Keeping Up Appearances: The Role of Motives and Utility Beliefs in Expressive Suppression

Abstract

We examined the role of impression management motives and utility beliefs in predicting suppression. In Study 1, 222 participants were assigned one of four motives (warmth, competence, pro-hedonic, control) during a job interview and reported their strategy use. In Study 2, 150 participants completed 9 days of experience sampling surveys assessing momentary emotion regulation motives and strategy use. In both studies, participants reported suppression utility beliefs. Lab results suggested a decreased preference for suppression when pursuing warmth motives over competence motives. This effect did not emerge when examining daily motives. In both studies however, people believed it was more useful to suppress negative than positive emotions, and more useful to suppress negative emotion when appearing competent than warm. Utility beliefs did not predict suppression or moderate motive-strategy links. Overall, there was mixed evidence about how suppression links to specific motives. Results suggest suppression is perceived as useful for certain impressions, but people do not act in accordance with these beliefs.

Expressive suppression, the behavioral inhibition of emotion expression (Gross, 1998), has somewhat of a bad reputation. It is not effective in changing how we feel (Webb, Miles, & Sheeran, 2012) and often harms our social relationships (English & Eldesouky, 2020). For instance, experiments show that people judge suppressors as cold and have a low desire to affiliate with them (Butler et al., 2003; Tackman & Srivastava, 2016). Meanwhile, individual differences in suppression are linked to negative social outcomes, such as lower social support (Gross & John, 2003) and relationship satisfaction (Impett et al., 2012), even months and years later (e.g., English, John, Srivastava, & Gross, 2012; Srivastava et al., 2009). However, little research has examined why people use this typically maladaptive emotion regulation (ER) strategy (Catterson, Eldesouky, & John, 2017; Eldesouky & English, 2019). Knowing why people use suppression is important because it will enhance the understanding of strategy implementation as well as the functions of modulating expressions. Suppression may be useful in certain contexts (e.g., Kalokerinos, Greenaway, Pedder, & Margetts, 2014), even if it is generally associated with negative outcomes. Furthermore, identifying why people use suppression is central to ER interventions, which sometimes aim to reduce suppression (e.g., Goldin et al., 2014). For instance, people may incorrectly believe that suppression is useful for achieving a particular motive, such as growing closer to others. However, by understanding people’s beliefs about suppression, we can aim to directly address incorrect beliefs to promote adjustment.

Emotion Regulation Motives and Suppression

Several studies suggest that one reason why people suppress their emotions is for impression management, or to influence others’ perceptions of their disposition and abilities (Jones & Pittman, 1982; Leary & Kowalski, 1990). Indirect evidence comes from research on social anxiety, where people are concerned with others’ perceptions of them (Schlenker & Leary, 1982). Undergraduates higher in trait anxiety report more habitual (Spokas, Luterek, & Heimberg, 2009) and daily suppression (Farmer & Kashdan, 2012). Meanwhile, people with social anxiety disorder suppress their anger more (Erwin, Heimberg, Schneier, & Liebowitz, 2003) and express their positive emotions less often than controls (Turk, Heimberg, Luterek, Mennin, & Fresco, 2005). There is also some direct evidence for the association between impression management motives and suppression from recent studies. English and colleagues (2017) found that undergraduates used suppression more on days when they pursued more impression management motives. These findings were then replicated in a community sample and extended to trait measures (Eldesouky & English, 2019), where people who often pursued impression management motives were also higher in habitual suppression use.

Notably, the link between impression management motives and suppression appears to be unique. Suppression is positively associated with impression management motives, but not with several other motivational contexts (e.g., pro-hedonic, pro-social; Eldesouky & English, 2019). Furthermore, impression management motives only consistently predict suppression, but not other strategies (e.g., reappraisal, distraction; Eldesouky & English, 2019; English et al., 2017). Modulating our emotional expressions is one way to manage our impressions on others because emotional expressions influence how people interpret our feelings (Keltner & Kring, 1998), intentions (Henley & LaFrance, 1984), and motivations (Zaki, Bolger, & Oschner, 2009). Suppression may be especially useful for impression management relative to antecedent-focused strategies, which primarily target emotional experience (e.g., reappraisal, distraction; English, Lee, John, & Gross, 2017). While antecedent-focused strategies can indirectly reduce expressions, suppression is more effective in doing so (Webb et al., 2012).

Impression Management Motives and Suppression

It is somewhat of a paradox that suppression is positively linked to impression management, given that use of this strategy can interfere with one’s social life (English & Eldesouky, 2020). One explanation for this paradox is that people only use suppression in certain situations or when they want to make certain impressions. As an initial investigation, we focus on the two most common impressions people are motivated to make on others: appearing warm (e.g., friendliness, trustworthiness) and competent (e.g., intelligence, efficacy; Leary, 1995; Nezlek, Schutz, & Sellin, 2007). Notably, the desire to make these particular impressions may be partially because they are also the dimensions people judge in others first (Fiske, Cuddy, & Glick, 2007). In addition to being common motives, we focus on warmth and competence because people often use different self-presentation strategies (e.g., engagement in a conversation) to achieve them (Jones & Pittman, 1982). While these motives are not mutually exclusive, this contrast can help us better delineate when suppression is more or less likely to be used.

Appearing warm.

Warmth is central to the formation and development of social relationships. It indicates whether someone is trustworthy, easy to interact with, and should be approached (Fiske et al., 2007). Suppression can lead to negative impressions of warmth (Tackman & Srivastava, 2016) because it disrupts the communication of relational interest via facial expressions (Simpson, Gangestad & Nations, 1996). Some of the behavioral cues associated with suppression, such as reduced responsiveness (Butler et al., 2003), also predict impressions of lower warmth (Carli, LaFleur, & Loeber, 1995). Moreover, the attachment literature suggests that people may habitually use suppression to avoid social connection (Shaver & Mikulincer, 2007). These attributes of suppression run contrary to the typical self-presentation strategies people engage in to appear warm, such as being more engaged in a conversation (e.g., Jones & Pittman, 1982). Therefore, if people want to appear warm, they might avoid using suppression.

Appearing competent.

In contrast to warmth, competence captures the ability to perform tasks well. It can also greatly impact how one is treated. For instance, people like and respect people they perceive as competent (Fiske, Cuddy, Glick, & Xu, 2002). Researchers have theorized that people might hide their emotional expression to appear “calm, cool, and collected” (Lyman & Scott, 1968; Richards, 2004), suggesting that suppression can serve as a protective or defensive strategy. This aligns with impression management research indicating that people want to show confidence, rather than vulnerability when striving to appear competent (e.g., Godfrey et al., 1986). In qualitative research, children and adolescents report hiding their emotions to appear competent to peers (e.g., Adler, Kless, & Adler, 1992; Dodge, Coie, Pettit, & Price, 1990), and doctors report hiding their emotions to appear competent to patients (e.g., Haas & Shaffir, 1977). These studies suggest that people might use suppression when wanting to appear competent.

Suppression Utility Beliefs

People are more likely to engage in behaviors when they believe it is useful for a given motive (Custers & Aarts, 2010; Dweck & Legget, 1998). For example, people who believe experiencing anger is useful for confronting others are more likely to upregulate their anger before confronting others (Tamir, Mitchell, & Gross, 2008). Relatedly, people are more likely to suppress their emotions if they believe that expressing their emotions will lead to social rejection (Spokas et al., 2009). Therefore, one reason why people might use suppression for some impression management motives and not others may be because of suppression utility beliefs, or beliefs about how useful suppression is.

Thus far, there are no published studies on beliefs about the utility of suppression. Therefore, one fundamental question is a descriptive one – what are people’s lay beliefs about suppression’s utility? The qualitative studies suggesting that people suppress their emotions to appear competent (e.g., Adler et al., 1992; Haas & Shaffir, 1977) may also suggest an underlying belief that suppression is useful for appearing competent. In the case of warmth, numerous studies indicate that suppression is harmful for appearing warm (e.g., Butler et al., 2003; Tackman & Srivastava, 2016). Therefore, to the extent that people are aware of this through learned experiences, they may form a belief that suppression is not useful for appearing warm.

In sum, most people may generally believe that suppression is at least more useful for appearing competent than for appearing warm. Lay beliefs aside, people are more likely to engage in a behavior if they think it is useful, even if it is not (Wrosch, Scheier, Miller, Schulz, & Carver, 2003). As an example, some people might use suppression when appearing warm, despite its harm (e.g., Tackman & Srivastava, 2016), simply because they believe it is useful. Therefore, people may be more likely to use suppression for a given motive as long as they believe it is useful for achieving that particular motive.

The Present Research

Our objective in the present research was to advance the understanding of why people use suppression by focusing on the role of motives and beliefs surrounding impression management. Prior research shows a positive association between general impression management motives and suppression use across trait and daily measures (Eldesouky & English, 2019; English et al., 2017). Importantly, this association between impression management and suppression appears to be unique. Suppression is not consistently tied to other motives and impression management (at least when measured broadly) is not consistently tied to other strategies. Here, we replicate dand extended the work on impression management and suppression in three important ways. First, because impression management is a broad construct, we investigated whether specific impression management motives (warmth, competence) differentially predicted suppression. Second, we examined beliefs about suppression’s utility for achieving different impression management motives. Third, we tested whether these beliefs predict suppression use.

We had three hypotheses: (1) people suppress less to appear warm and more to appear competent, (2) people believe that suppression is more useful for appearing competent than for appearing warm, and (3) people suppress more when trying to make a given impression if they believe suppression is useful for achieving that impression. To test these hypotheses, we conducted two methodologically diverse, but conceptually related studies. Study 1 employed an experimental design to test the causal relationship between specific impression management motives and suppression. Study 2 was an experience sampling study testing the naturalistic association between daily impression management motives and suppression. In both studies, beliefs about suppression’s utility for different impression management motives were assessed.

We took two distinct approaches to investigate the link between impression management and suppression: between-contexts and within-context. We define a between-contexts approach as comparing the use of a particular ER strategy across different contexts. In contrast, we define a within-context approach as comparing the use of a particular ER strategy to other ER strategies within a specific context. These approaches draw on general concepts from diary methods (Bolger, Davis, Rafaeli, 2003; Fleeson, 2004) where researchers seek to understand why a person’s behavior changes between situations and within a given situation. Prior studies on ER selection have taken a between-contexts approach, although they did not use this specific term (e.g., Sheppes et al., 2014). We apply a between-contexts approach to test whether people suppress their emotions relatively more when pursuing impression management motives versus other motives. Specifically, we focused on pro-hedonic motives (i.e., wanting to feel more positively or less negatively; Riediger et al., 2009) as a comparison because it is the most common ER motive and focuses on internal experience, rather than expression (Eldesouky & English, 2019; Gross, Richards, & John, 2006).

Meanwhile, taking a within-context approach can help determine whether people use suppression relatively more than other ER strategies when pursuing a given motive. We selected two comparison strategies that are distinct from suppression: reappraisal and distraction. These two strategies primarily target emotional experience, rather than expression (i.e., antecedent-focused; Gross, 1998), but they also differ from each other as reappraisal aims to change the interpretation of emotional stimuli, whereas distraction aims to redirect attention. Accordingly, the preference for using reappraisal versus distraction differs across contexts (i.e., high vs. low emotional intensity; e.g., Sheppes et al., 2014). Notably, distraction shares some similarities with suppression because it involves disengaging from an emotional stimulus (Millgram, Sheppes, Kalokerinos, Kuppens, & Tamir, 2019). Therefore, overall, taking both a between-contexts and within-context approach allows for a more comprehensive understanding of how ER strategy use fluctuates with motives.

Study 1

To test the impact of warmth and competence motives on suppression use, we randomly assigned participants to one of four motive conditions: warmth, competence, pro-hedonic, or control. Participants were instructed to pursue their assigned motive during a mock job interview. We chose a mock job interview because there are well-established paradigms (e.g., Kirschbaum Pirke & Hellahmmer, 1993) and it is a context where warmth and competence motives are both relevant (Cuddy, Glick, & Beninger, 2011). People may be motivated to appear warm given that it is a social interaction. Moreover, they may be motivated to appear competent given that they are being evaluated. Furthermore, we assessed people’s beliefs about the usefulness of suppressing their emotions when trying to appear warm and when trying to appear competent.

Method

Participants

The sample consisted of 222 undergraduates (59% female), ages 18–35 years (M = 19.78, SD = 1.68), from a mid-size Midwestern university. Their ethnic distribution was 42.5% White, 39.4% Asian/Asian-American, 15.8% African American, 8.5% Hispanic or Latin American, and 4.6% American Indian/Alaska Native or Middle-Eastern; some identified with multiple ethnicities. Participants received 1 course credit or $10 for participation. We determined sample size based on university participant subject pool availability within the semesters of data collection. Sensitivity analyses using G*Power (Version 3.1) showed that at 80% power, an alpha cutoff of .05, and four independent conditions, we could detect a small-to-medium effect size of h² = .05.

Design

We employed a between-subjects design with four ER motive conditions: warmth (n = 54), competence (n = 59), pro-hedonic motive (n = 52), and control (n = 56).

Procedure

For three minutes, participants prepared for the Trier Social Stress Task (TSST; Kirschbaum et al., 2008), a frequently used task where participants present themselves for their ideal job and try to convince interviewers to hire them. In preparation, participants could take notes, but they could not bring the notes to the interview. After preparation, participants received audio-recorded instructions to follow during the interview based on their assigned condition, adapted from Holoien and Fiske (2013). Instructions were delivered on a computer screen and via audio-recording to increase the likelihood that the instructions would be followed. Participants were randomly assigned to one of four conditions: control (“During the interview, please behave in such a way that you normally would to the interviewers”), warmth (“During the interview, please behave in such a way that you appear warm, friendly, and personable to the interviewers”), competence (“During the interview, please behave in such a way that you appear smart, competent, and in control to the interviewers”), or pro-hedonic (“During the interview, please behave in such a way that you feel calm. That means that you minimize any negative emotion that you may experience and try to feel more positively”). All participants were told that it was extremely important to follow these instructions for the sake of the study. Prior to the interview, participants completed a baseline emotion measure.

The interview lasted five minutes. It was conducted by two undergraduate interviewers who did not know the participant. The interviewers asked participants to discuss why they are a good candidate for their ideal job and prompted participants to continue their speech if they stopped talking (e.g., “You still have some time left. Please continue!”). As in the standard version of TSST, the interviewers did not ask specific questions, they kept a neutral emotional expression, and they avoided giving participants feedback (e.g., “Good job!”). Following the interview, participants completed measures of their impression management motive pursuit, ER strategy use, and suppression utility beliefs. We measured suppression utility beliefs after the TSST to avoid priming participants to think about suppression during the interview.

Measures

Emotion regulation strategy use.

Participants rated their suppression, reappraisal, and distraction use during the TSST. Suppression (four items; e.g., “I controlled my emotions by not expressing them;” “I kept my emotions to myself”) and reappraisal (six items; “I controlled my emotions by changing how I thought about the situation;” “When I was faced with a stressful situation, I made myself think about it in a way that helped me stay calm”) were measured using a version of the ERQ (Gross & John, 2003) modified for state use. We measured distraction using three items: One item borrowed from Sheppes and Meiran’s (2007) distraction instructions (“I focused my attention on something other than the situation”) and two additional items drawn from Wolgast, Lundh, and Viborg (2012; “I thought about pleasant things that have nothing to do with the situation;” “I thought about things to distract myself from my feelings”). Participants rated all items on a 7-point scale (1 = strongly disagree; 7 = strongly agree). Alphas were .65 for suppression, .80 for reappraisal, and .67 for distraction.

Suppression utility beliefs.

Participants indicated how useful they believe it is to suppress their emotions for competence and warmth motives. We selected emotions from the affective circumplex (Posner, Russell, & Peterson, 2005) to represent five positive emotions (proud, excited, happy, content, calm) and five negative emotions (embarrassed, frustrated, nervous, sad, bored) of varying arousal levels. An example competence item is “When you want to appear competent, how useful do you think it is to not show pride?” An example warmth item is “When you want to appear warm, how useful do you think it is to not show pride?” Participants rated the items on a 7-point scale (1 = not at all; 7 = a lot). We created separate composites of suppression utility beliefs for positive and negative emotions given that the perceived impact of inhibiting (and expressing) emotion may vary by valence. For warmth motives, alphas were .73 for the utility of suppressing positive emotions and .89 for the utility of suppressing negative emotions. For competence motives, alphas were .74 for the utility of suppressing positive emotions and .93 for the utility of suppressing negative emotions.

Impression management motive pursuit.

As a manipulation check, we measured self-reported motive pursuit during the TSST. Given that our primary focus was on impression management motives, participants only rated their pursuit of these motives. We used a variation of the items from Holoien and Fiske (2013) to assess the motivation to appear warm (3 items; “How warm did you try to appear to the interviewers?,” “How friendly did you try to appear?,” “How personable did you try to appear?”) and to appear competent (3 items; “How competent did you try to appear to the interviewers?,” “How smart did you try to appear?”, “How in control did you try to appear?”). Participants rated these items on a 7-point scale (1 = not at all; 7 = a great deal). Alphas were .90 for self-reported warmth motives and .80 for self-reported competence motives.

Results

We included 218 participants for our final analyses. Three participants were excluded because they did not take the study seriously (e.g., on their cell phone, completing survey measures in less than one minute). Another participant’s survey response was not recorded due to technical difficulties. The results did not significantly change when all participants were included in the analyses.

Analysis Plan

Table 1 shows descriptives and residual correlations for study variables by motive condition. We computed residual correlations between all study variables to account for potential variation due to condition assignment. To test our hypotheses, we conducted analysis of variance (ANOVA), which tested main effects (e.g., do beliefs differ across conditions?) and interactive effects (e.g., do beliefs interact with condition to predict suppression use?). We also conducted exploratory analyses involving gender and ethnicity, given the sample’s large percentage of females and Asian/Asian-Americans. These results are in the Supplementary Material. We used p = .05 as the absolute threshold. Therefore, any p-value above .05 was determined non-significant, whereas any p-value below .05 was determined significant. For post-hoc tests, we corrected for family-wise error with the Bonferroni correction (e.g., Cramer et al., 2016). We did not use this correction for the small number of a priori planned comparisons as they do not have the same risk for familywise error (e.g., Fenesi et al., 2014; Frane, 2015). The analyses were conducted in SPSS (Version 28).

Table 1.

Descriptives, Reliabilities, and Residual Correlations Between Study 1 Variables

	M(SD)	1.	2.	3.	4.	5.	6.	7.	8.	9.
Self-reported ER motives
1. Warmth motive	5.21(1.21)	1
2. Competence motive	5.36(1.11)	.37	1
ER strategy
3. Suppression	3.97(1.07)	−.14	.30	1
4. Reappraisal	4.30(1.10)	.09	.12	.28	1
5. Distraction	4.23(1.22)	−.03	.10	.01	.12	1
Suppression utility beliefs
6. Warmth (Pos)	3.12(1.19)	−.06	−.08	.10	.09	.00	1
7. Warmth (Neg)	5.16(1.56)	.07	.04	−.07	−.08	.04	−.65	1
8. Competence (Pos)	3.18(1.18)	−.05	−.07	.11	.04	.01	.63	−.46	1
9. Competence (Neg)	5.59(1.65)	.06	.14	−.00	.03	−.03	−.42	.58	−.65	1

Note. Values are based on the sample size of n = 218 who were included in the final analyses. M(SD) = means with standard deviations in parentheses; Pos = positive emotion; Neg = negative emotion. Significant effects (p < .05) are bolded.

Preliminary Analyses

Manipulation check.

There was a moderate positive correlation between the impression management motives (see Table 1), suggesting that people who wanted to appear warm also wanted to appear competent. We conducted a between-subjects ANOVA to test our manipulation check of impression management motives. The motivation to appear warm significantly varied across conditions in expected ways, F(3,214) = 7.17, p < .001, h² = .09. Specifically, participants in the warmth condition were more motivated to appear warm than those in all other conditions (competence: M_diff =.83, p < .01; pro-hedonic: M_diff =.95, p < .001; control: M_diff =.61, p = .03), indicating successful manipulation of warmth motives. However, participants did not vary across conditions in their motivation to appear competent, F(3,214) = 2.22, p = .08, h² = .03, indicating weak manipulation of competence motives.

Main Analyses

Overall strategy use between motives.

We tested overall strategy use between motive conditions (e.g., do all conditions use a given strategy equally?) using a between-subjects ANOVA. Overall, neither suppression (F(3,214) = 2.20, p = .08, h² = .03), reappraisal (F(3,214) = .38, p = .76, h² = .01), or distraction (F(3,214) = 1.03, p = .37, h² = .01) use significantly varied by condition; see Figure 1. However, given our primary interest in suppression and impression management, we also tested a theoretically-driven planned contrast: whether suppression use specifically differs between the two impression management motives. This analysis revealed that participants in the competence condition used suppression significantly more than participants in the warmth condition, t(214) = 2.19, p = .03, d = .30, as expected. To determine the source of the difference between impression management motives, we compared each of them to the control condition. Participants in the control condition used suppression more than those in the warmth condition, t(214) = 2.12, p = .03, d = .29, but similarly to those in the competence condition, t(214) = .02, p = .97, d = <.01, suggesting that focusing on appearing warm results in less use of suppression than usual.

Figure 1.

This figure represents the mean strategy use of each ER strategy by motive condition. The bars represent standard errors.

Relative strategy use within motives.

We tested relative strategy use within motive conditions using a mixed ANOVA (e.g., do all conditions use all strategies at relatively equal levels, or do some conditions use one strategy more than another?). There was an overall effect of strategy type (F(2,214) = 5.80, p < .01, h² = .03), such that participants across conditions used suppression less than distraction and reappraisal; see Figure 1. Meanwhile, there was no significant interaction between strategy type and condition, F(6,214) = 1.20, p = .30, h² = .03, suggesting that people used strategies similarly to one another across all conditions. In light of our focus on suppression and impression management, we also tested a planned contrast: whether suppression is used relatively more than other strategies in each of the two impression management motives. There was a significant difference in relative strategy use between the two impression management conditions. Within-subjects ANOVAs within each condition showed that participants in the warmth condition showed the same general trend of using suppression less than reappraisal (M_diff = .60, p < .001) and distraction (M_diff = .72, p < .001), whereas participants in the competence condition used suppression to a similar degree as reappraisal (M_diff = .23, p = .20) and distraction (M_diff = .11, p = .61).

Relative endorsement of suppression utility beliefs.

We investigated the relative endorsement of suppression utility beliefs (e.g., do people believe suppression is relatively useful for all motives?) using a mixed ANOVA. There was a significant difference between suppression utility beliefs (F(3,214) = 159.80, p < .001, h² = .52), demonstrating that people did not believe suppression was equally useful for all motives. As expected, participants believed that suppressing negative emotions was more useful for appearing competent than for appearing warm (M_diff = .43, p < .001). However, they believed that suppressing positive emotions was equally useful for appearing warm and for appearing competent (M_diff = .06, p = .99). Meanwhile, the mean-levels were higher for negative emotions than for positive emotions. Thus, participants rated suppressing negative emotions as being more useful than suppressing positive emotions; this was the case for both warmth motives (M_diff = 2.05, p < .001) and competence motives (M_diff = 2.48, p < .001).

Exploratory analyses revealed that relative differences in suppression utility beliefs did not vary by condition, F(9,214) = .90, p = .52, h² = .01. We did not hypothesize that there would be differences in suppression utility beliefs across conditions. However, to be consistent with our prior analyses in comparing the two impression management conditions, we explored relative differences in suppression utility beliefs within each of the two impression management conditions using a within-subjects ANOVA. Unlike the overall main effect, participants in the warmth condition believed suppression was equally useful when trying to appear warm and competent, regardless of whether those were positive emotions (M_diff = .21, p = 1.00) or negative emotions (M_diff = .21, p = 1.00). Similarly, participants in the competence condition also believed that suppression was equally useful for appearing warm and competent (positive emotions: M_diff = .05, p = 1.00; negative emotions: M_diff = .43, p = .25). However, consistent with the main effect, participants in the warmth condition believed that suppressing negative emotions was more useful than suppressing positive emotions when pursuing warmth motives (M_diff = 2.39, p < .001) and competence motives (M_diff = 2.39, p < .001). Meanwhile, participants in the competence condition also believed that suppressing negative emotions was more useful than suppression positive emotions for both warmth motives (M_diff = 1.60, p < .001) and competence motives (M_diff = 1.97, p < .001).

Suppression utility beliefs as a moderator of suppression use.

We ran a mixed ANOVA to test whether beliefs and their interaction with condition predict suppression use (e.g., do beliefs differentially predict suppression use across conditions?). We ran two separate models: one for beliefs about suppression’s utility for achieving warmth goals and one for beliefs about suppression’s utility for achieving competence goals. There was no main effect of suppression utility beliefs regarding warmth motives (positive emotions: F(1,218) = .83, p = .36, h² = .004; negative emotions; F(1,218) = .02, p = .89, h² > .01) or competence motives (positive emotions: F(1,218) = 3.43, p = .07, h² = .02; negative emotions; F(1,218) = 2.60, p = .11, h² = .01) on suppression use. This indicates that suppression utility beliefs did not predict suppression use overall. Furthermore, there were no significant interactions between assigned motivation (i.e., condition) and beliefs about the utility of suppression for achieving warmth motives (positive emotions: F(3,218) = .61, p = .61, h² = .01; negative emotions; F(3,218) = .72, p = .54, h² = .01) or competence motives (positive emotions: F(3,218) = .68, p = .57, h² = .01; negative emotions; F(3,218) = 1.11, p = .35, h² = .02). Therefore, suppression utility beliefs also did not predict suppression use when pursuing a particular motive.

Discussion

We conducted Study 1 to investigate the causal link between specific impression management motives (warmth, competence) and suppression use, as well as the potential moderating role of suppression utility beliefs. Comparisons between motives revealed partial support for our hypothesis: people used suppression less when focused on appearing warm, but they did not use suppression more when focused on appearing competent. Meanwhile, comparisons between strategies revealed that people used suppression less than antecedent-focused strategies when trying to appear warm, but they used all strategies equally when trying to appear competent.

Certain aspects of the study design may have also impacted the results. First, by focusing on a social evaluative task, participants in this context may have generally been more motivated to appear competent than to appear warm. Along those lines, self-reported competence motives during this task were higher than warmth motives, and only self-reported warmth motives varied across conditions. It is also possible that the manipulation was not powerful enough to result in behavioral change given that the interview was low-stakes. That is, there were no positive consequences for performing well or negative consequences for performing poorly.

In terms of suppression utility beliefs, people rated suppression as being more useful for trying to appear competent than for trying to appear warm, as expected. Notably, this effect only emerged for beliefs about suppressing negative emotions. Suppressing positive emotions was generally seen as unhelpful for making either type of impression. In contrast to our predictions, suppression beliefs were not linked to the extent to which people used suppression to achieve their assigned impression management motive. For example, people who reported stronger beliefs in suppression’s utility for appearing competence were not more likely to suppress their emotions in the competence condition than in other conditions. This pattern of findings may be due to manipulating motivation and assessing beliefs at the end of the task, thereby constraining behavior and reducing the impact of individual differences.

Study 2

In Study 2, we employed experience sampling to examine the naturalistic associations between impression management motives, suppression use, and suppression utility beliefs. Doing so allowed us to capture how psychological processes unfold in everyday life and took an idiographic approach to understanding individuals’ behaviors (Bolger, Davis, & Rafaeli, 2003; Conner, Tennen, Fleeson, & Barrett, 2009). At a basic level, one can compare between persons (Fleeson, 2004). For example, people who are more focused on competence motives might use suppression more than people who are less focused on competence motives. However, the repeated measures also allowed us to examine how a person’s own responses vary across situations, or within-person differences (Fleeson, 2004). For instance, a person might use suppression more when they are focused on competence motives than when they are focused on warmth motives. Furthermore, although suppression utility beliefs did not predict strategy use in the lab context in Study 1, these beliefs might be more relevant for suppression use in daily life, or drive people to use suppression more or less when pursuing certain motives. Given that global beliefs are a type of individual difference, they may have a greater impact on how people naturally regulate in day-to-day situations. Specifically, people may be more likely to use suppression in daily life when they believe suppression is useful for the motive they are pursuing.

Method

Participants

A sample of 150 undergraduates (68% female) were recruited from a mid-size Midwestern university, ages 18–27 years (M = 19.81, SD = 1.40). The ethnic distribution was 62.7% White, 24% Asian/Asian American, 13.3% African American, 7.3% Hispanic or Latin American, and 1.4% American Indian/Alaska Native or Middle-Eastern; the total is greater than 100% because some participants identified with more than one ethnicity. We determined the sample size using power analyses in G*Power 3 (Faul et al., 2007) at 80% power to reach an average effect size of d = .20, based on prior work demonstrating the link between daily ER motives and suppression use (English et al., 2017). Participants received 2 course credits or $25 as compensation. To increase response rates, participants who completed at least 80% of prompts were placed in a raffle to win a $50 Amazon gift card. Individuals needed to have a smartphone with internet access to participate.

Procedure

At baseline, participants completed an online survey with demographics, suppression utility beliefs, and other personality measures not relevant to the current study. On the following day, participants began the experience sampling portion of the study. This involved completing momentary measures of ER motives and ER strategy use for nine days. Each day participants were semi-randomly prompted at five time points in a 12-hour waking interval of their choice (e.g., 9am to 9pm), resulting in 45 prompts across the 9-day sampling period. When prompted, they received a text with the link to the survey. If they did not respond within 5-min, they received a reminder text. If they did not respond within 10-min, the link for that prompt expired.

Suppression utility beliefs.

Suppression utility beliefs were assessed using the same trait measure from Study 1. For warmth motives, alphas were .69 for the utility of suppressing positive emotions and .81 for the utility of suppressing negative emotions. For competence motives, alphas were .89 for the utility of suppressing positive emotions and .73 for the utility of suppressing negative emotions.

Experience Sampling Measures

At each prompt, participants were asked whether they regulated their emotions since the last text (yes or no). If they indicated that they did regulate their emotions, then they completed the measures below in regard to the most recent time they regulated. If they indicated that they did not regulate their emotions, then the survey ended.

Emotion regulation motives.

Participants selected the ER motives they had during that specific episode, indicating all motives that applied. This included two impression management motives – warmth (“to appear warmer [friendly, personable]”) and competence (“To appear more competent [smart, in control]”) – and a pro-hedonic motive (“to feel more positive or less negative”). We created the impression management motive items based on descriptions of warmth and competence motives in Holoein and Fiske (2013). We created the pro-hedonic motive based on items used to assess daily pro-hedonic motives in a study by Riediger and colleagues (2009).

Emotion regulation strategies.

Participants rated their use of suppression, reappraisal, and distraction during the episode. This included the stem “When regulating my emotions,” followed by the strategies. The suppression (“I kept my emotions to myself”) and reappraisal (“I changed how I thought about the situation”) items were measured using a slightly modified item from the Emotion Regulation Questionnaire (Gross & John, 2003) to indicate situational, rather than habitual strategy use. The distraction item was based on experimental instructions for distraction use in Sheppes and Meiran’s (2007) study: “I focused my attention on something other than the situation.” Participants rated all items on a 7-point agreement scale (1 = strongly disagree; 7 = strongly agree).

Results

Across the nine days, there were a total of 13,821 surveys. Participants completed an average of 33 surveys (M = 32.92, SD = 6.66). Across these surveys, there were 2,340 regulation instances throughout the nine-day period. The range of regulation instances was 0 (5.2% of participants) to 28 (.8% of participants). On average, participants reported at least five regulation instances (M = 5.46, SD = 5.02). We excluded seven participants for having no regulation instances, leaving 143 participants in the final analyses.

Analysis Plan

We conducted multi-level modeling (MLM) to account for the data’s two-level structure, with occasions (Level 1) nested within persons (Level 2). MLM allowed us to account for dependency of observations across occasions. In addition, it allowed us to examine individual variability in motives and strategy use, as well as their associations across occasions (Bryk & Raudenbush, 1992).

We began by testing whether daily suppression use varies by daily impression management motives. We did this in two ways. First, we examined associations between motives and overall strategy use. We conducted a model that included all motives (warmth, competence, pro-hedonic; dummy-coded as 0 = not pursued and 1 = pursued), as predictors of each ER strategy (suppression, reappraisal, distraction). In every model, we separated between-person and within-person effects of ER motives following recommendations by (Bolger & Laurenceau, 2013). To get between-person effects, we grand-mean centered the scores of a given motive across all participants, and then took the average of those scores within a person. To get within-person effects, we then subtracted the person’s grand-mean from their overall grand-mean score. Both types of effects are important for understanding how suppression operates in daily life. Therefore, we report and interpret both effects. We also highlight any places where the effects are not consistent across both levels, and why this might be. In general however, within-person effects are easier to detect as they have greater statistical power through repeated measures for a given individual (Bolger & Laurenceau, 2013).

We included random intercepts, but did not include random slopes because the models would not converge. Second, we tested whether ER motives predict the relative use of ER strategies (e.g., using suppression more than reappraisal). To do so, we adopted an MLM approach by Brans, Koval, Verduyn, Lim, & Kuppens (2013) that involves predicting overall strategy use from dummy variables that represented use of each strategy with one of the strategies serving as the reference group. This resulted in running a three-level model with strategies (Level 1) nested within occasions (Level 2) nested within persons (Level 3). This allowed us to determine whether participants used some strategies more than others, both in general and when pursuing a particular motive. We ran the model with suppression as the reference. A second model with reappraisal as the reference is in the Supplementary Material.

To test relative differences in suppression utility beliefs, we ran a repeated measures ANOVA. We then examined the moderating effects of suppression utility beliefs on the association between daily impression management motives and suppression using MLM. We conducted two separate MLM models: one focused on warmth motives and one focused on competence motives. Each model included the beliefs for a particular motive (z-scored) and their interactions with the relevant motive (e.g., belief for warmth motive × warmth motive). Furthermore, it contained the main effect of both warmth and competence motives as predictors to control for the association between them. Similar to Study 1, we also conducted exploratory analyses with gender and ethnicity, which are in the Supplementary Material. Although the use of MLM already reduces Type I error (Bryk & Raudenbush, 1992), we reduced Type I error further by using the False Discovery Rate (FDR) for all hypothesized tests (Radua & Albajes-Eizagirre, 2010), following guidelines by Benjamini and Hochberg (1995). We performed all analyses in SPSS (Version 26) and standardized all variables prior to running any models.

Preliminary Analyses

Table 2 shows descriptives, intra-class correlations (ICCs), and inter-correlations between the study variables. A little over half of the variance in both ER motives and strategies was due to between-person differences.

Table 2.

Descriptives, Intra-Class Correlation Coefficients, and Inter-Correlations Between Study 2 Variables

	M(SD)	ICC	1.	2.	3.	4.	5.	6.	7.	8.	9.	10.
ER motive
1. Warmth	30(28.78)	.55	1	.06	.29	.22	.26	.22	--	--	--	--
2. Competence	27(29.96)	.54	.74	.1	.33	.19	.21	.20	--	--	--	--
3. Pro-hedonic	45(37.15)	.54	.62	.62	1	.35	.35	.32	--	--	--	--
ER strategy
4. Suppression	4.53(1.71)	.65	.09	.05	−.06	1	.08	.01	--	--	--	--
5. Reappraisal	3.54(1.62)	.59	.03	.05	.11	−.29	1	.07	--	--	--	--
6. Distraction	3.69(1.82)	.58	−.06	.01	.06	−.04	.09	1	--	--	--	--
Suppression utility beliefs
7. Warm (Pos)	2.95(1.19)	--	.00	−.05	−.05	−.20	.09	−.01	1	--	--	--
8. Warm (Neg)	4.88(1.32)	--	−.11	−.07	.01	.22	−.04	.20	−.61	1	--	--
9. Competent (Pos)	3.22(1.25)	--	−.02	−.04	−.09	−.17	−.01	.−.01	.67	−.44	1	--
10. Competent (Neg)	5.15(1.52)	--	−.03	.01	.06	.18	.02	.10	−.56	.61	−.67	1

Note. Values are based on the sample size of n = 143 who were included in the final analyses. Correlations above the diagonal are within-person, whereas correlations below the diagonal are between-person. There are no within-person correlations with suppression utility beliefs since beliefs were measured at the between-person level. M(SD) = means and standard deviations; ICC = intra-class correlation coefficients; Pos = positive emotion; Neg = negative emotion. The mean and standard deviation for motives represent values for the aggregated proportion in which people pursued a motive when regulating their emotions; Significant effects (p < .05) are bolded.

Main Analyses

Overall strategy use between motives.

Contrary to our predictions, neither warmth motives nor competence motives were associated with daily suppression use. Thus, at the between-person level, participants who pursued warmth motives (b(SE) = .13(.07), 95% CI[−.01,.27], p = .25) more did not report lesser suppression and participants who pursued competence motives (b(SE) = −.05(.06), 95% CI =[−.19,.09], p = .75) more did not report greater suppression. Similarly, at the within-person level, participants did not report lesser suppression use on occasions when they pursued warmth motives (b(SE) = −.00(.02), 95% CI[−.04,.04], p = .99), or greater suppression on occasions when they pursued competence motives (b(SE) = .02(.02), 95% CI[−.02,.06], p = .58).

Reappraisal was also not associated with either warmth motives (between-person: b(SE) = −.01(.06), 95% CI[−.14,.12], p = .86; within-person: b(SE) = .00(.02), 95% CI[−.04,.05], p = .82) or competence motives (between-person: b(SE) = .01(.06), 95% CI[−.12,.14], p = .87; within-person: b(SE) = .00(.02), 95% CI[−.04,.05], p = .85). Distraction was negatively associated with warmth motives at the within-person level, b(SE) = −.05(.02), 95% CI[−.09,.01], p = .01: when people pursued warmth motives, they used distraction less. However, this effect was not found at the between-person level, b = −.12(.07), 95% CI[−.27,.03], p = .12, indicating that people who frequently pursued warmth motives were not more or less likely to use distraction than people who infrequently pursued these motives. Meanwhile, distraction was not associated with competence motives at either level (between-person: b(SE) = .05(.07), 95% CI[−.11,.21], p = .53; within-person: b(SE) = −.03(.02), 95% CI[−.07,.01], p = .19). And pro-hedonic goals did not significantly predict the use of any strategies at any level.

Relative strategy use within motives.

Results for relative differences in strategy use when pursuing each motive are presented in Table 3. There were no main effects of strategy type; therefore participants used all strategies at similar levels in general. In terms of interactions with impression management motives, there were no significant interactions between reappraisal and any motive at the between-person or within-person levels. Therefore, when pursuing either motive, people used reappraisal similarly to suppression. However, there was a significant interaction between distraction and warmth motives at the between-person level (b(SE) = −.22(.06), 95%CI[−.36,−.09], p = .02): participants who frequently pursued warmth motives used suppression less than distraction (see Figure 2); the interaction between distraction and warmth motives was not significant at the within-person level. And there were no significant interactions between distraction and competence motives at any level. As for the comparison motive (i.e., pro-hedonic), there were differences in relative strategy use at both the between-person and within-person levels. Participants who reported greater pursuit of pro-hedonic motives reported using suppression more than reappraisal (b(SE) = .15(.05), 95%CI[.04,.27], p < .01) and distraction (b(SE) = .15(.05), 95%CI[.04,.26], p < .01). Similarly, when pursuing pro-hedonic motives, participants used suppression significantly more than reappraisal (b(SE) = .12(.03), 95%CI[.04,.18], p = .001) and distraction (b(SE) = .09(.03), 95%CI[.02,.16], p < .01).

Table 3.

Multi-level Modeling Analyses for Effects of Daily ER Motives on Relative Strategy Use (Study 2)

	Relative strategy use
	b(SE) 95%CI	p
Intercept	−.15
Warmth motive
Between-person	.12(.05)[.01,.22]	.03
Within-person	−.00(.02)[−.05,.04]	.85
Competence motive
Between-person	−.07(.05)[−.17,.03]	.19
Within-person	.02(.02)[−.03,.06]	.40
Pro-hedonic motive
Between-person	−.09(.04)[−.18,.00]	.05
Within-person	−.05(.02)[−.10,−.00]	.03
ER strategy type
Reappraisal	−.23(.10)[−.44,−.04]	.22
Distraction	−.09(.09)[−.28,.11]	.62
ER strategy type × motive
Reappraisal × Warmth motive between-person	−.12(.06)[−.26,.01]	.25
Reappraisal × Warmth motive within-person	.01(.03)[−.05,.07]	.89
Reappraisal × Competence motive between-person	.07(.06)[−.05,.20]	.50
Reappraisal × Competence motive within-person	−.01(.03)[−.07,.05]	.89
Reappraisal × Pro-hedonic motive between-person	.15(.05)[.04,.27]	<.01
Reappraisal × Pro-hedonic motive within-person	.12(.03)[.04,.18]	.001
Distraction × Warmth motive between-person	−.22(.06)[−.36,−.09]	.02
Distraction × Warmth motive within-person	−.06(.03)[−.12,.00]	.25
Distraction × Competence motive between-person	.14(.06)[.01,.27]	.22
Distraction × Competence motive within-person	−.04(.03)[−.10,.01]	.46
Distraction × Pro-hedonic motive between-person	.15(.05)[.04,.26]	<.01
Distraction × Pro-hedonic motive within-person	.09(.03)[.02,.16]	<.01
Random effects
Level 1 residual	.94	<.001
Level 2 residual	.04	.01

Note. b(SE) 95%CI = standardized fixed effect estimates with standard errors in parentheses, followed by 95% confidence intervals; p = p-value. ER strategies were dummy coded as 0 = not pursued and 1 = pursued. Suppression was used as the reference group. ER motives, strategy type, and their interactions were entered as simultaneous predictors.

Figure 2.

This figure represents the mean-level use of suppression and distraction when pursuing warmth motives at the between-person level. Gray bands around each strategy line reflect 95% confidence intervals.

Relative endorsement suppression utility beliefs.

As in Study 1, there was a significant difference between suppression utility beliefs (F(3,148) = 92.20, p < .001, h² = .38). Participants rated suppressing negative emotions as being more useful for appearing competent than for appearing warm (M_diff = .28, p < .001). Unlike in Study 1, however, they also rated suppressing positive emotions as being more useful for appearing competent than for appearing warm (M_diff = .27, p < .001). Meanwhile, replicating Study 1 results, they believed that suppressing negative emotions was more useful than suppressing positive emotions for both warmth motives (M_diff = 1.93, p < .001) and competence motives (M_diff = 1.94, p < .001).

Suppression utility beliefs as a moderator.

Table 4 shows the results for testing suppression utility beliefs as a moderator of suppression use. As in Study 1, there were no main effects of utility beliefs for either warmth motives or competence motives on suppression use. That is, suppression utility beliefs were not associated with how much people used suppression on a daily basis. Furthermore, suppression utility beliefs did not moderate the associations between impression management motives and suppression use. That is, people who believed that suppression was more useful for achieving their motives did not report greater suppression when pursuing those motives in daily life. These findings held true regardless of the specific motive and whether the beliefs were about suppressing positive emotions or negative emotions.

Table 4.

Multi-level Modeling Analyses for Moderating Effects of Suppression Utility Beliefs on Suppression Use (Study 2)

	Suppression
	Warmth model		Competence model
	b(SE) 95%CI	p	b(SE) 95%CI	p
Intercept	−.05(.06)		−.06(.06)
Warmth motive
Between-person	.12(.07)[−.03,.27]	.11	.12(.07)[−.02,.26]	.09
Within-person	.01(.02)[−.04,.05]	.75	.00(.02)[−.04,.05]	.74
Competence motive
Between-person	−.07(.07)[−.22,.07]	.31	−.10(.07)[−.24,.05]	.18
Within-person	.03(.02)[−.02,.07]	.23	.03(.02)[−.01,.08]	.16
Suppression utility beliefs
Warmth (Pos)	−.10(.08)[−.27,.08]	.27	--	--
Warmth (Neg)	.11(.08)[−.06,.29]	.19	--	--
Competence (Pos)	--		−.11(.09)[−.30,.07]	.25
Competence (Neg)	--		.02(.09)[−.16,.20]	.81
Suppression utility beliefs × motive
Warmth (Pos) × Warmth motive between-person	.03(.05)[−.07,.14]	.80	--	--
Warmth (Neg) × Warmth motive between-person	.00(.06)[−.13,.14]	.99	--	--
Warmth (Pos) × Warmth motive within-person	−.00(.02)[−.05,.05]	.99	--	--
Warmth (Neg) × Warmth motive within-person	−.01(.02)[−.06,.04]	.89	--	--
Competence (Pos) × Competence motive between-person	--	--	−.09(.06)[−.22,.05]	.48
Competence (Neg) × Competence motive between-person	--	--	−.01(.05)[−.12,.11]	.88
Competence (Pos) × Competence motive within-person	--	--	.06(.03)[−.01,.12]	.25
Competence (Neg) × Competence motive within-person	--	--	.03(.02)[−.03,.09]	.48
Random effects
Level 1 residual	.82	<.001	.82	<.001
Level 2 residual	.16	.003	.17	.002

Note. b(SE) 95%CI = standardized fixed effect estimates with standard errors in parentheses, followed by 95% confidence intervals; p = p-value. Pos = positive emotion; Neg = negative emotion. ER motives were entered as simultaneous predictors in each model, but only beliefs relevant to a given model were entered in that model (e.g., utility beliefs for warmth were entered in the warmth model). Significant effects (p < .05) are bolded.

Discussion

The goal of Study 2 was to examine the naturalistic associations between impression management motives and suppression use in daily life, and to test suppression utility beliefs as a moderator of these links. We found almost no support for the hypothesis that impression motives differentially predict suppression use. Warmth motives were not negatively associated with suppression and competence motives were not positively associated with suppression. Furthermore, people did not use suppression less than other strategies when trying to appear warm, nor did they use suppression more than other strategies when trying to appear competent. A strategy preference only emerged for individual differences in warmth motives. Specifically, people who pursued warmth motives often tended to use suppression less than distraction (but not less than reappraisal). While the social consequences of distraction are largely known, perhaps distraction makes people seem more socially engaged than suppression (or people may hold such beliefs about these strategies). For example, people might be more focused in a conversation with partners when using distraction compared to when they are hiding their emotional expression.

In terms of suppression utility beliefs, the findings mostly replicated Study 1. Once again, beliefs about suppression’s utility were largely driven by valence as opposed to a specific motive. Thus, when people believed suppression was useful, they believed it was useful for both positive emotions and negative emotions. This was the case regardless of whether pursuing a warmth motive or a competence motive. At the same time, people believed that suppression was more useful for negative emotions than for positive emotions, perhaps because negative emotions can be more socially harmful (e.g., Chervonsky & Hunt, 2017). Further, suppression utility beliefs did not predict suppression use under different conditions (e.g., competence beliefs did not predict suppression use when competence motives were endorsed).

Although not the focus of our hypotheses, an interesting finding emerged when predicting overall distraction, which we can only speculate on. Warmth motives and distraction were negatively associated at the between-person level, suggesting that people who often want to appear warm are less likely to use distraction overall in daily life. The presence of this between-person effect suggests that individual differences in the motive to appear warm strongly predicts how much people use distraction. However, in situations where someone wants to appear warm, they are not any more or less likely to use it than they usually would.

General Discussion

In this paper, we investigated why people use suppression and the role of perceived utility beliefs. We focused on impression management motives given their positive association with trait and daily suppression use (Eldesouky & English, 2019b; English et al., 2017). In an experiment (Study 1) and an experience sampling study (Study 2), we tested whether specific impression management motives differentially predict suppression for people who believe suppression is useful for achieving those motives. Findings regarding the link between impression management motives and suppression were somewhat mixed across studies. Study 1 results suggested that people might suppress less when trying to appear warm than when trying to appear competent, and that they may prefer other strategies (reappraisal, distraction) over suppression. However, Study 2 findings showed that people did not suppress less when pursuing warmth motives (or suppress more when pursing competence motives) in daily life; they only used suppression less than distraction when trying to appear warm. Meanwhile, findings regarding suppression utility beliefs were largely consistent across studies. People reported believing that suppressing negative emotions was more useful than suppressing positive emotions. They also reported that suppressing negative emotions was more useful for competence motives than for warmth motives. However, people’s beliefs in suppression’s utility for achieving a certain impression did not predict suppression use when pursuing that motive,.

Impression Management Motives Predicting Strategy Use

Building on daily diary and global ER studies (Eldesouky & English, 2019b; English et al., 2017), we delineated between specific impression management motives when predicting suppression – warmth and competence. By providing a more nuanced perspective on impression management, as well as by manipulating motives, the current findings extend prior research suggesting that suppression is driven by social concerns (Eldesouky & English, 2019b; English et al., 2017). The mixed findings however, suggest minimal support for the idea that certain impression management motives impact suppression and highlight the need for additional replication. When people focused on appearing warm, they were less likely to use suppression (Study 1): but the association between warmth motives and suppression did not extend to daily life (Study 2). Meanwhile, there was no link between competence motives and suppression in the lab or in daily life (Study 2).

The findings in Study 1 regarding warmth motives may be somewhat unreliable given the multiple comparison tests that were run. Meanwhile, it may have been difficult to delineate effects of motives on suppression in daily life given the strong positive correlation between daily warmth and competence motives. It may be that people who use ER to try to make a good impression are motivated to look favorably across multiple dimensions (i.e., warmth, competence). Indeed, it would not be unusual for people to have multiple ER motives (Tamir, 2017). This sharpens the need to not only consider specific motives, but also the contexts in which people pursue those motives. For instance, in Study 1, we measured impression management motives during a mock job interview, where people are being evaluated. However, in Study 2, we had less information about the context in which people are pursuing impression management motives. Moreover, people likely pursued those motives across a broader range of contexts, which may have wiped out potential associations between specific motives and suppression. At the same time, given the relatively limited number of regulation instances there may not have been sufficient variation to detect within-person effects in our data. Some participants may have selected that they did not regulate to avoid answering additional questions. Alternative sampling schemes (e.g., event-based designs) may be useful for acquiring more regulation instances and determining how specific motives guide daily strategy use. For example, participants could complete a survey based on each time they try to appear warm.

Attempts to continue exploring the link between specific motives and suppression will still be important because they can help us better understand when people are more (or less) likely to use suppression, as well as other strategies. For instance, future studies may want to consider both the contextual conditions and individual differences that make people more or less likely to suppress their emotions when pursuing given motives. For example, what situations make suppression when trying to appear warm less likely? Relatedly, what are the personality characteristics that make people less likely to use suppression to appear warm? Such findings can help determine whether people know how to regulate their emotions in adaptive ways. Given that suppression has negative consequences for warmth impressions (Butler et al., 2003; Tackman & Srivastava, 2016), avoiding suppression when trying to appear warm could be considered functional. In this instance, leaning away from suppression and towards reappraisal or distraction, two strategies that have been shown to effectively change emotional experience (Webb et al., 2012), could also be considered functional. Positive emotion can promote social bonding and connection (Fredrickson, 2001), while certain negative emotions can be social destructive (e.g., Chervonsky & Hunt, 2017), so strategies that are better able enhance emotional well-being could also be beneficially for appearing warm. It will also be critical to examine how impression management motives may shape other strategies that target emotional expression besides suppression, such as exaggeration or masking.

Suppression Utility Beliefs for Impression Management

There is growing interest in beliefs about ER (Ford & Gross, 2019) and the perceived value of emotions (Becerra et al., 2020). The present research builds on this nascent literature by measuring utility beliefs regarding a specific ER strategy. Our analyses showed that at the individual level, people’s beliefs about suppression’s utility tended to correlate with one another positively. For example, if someone believed suppressing positive emotions was useful, then they also believed that suppressing negative emotions was useful. However, on average, people believed that suppressing negative emotions was relatively more useful than suppressing positive emotions. This demonstrates that there were clearly distinct beliefs about suppression’s utility based on valence. Moreover, they suggest that when examining the impact of utility beliefs on suppression, future work should consider the valence of the emotion being suppressed.

When considering motives, people believed it was more useful to suppress negative emotions for competence motives than for warmth motives, consistent with research suggesting that suppression helps people appear “calm, cool, and collected” (Lyman & Scott, 1968; Richards, 2004). Meanwhile, people rated suppressing positive emotions for warmth motives as being the least useful, a belief that aligns with research indicating positive emotions facilitate social relationships (e.g., Fredrickson, 2001). Thus, people may be less likely to suppress their positive emotions than negative emotion when focused on appearing warm. Notably, this pattern of suppression utility beliefs suggest that people are at least somewhat aware of the social impact of this strategy. More work is needed however, to delineate utility beliefs of antecedent-focused strategies (e.g., reappraisal, distraction) and how they differ from utility beliefs of suppression.

Interestingly, people did not act in accordance with their beliefs about suppression, at least regarding its utility for impression management (Studies 1 and 2). That is, even if people believed suppression was useful for a given motive they did not use suppression more when pursing that motive. According to the extended process model of ER (Gross, 2015), people evaluate a strategy’s utility and consider its potential tradeoffs when selecting a strategy. If a person deems a strategy as useful and favors it over other strategies, then that strategy will be activated. One potential reason that we did not find support of this idea is that we did not consider the emotion being regulated when assessing motives, at least in Study 2. Alternatively, it may be that people’s beliefs about other aspects of suppression play a more fundamental role in predicting its use, such as beliefs about capability. For instance, prior work showed that people are more likely to use reappraisal (Goldin et al., 2012) and suppression (Kivity & Huppert, 2018) if they believe they are capable of using it. Building on these findings, future work could explore whether multiple belief types interact to predict strategy use. For example, selection of a given strategy may require people to believe it is useful and that they are capable of deploying it.

Limitations and Future Directions

While our study had various conceptual and methodological strengths, we recognize that it also has some important limitations. One limitation is that we did not successfully manipulate competence motives in Study 1. Although it can be relatively easy to deliver motive instructions (e.g., via bolded text, audio), it may be difficult to have people follow them effectively, particularly if the motive goes against a participant’s typical tendencies. A promising alternative is to deploy situations where a more meaningful outcome depends on achieving a particular motive. For instance, people might be more motivated to appear competent during a job interview if a job opportunity is at stake. Along these lines, experimental paradigms may be able to leverage additional incentives to increase people’s motivation to pursue their assigned motive and make greater efforts to effectively manage their emotions to meet it.

At the same time, it will be critical to identify contexts where all of the manipulated motives are equally relevant, so that people do not generally gravitate more towards a particular motive. For instance, the mock job interview in Study 1 may have led to a heightened motivation to appear competence in everyone, more than the desire to appear warm. Relatedly, while we considered two of the major impressions that people are motivated to make (e.g., Leary, 1995), researchers should also consider how suppression impacts other impression management motives (e.g., showing self-control, confidence). Examining other impressions is important because people may want to make other types of impressions depending on the context. Furthermore, it will be essential to continue investigating ER processes in social interactions specifically given that people are more likely to regulate their emotions when they are with others than when alone (Heiy & Cheavens, 2014; Williams, Morelli, Ong, & Zaki, 2018). We attempted to create a controlled social interaction by using interviewers (Study 1). However, little is known about nuanced aspects of the social context in which suppression is deployed in daily life. Therefore, it may be useful for future studies to consider how effects of motives on strategy use can interact with social factors, such as who people are interacting with. Researchers can then target specific relevant factors in laboratory settings to examine their effects on strategy use.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon request.