Frontline managers’ task-related emotion regulation,emotional intelligence,and daily stress

Abstract

Research has focused on employee emotion regulation as a stable dispositional tendency. Yet effective and healthy emotion regulation requires flexibly choosing between different regulation strategies in response to various workplace situational demands. In this study, we investigate the between- and within-person emotion regulation differences of 83 frontline managers across 10 working days. Using affective events theory, we examine managers’ use of three main emotion regulation strategies (cognitive reappraisal, suppression, expression) in response to the negative affect they experience while engaging in various tasks, and the consequences for their daily stress. The moderating effects of four emotional intelligence abilities are also examined. Our results demonstrate negative emotions associated with work tasks are regulated in ways that are determined by stable, situational, and personal factors. Practical implications for organizations are considered.

JEL Classification: JEL code - D23

Keywords

Emotional intelligence emotional regulation frontline managers negative affect stress

1. Introduction

Emotion regulation (ER; i.e. modifying the experience and expression of emotions; Gross and John, 2003) is critical to workplace functioning. Research shows how effectively and appropriately individuals increase, decrease, or maintain the intensity of their emotions (e.g. anger, pride) in the workplace affects their psychological and physiological well-being, and work performance (Grandey et al., 2013; Lawrence et al., 2011). Burgeoning literature in the last two decades shows individual differences in ER strategy preferences (Gross and John, 2003), its neurophysiological underpinnings (e.g. Hobson et al., 2014), and conceptualization at multiple levels (e.g. Troth et al., 2018). Yet a much smaller body of research investigates how emotions are regulated in situ (e.g. Brans et al. (2013) examined ER in a non-work setting), especially in the work context (e.g. Totterdell and Holman (2003) examined emotional labor in service interactions). Thus, the first major aim of our study is to extend understanding about how workers, specifically frontline managers (FLMs), regulate their negative affective experiences and expressions when going about their daily job tasks (e.g. administration, operations, corrective communication). Thus, in addition to viewing ER as a stable individual difference, we explore how the use of three major ER strategies (i.e. cognitive reappraisal, expression, suppression) varies when FLMs regulate the negative affect (NA) they experience performing specific tasks (i.e. task-related NA). This contributes to workplace ER research as scholars recognize employee experiences are complex and comprise both stable (individual level) and situational (within-person level) components (Egloff et al., 2006).

The second aim is to determine the extent FLMs’ emotional intelligence (EI) abilities, at the separate ability level (awareness of own and others’ emotions; control and use of emotions; Law et al., 2004), moderate: (1) individual differences in ER strategy use and (2) within-person variability in situational ER use tied to the NA they experience while performing specific tasks (i.e. task-related NA). We propose that the extent task-related NA is associated with different ER strategies is contingent on EI. Indeed, the notion of higher EI enabling individuals to be more flexible in their ER (i.e. the ability to adapt their ER to environmental circumstances) has gained traction recently (Aldao et al., 2015) with appeals for future scholars to investigate whether specific facets of EI relate differently to particular workplace outcomes. Thus, our study builds on Totterdell and Holman (2003) showing customer service employees with higher EI use positive refocus (conceptualized as a form of ER) more frequently at work. Our third aim is to examine how the NA experienced by managers while engaging in tasks (i.e. task-related NA) and ER strategy usage relates to the stress experienced at the end of the work day. Stress is an important indicator of well-being and linked to ER and EI (Lawrence et al., 2011). Ascertaining how these factors exacerbate or ameliorate FLMs’ stress levels has important practical implications given this cohort reports high levels of strain due to heavy and varied workloads and people management challenges.

Affective events theory (AET) underpins our research (Weiss and Cropanzano, 1996). The myriad tasks that FLMs perform (e.g. chairing meetings, juggling tight budgets; Hales, 2005) are viewed as affect-eliciting events (i.e. eliciting anger, fear, pride, etc.) that give rise to ER strategy use with consequences for job-related attitudes and well-being. As specified, our particular focus is on the regulation of negative affect (NA) associated with tasks. This aligns with Dasborough’s (2006) view of work events triggering negative emotions as “hassles” with particularly deleterious consequences for individuals. Thus, we contribute to theory on the between- and within-person use of ER strategies in response to task-related NA, and the moderating influence of separate EI abilities.

There are also potential practical implications from our study. Research shows the positive effects of job design (e.g. downtime from emotionally challenging tasks; task feedback) on workplace affective experiences that in turn influence ER responses (Saavedra and Kwun, 2000). It also shows how EI training promoting emotional awareness and understanding (EI components) optimizes ER strategy usage in response to various affect-eliciting tasks (Aldao et al., 2015) and subsequent workplace functioning (Jordan et al., 2002). Thus, better understanding the nature of ER at work and the role of specific facets of EI in this process enables a more fine-grained and accurate use of these intervention strategies.

2. Theoretical background and hypotheses

2.1. Conceptualizing FLMs’ daily work as affective events

As “first-tier” managers, FLMs’ daily tasks typically relate to routine production and service delivery within an organization and the management of employees using day-to-day HR practices (Nehles et al., 2006). These managers also provide an important conduit between top-level strategic policy and its implementation on the ground with staff (Purcell and Hutchinson, 2007). The FLM role has become increasingly crucial to organizational performance over the last two decades with the delayering of organizations (Hales, 2005; Townsend, 2013). Indeed, existing research shows FLMs perform a range of daily tasks that invoke emotional responses. These include complex interactions with employees and clients such as giving performance feedback, addressing budgetary concerns, and representing multiple stakeholders’ interests (Hales, 2005). Yet despite a reasonable understanding of what their job content entails, we know little about how FLMs experience and handle their emotions in response to multiple job demands or what the consequences of this are. FLMs represent a specific workplace cohort suited to our study of task-related ER and also contrast with most leadership research focusing on middle or senior managers whose work deals less with the day-to-day demands of the frontline role.

As mentioned, AET underpins our conceptualization of the tasks that FLMs perform as affect-eliciting events (experiences) of varying intensity. AET captures within-person changes (Dasborough, 2006) by proposing characteristics of the workplace environment prompt the occurrence of certain work events, which lead to specific emotions (affective reactions), which in turn shape the work attitudes and behaviors of employees (Weiss and Cropanzano, 1996). In this study, we apply AET to examine FLMs’ daily tasks that give rise to NA (hassles) and the consequences of this affective state for their use of particular ER strategies. Although many organizations have emotional display rules restricting NA expression (and encourage positive affective expressions; Humphrey et al., 2008), workers generally report experiencing more negative (as opposed to positive) affective events at work that relate to poorer outcomes over time such as lowered commitment (Dasborough, 2006). We also use AET to examine how FLMs’ separate EI abilities moderate their use of ER strategies in response to their NA elicited from particular tasks. This responds to calls by Cropanzano et al. (2017) for researchers to recognize that individual dispositions, such as EI, impact how actors respond to affective events. Overall, we contend FLMs with higher EI are more able to appropriately adjust their ER strategy in response to NA that is elicited when performing daily tasks. We also argue these processes relate to their daily levels of stress. Figure 1 summarizes the research model and hypotheses guiding this study.

2.2. ER at work

ER has been conceptualized using various theoretical perspectives including Gross’ ER process model (Gross, 1998) and emotional labor models that focus on ER (surface and deep acting) in terms of display rules (Grandey et al., 2013). We adopt the process model as it encompasses strategies involved in “the process by which individuals influence which emotions they have, when they have them and how they experience and express these emotions” (Gross, 1998: 275). In our view, this has particular resonance for the regulation of task-related NA. The strategies in the process model are either antecedent-focused (prior to the full development of an emotional experience) or response-focused strategies (after the full intensity of the emotion has been experienced but not expressed). Gross identified four antecedent strategies comprising situation selection, situation modification, attentional deployment, and cognitive change together with a fifth response-focused strategy of response modulation (i.e. suppression). Authentic emotion expression is now also viewed as a major type of ER strategy employed to maintain felt emotion (Humphrey et al., 2015).

We focus our study on FLMs’ use of cognitive reappraisal, suppression, and expression as these mirror the distinction between antecedent- and response-focused strategies central to the process theory of ER. Cognitive reappraisal and suppression, the most commonly examined antecedent- and response-focused strategies, also show markedly different outcomes in response to NA regulation (Gross and John, 2003). Furthermore, all three strategies are common in everyday work life and, taken together, fulfill criteria for the inclusion of specific ER strategies for research (Gross and John, 2003).

More specifically, cognitive reappraisal involves cognitively transforming a situation so as to alter its emotional impact and is the most commonly used ER strategy to reduce negative emotion (Gross, 1998). Individuals using this strategy are more likely to take in the facts and critical elements of the situation and formulate a plausible narrative to make a different emotional impact (Williams, 2007). When used consistently to regulate NA, reappraisal decreases levels of negative emotional experience (Ray et al., 2010). For instance, instead of using suppression to quash anxious feelings when giving corrective feedback to staff, an FLM might cognitively reframe giving corrective feedback as a positive development opportunity, or chance for connection, to elicit positive feelings.

Emotion suppression, a response-modulation strategy, involves the active reduction of emotion-expressive behavior even though the individual is inwardly experiencing the emotion (Butler et al., 2007). It is also a strategy generally associated with self-protection and negative emotions like fear and anger. Suppression is draining of cognitive resources needed for memory, decision-making, and social functioning (Lawrence et al., 2011) as it represents an active effort to inhibit the expressive component of an emotional response. Angry encounters, customer conflict, and hostile work environments are associated with suppression strategies (Diefendorff et al., 2008). For example, an FLM might be more inclined to “bite their tongue” or maintain “a poker face,” instead of venting their feelings, if experiencing anger or frustration dealing with a challenging client. This could be an effort to avert damage to an ongoing client relationship. As with cognitive reappraisal, we expect there to be stable individual differences in FLMs’ use of suppression overall and in response to task-related NA.

Expression is another valid and effective form of ER (Gross, 2015; Humphrey et al., 2015). Individual expression of negative emotions relates to positive outcomes such as improved interpersonal functioning and increased motivation toward achieving short-term intrapersonal goals (e.g. confronting injustice; Tamir et al., 2008). Geddes and Callister’s (2007) dual threshold model explains how, in the case of anger, the controlled expression of negative emotion (e.g. slight de-amplification of experienced anger) plays a role in producing successful outcomes if the genuine emotion expressed is not so excessive as to interfere with the receptivity of the message. De-amplified expressions of anger involve emotion expressions of low intensity (e.g. frustration) that are appropriate to the display norms of anger expression for that organizational context. While our focus is not on anger per se, we expect that FLMs’ reasonable expressions of task-related NA will comprise an effective ER strategy that falls within the expression and impropriety thresholds articulated by Geddes and Callister (2007). For example, an FLM who tends to adopt this strategy is more inclined to express frustration to an employee about an overdue report. FLMs also have more status than their subordinates in completing their daily tasks. Thus, the emotional display rules around expression (of both positive and negative emotions) are more acceptable and normative for FLMs (Humphrey et al., 2008). As with reappraisal and suppression, we expect stable individual-level differences in FLMs’ use of expression in response to NA attached to job tasks.

These three main strategies are also distinct from each other in terms of modality (cognitive vs behavioral) and engagement (attention directed toward vs away from the affective experience; Parkinson and Totterdell, 1999). Reappraisal and expression are both engagement forms of ER, while suppression is a form of disengagement. Furthermore, reappraisal is conceptualized as a cognitive engagement ER strategy (involving the reinterpretation of a situation to alter one’s emotional response) that is more effortful (Sheppes et al., 2011), expression is viewed as a behavioral-engagement strategy that is less taxing overall (Humphrey et al., 2015; Lawrence et al., 2011), while suppression is considered a behaviorally effortful strategy (Brans et al., 2013).

Hypothesis 1: At the individual level, FLMs will differ in their average use of (1) cognitive reappraisal, suppression, and expression, and (2) ER strategies in response to task-related NA.

2.3. Task-related ER at work

While individuals have stable tendencies toward the ER strategies they adopt overall, healthy ER strategy usage also requires choosing between regulation strategies in a manner that is responsive to different situation demands (Sheppes et al., 2014). However, current ER research is predominantly based on studies of trait (stable) ER with questionnaires assessing individual preferences in response to generalized emotion-eliciting events that are decontextualized (Gross, 2015). If trait questionnaires are not used, research tends to comprise experimental studies instructing participants to use a single strategy in a contrived context (Sheppes et al., 2014). These provide important knowledge about the causes and consequences of ER but fail to capture the multiple ER strategies individuals employ simultaneously when confronted with an emotional event (Webb et al., 2012). Finally, studies that have investigated the dynamic situational relationships between affect and ER in daily life mostly rely on diary reports at the end of each day that are susceptible to memory biases (Nezlek and Kuppens, 2008). In our study of FLMs, we employ experience sampling methodology (ESM; Mehl and Conner, 2012) to overcome memory distortions, investigate multiple ER strategy use, and respond to calls to focus on the spontaneous selection of regulation strategies in real time and work contexts (Egloff et al., 2006; Gross, 2015).

Researchers also seldom tie ER to specific events and appraisals but rather ask participants to reflect, in an abstract manner, about how they regulate their emotions over a specific period. The challenge in obtaining interpretable relationships between a contextualized predictor (e.g. NA in response to a task) and the criterion (e.g. ER strategy in response to task-related NA) is in ensuring the appropriate operationalization and functional equivalence of the constructs (Arthur et al., 2007). Tse et al. (2018) note this is a significant issue in AET research, whereby an emotional response or other outcome variable is untied to the specific affective event. In response, we focus on emotions and ER specifically tied to FLM tasks as we do not expect all ER responses in general across the working day to be conceptually or operationally equivalent.

Broadly, in terms of within-person use of ER strategies (i.e. at the task/event level), it is expected that higher NA intensity will be positively associated with FLMs’ use of cognitive reappraisal, suppression, and expression strategies. This corresponds with existing research showing greater NA intensity is associated with a greater need to regulate (Barrett et al., 2001) and that when possible, participants report using multiple ER strategies in response to brief stressors (Aldao and Nolen-Hoeksema, 2013). In sum, besides expecting FLMs’ use of cognitive reappraisal, suppression, and expression “on the job” to vary on average between individuals and in response to task-related NA intensity (i.e. representing their more stable ER tendencies in the workplace), we also expect greater situational use of ER in response to task-related NA at the within-person level.

Hypothesis 2: At the within-person level, higher task-related NA intensity increases situational ER strategy use (cognitive reappraisal, suppression, expression) by FLMs.

2.4. The role of EI in ER

A key aim of this study is to examine whether FLMs’ use of ER in response to task-related NA is moderated by their EI abilities. EI is the ability to engage in information processing about one’s own and others’ emotions and the ability to use this information as a guide to thinking and behavior (Mayer and Salovey, 1997). It is broadly accepted that EI comprises four branches: awareness of emotions, acquiring emotional knowledge, using emotions in decision-making, and managing emotions (Mayer and Salovey, 1997). In terms of the EI and ER relationship, scholars assert ER strategies operate in concert with EI (Mayer and Salovey, 1995, 1997), and a growing group of researchers (e.g. Barrett and Gross, 2001; Geddes and Callister, 2007; Lawrence et al., 2011; Troth et al., 2018) view EI as an antecedent to ER behavior. Scholars argue that individuals higher in EI abilities share a greater capacity to choose and effectively utilize an ER strategy to produce discrete emotional experiences and expressions that will achieve their instrumental goals in a given context (Troth et al., 2018). That is, individuals with greater ability to recognize and manage their own and others’ emotions are more likely to engage in a more adaptive and greater range of ER strategies in response to affective events (Lawrence et al., 2011; Williams, 2007). However, there is little empirical work examining the relationship between EI and ER at the separate EI ability and ER strategy level.

We adopt Law et al.’s (2004) approach to EI grounded in Mayer and Salovey’s (1997) model. It provides a validated and efficient self-report way to evaluate individual EI (Kong, 2017). Law and colleagues delineate four EI abilities. These are as follows: (1) appraisal and expression of emotion (ability to understand and express own emotions), (2) appraisal and recognition of emotion in others (ability to perceive and understand others’ emotions), (3) control of emotion (ability to return to a normal psychological state after disruption), and (4) use of emotion to facilitate performance (ability to use one’s emotions constructively). At the individual level, we expect these four separate EI abilities to moderate FLMs’ habitual (average) use of ER in response to task-related NA across the 10-day study. At the within-person/situational level, we also expect these EI abilities to moderate an FLM’s ER strategy use in response to task-related NA events (i.e. a cross-level moderation). It is important to note here that the EI ability of control (sometimes also labeled regulation) delineated by Law et al. is an EI construct completely distinct from ER, despite the terminology used (see Troth et al., 2018 for discussion). In essence, Mayer and Salovey argue that EI abilities provide individuals with the capacity to choose and effectively utilize subsequent ER strategies.

As discussed, Totterdell and Holman (2003) found employees scoring higher on a global EI measure use higher levels of positive refocus (a form of ER) in a customer service context but do not differ in their perspective taking (some similarity to cognitive reappraisal) or faking of emotions (some similarity to suppression). We reason it is additionally meaningful to examine the moderating influence of separate EI abilities on ER to elucidate specific effects. We also expect the moderating influence of EI abilities to be more salient for ER responses to NA (as opposed to positive emotions). There is some evidence NA and EI are inversely related (Shi and Wang, 2007) and that people with a higher level of EI (and its branches) perceive negative emotion (e.g. in the form of anger) more objectively and are better able to regulate their NA (García-Sancho et al., 2016).

At the individual level, we expect FLMs higher in EI with a greater ability to recognize their own and others’ emotions to engage in more cognitive reappraisal and expression in response to task-related NA. These ER strategies are generally considered more functional as they are associated with reduced NA experiences that, if otherwise allowed to accumulate over time, lead to poorer health and performance outcomes (Gross and John, 2003; Humphrey et al., 2015). We reason FLMs more attuned to their own and others’ feelings will recognize earlier and more accurately if and when negative emotions in workplace situations are likely to escalate. Such recognition is also likely to prompt FLMs with these higher EI abilities to take the important first step of regulating their own NA before it becomes full blown, that is, by either utilizing the antecedent strategy of cognitive reframing or engaging in muted expression. FLMs with lower ability to recognize their own and others’ emotions, by contrast, are expected to typically use the less functional strategy of suppression in response to task-related NA (Gross and John, 2003). More specifically, FLMs less able to identify their own or others’ felt emotion, and/or accurately identify or understand the source of NA until it becomes full blown, are anticipated to report greater use of suppression, a response-modulation strategy.

Hypothesis 3: At the individual level, FLMs with higher levels of EI abilities (recognition of own/others’ emotions; control and use of emotions) will report greater average use of cognitive reappraisal and expression ER strategies, and lower average use of the suppression ER strategy, in response to higher task-related NA intensity.

The concept of ER flexibility (Aldao et al., 2015; Gross, 2015), defined as the ability to match ER strategy to environmental circumstances, is particularly relevant to this study. We suggest that EI also influences the capacity to choose more appropriate and effective ER strategies for the situation. FLMs higher in EI should have a greater ability to adapt their ER strategies to match the affective intensity of the eliciting task in situ (Gross, 2015; Lawrence et al., 2011; Troth et al., 2018). Thus, when performing tasks associated with higher levels of NA, it is conceivable that the moderating influence of EI on dynamic situational ER strategy usage is distinct from the EI and ER relationship at the individual level. The importance of teasing out these dynamic and stable components of workplace phenomena is underscored as follows (Egloff et al., 2006). A high EI FLM (who typically prefers expression as a form of ER) might choose to respond to an angry or distraught employee with a different ER response. Such an individual, adept at recognizing their own NA and appraising the negative emotion of another, is more likely to foresee the interaction escalating (with other negative long-term consequences) if their own high levels of NA are expressed. Instead, the FLM high on these two EI abilities in this situation might be more inclined to reframe their NA experience before it escalates, or otherwise engage in suppression (if high NA is already felt). FLMs higher in EI abilities are also more likely to correctly identify the particular emotion they and/or others are experiencing (e.g. embarrassment as opposed to anger, stress as opposed to enthusiasm) which makes it easier to decide on the most reasonable ER strategy for that situation. The use of ESM to investigate the dynamic use of ER in response to varying levels of NA intensity provides a valid way forward for scholars to integrate EI and ER into a coherent process. To the best of our knowledge, there are no prior empirical tests of the relationship between separate EI abilities and the dynamic situational use of specific ER strategies at the within-person level.

Hypothesis 4: At the within-person level, FLMs with higher EI (recognition of own/others’ emotions, control and use of emotions) will be more likely to use cognitive reappraisal or suppression for higher levels of NA (i.e. cross-level moderation) and be less likely to use expression.

2.5. Task-related NA, ER, and daily stress

In this study, we also examine how FLMs’ use of ER in response to task-related NA relates to their daily stress levels. The overarching thesis of AET is that affect-eliciting events give rise to important outcomes for individuals at work (Weiss and Cropanzano, 1996). As mentioned, FLMs’ daily stress is an important indicator of individual well-being linked to NA and ER (e.g. Lawrence et al., 2011). When dealing with intense negative emotions, the use of suppression is generally thought to result in non-functional outcomes, compared to the use of reappraisal or expression strategies which lead to more functional outcomes (see Gross and John, 2003; Humphrey et al., 2015). Thus, we predict

Hypothesis 5: At the within-person level, FLMs’ use of suppression positively relates to daily stress; use of cognitive reappraisal and expression negatively relates to daily stress.

Hypothesis 6: At the within-person level, ER strategies mediate the relationship between task-related NA intensity and end-of-day stress.

3. Method

3.1. Participants and procedure

Participants were 83 FLMs from several large Australian organizations in health, government, retail, and higher education. Recruitment was via workplace e-mail broadcasts requesting volunteers for a study on the daily work experiences of individuals holding FLM positions. The term FLM was explained as those holding first-level (or frontline) positions in organizations and with staff management responsibilities that are related to broad HR and industrial relations functions (e.g. performance management; staffing). Five participants (6%) were aged between 20 and 30 years, 20 (24%) between 31 and 40 years, 26 (31%) between 41 and 50 years, 24 (29%) between 51 and 60 years, and 8 (9%) were over the age of 60. Forty-one (49.4%) of the participants were women and most of the sample reported English as their first language (90%).

Participants initially completed a baseline survey on general demographics and EI. Subsequent data were collected using an experience sampling design. Participants were signaled via e-mail at intervals to complete an online survey three times per day for 10 consecutive working days (i.e. 30 maximum occasions). At each signal, participants described the task they were currently involved in and answered questions assessing the NA they associated with performing this specific task. Next, they were assessed on their ER strategy use with respect to the specific task-related NA. A total of 963 signal-level reports were received with a response rate of 39% and an average of 11.6 responses (Min = 2; Max = 29) per person. This is analogous with response rates found by Fisher et al. (2013). Eighty-eight FLMs initially completed the baseline survey. Five of these were subsequently dropped from the sample due to insufficient interval data to give a retention rate of 94%.

3.2. Measures

3.2.1. EI

EI was assessed using the 16-item Wong Law EI Scale (WLEIS; Law et al., 2004). Responses ranged from 1 “totally disagree” to 7 “strongly agree.” The four subscales are awareness of own emotions (e.g. I have a good understanding of my own emotions), appraisal of other’s emotions (e.g. I have a good understanding of the emotions of people around me), emotional control (e.g. I have good control of my own emotions), and use of emotion (e.g. I always tell myself I am a competent person).

3.2.2. Task

At each signal, FLMs were asked to indicate the work-related task they were currently engaged in from a list of seven options. These were as follows: (1) operational duties (i.e. performing the typical tasks, duties, and responsibilities of the individual’s core job as a nurse or engineer, etc.); (2) administrative/managerial duties (i.e. bureaucratic tasks such as budgeting, timetabling, staffing); (3) demonstrating duties (i.e. demonstrating/modeling good work practices to guide staff); (4) positive/motivational communication tasks (e.g. praising, listening, or mentoring staff); (5) correctional communication tasks (i.e. difficult conversations about poor performance, attitudes, and behaviors); (6) information sharing (i.e. exchange of information, problem-solving, updates on new processes/products); and (7) external influencing duties (i.e. attending committees with participants from other departments/teams; working with external stakeholders). These task types were identified in a pilot interview study with a separate sample of FLMs and correspond with other research examining FLM roles (Hales, 2005; Hutchinson and Purcell, 2010).

3.2.3. NA

At each signal, participants were asked to rate how much the task they were currently performing made them feel stressed, tense, frustrated, angry, and sad. These items were from the NA subscale of the Positive and Negative Affect Schedule (PANAS; Watson et al., 1988). Participants indicated the degree they were experiencing each emotion on a 7-point response scale anchored 1 “not at all” and 7 “very much.”

3.2.4. ER strategies

Task-related cognitive reappraisal, suppression, and expression were assessed at each signal. Cognitive reappraisal was assessed with two items from the Emotion Regulation Questionnaire (ERQ; Gross and John, 2003). An example is, “I tried to change the way I thought about this activity to modify my feelings.” Emotion suppression was also assessed with two items from the ERQ. An example is, “I did not express my emotions.” Emotion expression was assessed with three items from the Behavioral Expressivity Questionnaire (BEQ; Gross and John, 1997). An example is, “I was able to express my genuine feelings during this activity.” It is common to use a small number of items to assess variables in experience sampling studies to reduce response burden (Fisher et al., 2013). For each task, participants rated the extent they engaged in each item on a 5-point response scale anchored 1 “not at all” and 5 “very much.”

3.4.5. Daily stress

Stress was assessed with a 4-item scale (Robinson et al., 2012) at the end of each workday. An example item is, “I had a lot of responsibilities.” Participants indicated the extent they felt on a 4-point scale anchored 1 “not at all true today” and 4 “very much true today.”

3.3. Data analysis approach

To test Hypotheses 1–4, multilevel path analysis was used with model comparisons to test progressively more complex models (Bliese and Ployhart, 2002). Multilevel path analysis allowed us to concurrently test all three ER strategies at both the within-person (i.e. Level 1, task) and individual level (i.e. Level 2, between-person). Level 1 captures variation in FLMs’ ER responses to NA associated with each task, while Level 2 captures the dispositional tendency of individual FLMs to adopt each of the three ER strategies on average across 10 working days. At each stage, a Satorra–Bentler chi-square difference test was used to determine whether each step was a better overall fit to the data (Satorra and Bentler, 2010).

In the first step, we conducted a variance component model that allowed separation of variance in ER strategy use at the within-person task level (Level 1) and ER strategy use at the individual level (Level 2). As our baseline model, we fixed the covariance between ER strategies to zero. In the second step, we added task-related NA intensity as a predictor of the task-level variance in ER strategy use at the within-person task level (Level 1). We also allowed covariance in ER use to account for concurrent use of ER strategies.

In the third step, we tested whether EI abilities predicted variation between FLMs in their use of the three ER strategies at the individual level (Level 2). We controlled for between-person differences in NA intensity using group mean centering (Raudenbush and Bryk, 2002). In the fourth step, we tested (1) whether there were differences between FLMs in their dynamic use of the three ER strategies in response to higher or lower than average NA intensity, and (2) whether these differences are moderated by the four EI abilities. We estimated a random slopes model that (1) modeled variability in the relationship between NA intensity and the three ER strategies and (2) predicted the variability in this relationship by FLMs’ EI abilities (a cross-level moderation model).

4. Results

4.1. Preliminary analysis

Before testing the hypotheses, we examined the between- and within-person variance of the task-related measures across the 10 days by estimating the null model for each variable. The proportion of within-person variance ranged from 38% to 73% supporting use of multilevel analysis. Importantly, task-related NA and the three ER strategies showed significant within-person variation, supporting our expectation that the level of NA and ER varies by task for each individual FLM (Table 1).

Table 1.

Percentage of within-person variance among variables.

Task-related/daily variables	Intercept (b₀₀)	Within-person variance (e²)	Between-person variance (R²)	% of within-person variance
Task-related NA	2.14**	0.83**	1.16**	73
Reappraisal	2.49**	0.55**	0.89**	38
Expression	2.50**	0.50**	0.45**	53
Suppression	2.64**	0.77**	0.47**	62
Daily stress	2.58**	1.51**	1.42**	52

NA: negative affect.

p < .01.

Tables 2 and 3 report means, standard deviations, and zero-order correlations for variables at the within- and between-person levels. At the within-person level, Table 2 shows FLMs most frequently report undertaking operational and administrative tasks at each signal and report engaging in corrective tasks least frequently. Furthermore, undertaking corrective communication positively relates to task NA, while demonstration tasks negatively relate to task NA. At the within-person level, Table 3 also shows task NA is associated with greater use of reappraisal, suppression, and daily stress. Both reappraisal and suppression are also positively associated with daily stress. At the individual (between) level, task NA is positively associated with reappraisal, suppression, and stress. It is also negatively associated with awareness of own emotions and control of emotions. Awareness of others’ emotions positively relates to individual differences in reappraisal. There were no significant correlations with gender or age.

Table 2.

Descriptive statistics and correlations for task type at each signal with task NA.

	Mean	%	SD	1.	2.	3.	4.	5.	6.	7.
FLM tasks
1. Operational	0.23^a	23.00	0.42	–
2. Admin	0.20^a	19.70	0.40	−0.27*	–
3. Demonstrate	0.02^a	2.40	0.15	−0.09*	−0.08*	–
4. Positive	0.02^a	1.93	0.14	−0.08*	−0.07*	−0.02	–
5. Corrective	0.01^a	1.06	0.10	−0.06*	−0.05*	−0.02	−0.01	–
6. Inform	0.09^a	9.01	0.29	−0.17*	−0.16*	−0.05*	−0.04*	−0.03	–
7. External	0.04^a	3.70	0.19	−0.11*	−0.10*	−0.03	−0.03	−0.02	−0.06**	–
Affective response
8. Task NA	1.83	–	0.87	0.01	0.00	−0.05*	−0.04	0.15*	−0.03	0.03

NA: negative affect; FLM: frontline manager.

N = 963 observations. Statistics represent the maximum likelihood estimated means, standard deviations, and correlations of the data.

Mean task preference for each signal.

p < .05; **p < .01.

Table 3.

Multilevel descriptive statistics.

Within-person	Mean			SD		1.		2.		3.		4.
1. Task NA	0			0.87		–
2. Reappraisal	0			0.74		0.18**		–
3. Expression	0			0.87		0.02		−0.01		–
4. Suppression	0			0.70		0.14**		0.27**		−0.06		–
5. Daily stress	0			0.92		0.63**		0.26**		0.01		0.23**
Between-person	Mean	SD	1.	2.	3.	4.	5.	6.	7.	8.	9.	10.	12.
1. Gender	–	–	(–)
2. Age	–	–	−0.09	()
3. Task NA	1.83	1.17	0.01	−0.07	(0.86)
4. EI—Own aware	5.72	0.99	0.14	0.08	−0.18*	(0.94)
5. EI—Other aware	5.42	1.03	0.19	0.06	−0.08	0.47**	(0.94)
6. EI—Control	5.43	1.02	0.15	0.10	−0.20*	0.67**	0.49**	(0.90)
7. EI—Use	5.70	0.95	0.03	0.07	−0.16	0.57**	0.42**	0.56**	(0.86)
8. Reappraisal	2.48	0.95	0.03	0.04	0.25*	0.05	0.33*	0.03	0.14	(0.92)
9. Expression	2.51	0.67	0.13	0.09	0.17	−0.01	0.16	−0.02	−0.05	0.34**	(0.63)
10. Suppression	2.66	0.70	−0.04	0.10	0.22*	−0.01	0.10	−0.08	−0.11	0.72***	0.23	(0.63)
11. Stress	2.64	0.92	−0.02	0.03	0.24*	0.09	0.09	0.03	−0.06	0.11	−0.06	0.11	(0.81)

NA: negative affect; EI: emotional intelligence.

N = 475–963 observations and N = 83 individuals. Statistics represent the maximum likelihood estimated means, standard deviations, and correlations of the data. Intraclass correlations for cognitive reappraisal = .62, expression = .48, suppression = .39; task NA = .58; and stress = .48.

p < .05; **p < .01; ***p < .001.

We also conducted multiple confirmatory factor analysis (CFA) to verify the discriminant validity of the measures. First, we tested a model whereby within-person measures were included at within and individual levels, while EI abilities were included at the individual level only. This model displayed an acceptable fit (χ²(826) = 1466.83, p < .001; comparative fit index (CFI) = .95, Tucker–Lewis index (TLI) = .94, root mean square error of approximation (RMSEA) = .05, SRMR_withn = .05, SRMR_between = .06), with better fit indices than two alternative models. The first alternative model loaded all within-person variables on one within-person factor (χ²(869) = 2947.67, CFI = .71, TLI = .67, RMSEA = .08, SRMR_within = .14, SRMR_between = .08). The second alternative model loaded all individual-level variables on one between-person factor (χ²(859) = 4629.86, CFI = .51, TLI = .46, RMSEA = .10, SRMR_within = .19, SRMR_between = .08). Overall, the results support the separation of focal variables.

4.2. Testing the effects between task NA and ER strategies

Hypotheses 1 to 4 were tested by conducting multilevel path analysis with a maximum likelihood robust estimator using MPlus 7.4 (Muthén and Muthén, 2012). We established model fit in four steps by calculating the differences in the log likelihood ratio for each model (Finch and Bolin, 2017). The addition of task-related NA intensity as a within-person level predictor significantly improved the model (Model 1—Model 2, ∆χ² = 121.92, ∆df = 9, p < .001), as did the addition of EI as an individual-level (between-person) predictor of ER strategy use (Model 2—Model 3, ∆χ² = 24.12, ∆df = 12, p < .05). Furthermore, the estimation of a random slopes model that allowed individual-level variation in the relationship between task-related NA intensity and ER strategy use resulted in significantly better model fit over the fixed slopes of Model 3 (Model 3—Model 4, ∆χ² = 188.05, ∆df = 18, p < .001). Figure 2 shows the results of the final random intercepts, random slopes path analysis (Model 4), depicting the significant paths only.

In support of Hypothesis 1a, Figure 2 shows there is significant variation between FLMs in their individual-level use of all three ER strategies (γ₀₁ = .43, p < .001 for suppression; γ₀₁ = .45, p < .001 for expression; γ₀₁ = .75, p < .001 for cognitive reappraisal). In accordance with Hypothesis 1b, there is also significant variation between FLMs in the relationship between task-related NA intensity and their average use of cognitive reappraisal (γ₁₁ = .06, p < .05), suppression (γ₁₁ = .04, p < .05), and expression (γ₁₁ = .07, p < .05) across the 10 days. In partial support of Hypothesis 2, higher task NA is positively related to reappraisal (γ = .20, p < .001) and suppression (γ = .16, p < .001), but not expression (γ = .03, ns) at the within-person level. While not specifically tied to a hypothesis or shown in Figure 2, further analysis revealed a significant negative relationship between the intercept and slopes for suppression (γ = −.06, p < .01) and expression (γ = −.06, p < .01). This means FLMs reporting higher levels of task NA on average over the study period showed smaller increases in suppression and expression when situational task NA was above average for them.

4.3. Testing the moderating effects of EI on ER strategies

To test Hypotheses 3, we explored the moderating effect of the four EI abilities on FLMs’ individual-level (average) use of ER strategies in response to task-related NA over the study period. As shown in Figure 2, the results indicate that FLMs’ awareness of other’s emotions predicted significantly higher subsequent use of both cognitive reappraisal (b = .39, p < .001) and expression (b = .16, p < .05) in response to task-related NA, partially supporting Hypothesis 3. None of the EI abilities moderated individual-level suppression in response to task-related NA.

Figure 2 also displays the results of the cross-level moderation analysis to test Hypothesis 4. In tentative support for Hypothesis 4, there is some evidence of a cross-level moderation effect for one aspect of EI on the within-person use of cognitive reappraisal in response to varying levels of task-related NA intensity. Specifically, an FLM’s rating of their awareness of others’ emotions was marginally related to subsequent higher use of cognitive reappraisal at higher levels of NA (b = .07, p < .08). Examining the simple slopes of this relationship (Figure 3), we note that those FLMs with below-average ability to recognize others’ emotions tended to use cognitive reappraisal less at higher levels of NA. Conversely, those with above-average ability to recognize others’ emotions tended to maintain use of cognitive reappraisal at higher levels of NA intensity.

4.4. Testing mediation effects

As shown in Table 4, process analysis (Hayes, 2017) indicates that the direct effect of task-related NA intensity (direct effect = .18, SE = .03, p < .001) and suppression (direct effect = .03, SE = .02, p < .001) on daily stress is significantly positive. In contrast, the direct effect of expression on daily stress is significantly negative (direct effect = −.05, SE = .02, p < .001), providing partial support for Hypothesis 5. Table 4 also shows the indirect effect of task-related NA intensity on daily stress through the ER strategy of suppression is significantly positive (indirect effect = .01, SE = .01, 95% confidence interval (CI) = (.01, .04)) and through expression is significantly negative (indirect effect = −.02, SE = .01, 95% CI = (−.04, −.01)). Thus, Hypothesis 6 is partially supported.

Table 4.

Mediating effects of emotional regulation strategies on daily stress.

Path analysis	Mediators
Path analysis	Reappraisal		Suppression		Expression
	β	SE	β	SE	β	SE
Task NA to ER strategy	0.49**	0.11	0.63**	0.10	0.32**	0.09
ER strategy to daily stress	0.02	0.02	0.03*	0.02	−0.05**	0.02
Bootstrapping
Indirect effect through mediator	0.01	0.01	0.01*	0.01	−.02*	0.01
BC 95% CI^a	[−0.0101, 0.0338]		[0.0001, 0.0363]		[−0.0410, −0.0037]
Effect of task NA on stress	β	SE
Total effect	0.20**	0.01
Direct effect	0.18**	0.03

NA: negative affect; ER: emotion regulation.

N = 475 observations.

IV = task NA; DV = daily stress.

BC = bias-corrected; CI = confidence interval (5000 bootstrap samples).

p < .05; **p < .01.

5. Discussion

Our study aimed to examine in a workplace setting the stable (individual-level) and situational (within-person) use of three specific ER strategies (i.e. cognitive reappraisal, suppression, expression) in response to the NA associated with tasks over a period of 10 working days. A related goal was to investigate whether ER strategy usage mediates the harmful effects of task-related NA on FLMs’ daily stress levels. In this regard, our study contributes to the body of ER research with its predominant focus on individual differences in ER as stable and trait-like properties (Webb et al., 2012). Another key objective was to investigate the moderating influence of the four separate EI abilities on ER strategy usage “on the job” (in response to calls by Lawrence et al., 2011; Troth et al., 2018). The focus on FLMs as a distinct cohort uniquely involved in the daily juggle of tasks that elicit emotion is also novel. Extending existing research (e.g. Brans et al., 2013; Totterdell and Holman, 2003) the results show that FLMs’ ER strategy usage at work comprises both stable and situational/dynamic components (Hypotheses 1 and 2) that is moderated by FLMs’ EI, more specifically awareness of other’s emotions (Hypotheses 3 and 4). Finally, our study showed how FLMs regulate their NA arising from job tasks has consequences for their daily stress levels (Hypotheses 5 and 6).

5.1. Theoretical implications

Several theoretical implications stem from these findings. Our study builds on previous research to show how situational ER and its consequences in the form of daily stress can be understood using AET (Ashkanasy and Humphrey, 2011). The utility of conceptualizing the repeated daily task appraisals of FLMs as affective events also builds on recent work by key scholars in the AET literature (Cropanzano et al., 2017). By focusing our investigation on task events, the associated NA intensity, and ER strategies elicited by such appraisals, we were able to examine ER at the level of an affective event and capture some of its more dynamic characteristics. This approach has largely been missing from past ER research (Tse et al., 2018), although we did acknowledge some critical ESM research on which we build (Brans et al., 2013; Totterdell and Holman, 2003). Thus, our study underscores the importance of obtaining interpretable relationships between a contextualized predictor (NA in response to a task) and the criterion (ER strategy in response to task-related NA) when using AET (Arthur et al., 2007).

Our study also extends research on stable forms of ER strategy usage (Gross and John, 2003) in two ways. One, by investigating individual differences in the workplace and FLM domain. Two, by broadening the focus beyond the commonly examined cognitive reappraisal and suppression strategies to include expression (as advocated by Brans et al., 2013; Humphrey et al., 2015). Results show that as a group FLMs habitually or typically use expression and suppression ER strategies on the job, although there is some variability between managers. As both strategies capture different types of behavioral responses to NA experiences (Parkinson and Totterdell, 1999), one explanation is that within the “normal bounds” of felt NA, FLMs on average are more inclined to express their feelings (the least effortful strategy) or “hold their tongue” when dealing with the task at hand. We reason that, when NA is nonextreme, suppression and expression might represent more expeditious habitual ER strategies for FLMs (at least in the short term) to ensure job completion and a quick transition to the next task when compared to a habitual form of cognitive reappraisal. That is, the average FLM conserves their cognitive resources for more negatively charged tasks (as our cross-moderation findings suggest). It has also previously been established that strategies such as social sharing (a behavioral-engagement approach) can provide emotional relief (Brans et al., 2013) and is a less effortful choice (Humphrey et al., 2015).

While not specifically hypothesized, we also found unpredicted relationships between the habitual (individual-level) and dynamic (within-person) use of ER strategies. FLMs with higher habitual use (average) of suppression and expression were less likely to increase their use of suppression and expression for tasks at higher levels of NA. This is an important finding as it supports our arguments about the importance of partitioning out the stable (i.e. habitual) and situational (i.e. task-specific) components of ER. The dual threshold model (Geddes and Callister, 2007) provides one explanation for these results. It is possible some FLMs were already operating at the upper limits of the propriety expression threshold for their position (for average task-related NA). Cognizant of crossing the impropriety threshold if they authentically expressed higher felt negative emotion, these FLMs might have switched to a different ER strategy (e.g. reappraisal) to downplay the likelihood of such expressions and their possible deleterious outcomes. Indeed, Brans et al. (2013) suggest that expression, at least compared with reappraisal, is especially sensitive to contextual factors such as the intensity of the emotion-eliciting event. We are less clear about the finding for suppression, although we consider a ceiling effect as one possible explanation. Clearly, these explanations were untested in this study and require further investigation to be substantiated.

Overall, our results demonstrate that workplace ER is complex and needs to be considered in terms of its stable and situational/dynamic components in real-time contexts (Egloff et al., 2006). They also extend a small but growing body of research suggesting individuals to different degrees flexibly adapt their use of ER strategies depending on situational demands (Gross, 2015; Lawrence et al., 2011; Sheppes et al., 2014). It is clear that individual ER responses do not operate in a vacuum and are contingent on the current task and the associated NA it elicits.

Our second aim was to investigate the influence of EI on the habitual and dynamic use of ER strategies. FLMs higher on the EI component of understanding (or appraising) others’ emotions reported higher habitual use of cognitive reappraisal and expression strategies for typical levels of task NA. Based on the original arguments put forward by Mayer and Salovey (1997), it follows that those FLMs more astute at accurately reading and understanding another’s emotions (e.g. recognizing that a subordinate is unhappy with a work assignment) will be more actively involved in the ER process. Indeed, expression and cognitive reappraisal represent ER strategies whereby attention is directed toward the affective experience (Parkinson and Totterdell, 1999) and are viewed as more effective and functional habitual ER strategies when compared with habitual suppression (Gross and John, 2003; Humphrey et al., 2015). We also found evidence that the same EI component of understanding (or appraising) others’ emotions increases (i.e. moderates) the dynamic use of cognitive reappraisal for tasks with higher NA. For us, this aligns with the view purported by Mayer and Salovey (1995) and subsequent others (Aldao et al., 2015; Pẽna-Sarrionandia et al., 2015) that individuals with higher levels of EI will be more effective and adaptive (or flexible) in their choice of ER strategy in response to an emotion-eliciting event. Such individuals are able to more successfully manage their emotional experiences to achieve instrumental goals in a given context. In other words, they have a greater ability to match ER strategy to environmental circumstances.

We are somewhat unclear why the other EI components did not emerge as significant moderators of the task NA and ER strategy relationship. We suspect EI pertaining to understanding others’ emotions is most salient when FLMs are faced with tasks involving the supervision or management of others (although this is untested). Managers’ report these more advanced interpersonal tasks (e.g. giving correctional feedback; allocating difficult work tasks to others) as most challenging and most likely to increase their levels of arousal (Atwater et al., 2001). Other individuals in these types of interactions are also more likely to express higher than usual levels of emotion that, in turn, impact FLMs’ NA levels via emotional contagion processes (Johnson, 2008). It seems reasonable that FLMs more capable of identifying and understanding others’ feelings during such tasks are better placed to choose the most effective and appropriate form of ER strategy in this context. Our results provided some preliminary support in this regard finding a positive correlation between the FLM task of giving corrective communication and task-related NA intensity. It could be for this particular task that situational cognitive reappraisal (Ray et al., 2010) would enable FLMs to reduce their own levels of NA and to proceed with the task at hand. We acknowledge that this cross-level moderation result for reappraisal needs to be replicated in further research. We are also mindful that the FLMs in our sample generally experienced low levels of NA and speculate that the other three EI components could emerge as salient in another group of participants experiencing much higher NA.

5.2. Practical implications

Several practical implications are drawn. The overarching implication is that, given the habitual and dynamic aspects of ER strategies, there are multiple points of intervention available for organizations regarding the effective use of situational ER in response to negative events. This includes the FLM themselves and others who have to interact with them (e.g. subordinates). While the typical FLM will on average employ suppression and expression strategies, the dynamic variability observed in both ER responses suggests that ER is not a rigid response. ER strategy use might be modifiable through training (see Wadlinger and Isaacowitz, 2011 for ER training options for organizations) and via modification of the organizational emotional display rules explicitly communicated to FLMs (see Humphrey et al., 2015).

Second, the results point to the potential value in EI training (Jordan et al., 2002) to optimize ER strategy selection, especially for tasks associated with higher NA (e.g. encouraging higher use of cognitive reappraisal during corrective communication with subordinates). However, the typically blanket assertion to boost EI needs a more nuanced focus. Our study indicates it is the specific capacity to recognize and understand others’ emotions that contributes to habitual and situational ER. The complex construct of EI cannot be reduced to a simple single factor without reducing our understanding of the construct and diminishing its effectiveness in terms of selection decisions or training. Finally, the findings show that ER has a dynamic component. The selection of FLMs (or others) on traits alone will not necessarily account for these dynamic situational factors in jobs where effective ER is imperative. Selection based on the situational use of ER connected to salient affective events for the job role could provide incremental validity for predicting outcomes over and above traditional trait measures.

5.3. Limitations and future directions

We acknowledge limitations in the design of this study and make recommendations for future research. First, we relied on self-reports to assess the experience and regulation of emotions. These data are prone to memory biases and social desirability motivations. We were able to lessen some of these limitations by using ESM which reduces the time between experience and assessment (Mehl and Conner, 2012). A self-report measure of EI was also chosen in favor of an ability measure (e.g. MSCEIT (Mayer–Salovey–Caruso Emotional Intelligence Test); Mayer et al., 2002) because of its quick and simple administration yet preservation of sufficient validity and reliability requirements. Furthermore, while a strength of this study was the clear tying of ER strategy to specific work task appraisals, we did not have sufficient variation in our task-type data (especially tasks with high NA intensity) to meaningfully test our full research model. That is, we were unable to identify task-type differences in terms of ER use and the moderating effects of EI abilities. As mentioned, the most frequent tasks reported were operational and administrative tasks with relatively low levels of NA intensity. Rather than collecting data at intervals about discrete task events, we also recommend in the future asking participants to report only on tasks eliciting high emotional intensity as they occur.

Our findings suggest several future directions for scholars. First, our focus was on affective “hassles” or ER associated with appraisal ratings of tasks associated with NA. It might also be useful to examine ER responses to “uplifts” or positive affect appraisals of events. Second, scholars might consider collecting information about the goals FLMs were pursing in their various use of ER strategies. Research is emerging to show that people do not always regulate their emotions to increase pleasant and decrease unpleasant emotions (Tamir et al., 2008). Rather, goals in ER typically involve attaining instrumental goals rather than hedonistic ones. Studying when and how different ER strategies are used in the context of workplace goal pursuits seems fruitful. We also propose investigating the additive effects of affective events (hassles and uplifts) over time on subsequent affective event experiences as worthwhile. AET researchers (Cropanzano et al., 2017) argue that discrete affective events are likely to accumulate and have additive (even synergistic) effects such that the momentary emotional impact of a negative emotive task could build on other similar events, to have an overall effect (in this case on situational ER).

We acknowledge our design does not establish the temporal order of the relationship between NA and ER. Rather our analysis focused on variations over time in the relationship between NA and ER measured at the same time point. While the design and wording of our instruments specifically instructed participants to rate their NA with respect to a task, and then subsequently rate how they regulated these task-related emotions, we did not analyze lagged effects within this study. In investigating how a single situationally anchored relationship varies over time and person, our design limited our ability to establish the temporal precedence of NA on ER. The particular features of our design, however, do reduce the plausibility of a reverse-causality (responses to ER items causing responses to NA items) critique. A reverse-causality argument would also suggest that higher cognitive reappraisal led to higher NA (i.e. it is either an ineffective strategy, contradicting other studies, or that individuals attempt to reappraise to increase the amount of negative emotion they experience). It is also important to highlight that we measured EI several days prior to ER and so the ER and EI reverse-causal argument is less plausible. Nevertheless, this is an important question and we recommend future researchers look to techniques such as cross-lagged analysis or event-contingent experience sampling to establish the temporal precedence of NA to ER in situ. Finally, there is also a need to extend this research to within-person mediation designs with other well-being (beyond daily stress) and performance measures as subsequent within-person outcomes of situational ER. Our findings regarding daily stress outcomes suggest this would be fruitful. Also connected to this idea is the relatively under-researched notion that some EI abilities might have unintended negative consequences (see Foo et al., 2004). We note our initial prediction was that FLMs higher in EI would be more likely to use situational suppression in the face of high task NA which would then lead to more daily stress (moderated mediation). Although we found no support for this, it important for researchers to appreciate that EI is not a panacea and separate EI abilities may affect outcomes differently and even unfavorably in some instances.

5.4. Conclusion

This study extends understanding of the use of habitual (individual-level) and dynamic/situational (within-person-level) forms of ER in real-time work settings with a sample of FLMs. It also responds to calls to increase knowledge of the relationship between ER and specific EI abilities. It is clear FLMs regulate NA related to their job tasks in various ways that are determined by stable, situational, and personal factors. While on average the FLMs in our sample used the expression and suppression forms of ER, they were less likely to do so when specific tasks elicited higher levels of NA. FLMs more able to understand others’ emotions were also inclined to use the more functional forms of expression and reappraisal, and to increase their use of situational reappraisal when faced with higher levels of task-related NA. Organizations looking to improve FLMs’ daily affective experiences would be wise to investigate this aspect of organizational behavior to address well-being and performance issues, especially given our demonstrated associations with stress.

Footnotes

Final transcript accepted 15 May 2022 by Weiting Zhang (AE International Business).

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was funded by Centre for Workplace Leadership (CWL), University of Melbourne, Australia.

ORCID iDs

Ashlea C Troth

Keith Townsend

References

Aldao

Nolen-Hoeksema

(2013) One versus many: Capturing the use of multiple emotion regulation strategies in response to an emotion-eliciting stimulus. Cognition & Emotion 27: 753–760.

Aldao

Sheppes

Gross

(2015) Emotion regulation flexibility. Cognitive Therapy and Research 39: 263–278.

Arthur

Bell

Edwards

(2007) A longitudinal examination of the comparative criterion-related validity of additive and referent-shift consensus operationalizations of team efficacy. Organizational Research Methods 10: 35–58.

Ashkanasy

Humphrey

(2011) A multi-level view of leadership and emotions: Leading with emotional labor. In: Bryman

Collinson

Grint

, et al. (eds) The SAGE Handbook of Leadership. London: SAGE, pp. 363–377.

Atwater

Carey

Waldman

(2001) Gender and discipline in the workplace: Wait until your father gets home. Journal of Management 27: 537–561.

Barrett

Gross

(2001) Emotional intelligence: A process model of emotion representation and regulation. In: Mayne

Bonanno

(eds) Emotions: Current Issues and Future Directions. New York: The Guilford Press, pp. 286–310.

Barrett

Gross

Christensen

, et al. (2001) Knowing what you’re feeling and knowing what to do about it: Mapping the relation between emotion differentiation and emotion regulation. Cognition & Emotion 15: 713–724. https://doi.org/10.1080/02699930143000239

Bliese

Ployhart

(2002) Growth modeling using random coefficient models: Model building, testing, and illustrations. Organizational Research Methods 5: 362–387.

Brans

Koval

Verduyn

, et al. (2013) The regulation of negative and positive affect in daily life. Emotion 13: 1–14.

10.

Butler

Lee

Gross

(2007) Emotion regulation and culture: Are the social consequences of emotion suppression culture-specific? Emotion 7: 30–48.

11.

Cropanzano

Dasborough

Weiss

(2017) Affective events and the development of leader-member exchange. Academy of Management Review 42: 233–258.

12.

Dasborough

(2006) Cognitive asymmetry in employee emotional reactions to leadership behaviors. The Leadership Quarterly 17: 163–178.

13.

Diefendorff

Richard

Yang

(2008) Linking emotion regulation strategies to affective events and negative emotions at work. Journal of Vocational Behavior 73: 498–508.

14.

Egloff

Schmukle

Burns

, et al. (2006) Spontaneous emotion regulation during evaluated speaking tasks: Associations with negative affect, anxiety expression, memory, and physiological responding. Emotion 6: 356–366.

15.

Finch

Bolin

(2017) Multilevel Modeling Using Mplus. Boca Raton, FL: CRC Press.

16.

Fisher

Minbashian

Beckmann

, et al. (2013) Task appraisals, emotions, and performance goal orientation. Journal of Applied Psychology 98: 364–373.

17.

Foo

Elfenbein

Tan

, et al. (2004) Emotional intelligence and negotiation: The tension between creating and claiming value. International Journal of Conflict Management 15: 411–429.

18.

García-Sancho

Salguero

Fernández-Berrocal

(2016) Angry rumination as a mediator of the relationship between ability emotional intelligence and various types of aggression. Personality and Individual Differences 89: 143–147.

19.

Geddes

Callister

(2007) Crossing the line(s): A dual threshold model of anger in organizations. Academy of Management Review 32: 721–746.

20.

Grandey

Diefendorff

Rupp

(2013) Bringing emotional labor into focus: A review and integration of three research lenses. In: Grandey

Diefendorff

Rupp

(eds) Emotional Labor in the 21st Century: Diverse Perspectives on Emotion Regulation at Work. New York: Psychology Press/Routledge, pp. 3–27.

21.

Gross

(1998) The emerging field of emotion regulation: An integrative review. Review of General Psychology 2: 271–299.

22.

Gross

(2015) Emotion regulation: Current status and future prospects. Psychological Inquiry 26: 1–26.

23.

Gross

John

(1997) Revealing feelings: Facets of emotional expressivity in self-reports, peer ratings, and behavior. Journal of Personality and Social Psychology 72: 435–448.

24.

Gross

John

(2003) Individual differences in two emotion regulation processes: Implications for affect, relationships, and well-being. Journal of Personality and Social Psychology 85: 348–362.

25.

Hales

(2005) Rooted in supervision, branching into management: Continuity and change in the role of first-line manager. Journal of Management Studies 42: 471–506.

26.

Hayes

(2017) Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach. New York: The Guilford Press.

27.

Hobson

Saunders

Al-Khindi

, et al. (2014) Emotion down-regulation diminishes cognitive control: A neurophysiological investigation. Emotion 14: 1014–1026.

28.

Humphrey

Ashforth

Diefendorff

(2015) The bright side of emotional labor. Journal of Organizational Behavior 36: 749–769.

29.

Humphrey

Pollack

Hawver

(2008) Leading with emotional labor. Journal of Managerial Psychology 23: 151–168.

30.

Hutchinson

Purcell

(2010) Managing ward managers for roles in HRM in the NHS: Overworked and under-resourced. Human Resource Management Journal 20: 357–374.

31.

Johnson

(2008) I second that emotion: Effects of emotional contagion and affect at work on leader and follower outcomes. The Leadership Quarterly 19: 1–19.

32.

Jordan

Ashkanasy

Härtel

, et al. (2002) Workgroup emotional intelligence: Scale development and relationship to team process effectiveness and goal focus. Human Resource Management Review 12: 195–214.

33.

Kong

(2017) The validity of the Wong and Law Emotional Intelligence Scale in a Chinese sample: Tests of measurement invariance and latent mean differences across gender and age. Personality and Individual Differences 116: 29–31.

34.

Law

Wong

Song

(2004) The construct and criterion validity of emotional intelligence and its potential utility for management studies. Journal of Applied Psychology 89: 483–496.

35.

Lawrence

Troth

Jordan

, et al. (2011) A review of emotion regulation and development of a framework for emotion regulation in the workplace. In: Perrewe

Ganster

(eds) Research in Occupational Stress and Wellbeing. Bingley: Emerald Publishing, pp. 197–263.

36.

Mayer

Salovey

Caruso

(2002) MSCEIT: Mayer-Salovey-Caruso Emotional Intelligence Test. Toronto, ON, Canada: Multi-Health Systems.

37.

Mayer

Salovey

(1995) Emotional intelligence and the construction and regulation of feelings. Applied and Preventive Psychology 4: 197–208.

38.

Mayer

Salovey

(1997) What is emotional intelligence? In: Salovey

Sluyter

(eds) Emotional Development and Emotional Intelligence. New York: Basic Books, pp. 3–31.

39.

Mehl

Conner

(2012) Handbook of Research Methods for Studying Daily Life. New York: The Guilford Press.

40.

Muthén

(2012) Software Mplus Version 7. Los Angeles, CA: Muthén & Muthén.

41.

Nehles

van Riemsdijk

Kok

, et al. (2006) Implementing human resource management successfully: A first-line management challenge. Management Review 17: 256–273.

42.

Nezlek

Kuppens

(2008) Regulating positive and negative emotions in daily life. Journal of Personality 76: 561–580.

43.

Parkinson

Totterdell

(1999) Classifying affect-regulation strategies. Cognition and Emotion 13: 277–303.

44.

Peña-Sarrionandia

Mikolajczak

Gross

(2015) Integrating emotion regulation and emotional intelligence traditions: A meta-analysis. Frontiers in psychology 6: 1–27. https://doi.org/10.3389/fpsyg.2015.00160

45.

Purcell

Hutchinson

(2007) Frontline managers as agents in the HRM—Performance causal chain: Theory, analysis and evidence. Human Resource Management Journal 17: 3–20.

46.

Raudenbush

Bryk

(2002) Hierarchical Linear Models: Applications and Data Analysis Methods. Newbury Park, CA: SAGE.

47.

Ray

McRae

Ochsner

, et al. (2010) Cognitive reappraisal of negative affect: Converging evidence from EMG and self-report. Emotion 10: 587–592.

48.

Robinson

Moeller

Buchholz

, et al. (2012) The regulatory benefits of high levels of affect perception accuracy: A process analysis of reactions to stressors in daily life. Emotion 12: 785–795.

49.

Saavedra

Kwun

(2000) Affective states in job characteristics theory. Journal of Organizational Behavior 21: 131–146.

50.

Satorra

Bentler

(2010) Ensuring positiveness of the scaled difference Chi-Square Test statistic. Psychometrika 75: 243–248.

51.

Sheppes

Scheibe

Suri

, et al. (2011) Emotion-regulation choice. Psychological science 22: 1391–1396. https://doi.org/10.1177%2F0956797611418350

52.

Sheppes

Scheibe

Suri

, et al. (2014) Emotion regulation choice: A conceptual framework and supporting evidence. Journal of Experimental Psychology: General 143: 163.

53.

Shi

Wang

(2007) Validation of Emotional Intelligence Scale in Chinese university students. Personality and Individual Differences 43: 377–387.

54.

Tamir

Mitchell

Gross

(2008) Hedonic and instrumental motives in anger regulation. Psychological Science 19: 324–328.

55.

Totterdell

Holman

(2003) Emotion regulation in customer service roles: Testing a model of emotional labor. Journal of Occupational Health Psychology 8: 55–73.

56.

Townsend

(2013) To what extent do line managers play a role in modern industrial relations? Asia Pacific Journal of Human Resources 51: 421–436.

57.

Troth

Lawrence

Jordan

, et al. (2018) Interpersonal emotion regulation in the workplace: A conceptual and operational review and future research agenda. International Journal of Management Reviews 20: 523–543.

58.

Tse

HHM

Troth

Ashkanasy

, et al. (2018) Affect and leader-member exchange in the new millennium: A state-of-art review and guiding framework. The Leadership Quarterly 29: 135–149.

59.

Wadlinger

Isaacowitz

(2011) Fixing our focus: Training attention to regulate emotion. Personality and Social Psychology Review 15: 75–102.

60.

Watson

Clark

Tellegen

(1988) Development and validation of brief measures of positive and negative affect: The PANAS Scales. Journal of Personality and Social Psychology 54: 1063–1070.

61.

Webb

Miles

Sheeran

(2012) Dealing with feeling: A meta-analysis of the effectiveness of strategies derived from the process model of emotion regulation. Psychological Bulletin 138: 775–808.

62.

Weiss

Cropanzano

(1996) Affective events theory: A theoretical discussion of the structure, causes and consequences of affective experiences at work. In: Staw

Cummings

(eds) Research in Organizational Behavior. Greenwich, CT: JAI Press, pp. 1–74.

63.

Williams

(2007) Building genuine trust through interpersonal emotion management: A threat regulation model of trust and collaboration across boundaries. Academy of Management Review 32: 595–621.