate their emotions in instrumental ways that are consistent with how they also typically think and behave. For instance, conscientious individuals are hardworking and motivated (e.g., Jackson, Wood, Bogg, Walton, Harms, & Roberts, 2010). Open individuals are also achievement-oriented in that they are more likely to choose investigative careers and go back to school in middle-age (e.g., George, Helson, & John, 2011). Thus, we hypothesize that performance goals are positively linked to conscientiousness and openness. Notably, conscientiousness is associated with more performance-related major life goals (e.g., achievement; Bleidorn et al., 2010; Reisz et al., 2013).

Meanwhile, more extraverted and agreeable people value their social experiences (Asendorpf & Wilpers, 1998) and have more satisfying relationships (e.g., Demir & Weitekamp, 2007). Furthermore, individuals higher in agreeableness strongly value social harmony (Graziano & Tobin, 2009) and behave in ways to maintain harmony. For instance, highly agreeable individuals are more likely to suppress their negative emotions when interacting with others compared to individuals low in agreeableness (Tobin, Graziano, Vanman, & Tassinary, 2000). Thus, we hypothesize that pro-social goals are positively associated with extraversion and agreeableness. Indeed, outside the context of emotion regulation, people higher in these traits pursue pro-social life goals more (e.g., communion; Reisz et al., 2013; Roberts & Robins, 2010). In contrast to pro-social goals, a different trait profile may be relevant for impression management goals. Individuals higher in neuroticism have greater rejection sensitivity (Downey & Feldman, 1996) and social anxiety (Leary, Kelly, Cottrell, & Schreindorfer, 2013). They are also less likely to present their actual selves to others (Trapnell & Campbell, 1999). Thus, we expect that impression management goals are positively associated with neuroticism.

Kalokerinos et al. (2017) examined how the Big Five predict daily instrumental goals, including performance and social goals. Performance goals were not related to any of the Big Five, while social goals were predicted by neuroticism. In general, it may have been difficult to detect associations with the Big Five because emotion regulation goals were assessed in a limited range of situations (i.e., a daily negative event). In addition, the researchers measured social goals at a broad level. Certain goals might only be pursued in certain situations. Thus, the chances of picking up on certain goals and their associations with other constructs are reduced with fewer situations. Perhaps the social goals captured in daily negative events mostly tap into impression management concerns. This may be why they found a positive link with neuroticism and no association with extraversion or agreeableness, which may be tied to pro-social concerns.

The Present Research

Prior research has largely focused on the behavioral aspect of emotion regulation, namely, how people regulate, rather than the motivational aspect of emotion regulation, namely, why people regulate. Even less attention has been given to individual differences in the reasons why people regulate, and the factors that predict them. Given the critical implications of individual differences in emotion regulation goals for regulatory behavior and well-being, we conducted two well-powered and complementary studies to investigate how the Big Five traits predict emotion regulation goals. Study 1 had two independent samples of undergraduates and community adults who completed a new global measure of emotion regulation goals. Study 2 was a 9-day diary study with community adults who completed daily goal assessments.

We included non-college samples to increase the generalizability of our findings, especially since most research on emotion regulation goals has been conducted with undergraduates. Although we do not have predictions about age as a moderator, there are well-established age differences in the Big Five (Roberts & DelVecchio, 2000) and potential age differences in emotion regulation goals (Riediger et al., 2009). We took a multi-method approach because global and daily measures of individual differences each offer important advantages. Thus far, no study has assessed global emotion regulation goals. However, global measures are useful for assessing trait-level differences and can be used to reliably predict important affective, social, and health outcomes (Paulhus & Vazire, 2007). At the same time, there are limitations to global measures (e.g., Robinson & Clore, 2002). For instance, they can be less accurate because they draw more on self-perceptions than actual behavior (cf. Finnegan & Vazire, 2017 on how daily measures can sometimes be less accurate than global measures). Daily measures may allow us to better capture how emotional processes naturally and dynamically unfold in everyday life. When assessing daily goals, we also expand on Kalokerinos et al.’s (2017) study by assessing goals generally across the day as opposed to in a specific event (i.e., negative), distinguishing between specific social goals, and assessing hedonic goals in addition to instrumental goals. A summary of our hypotheses is in Table 1.

Table 1.

Summary of Hypotheses and their Level of Support Across Studies 1 and 2

	Emotion regulation goal
	Pro-hedonic		Contra-hedonic		Performance		Pro-social		Impression management
Big Five	Hyp	Result	Hyp	Result	Hyp	Results	Hyp	Results	Hyp	Results
Extraversion	+	Partially supported; Study 1 undergraduates					+	Not supported
Agreeableness	+	Fully supported					+	Fully supported
Conscientiousness					+	Not supported
Neuroticism			+	Mostly supported; Study 1 undergraduates and Study 2					+	Fully supported
Openness	+	Partially supported; Study 1 undergraduates	+	Not supported	+	Fully supported

Note. Hyp = hypothesis; + = positive association. Blank spaces indicate that we did not have a hypothesis regarding the association between a given emotion regulation goal and Big Five trait. Not supported = no support in any sample across either study; Partially supported = support in at least one sample; Mostly supported = supported in at least two samples; Fully supported = supported in both studies in all samples. Study 1 consisted of two samples (undergraduates and community adults) and used global measures. Study 2 consisted of one sample of community adults and used daily measures.

Study 1: Global Emotion Regulation Goals

Participants and Procedure

Sample 1: Undergraduates.

We recruited a sample of 394 undergraduates (69.3% female) ages 18–25 years (M = 19.63 years, SD = 1.30) enrolled in psychology courses; 55.7% were European/European-American, 26.7% were Asian/Asian-American, 6.7% were Latino, 5.2% were African-American, and 5.7% were Multi-racial or Other. They completed an online survey measuring individual differences in emotion regulation goals and the Big Five. Not relevant to this paper, they also described a time when they regulated their emotions and completed other emotion regulation measures. They received one course credit for participation.

Sample 2: Community Adults.

We recruited a sample of 302 adults (50.3% female) from Amazon Mechanical Turk ages 19–74 years (M = 35.68 years, SD = 11.15); 74.5% were European/European-American, 10.3% were African-American, 9.9% were Asian/Asian-American, 6.6% were Latino, and 3.3% were Other; the total is larger than 100% because some people identified with more than one ethnicity. Participants completed an online survey with measures of emotion regulation goals and the Big Five. Not relevant to the current study, they also completed other measures of personality and well-being. They were given a $2 payment.

We determined our sample sizes based on sufficient power needed to reach a small effect size (r = .20), given that the effect size for links between the Big Five and major life goals, which are broader than emotion regulation goals, is moderate (r = .40; e.g., Roberts & Robins, 2000).

Measures

Emotion regulation goals.

There is no existing measure for individual differences in emotion regulation goals. Thus, we created the Emotion Regulation Goals Scale (ERGS) to assess global pursuit of pro-hedonic, contra-hedonic, performance, pro-social, and impression management goals. We drew on the experience sampling study by Riediger et al. (2009) and the daily diary study by (English et al., 2017) to create pro-hedonic and contra-hedonic goal items. In these studies, participants or independent coders rated the desire to increase, maintain, or decrease emotional experiences. The pro-hedonic goal composite captured the desire to feel positively by aggregating across the desire to increase or maintain positive emotions, and decrease negative emotions. In contrast, the contra-hedonic goal composite captured the desire to feel negatively by aggregating across the desire to increase or maintain negative emotions, and decrease positive emotions. We adapted the general phrasing from these studies (e.g., modified “increase” to “feel more”) to create similar items, but we did not use the same specific emotions. Instead, we included examples of two common positive emotions (joy, contentment) and two common negative emotions (anger, sadness) to facilitate participant understanding. Notably, the most widely used emotion regulation measures, such as the Emotion Regulation Questionnaire (Gross & John, 2003), have also used common positive and negative emotions as examples within their items. To more holistically capture different dimensions of emotion (Rusting & Larsen, 1995), we listed examples that not only varied in their valence (i.e., positive, negative), but also in their arousal. Thus, we gave examples of high arousal (joy, anger) and low arousal emotions (contentment, sadness). To measure performance, pro-social, and impression management goals¹, we drew on English et al.’s single goal items (2017) and created additional items. We had done a pilot study (N = 151) in which an age-diverse sample of adults described situations when they regulated their emotions in a variety of social contexts (e.g., with family members, friends). We coded these descriptions for performance, pro-social, and impression management goals, and extracted common phrases as items. The final scale consisted of 18 items (see Appendix A), with five sub-scales (one per goal), which were rated using a 7-point scale (1 = never; 7 = always).

Big Five traits.

Sample 1 completed the 44-item Big Five Inventory (BFI; John & Srivastava, 1999) and Sample 2 completed the 60-item Big Five Inventory-2 (BFI-2; Soto & John, 2017). Both measures reliably assess the Big Five traits with one sub-scale per trait: extraversion, neuroticism, conscientiousness, agreeableness, and openness; the BFI-2, a revised version of the BFI with improved psychometric properties, uses the terms negative emotionality and open-mindedness to describe neuroticism and openness, respectively. Each sample rated its respective measure on a 5-point agreement scale (1 = disagree strongly; 5 = agree strongly).

Results and Discussion

Emotion Regulation Goals Factor Structure

Although this is not a scale paper, it is important to describe the psychometric properties of our global emotion regulation goals measure. To evaluate the factor structure of the emotion regulation goals, we followed recommended best practices (Jolliffe, 1986; Thompson, 2004). With both samples, we first conducted an exploratory principal components analysis with an oblique rotation using oblimin. We chose this statistical technique because it allows the factors to correlate. We expected that people often pursue a variety of emotion regulation goals. It is common for individual difference sub-scales to correlate, including the Big Five (e.g., Soto & John, 2017). We relied on common guidelines suggesting a primary loading of at least .40 and cross-loadings less than .30 (Floyd & Widaman, 1995). Table 2 shows the oblimin-rotated loadings for each goal factor as well as alpha reliabilities, means and standard deviations, and correlations between goals¹.

Table 2.

Oblimin-Rotated Loadings for the Five Emotion Regulation Goal Factors after Principal Components Analysis, Separately in Samples 1 and 2 (Study 1)

	Pro-hedonic		Contra-hedonic		Performance		Pro-social		Impression management
Emotion regulation goal items	S1	S2	S1	S2	S1	S2	S1	S2	S1	S2
Pro-hedonic (3 items)			−.04	.06	.48*	.36*	.42*	.44*	.23*	.26*
To feel more positive emotion (e.g., joy, contentment)	.86	.93	.07	.03	.03	.03	.09	.02	−.08	.00
To feel less negative emotion (e.g., anger, sadness)	.79	.20	.02	.13	.04	.48	.14	.07	.00	.16
To keep feeling positive emotion (e.g., joy, contentment)?	.73	.86	.04	.05	.00	.06	.12	.03	.04	.02
Contra-hedonic (3 items)					.04	−.05	.05	.20*	.05	.27*
To feel less positive emotion (e.g., joy, contentment)?	.00	.04	.87	.83	.01	.07	.00	.00	−.01	.02
To keep feeling negative emotion (e.g., anger, sadness)?	.00	.03	.84	.84	.10	.01	.01	.01	−.02	.07
To feel more negative emotion (e.g., anger, sadness)?	.02	.05	.82	.89	.14	.05	.00	.05	−.00	.00
Performance (3 items)							.30*	.42*	.22*	.27*
To concentrate on your work or what you’re doing?	.11	.06	.01	.07	.97	.85	.05	.05	.09	.08
To stay focused on a task you’re working on?	.00	.02	.14	.06	.90	.90	.02	.07	.01	.08
To avoid being distracted by how you’re feeling?	.20	.08	.10	.17	.70	.84	.12	.03	.08	.00
Pro-social (5 items)									.62*	.52*
To make someone else feel good?	.02	.07	.05	.01	.08	.03	.87	.93	.02	.06
To cheer someone else up?	.02	.06	.04	.04	.05	.07	.86	.96	.08	.10
To avoid ruining someone else’s mood?	.04	.12	.01	.03	.01	.09	.80	.84	.02	.03
To avoid drifting apart from others?	.05	.00	.16	.20	.08	.18	.64	.53	.10	.16
To maintain a close relationship with others?	.19	.00	.04	.17	.00	.00	.49	.53	.23	.44
Impression management (4 items)
To avoid being rejected by others?	.01	.00	.02	.13	.02	.02	.13	.10	.92	.83
To have others approve of you?	.19	.06	.05	.00	.06	.05	.02	.03	.83	.90
To avoid making an unfavorable impression on others?	.09	.15	.08	.01	.09	.13	.08	.04	.77	.86
To make a positive impression on others?	.10	.07	.08	.07	.02	.10	.10	.09	.73	.87
Sum of squared loadings in oblimin analysis	696.75	624.9	701.13	585.92	673.83	696.75	1038.55	602.06	1408.49	830.81
Alpha reliability of scale based on this preliminary factor	.74	.66	.80	.84	.83	.84	.84	.89	.85	.89
Means and standard deviations	4.95 (1.05)	4.46 (1.09)	2.29 (1.15)	2.19 (1.28)	4.86 (1.18)	4.48 (1.19)	4.98 (.96)	4.22 (1.26)	5.10 (1.03)	3.75 (1.38)

Note. S1 = Sample 1; S2 = Sample 2. Absolute loadings of .40 or greater are bolded. Items in each factor are listed in descending order based on S1 loading size. Correlations between subscales for each sample are in the right-hand corner of the table on the rows with bolded goal labels.

^*p < .05.

We first conducted our analyses with Sample 1 and then attempted to replicate our findings in Sample 2. Across both samples, the eigenvalues suggested the presence of five factors, which were clearly interpretable. The five factors accounted for a substantial portion of the total variance (Sample 1: 69.50%; Sample 2: 74.67%). As is customary, we report more detailed statistics for our replication sample. The first five eigenvalues in Sample 2 were 6.29, 2.61, 1.84, 1.47, and 1.21. The next eigenvalue was .80, followed by lower sizes. The first factor was defined by pro-social items (mean loading = .76). The second factor was defined by contra-hedonic items (mean loading = .85). The third factor was defined by impression management items (mean loading = .86). The fourth factor was defined by performance items (mean loading = .86). The fifth factor was defined by pro-hedonic items (mean loading = .66). The intended loadings were generally larger than any cross-loadings (mean cross-loading across factors = .08).

Most goals were positively correlated with one another, except for contra-hedonic goals. (r = −.05 to .52). The lowest correlation was between contra-hedonic and performance goals, while the highest correlation was between the social goals. Despite these correlations, the analyses revealed distinct factors for each of the five emotion regulation goals. We conducted a confirmatory factor analysis on these data using R Version 3.4.3 and found that the data fit a five factor structure well, χ²(109, N = 302) = 416.9, p = .00, RMSEA = .09, 90% CI[.08, .11], CFI = .90, SRMR = .09. We used common thresholds to determine goodness of fit (Hu & Bentler, 1999). We also correlated the five emotion regulation goal factors across the two samples and found them to be highly correlated (r = .92 to .99 between similar factors), demonstrating strong replication. The range of correlations between the items within each sub-scale across the two samples were as follows: pro-hedonic (r = .24-.66), contra-hedonic (r = .57-.71), performance (r = .52-.84), pro-social (r = .37-.90), and impression management (r = .51-.83).

Links Between the Big Five and Global Emotion Regulation Goals

To examine links between the Big Five and emotion regulation goals we conducted regression analyses with the Big Five as simultaneous predictors of each goal. This allowed us to account for any potential overlap between the Big Five and to better isolate the links between each trait and goal. Our results in Table 3 show the systematic relations with the Big Five. We collapsed across demographics because preliminary analyses revealed little evidence for moderation by demographics, including age, gender, and ethnicity.

Table 3.

Regression Analyses Predicting Global Emotion Regulation Goals from the Big Five (Study 1)

	Emotion regulation goal
	Pro-hedonic		Contra-hedonic		Performance		Pro-social		Impression management
Big Five	S1	S2	S1	S2	S1	S2	S1	S2	S1	S2
Extraversion	.15* [.04,.21]	.07 [−.09,.29]	.06 [−.03,.14]	.12 [−.01,.43]	−.04 [−.13,.05]	−.02 [−.24,.18]	.07 [−.01,.13]	.08 [−.07,.36]	−<.01 [−.08,.07]	−.02 [−.28,.20]
Agreeableness	.25* [.19,.43]	.32* [.30,.71]	−.17* [−.35,−.09]	−.16* [−.55,−.07]	.05 [−.06,.21]	.13* [.01,.47]	.30* [.22,.44]	.41* [.52,.99]	.24* [.16,.40]	.08 [−.08,.42]
Conscientiousness	.08 [−.01,.19]	−.04 [−.27,.14]	−.14* [−.27,−.05]	−.14* [−.49,−.01]	.09 [−.01,.23]	.08 [−.08,.37]	−.06 [−.15,.03]	−.17* [−.53,−.05]	−.06 [−.16.03]	−.25* [−.72,−.20]
Neuroticism	.10* [.01,.20]	.04 [−.12,.23]	.15* [.05,.27]	−.01 [−.23,.18]	.13* [.03,.26]	.10 [−.05,.34]	.13* [.03,.21]	.10 [−.05,.35]	.28* [.17,.37]	.15* [.01,.47]
Openness	.11* [.02,.27]	−.06 [−.29,.08]	−<.01 [−.14,.12]	−.22* [−.62,−.19]	.16* [.09,.37]	.14* [.03,.45]	.08 [−.01,.20]	−.03 [−.28,.14]	.07 [−.02,.21]	−.03 [−.29,.16}

Note. Values reflect standardized beta coefficients with 95% confidence intervals listed in brackets. S1 = Sample 1; S2 = Sample 2. Bolded cells indicate significant effects that replicate findings across both samples.

^*p < .05.

For hedonic goals, we had full support for the hypothesis that pro-hedonic goals are positively associated with agreeableness. We only had support in Sample 1 for the following hypotheses: pro-hedonic goals are positively associated with extraversion and openness, and contra-hedonic goals are positively linked to neuroticism. There was no evidence that openness positively predicted contra-hedonic goals. For instrumental goals, there was full support for the predicted positive associations between performance goals and openness, pro-social goals and agreeableness, and impression management goals and neuroticism. However, our hypotheses that there are positive links between performance goals and conscientiousness, or extraversion and pro-social goals were not supported. Notably, there were also some unexpected findings: in both samples, contra-hedonic goals were negatively linked to agreeableness and conscientiousness. We attempted to replicate our findings in Study 2, using daily emotion regulation goals.

Study 2: Daily Emotion Regulation Goals

Method

Participants and Procedure

The sample had 136 married couples (N = 272), ages 23–85 years (M = 53.24, SD = 18.23) who were recruited for a larger study on emotion regulation in adulthood (Eldesouky & English, 2018a); 83.6% were European/European-American, 9.3% were African American, 1.9% were Hispanic or Latin-American, and 5.2% were Multi-racial or Other. Eligibility criteria were having internet access, being married to someone who was not more than 10 years older, and not having significant cognitive impairment (as screened via the Mini-Mental Exam; Folstein, Folstein, & McHugh, 1975). Participants completed a global measure of the Big Five in an initial laboratory session. For the next nine days, they completed a 5-min survey of their daily emotion regulation goals every evening on their personal computer. They were compensated with $10 in the laboratory session and $20 for the daily diary portion of the study. We did not do an a priori power analysis to determine our sample size. However, we conducted a sensitivity analysis in GPower Version 3 and found that we could detect two-tailed correlations of at least r = .17 at 80% power. This effect size falls within the range we found in Study 1 (r = .10 to .28).

Measures

Daily emotion regulation goals.

Participants rated how much they pursued pro-hedonic, contra-hedonic, performance, pro-social, and impression management goals in regards to when they regulated their emotions on each day. Similar to the phrasing we adapted from Riediger et al. (2009) in Study 1, participants rated how much they tried to upregulate (i.e., feel more) and downregulate (i.e., feel less) various emotional experiences. We assessed the upregulation and downregulation of a wider range of emotions than Riediger et al. (2009), including six positive emotions (excited, content, enthusiastic, relaxed, happy, calm) and six negative emotions (lonely, bored, sluggish, sad, angry/frustrated, anxious/nervous). Pro-hedonic goals were an aggregate of attempts to upregulate positive emotional experience and downregulate negative emotional experience (12 items; α = .95), while contra-hedonic goals were an aggregate of downregulating positive emotional experience and upregulate negative emotional experience (12 items; α = .93). Participants also rated how often they managed their emotions for instrumental reasons using the following single items: “to accomplish a task” (performance goals), “to make someone else feel better” (pro-social goals), and “to avoid making a bad impression” (impression management goals). These items were based on English et al.’s (2017) daily items and they mapped onto at least one item from each subscale from the ERGS in Study 1. Participants rated all items on a 7-point scale (1 = not at all; 7 = a great deal).

Big Five traits.

We used the 44-item Big Five Inventory (John & Srivastava, 1999).

Results and Discussion

Analysis Plan

There were 2,289 observations with about five surveys per participant (M = 4.89, SD = 2.59)². All participants were included in our analyses. To account for the nested structure of the data, we conducted multi-level modeling (MLM)³. The variance at each level was as follows: .54 for days (Level 1), .45 for persons (Level 2), and .01 for couples (Level 3). In cases where there is a level with little variance, it is recommended to remove that level from analyses (Snijders & Bosker, 1999). Thus, we conducted two-level models, collapsing across couples (with days nested within persons). The intraclass correlation coefficients showing variance between-persons for the emotion regulation goals at the person-level suggest that it is reasonable to examine individual differences using this data: performance = .57, impression management = .59, pro-social = .60, pro-hedonic = .78, and contra-hedonic = .87⁴. Goals were moderately correlated between-persons (rs = .48-.73); the lowest correlation was between contra-hedonic and pro-social goals and the highest correlation was between the social goals (as in Study 1).

We modeled the degree to which each emotion regulation goal (Level 1) was a function of time (Level 1) and all of the Big Five traits (grand-mean centered; Level 2). We used maximum likelihood (ML) estimation, which effectively accounts for missing data by using all cases to calculate estimates (Rubin, 1976). We only included random intercepts because models with random slopes failed to converge. An autoregressive covariance structure was used to control for associations between observations that were close in time. As recommended for MLM (Edwards et al., 2008), we calculated semi-partial R² (R_β²) as our effect size. It captures the proportion of variance explained by each predictor. We report the intercepts (to represent mean goal pursuit), unstandardized coefficients, 95% confidence intervals, and R_β² values in Table 4. In preliminary analyses, there was little evidence for moderation by demographics, and thus, these variables are not included in the main analyses.

Table 4.

Multi-level Modeling Analyses Predicting Daily Emotion Regulation Goals from the Big Five (Study 2)

	Emotion regulation goal
	Pro-hedonic		Contra-hedonic		Performance		Pro-social		Impression management
	B(SE) 95% CI	Rβ2	B(SE) 95% CI	Rβ2	B(SE) 95% CI	Rβ2	B(SE) 95% CI	Rβ2	B(SE) 95% CI	Rβ2
Intercept	3.21		1.45		3.74		3.54		2.65
Day	−.05(.00)*[−.07,−.04]	.08	−.02(.00)*[−.03,−.01]	.06	−.09(.01)*[−.11,−.06]	.07	−.08(.01)*[−.11,−.06]	.05	−.04(.01)*[−.06,−.02]	.03
Big Five
Extraversion	.18(.09)[−.00,.37]	.01	.03(.03)[−.02,.09]	<.01	.04(.10)[−.16,.24]	<.01	.14(.09)[−.04,.34]	<.01	.01(.09)[−.17,.20]	<.01
Agreeableness	.33(.16)*[.01,.65]	.01	.03(.05)[−.02,.09]	<.01	.25(.17)[−.09,.59]	<.01	.51(.16)*[.18,.84]	.03	.18(.16)[−.12,.50]	<.01
Conscientiousness	.13(.12)[−.11,.38]	<.01	.04(.04)[−.03,.12]	<.01	.22(.14)[−.04,.50]	<.01	.03(.13)[−.23,.29]	<.01	.06(.12)[−.18,.31]	<.01
Neuroticism	.39(.10)*[.19,.60]	.06	.11(.03)*[.04,.17]	.05	.40(.11)*[.18,.62]	.05	.38(.10)*[.17,.59]	.05	.37(.10)*[.17,.57]	.05
Openness	.16(.12)[−.07,.40]	<.01	−.02(.03)[−.09,.05]	<.01	.26(.13)+[−.00,.52]	.01	.14(.12)[−.10,.39]	<.01	.13(.12)[−.10,.37]	<.01

Random effects	B(SE) 95% CI	Z	B(SE) 95% CI	Z	B(SE) 95% CI	Z	B(SE) 95% CI	Z	B(SE) 95% CI	Z
Level 2 residual	1.52(.14)*[1.25,1.84]	10.26	.13(.01)*[.11,.17]	9.43	1.67(.17)*[1.35,2.05]	9.43	1.50(.15)*[1.22,1.85]	9.50	1.43(.14)*[1.17,1.74]	9.79
Level 1 residual	.87(.03)*[.81,.95]	25.08	.20(.00)*[.19,.22]	28.98	2.45(.08)*[2.29,2.62]	2.39	2.08(.07)*[1.94,2.23]	28.78	1.70(.05)*[1.59,1.82]	29.27
Autocorrelation	.26(.02)*[.20,.31]	9.09	.09(.02)[.04,.15]	3.36	.06(.02)*[.01,.12]	9.43	.09(.02)*[.04,.15]	3.37	.07(.02)*[.01,.12]	2.64

Note. B(SE) = unstandardized fixed effect estimates with standard errors in parentheses; 95% CI = 95% confidence intervals; R_β² = semi-partial R_β² effect size; Z = Wald’s Z. Intercept = mean pursuit of each goal across days. Day coded so 0 = day 1 of assessment. Big Five traits were entered as simultaneous predictors of each goal.

^*p < .05;

⁺p = .05.

Links Between the Big Five and Daily Emotion Regulation Goals

As in Study 1, people higher in agreeableness reported frequently pursuing pro-hedonic goals and people higher in neuroticism reported often pursuing contra-hedonic goals (Sample 1). Pro-hedonic goals were not positively linked to extraversion or openness. In terms of instrumental goals, people higher in neuroticism reported often pursuing impression management goals and people who were more open reported pursuing performance goals more, as expected. There were no links between performance goals and conscientiousness, or extraversion and pro-social goals. The unexpected global negative associations that contra-hedonic goals shared with agreeableness and conscientiousness from Study 1 did not emerge at the daily level.

General Discussion

Overall, we found that the Big Five are not only key predictors of individual differences in how people regulate their emotions, but also why they regulate them. Given that not all traits predicted all emotion regulation goals, there appear to be unique and meaningful trait profiles for emotion regulation goals. We conducted studies using global and daily measures of emotion regulation goals, and while the findings across these studies did not always perfectly align, there was still some convergence; see Table 1. Thus, these results provide some initial evidence for systematic relations between the Big Five and emotion regulation goals.

The Big Five and Emotion Regulation Goals

Hedonic goals.

Across all studies and samples, pro-hedonic goals were positively associated with agreeableness. Thus, while extraversion and neuroticism are the main affect-laden traits (Larsen & Ketelaar, 1991), the other Big Five can still play a fundamental role in emotional processes. The least consistent predictors of pro-hedonic goals were extraversion and openness. These associations were only present in Sample 1 in Study 1, which consisted of undergraduates. Thus, there may be an important moderating role of sample type. Notably, the studies we used to make the predictions about pro-hedonic goals also used undergraduates (e.g., Roberts & Robins, 2000; Tamir, 2009). The Big Five might be stronger predictors of emotional experiences and the desire to influence them in younger adults than in older adults. Perhaps one reason is because emotional experience (Carstensen et al., 2000) and emotion regulation (Eldesouky & English, 2018a) are more stable in older age. To the extent that emotional processes are more stable, it might be more difficult to differentially predict them in older samples. Meanwhile, we want to make a point about also attending to measurement. Contra-hedonic goals were predicted by agreeableness and conscientiousness at the global level, but not at the daily level. Perhaps this is because daily contra-hedonic goals were reported relatively infrequently. Thus, goal frequency might affect the ability to detect potential associations with the Big Five or other predictors.

Instrumental goals.

In terms of instrumental goals, our most consistent findings were that pro-social goals were positively linked to agreeableness and impression management goals were positively associated with neuroticism. These results broadly demonstrate the utility of distinguishing between social goals by indicating that different traits are associated with distinct social goals. Building on Kalokerinos et al. (2017), our findings also suggest that the link between social goals and neuroticism may be driven by impression management concerns. Similar to their study, we did not find a link between social goals and extraversion, even after distinguishing between social goals. Thus, while extraverts have more satisfying relationships (e.g., Demir & Weitekamp, 2007) and value them more (Asendorpf & Wilpers, 1998), only agreeable people regulate their emotions to help maintain relationships. Replicating Kalokerinos et al. (2017), we also never found links between performance goals and conscientiousness even though we assessed broad daily goals and used global measures. Thus, conscientious people only seem to pursue performance goals outside the context of emotion regulation more often (Roberts & Robins, 2000). Although performance goals were not associated with conscientiousness, they were positively linked to openness. Perhaps this is because performance goals encompass many activities, including those that are work-related, as well as those involving creativity. Thus, it may be important to delineate between performance goals, much like social goals.

Broad patterns.

While our focus was on specific associations between traits and goals, some of the Big Five traits were tied to more goals than others, such as neuroticism and agreeableness (Sample 1 in Study 1; Study 2). People most often regulate their emotions when they feel negatively (e.g., Barrett, Gross, Christensen, & Benvenuto, 2001; Gross et al., 2006). Given that neurotic people experience greater levels of negative affect (Watson & Clark, 1992), they might have a greater need to regulate, and thus, also pursue more goals. Meanwhile, people tend to regulate their emotions when they are with others (English et al., 2017; Heiy & Cheavens, 2014) and agreeable people are especially likely to seek out others to regulate their emotions (Williams, Morelli, Ong, & Zaki, 2018). Moreover, we found that agreeable people are also more likely to pursue pro-social goals. Taken together, these studies may help explain why agreeable people reported pursuing more emotion regulation goals. In fact, individuals higher in agreeableness might even have greater emotion regulation goal flexibility as they attempt to reach their own goals and accommodate the goals of others. Meanwhile, many of our findings aligned with prior studies on goals in general (e.g., agreeableness and pro-social goals; Roberts & Robins, 2010), which suggests a common underlying structure connecting goals across different domains. Notably, our effect sizes were smaller than effect sizes for broader goals outside emotion regulation (e.g., major life goals; Roberts & Robins, 2000), but were similar to associations between the Big Five and other emotional processes (e.g., emotion regulation strategies; Gross & John, 2003).

Future Directions and Limitations

Our current paper builds on past work on the Big Five and emotion regulation goals by examining global emotion regulation goals for the first time, assessing a range of hedonic and instrumental goals in daily life, and investigating these associations in non-college student samples. However, there are some important limitations to our work. First, we used findings on goals in general to predict how personality predicts similar types of goals. However, global goals may not always translate into the domain of emotion regulation. For instance, someone might value doing well on tasks (i.e., performance), but not typically engage in emotion regulation to achieve those tasks. Thus, it will be important to continue measuring emotion regulation goals explicitly instead of only relying on goals in general. At the same time, we suggest examining how much overlap there is between goals across related domains. Second, we did not explore the role of consciousness in emotion regulation goals. Although emotion regulation goals can be conscientious or unconscious (Mauss & Tamir, 2014), global and daily responses primarily capture how much goals are consciously pursued or valued. These reports may also be more accurate for certain individuals, such as those with greater emotional awareness. Notably, these issues surrounding consciousness are likely to be a challenge for anyone assessing psychological constructs via self-report, including emotion regulation.

Third, while we examined an important set of theoretically-derived and commonly pursued goals, there are diverse ways to measure them and additional sub-types to evaluate. We measured hedonic goals using positive and negative emotions, which are pleasant and unpleasant, respectively (Rusting & Larsen, 1999). However, some researchers propose that pleasure and pain are distinct from emotion (e.g., finding negative emotions pleasant; Tamir, 2016). Thus, it may be useful to test whether links between hedonic goals and the Big Five replicate when a pleasure vs. pain distinction is used. Meanwhile, we assessed hedonic goals by assessing the up- and down-regulation of emotional experiences. However, by definition, people need to up- or down-regulate an emotion to reach instrumental goals (e.g., feel negatively to achieve a task). Therefore, it will be important for researchers to identify the best ways for assessing pure hedonic goals (i.e., the desire to feel emotions as an end-state). Moreover, instrumental goals are a broad category going beyond performance and social goals. Traits that were not consistently related to the goals in our paper (e.g., extraversion), might be more strongly tied to other goals (i.e., epistemic goals, or wanting to learn more about the world/oneself; Tamir, 2016).

Fourth, while the Big Five are the most common trait taxonomy (Goldberg, 1993; John et al., 2008), there are many other dimensions of personality (McAdams & Pals, 2006). Future research should examine how other personality dimensions, such as life narratives or attachment style, might predict emotion regulation goals. Taking more of a dynamic approach to understanding individual differences in emotion regulation goals may be especially fruitful. For example, a longitudinal design could be used to examine how personality predicts changes in the types and range of emotion regulation goals people pursue over time, as it does for major life goals (e.g., Ludtke et al., 2009; Roberts et al., 2004). Additionally, experience sampling methods could be used to investigate whether emotion regulation goals predict daily changes in personality, as do goals in broader domains (e.g., McCabe & Fleeson, 2006).

Fifth, implications of the links between the Big Five and emotion regulation goals for regulatory behavior and well-being should be explored. Personality predicts the spaces people choose and how they influence the spaces around them (Graham, Gosling, & Travis, 2015). Emotion regulation goals can be used to understand why people select and interact with their environments the way they do. For instance, given that conscientious individuals pursue contra-hedonic goals less frequently, they might actively avoid situations that make them feel negatively. Notably, regulating in a contra-hedonic manner to achieve instrumental goals can sometimes be beneficial (e.g., Tamir, 2009). However, contra-hedonic goals can be maladaptive if pursued chronically as an end-state (i.e., the definition we focused on). Indirect evidence for this idea comes from a study showing that depressed individuals have a greater desire to feel sad than healthy controls (Millgram et al., 2015). In this case, reduced pursuit of contra-hedonic goals by those higher in conscientiousness may help explain why these individuals experience lower levels of negative affect (Fayard, Roberts, Robins, & Watson, 2011) and are less likely to experience negative life events (Kendler, Karkowki, & Prescott, 1999) or have affective disorders (Kotov, Gamez, Schmidt, & Watson, 2010). Along these lines, the Big Five might not only relate to emotion regulation goal pursuit, but also goal success. For example, because conscientious people tend to be driven and successful (Jackson et al., 2010), they might be more likely to use effective strategies for their goals. Thus, certain traits might not only be linked to a healthier profile of emotion regulation goals, but also more flexible strategy use.

Conclusion

We used the Big Five traits to identify who is more or less likely to pursue certain types of emotion regulation goals. Whereas past research focused on links between the Big Five and how people regulate their emotions (i.e., strategy use), the present results show that the Big Five are also systematically linked to individual differences in the reasons why people regulate them. These findings have implications for understanding the connections between traits and various aspects of emotion regulation, and consequentially can help elucidate how personality affects well-being. Individual differences in emotion regulation goals likely have a broader range of consequences for adjustment (e.g., social functioning, cognitive performance, physical health).

Appendix A. Emotion Regulation Goals Scale

Directions:

People often try to manage (or regulate) their emotions, including what they are feeling on the inside and what they are showing on the outside. The questions below are focused on reasons you might regulate your emotions (i.e., why you try to manage your emotions). Please indicate how often you generally regulate for the following reasons. For each reason, please use the following scale:

When you are regulating your emotions, how often do you do so because you want…

______ 1. To feel less negative emotion (e.g., anger, sadness)?

______ 2. To feel more positive emotion (e.g., joy, contentment)?

______ 3. To keep feeling positive emotion (e.g., joy, contentment)?

______ 4. To feel more negative emotion (e.g., anger, sadness)?

______ 5. To feel less positive emotion (e.g., joy, contentment)?

______ 6. To keep feeling negative emotion (e.g., anger, sadness)?

______ 7. To avoid being distracted by how you’re feeling?

______ 8. To concentrate on your work or what you’re doing?

______ 9. To stay focused on a task you’re working on?

______ 10. To make someone else feel good?

______ 11. To avoid ruining someone else’s mood?

______ 12. To avoid drifting apart from others?

______ 13. To cheer someone else up?

______ 14. To maintain a close relationship with others?

______ 15. To avoid being rejected by others?

______ 16. To avoid making an unfavorable impression on others?

______ 17. To have others approve of you?

______ 18. To make a positive impression on others?

Scoring Instructions:

Items should be presented in random order. There are five emotion regulation goal sub-scales. The items for each subscale should be averaged. There are no reverse-keyed items.

Pro-hedonic (3 items): 1, 2, 3

Contra-hedonic (3 items): 4, 5, 6

Performance (3 items): 7, 8, 9

Pro-social (5 items): 10, 11, 12, 13, 14

Impression management (4 items): 15, 16, 17, 18