How to Cope with an Abusive Leader? Examinations of Subordinates’ Affective Reactions,CWB-O and Turnover Intentions

Abstract

When subordinates experience abusive supervision, they often respond with “fight” (e.g., organizationally-directed counterproductive work behavior: CWB-O) or “flight” reactions (e.g., turnover intentions). Drawing on cognitive appraisal theory, we propose that negative (NA) and positive affect (PA) explain these distinct responses and that coworker emotional support differentially moderates the relationships of affective states with CWB-O and turnover intention. That is, we expect that emotional support exacerbates the mediating effect of NA, whereas it weakens the mediating effect of PA. We tested the hypothesized model in two time-lagged studies, one among subordinates (N = 162), and the other among supervisor-subordinate dyads (N = 255 dyads). Results confirmed that the relation between abusive supervision and CWB-O was mediated by NA, and that the relation between abusive supervision and turnover intentions was mediated by PA. Coworker emotional support strengthened the mediating effect of NA, but it did not attenuate the mediating effect of PA. These results only hold for CWB-O, not for supervisor-rated CWB-O. Explanations, implications as well as limitations and ideas for future research are discussed.

Keywords

abusive supervision negative affect positive affect turnover intentions organizationally-targeted counterproductive work behaviors coworker emotional support

Abusive supervision, which refers to “subordinates’ perception of the extent to which supervisors engage in the sustained display of hostile verbal and non-verbal behaviors, excluding physical contact” (Tepper, 2000, p. 178), has received ongoing attention and scrutiny from both scholars and practitioners over the last two decades (Fischer et al., 2021). Threatening or stressful situations, such as the experience of abusive supervision (Vogel & Bolino, 2020), trigger the so called “fight or flight” response (Lazarus & Folkman, 1984). In line with the “fight” response, previous studies have shown that employees respond to abusive supervision by engaging in counterproductive work behaviors that harm the organization (referred to as CWB-O). There is also evidence for the “flight” response, whereby employees consider leaving the organization when subjected to an abusive supervisor (i.e., higher turnover intentions). The current study focuses on CWB-O and turnover intentions as representatives of the “fight” and “flight” responses, and aims to disentangle the differential mechanisms underlying these responses.

Despite the accumulation of knowledge about the consequences of abusive supervision, there is a lack of empirical evidence explaining why employees respond differently to the abuse (Oh & Farh, 2017). Drawing on Cognitive Appraisal Theory (Lazarus, 1966) and its application to emotions (Ellsworth & Scherer, 2003; Frijda, 2009; Moors et al., 2013; Scherer, 1999), we propose that abusive supervision is appraised as demeaning and harmful, which may elicit a variety of negative emotions. However, specific, intense and short-lived emotions occurring shortly after the abusive episode may, over time, accumulate and evolve into more general negative affective states, including increased negative affect (NA) and decreased positive affect (PA). These differential affective responses to abusive supervision are hypothesized to subsequently result in “fight” or “flight” reactions as a way for subordinates to cope with the situation (Barsade & Gibson, 2007; Lazarus & Folkman, 1984). CWB-O is usually directly driven by negative emotions (Spector & Fox, 2005), and thus we posit that NA mediates the relation between abusive supervision and CWB-O (i.e., angry employees are likely to engage in deviant behavior). In contrast, PA is more strongly associated with subordinates’ overall evaluation of the working environment (Breckler & Wiggins, 1989), and therefore we propose that PA mediates the relation between abusive supervision and subordinates’ turnover intentions (i.e., unhappy employees tend to leave the organization).

Lastly, we posit that abused subordinates’ “fight” or “flight” responses may also depend on the secondary appraisal of their available resources to cope with the abuse (Lazarus & Folkman, 1984). Coworker emotional support is an important resource that abused subordinates can rely on to help them cope with the situation. Appraisal theory originally established coworker emotional support as a stress buffer (Lazarus & Folkman, 1984), but this proposition has yielded mixed results (e.g., Fischer, et al., 2021). The current study proposes that coworker emotional support plays a differential moderating role. Specifically, we propose that, on the one hand, high coworker emotional support validates and emphasizes the abused employees’ experiences and feelings (Boren, 2014), and thus strengthens the mediating effect of NA between abusive supervision and CWB-O. On the other hand, coworker emotional support is a key resource in subordinates’ work environment, and therefore weakens the mediating roles of PA between abusive supervision and turnover intentions.

Overall, our study contributes to the literature in the following ways. First, it goes beyond the initial short-lived negative emotions evoked by abusive supervision and provides a more comprehensive picture of affective reactions to abusive supervision. Second, by distinguishing the differential mediating roles of NA and PA, we help to explain how different affective reactions relate to the abused subordinates’ “fight” or “flight” responses. In addition, we address the conflicting moderating results of coworker emotional support in the secondary appraisal process and try to reveal the double-edged effects of coworker emotional support. Lastly, in a field dominated by cross-sectional studies (Tepper et al., 2017), the current research questions are examined in two time-lagged, studies, of which one relied on multi-source ratings. The conceptual moderated mediation model is represented in Figure 1.

Figure 1.

Theoretical model. Note. CWB-O = organization-targeted counterproductive work behavior.

Theoretical Framework and Hypothesis Development

Cognitive Appraisal Theory and its Applications to Emotions

The current study draws upon Cognitive Appraisal Theory (Lazarus, 1966; Lazarus & Folkman, 1984), and specifically on its application to emotions (Ellsworth & Scherer, 2003; Frijda, 2009; Moors et al., 2013) to develop the hypotheses. According to cognitive appraisal theory, emotional reactions to an event are determined by how we appraise the event. First, through primary appraisal, people evaluate the importance of the event to their personal well-being (e.g., Lazarus, 1991; Smith & Ellsworth, 1985). The more relevant an event is to an individual’s well-being, the more intense the emotional reaction will be. Next, by means of secondary appraisal, people evaluate their resources and options to cope with the event.

Despite the fact that prior studies have established the relations between abusive supervision and discrete emotions among subordinates, they largely focused on negative emotions. For example, Peng et al. (2019) proposed that anger as a reaction to abusive supervision is associated with deviant behaviors, whereas shame and fear are related to increased turnover intentions. However, seldom did literature pay attention to affective states, especially PA. Although it is interesting to examine employees’ immediate discrete emotions in response to abusive supervision, the sustained nature of abusive supervision implies that it may have a more lasting effect on employees’ affective states. As appraisals and emotions are processes that develop over time and across situations (Lazarus & Folkman, 1987), we argue that discrete emotions elicited by abusive supervision would dwell in subordinates’ memories, and over time, evolve into a more general negative affective state (e.g., Lerner & Keltner, 2001; Peng et al., 2019). Also, being continuously subjected to psychological abuse may not only lead to an accumulation of negative emotions, but it may also deprive employees to perceive and experience events that elicit positive emotions. In line with this reasoning, research has shown that employees are less satisfied with their job and with life in general when repeatedly exposed to an abusive supervisor (see e.g., Mackey et al., 2017; Martinko et al., 2013).

Following the two-process model of appraisal, we argue that abusive supervision is related to CWB-O and turnover intentions via increased NA and decreased PA, respectively. Additionally, we hypothesize that these relations are moderated by an important resource that affects how employees cope with abusive supervision (i.e., secondary appraisal), namely coworker emotional support.

The Mediating Effect of NA between Abusive Supervision and CWB-O

NA generally represents a state of distress and a variety of aversive feelings such as anger, guilt, and fear. Higher NA is associated with a higher intensity of the feelings mentioned above, while lower NA comprises a state of calmness and serenity (Watson et al., 1988). PA reflects an individual’s affective experience of feelings such as being enthusiastic, alert, and active. Higher PA indicates a state of high energy, while lower PA is related to lethargy and sluggishness. There has been an ongoing debate about whether NA and PA are opposite ends of the same continuum or two independent, unipolar dimensions (e.g., Russell & Carroll, 1999; Watson et al., 1999). The current study adopts the latter perspective (Barsade & Knight, 2015; Thoresen et al., 2003). That is, evidence indicates that individuals’ memory of PA in the last year is almost completely independent of their memory of NA (Diener & Emmons, 1984), and that PA and NA operate on different biological, including brain, mechanisms (Polk et al., 2005). Thus, high NA does not necessarily indicate low PA, nor vice versa.

Abusive supervision is related to a high intensity of aversive feelings (i.e., high NA), such as anger, fear, and shame (Oh & Farh, 2017). For example, it can interfere with personal goals and with subordinates’ needs for security, relatedness, and growth (Tepper, 2007), and can thus trigger anger and fear (Berkowitz & Harmon-Jones, 2004; Oh & Farh, 2017); it can also threaten subordinates’ self-worth and is thus related to feelings of shame (Hannah et al., 2016; Patrizia et al., 2017) and; it can thwart subordinates’ coping capacity and is related to sense of powerlessness, fear, and distress (Li et al., 2017). As abusive supervision is part of supervisors’ behavioral pattern, the elicited discrete negative emotions in subordinates may accumulate and evolve into a more permanent negative affective state. Indeed, victims of abusive supervision not only tend to re-experience the negative emotions elicited by abusive supervision, but also retain those aversive feelings in memory longer than most other work experiences Peng et al., 2019). According to Vogel and Bolino (2020), abusive supervision may even linger in the victim’s memory as a nightmare and trap the victim in a long-term depressive state (i.e., negative affect) even when the abusive supervisory relationship has ceded to exist.

Cognitive appraisal theorists have established coping styles to have an affective regulation effect (Lazarus & Folkman, 1984). As NA may nurture subordinates’ retaliation (Spector & Fox, 2002), we propose that NA is further positively related to subordinates’ “fight” response in the form of CWB-O. From the perspectives of subordinates, engaging in CWB-O can be seen as an adaptive way to cope with the experience of abusive supervision and to ameliorate the resulting influence on subordinates’ increase in NA (Zhang et al., 2019). In other words, subordinates engage in CWB-O in order to cope with the experienced abusive supervision and with the resulting increased NA. Indeed, various meta-analyses have found that NA is positively related to CWB-O (Dalal, 2005; Kaplan et al., 2009). The current study aims to replicate the positive relations between NA and CWB-O, and proposes that the relation between abusive supervision and CWB-O is mediated by NA.

Hypothesis 1: NA mediates the positive relation between abusive supervision and CWB-O.

The Mediating Effect of PA Between Abusive Supervision and Turnover Intentions

It is intuitive that abusive supervision may elicit negative emotions, but there are also indications that abusive supervision reduces the experience of positive emotions. First, exposure to supervisors’ mistreatment can discourage employees, and decrease their motivation to be highly dedicated to and enthusiastic about their work (Li et al., 2017). Second, because of the sustained nature of abusive supervision, it is likely that subordinates lose hope and optimism that their situation may improve (Bamberger & Bacharach, 2016). Finally, there may not be much for subordinates of abusive supervisors to be happy about as indicated by reduced satisfaction with their job and life in general (Tepper et al., 2017). As the abuse continues and subordinates are deprived from the experience of positive emotions on a daily basis, this likely results in lethargy and sluggishness (i.e., low PA). Following this logic, we propose that abusive supervision is associated with decreased PA.

Furthermore, we propose that decreased PA reminds individuals of their dissatisfaction and promotes “flight” reactions, such as turnover intentions (George & Jones, 1996; Judge et al., 2012). Considering that actual turnover is constrained by various intricate factors, such as salary, geographical locations of workplaces, and employment rates, and given that turnover intentions is one of the best predictors of turnover (e.g., Hom et al., 1992; Tett & Meyer, 2006), the current study focuses on turnover intentions as a proxy of actual turnover. Turnover intentions are rarely born out of a single disagreeable event. Instead, they are usually formed through a series of cognitive appraisals of emotions, and influence employees’ general attitudes towards the organizations (Breckler & Wiggins, 1989; Watson & Spence, 2007).

As described above, repetitive exposure to abusive supervision may decrease subordinates’ experience of PA over time. PA, associated with individual cognitive flexibility and open mindedness (Fredrickson & Losada, 2005), is formed through a series of emotional reactions and affect-based attitudinal evaluations, and thus reminds individuals of their dissatisfaction and therefore promotes turnover intentions (Judge et al., 2012). Further, experiencing PA is generally desirable, and individuals who experience low levels of PA at work might actively seek for opportunities to increase their own levels of experienced PA. In other words, experiencing low levels of PA might be associated with increased turnover intentions. Based on these arguments, we hypothesize:

Hypothesis 2: PA mediates the positive relation between abusive supervision and turnover intentions.

The Moderating Roles of Coworker Emotional Support

When abused by a supervisor, coworkers’ emotional support becomes an important resource that helps employees to cope with the abuse. Coworker emotional support refers to perceived emotional assistance, such as empathy, consolation and a sense of being valued (Cohen & Wills, 1985), and it is characterized by caring and listening sympathetically to those who may seek or need support.

Despite the fact that coworker emotional support is conventionally considered a buffering factor in the framework of the cognitive appraisal theory of emotions and coping (Cohen & Wills, 1985; Fenlason & Beehr, 1994), empirical studies have yielded mixed results (Fischer et al., 2021). Whereas some found that social support exacerbated the negative effects of abusive supervision (Wu & Hu, 2009), others found no effect (Poon, 2011) or even a reversed effect (Caesens et al., 2019). To integrate these seemingly conflicting results, we propose that coworker emotional support may be a double-edged sword. That is, coworker emotional support may influence subordinates’ secondary appraisal of the affective states resulting from abusive supervision differently. On the one hand, high coworker emotional support may help subordinates feel that they are understood or cared for (Ng & Sorensen, 2008), and thus buffer the negative indirect relation of abusive supervision with turnover intentions via PA (Van Emmerik et al., 2007). On the other hand, feeling supported seldom helps subordinates to forget or forgive the insults and belittling from their supervisors (Chiaburu & Harrison, 2008). Even worse, when coworkers stand by the subordinates’ side, it may reinforce employees’ experience of NA and thereby strengthen the positive indirect relation of abusive supervision with CWB-O via NA.

We propose that high coworker emotional support strengthens the relation between abused subordinates’ NA and CWB-O. When experiencing high emotional support, coworkers may explicitly acknowledge, elaborate, legitimize, and contextualize the distressed others’ feelings and perspective (Burleson, 2008; High & Dillard, 2012). They may ruminate together about the toxic behaviors of the supervisors and about the problems experienced by the subordinates. In the secondary appraisal stage, coworker emotional support may consolidate and sometimes exaggerate the abused subordinates’ feelings, and thus may strengthen subordinates’ belief to blame the supervisors. When subordinates feel validated and supported by their coworkers, we argue that they are more likely to transform their NA into CWB-O.

Hypothesis 3: Coworker emotional support moderates the positive relation between NA and CWB-O, in a way that the higher level of coworker emotional support, the stronger the positive relation between NA and CWB-O.

Hypothesis 4: Coworker emotional support moderates the mediating role of NA in the relation of abusive supervision with CWB-O, in a way that the higher level of coworker emotional support, the stronger the indirect relation of abusive supervision with CWB-O via NA.

Next, turnover intentions result from a cognitive appraisal process that includes the overall environment (Breckler & Wiggins, 1989). It is formed through a series of evaluations and comparisons of a variety of factors, which may not be limited to supervisors. In the situation of abusive supervision, the leader is antagonistic, however, high co-worker emotional support may be appraised as an important resource in the secondary appraisal stage that promotes retention. According to the cross-domain buffering hypothesis (Duffy et al., 2002), social support from one domain (e.g., coworker emotional support) may cushion the negative effect of social undermining from another domain (e.g., abusive supervision). High coworker emotional support means coworkers may listen to, comfort, and affirm the abused subordinates. It may not completely counteract the relation between abusive supervision and PA, as aggression has more potent influence than support on subordinates’ affect (Rozin & Royzman, 2001; Tews et al., 2019), but it can elicit the feeling in subordinates that they are not alone, and thus buffer the positive relation of PA with turnover intentions and the indirect relation of abusive supervision with turnover intentions via PA.

Hypothesis 5: Coworker emotional support moderates the negative relation between PA and turnover intentions, in a way that the higher level of coworker emotional support, the weaker the negative relation between PA and turnover intentions.

Hypothesis 6: Coworker emotional support moderates the mediating role of PA in the negative relation of abusive supervision with turnover intentions, in a way that the higher level of coworker emotional support, the weaker the indirect relation of abusive supervision with turnover intentions via PA.

Overview of Studies

We tested our hypotheses in two studies. In Study 1, we performed a two-wave survey study among employees. In Study 2, we attempted to replicate results from Study 1 with a more rigid three-wave, multi-source survey study involving ratings from both supervisors and subordinates.

Study 1: Method

Procedure and Participants

We recruited participants in China via the snowball sampling method with the help of classmates who were psychology majors and our laboratory members. They advertised the study among their acquaintances. All participants took part in this study voluntarily and were assured full confidentiality. A research assistant sent the survey link to each participant through Wechat, the most popular social media platform in China. Data were collected at two time points. At Time 1 (T1), participants were asked to report their experiences of abusive supervision and their demographic information. Approximately one month later, at Time 2 (T2), participants who participated at T1 were invited to complete a follow-up survey, which included measures of NA, PA, turnover intentions, CWB-O, and perceived coworkers’ emotional support. All questionnaires not available in Chinese were translated and back-translated independently by three English majors with a psychology background. Differences in translations were reconciled by comparing and discussing their respective translations to ensure that all the survey items were accurately translated from English to Chinese (Brislin, 1970). To encourage participation, participants were compensated with a small monetary incentive for each questionnaire (3 yuan for T1 survey and 5 yuan for the T2 survey, which corresponded to roughly to $1.17 in total).

We received completed surveys from 290 participants at T1 and from 231 participants at T2. We included attention check items within each of the surveys (e.g., “When you get to this item, please answer ‘Strongly agree’”). Participants who failed the attention check items (53 participants at T1 and 16 participants at T2) were not included in the final sample. Complete data for all variables in our theoretical model were available for 162 participants. Analysis of variances (ANOVA) tests demonstrated that no significant differences emerged for any of the demographic and study variables between participants who completed both the T1 and T2 surveys and those participants who only completed the T1 survey.

Among the final sample, 62.11% were female. Most participants were in the age group of 20–25 years old (59.88%). The majority of the participants (59.01%) had up to two years of work experience, and more than three quarters of the participants (77.64%) held at least a university degree. Participants worked in a variety of organizations including private sectors (47.83%), government or government-owned units (36.64%), and self-owned businesses (3.73%).

Measures

Abusive Supervision (T1)

Abusive supervision was measured using the 15-item abusive supervision scale developed by Tepper (2000). This measure assesses the frequency with which subordinates experience nonphysical, abusive acts by their supervisor. Participants responded using a 5-point scale ranging from 1 (never) to 5 (very often). A sample item includes “My leader puts me down in front of others”. In the current study, Cronbach’s alpha was .92.

Affective States (T2)

We used the 18-item measure adapted by Qiu et al. (2008) from the Positive and Negative Affect Scales (PANAS) originally constructed by Watson et al. (1988) to assess positive and negative affect. NA and PA were measured with nine items each. Participants reported how they had felt during the last month on a 7-point scale ranging from 1 (very slightly or not at all) to 7 (very much). Sample items for NA include “nervous” and “upset”. Sample items for PA include “happy” and “inspired”. In the current study, Cronbach’s alpha was .90 for NA and .96 for PA.

Coworker Emotional Support (T2)

To measure coworker emotional support, we used the four-item measure that Qian (2001) adapted from Zimet et al.’s (1988) Perceived Social Support Scale (PSSS). All items are rated on a 7-point scale ranging from 1 (very strongly disagree) to 7 (very strongly agree). A sample item is “I can share my joys and sorrows with my coworkers”. Cronbach’s alpha was .91.

Turnover Intentions (T2)

We measured turnover intentions with the four-item Turnover Intention Scale developed by Mobley and colleagues (1978). Participants responded using a 5-point scale ranging from 1 (strongly disagreed) to 5 (strongly agree). A sample item includes “It is very possible that I will leave the organization in half a year”. In the current study, Cronbach’s alpha for this scale was .87.

Organizationally-Directed Counterproductive Work Behavior (CWB-O) (T2)

We measured CWB-O with an adapted version of the scale developed by Bennett and Robinson (2000). Employees also engage in higher levels of counterproductive work behavior directed at other individuals in the organization (CWB-I) when experiencing abusive supervision, but we intentionally do not examine CWB-I in the current study because it does not distinguish behaviors directed at the supervisor or at coworkers, which may cause mixed results. Among the twelve original items measuring CWB-O, one item related to drug abuse was removed because it was not applicable to the Chinese context. Participants were asked to indicate the extent to which they had engaged in each of the behaviors targeted at their organization. The resulting measure consisted of 11 items and participants rated each item on a 7-point scale from 1 (never) to 7 (very often). A sample item was “Spent too much time fantasizing or daydreaming instead of working”. In the current study, Cronbach’s alpha for this scale was .86.

Analytical Strategy

First, before the hypotheses testing, we conducted confirmatory factor analyses (CFA) with all items in Mplus 7.4 to examine whether the scales used in the current study represent distinct constructs. Then, to test a parsimonious model with latent variables, we created three parcels for each latent construct using a factorial algorithm (Rogers & Schmitt, 2004), except for the four-item coworker support scale and the four-item turnover intention scale. This strategy has the advantage of preserving common latent variable variance while minimizing unrelated specific variance (Little et al., 2013). We also conducted CFA with the item parceling. Next, we employed structural equation modeling (SEM) using maximum likelihood estimation to test the hypotheses proposed in this study. The mediation effects of NA and PA were tested in Model 1, then in Model 2 we entered the interaction terms using the latent moderated structural equations (LMS; Klein & Moosbrugger, 2000) using the XWITH command in Mplus to test the moderation and moderated mediation effects (Stine, 1989). We bootstrapped 10000 samples to obtain bias-corrected confidence intervals (CIs) of the indirect effects and of the moderated mediation effects (Stine, 1989). When plotting simple slopes, we first computed the mean, variance and standard deviation of the latent variable of coworker emotional support, and then used one standard deviation below and above the mean of coworker emotional support to compute simple slopes in the “model constraint” instruction. Furthermore, we also used the Johnson-Neyman technique to create confidence bands for the conditional indirect effects (Hayes, 2018; Preacher et al., 2007).

Study 1: Results

When conducting a CFA in Mplus, we used modification indices because it can improve model fit while providing remedies for discrepancies between proposed and estimated models (Bollen, 1989). Modifications can be made to items if they are theoretically justifiable, few in number, and they do not have a major impact on estimates of other parameters in the model (Byrne, 2013). CFA result showed that the measurement model, which consisted of six factors including abusive supervision, NA, PA, coworker emotional support, turnover intentions, and CWB-O, exhibited an acceptable fit to the data (χ2 (1181) = 1843.57, SRMR = .08, CFI = .90, TLI = .89, and RMSEA = .06). We further compared the hypothesized model with several alternative measurement models. The six-factor model fitted better than any other models. Therefore, measures of the studied variables had good discriminant validity. Table 1 shows the means, standard deviations, and correlations among the studied variables.

Table 1.

Descriptive Statistics and Correlations Among Variables of Study 1.

	M	SD	1	2	3	4	5	6	7	8	9	10
1. Gender ^a	1.62	.49	-
2. Age	3.14	1.39	.09	-
3. Dyadic Tenure	2.47	1.28	−.01	.71**	-
4. Education	1.99	.71	−.09	−.17*	−.10	-
5. Abusive supervision	1.39	.49	−.03	.04	.14†	.03	(.92)
6. NA	2.43	1.10	−.03	−.26*	−.21*	.01	.35**	(.90)
7. PA	4.36	1.53	.14†	.07	.02	−.06	−.23**	−.36**	(.96)
8. Coworker support	4.68	1.40	.20**	−.07	.06	.12	−.19*	−.15†	.30**	(.91)
9. Turnover intentions	2.88	1.03	−.16*	−.19*	−.12	.10	.23**	.37**	−.51**	−.09	(.87)
10. CWB-O	1.40	.56	−.25**	−.14	−.09	.14†	.34**	.41**	−.16*	−.03	.22**	(.86)

Note: N = 162. ^a 1 = male, 2 = female. NA = negative affect. PA = positive affect. Coworker support = Coworker emotional support. CWB-O = organization- targeted counterproductive work behavior. **p < .01, *p < .05.

Test of Mediation Effects

Hypothesis 1 proposed that NA mediates the positive relation of abusive supervision with CWB-O, and Hypothesis 2 proposed that PA mediates the positive relation of abusive supervision with turnover intentions. As shown in Table 2, bootstrapping results showed that the indirect relation of abusive supervision with CWB-O via NA was positive and significant (indirect effect = .13, 95% CI = [.06, .22]), supporting Hypothesis 1. Regarding the mediating role of PA, the indirect relation of abusive supervision with turnover intentions via PA was positive and significant (indirect effect = .12, 95% CI = [.02, .22]). Therefore, Hypothesis 2 was supported. In line with our expectations, in the combined mediation model, NA mediated the relation of abusive supervision with CWB-O but not with turnover intentions, whereas PA mediated the relation of abusive supervision with turnover intentions but not with CWB-O.

Table 2.

Test of Mediation and Moderated Mediation Effects for Study 1.

Model 1: Mediating effects	NA		PA		CWB-O		Turnover intentions
Model 1: Mediating effects	β	SE	β	SE	β	SE	β	SE
Abusive supervision	.36**	.09	−.24**	.10	.23	.16	.06	.07
NA					.36**	.09	.18^†	.10
PA					.03	.07	−.50**	.07
R²	.13*		.06		.24*		.33**
Model 2: Moderated mediating effects	NA		PA		CWB-O		Turnover intentions
Model 2: Moderated mediating effects	β	SE	β	SE	β	SE	β	SE
Abusive supervision	.37**	.07	−.24**	.08	.25**	.08	.07	.08
NA					.38**	.08	.17*	.09
PA					−.01	.08	−.52**	.07
CES					.11	.08	.04	.08
NA × CES					.17*	.09	−.03	.08
PA × CES					−.13	.09	−.10	.08
R²	.14**		.06		.33**		.35**

Note: N = 162. PA = positive affect, NA = negative affect, CWB-O = organization- targeted counterproductive work behavior. Coefficients are standardized path coefficients between the latent variables. The mediating effect was tested with Bootstrap = 10000. **p < .01, *p < .05.

Test of Moderated Mediation Hypotheses

Further, as shown in Table 2, we found that coworker emotional support significantly interacted with NA to predict CWB-O (β = .17, 95% CI = [.00, .33]). In particular, we operationalized high and low levels of coworker emotional support as one standard deviation above and below the mean. In line with our expectations, the relationship between NA and CWB-O was stronger for high coworker emotional support (β = .30, 95% CI = [.15, .44]) than for low coworker emotional support (β = .12, 95% CI = [−.00, .23]), supporting Hypothesis 3. Figure 2 illustrated the simple slopes for the relation of NA with CWB-O under different levels of coworker emotional support.

Figure 2.

Moderating effect of coworker emotional support in the relation of NA and CWB-O (study 1). Note. CWB-O = predicted latent variable mean of organization-targeted counterproductive work behavior. NA = negative affect.

We then tested the conditional indirect effect of abusive supervision on CWB-O and on turnover intentions. Bootstrap analysis revealed that, compared with low coworker emotional support (indirect effect = .10, 95% CI = [−.01, .20]), the conditional indirect effect of abusive supervision on CWB-O via NA was stronger for high coworker emotional support (indirect effect = .24, 95% CI = [.08, .40]). Therefore, Hypothesis 4 was supported. Coworker emotional support and the interaction term accounted for 9% of additional variance over and above the mediation model in predicting CWB-O. As shown in Figure 3, the Johnson- Neyman technique indicates the conditional indirect effect of abusive supervision on CWB-O via NA (Hayes, 2018; Preacher et al., 2007).

Figure 3.

Indirect effect of abusive supervision on CWB-O through negative affect using the johnson-neyman technique (study 1). Note. CWB-O = organization-targeted counterproductive work behavior.

However, the interaction between PA and coworker emotional support was not significant in predicting turnover intentions (β = −.10, 95% CI = [−.26, .06]). Further, there was no significant difference in the conditional indirect effect of abusive supervision on turnover intention via PA across different levels of coworker emotional support (high coworker emotional support: indirect effect = .32, 95% CI = [.08, .57]; low coworker emotional support: indirect effect = .22, 95% CI = [.04, .40]). Thus, Hypotheses 5 and 6 were not supported.

Supplemental Analyses

To further examine the robustness of our model, we conducted several supplemental analyses. First, we addressed the common method issue in the study. As proposed by Podsakoff et al. (2003), procedural remedies such as a time-lag between different measures and different rating sources would help reduce common method bias. The time-lagged design in our study therefore reduces common method bias concerns. Also, we used latent variables in the analyses, which controls for measurement error. Further, we employed the covariance method by creating a new construct, named common latent factor (e.g., Eichhorn, 2014). All items load on a common latent factor, and the variance of the common latent factor was constraint to “1” to indicate that the total variance can share its variance with all observations involved in the measurement model (e.g., Afthanorhan et al., 2021). The result showed that the differences of standardized regression weights of constraint and unconstrained models were smaller than .20 for each latent variable (e.g., Afthanorhan et al., 2021), and the hypothesized paths remained significant when including the common latent factor in the measurement model. Thus, we assume that the current results were not substantially contaminated by common method bias.

Second, we ran an alternative model to test whether coworker emotional support moderates the relations between abusive supervision and affective states (i.e., NA and PA). The result showed that the interaction between abusive supervision and coworker emotional support did not significantly predict NA (β = −.03, p = .666) nor PA (β = .01, p = .889).

Third, we also tested our hypotheses when including subordinates’ age, gender, education, and tenure as control variables. We did not report results with control variables in the main analyses as including them may decrease the statistical power of our tests (Britt & Weisburd, 2010). However, in the supplementary materials, we report results including control variables, and the results showed that the pattern of our results remains essentially unchanged.

Study 2: Method

Procedure and Participants

In study 2, we collected multi-source data at three time points through Credamo, an online data collection platform whose participant pool contains supervisor-subordinate dyads. At T1, we sent survey links to the subordinates in the participant pool to ask them to rate their direct supervisors’ abusive supervision and to provide demographic information. A total of 352 subordinates completed the T1 survey. Approximately one month later, we invited those participants who completed the T1 survey to participate in the T2 survey. At T2, subordinates were asked to rate their NA, PA, and coworker emotional support. A total of 326 subordinates (92.6% response rate) returned the completed survey. Another month after T2, the Time 3 (T3) survey was sent to the same subordinates, and they were asked to rate their turnover intentions and CWB-O. Additionally, supervisors rated subordinates’ CWB-O at T3, and provided information on their own gender. Totally, 267 subordinates (81.9% response rate) completed the third survey, and 232 supervisors (71.2% response rate) completed the survey.

After matching the multi-wave and multi-source data via an identification code, and after removing thirteen participants who failed the attention checks (i.e., “please choose 4 for this item”), our final sample comprised a total of 225 valid dyadic records. Participants were ensured confidentiality throughout the entire data-collection process. Each of the subordinate and supervisor was compensated with $0.78 per survey, and subordinates who completed three waves of surveys were rewarded an extra $0.78 (for a total maximum of $3.12).

Among the final subordinate sample, 50.2% were male. Our sample was relatively young, with 52.9% of all subordinates being between 20–30 years old, 39.1% between 31–40 years old, and 4.4% between 41–50 years old. The majority of participants (78.2%) had obtained a bachelor’s degree, about 14.3% of them had graduated from technical school, and 5.3% had obtained a master’s degree or an even higher degree. They were employed in a variety of industries, including service (32.0%), health and technology (23.6%), education (23.1%), and manufacturing (21.3%). Subordinates’ average organizational tenure was 7.36 years (SD = 7.38), and average dyadic tenure with their supervisor was 4.38 years (SD = 2.98). Among the supervisor sample, 66.7% were male.

Measures

All study variables (i.e., abusive supervision, NA, PA, coworker emotional support, turnover intentions, and CWB-O) were assessed with the same back-translated measures as in Study 1. The only difference, next to the fact that we used three measurement points in Study 2, was that subordinates’ CWB-O was rated by both subordinates and supervisors. Cronbach’s alpha values ranged between .84 and .94 (see Table 3).

Table 3.

Descriptive Statistics and Correlations Among Variables of Study 2.

	M	SD	1	2	3	4	5	6	7	8	9	10	11	12	13
1. Gender ^a	1.50	.50	-
2. Age	2.50	.70	.07	-
3. Education	2.87	.72	−.09	−.26**	-
4. Dyadic Tenure	4.38	2.98	.09	.68**	−.17*	-
5. Negative affectivity	2.41	.90	.20	−.03	−.06	−.15*	(.86)
6. Positive affectivity	5.37	.86	−.05	−.06	−.01	.02	−.44**	(.91)
7. Abusive supervision	1.62	.54	−.14*	−.16*	.12	−.13	.54**	−.41**	(.91)
8. NA	2.13	.99	−.03	−.14*	.03	−.14*	.56**	−.43**	.32**	(.93)
9. PA	5.36	1.05	−.02	.03	−.05	.08	−.39**	.60**	−.25**	−.53**	(.94)
10. Coworker support	5.20	1.19	.08	.01	.03	.10	−.42*	.55**	−.27**	−.24*	.31**	(.87)
11. Turnover intentions	1.96	.84	.01	−.01	.10	−.06	.42**	−.40**	.34**	.32**	−.36**	−.38**	(.84)
12. Supervisor-rated CWB-O	1.37	.45	−.01	−.06	.05	−.05	.39**	−.27**	.21**	.36**	−.17**	−.22**	.32**	(.89)
13. CWB-O	1.42	.48	−.10	−.19**	.17**	−.16**	.44**	−.30**	.30**	.39**	−.29**	−.31**	.49**	.66**	(.87)

Note: N = 225. ^a 1 = male, 2 = female. Coworker support = coworker emotional support, NA = negative affect, PA = positive affect, CWB-O = organization-targeted counterproductive work behavior. **p < .01, *p < .

Study 2: Results

To test the factor structure and discriminant validity of our measures, we conducted CFAs involving seven key variables (i.e., abusive supervision at T1; NA, PA, and coworker emotional support at T2; turnover intentions, and both supervisor- rated CWB-O and CWB-O at T3). Results demonstrated that a seven-factor structure had a good fit to the data (χ² (1415) = 2275.11, SRMR = .06, RMSEA = .05, CFI = .90, TLI = .90) and fitted better than any of the other models that combined similar factors or factors measured at the same time points. Table 3 shows the means, standard deviations, and correlations among the studied variables.

Test of Mediation Effects

In line with Study 1, we ran a saturated model which tested the mediating effects of NA and PA for the relations of abusive supervision with both CWB-O and turnover intentions. As shown in Table 4, we found that NA did not mediate the relation between abusive supervision and supervisor-rated CWB-O (indirect effect = .04, 95% CI = [−.03, .10]). However, NA mediated the relation between abusive supervision and CWB-O (indirect effect = .09, 95% CI = [.01, .17]). Thus, the results indicate that Hypothesis 1 was supported when CWB-O was rated by subordinates themselves, but not when rated by their supervisor. Hypothesis 2 posited that PA mediates the relation between abusive supervision and turnover intentions. Results revealed that PA did mediate this relation (indirect effect = .07, 95% CI = [.02, .12]). Therefore, Hypothesis 2 was supported.

Table 4.

Test of Mediation and Moderated Mediation Effects for Study 2.

Model 1: Mediating effects	NA		PA		Supervisor-rated CWB-O		CWB-O		Turnover intentions
Model 1: Mediating effects	β	SE	β	SE	β	SE	β	SE	β	SE
Abusive supervision	.37**	.08	−.20**	.07	.52**	.09	.54**	.09	.12	.11
NA					.10	.08	.24*	.08	.24*	.10
PA					−.09	.06	−.01	.09	−.36**	.09
R²	.33**		.32**		.19**		.33**		.22**
Model 2: Moderated mediating effects	NA		PA		Supervisor-rated CWB-O		CWB-O		Turnover intentions
Model 2: Moderated mediating effects	β	SE	β	SE	β	SE	β	SE	β	SE
Abusive supervision	.37**	.08	−.20**	.07	.51**	.10	.31**	.07	.11	.10
NA					.13	.07	.25**	.07	.20	.13
PA					−.11	.09	−.04	.08	−.20*	.07
CES					−.08	.13	.04	.12	−.37	.24
NA × CES					.02	.02	.12**	.05	−.05	.05
PA × CES					−.01	.02	.06	.05	.07	.06
R²	.33**		.33**		.21**		.39**		.29**

Note: N = 225. NA = negative affect. PA = positive affect. CES = coworker emotional support. CWB-O = organization-targeted counterproductive work behavior. Coefficients are standardized path coefficients between the latent variables. **p < .01, *p < .05.

Test of Moderated Mediation Hypotheses

Hypothesis 3 posited that coworker emotional support positively moderates the relation between NA and CWB-O. As shown in Table 4, the interactive effect of NA and coworker emotional support on supervisor-rated CWB-O was not significant (β = .02, 95% CI = [−.03, .07]), but was significant on CWB-O (β = .12, 95% CI = [.06, .18). The relation between NA and CWB-O was stronger for high coworker emotional support (β = .28, 95% CI = [.20, .36]) than for low coworker emotional support (β = .07, 95% CI = [.02, .12]). Figure 4 depicts the simple slope for the interactive effect of NA and coworker emotional support on CWB-O.

Figure 4.

Moderating effect of coworker emotional support in the relation of NA and CWB-O (study 2). Note. CWB-O = organization-targeted counterproductive work behavior. NA = negative affect.

Furthermore, the moderated mediating effect of NA between abusive supervision and supervisor-rated CWB-O was not significant (β = .01, 95% CI = [−.03, .05]). Coworker emotional support moderated the mediating effect of NA between abusive supervision and CWB-O (β = .04, 95% CI = [.02, .06]), in a way that compared with low coworker emotional support (indirect effect = .05, 95% CI = [−.02, .13]), the conditional indirect effect of abusive supervision on CWB-O via NA was stronger for high coworker emotional support (indirect effect = .14, 95% CI = [.02, .26]). The moderated mediation explained 6% additional variance over and above the mediating model in predicting CWB-O. As shown in Figure 5, we adopted the Johnson- Neyman technique to indicate the conditional indirect effect of abusive supervision on CWB-O via NA (e.g., Preacher et al., 2007). Therefore, Hypotheses 3 and 4 were supported only when CWB-O was rated by subordinates themselves, instead of by their supervisors.

Figure 5.

Indirect effect of abusive supervision on CWB-O through negative affect using the johnson-neyman technique (study 2). Note. CWB-O = organization-targeted counterproductive work behavior.

Hypothesis 5 posited that coworker emotional support negatively moderates the relation between PA and turnover intentions, and Hypothesis 6 proposed that coworker emotional support moderates the mediating effect of PA for the relation between abusive supervision and turnover intentions. However, we did not find evidence for the moderating effect of coworker emotion support (β = .07, 95% CI = [−.02, .17]) between PA and turnover intentions, nor for the moderated mediating effect of PA (β = −.02, 95% CI = [−.09, .05]) in the relation of abusive supervision and turnover intentions. Therefore, consistent with Study 1, Hypotheses 5 and 6 were not supported.

Supplemental Analyses

In accordance with Study 1, we conducted supplemental analyses to test the robustness of our model. We added the common latent construct and compared the model including the common latent construct with the model excluding this factor. The result showed that the differences of standardized regression weights in the two models were less than .20 for each latent variable, and the hypothesized paths remained significant after including the common latent construct in the measurement model. Therefore, we speculate that common method bias did not severely affect the current results.

Second, we discovered that coworker emotional support did not moderate the relation of abusive supervision with NA (β = −.01, p = .863), neither did it moderate the relation of abusive supervision with PA (β = −.10, p = .304). Therefore, coworker emotional support did not moderate the relations of abusive supervision with subordinates’ affective responses.

Third, we tested a model including the control variables of subordinates’ age, gender, education, and dyadic tenure. Results were generally comparable with regard to the magnitude of regression coefficients and significance levels reported in the main analyses above.

Discussion

Subordinates tend to respond to abusive supervision with “fight” or “flight” strategies, which were operationalized as CWB-O and turnover intentions, respectively, in the current study. Through the lens of cognitive appraisal theory (Lazarus, 1966) and its applications to affect (e.g., Ellsworth & Scherer, 2003 Frijda, 2009), we examined NA and PA as two independent mediators for the relations of abusive supervision with CWB-O and turnover intentions. The current study demonstrated that abusive supervision indirectly relates to subordinates’ CWB-O through NA, whereas abusive supervision indirectly predicted turnover intentions via PA. Further, coworker emotional support strengthens the indirect positive relation of abusive supervision with CWB-O via NA, showing that the mediating effect is stronger when coworker emotional support is high (vs. low). These results only hold when CWB-O was assessed with subordinate-reports, but not when it was rated by the supervisor. Contrary to expectations, coworker emotional support did not moderate the relation of PA with turnover intentions nor the indirect, negative relation of abusive supervision with turnover intentions via PA.

Theoretical Implications

Going beyond resource and justice theories, the current study examined subordinates’ “fight” (i.e., CWB-O) or “flight” (i.e., turnover intentions) responses to abusive supervision from the perspective of affective processes under the framework of cognitive appraisal theory. Despite the fact that some studies have tested how discrete negative emotions mediate the relations of abusive supervision with CWB-O and turnover intentions (e.g., Peng et al., 2019), our study contributes insights about subordinates’ more enduring affective responses after experiencing abusive supervision. Specifically, our study provided the empirical evidence for the distinct mechanisms of NA and PA for the relations of abusive supervision with subordinates’ “fight” or “flight” responses. Previous literature has shown that NA and PA are evolutionarily adaptive responses (Fredrickson & Losada, 2005). Exposure to a threatening situation such as abusive supervision increases subordinates’ NA, which narrows their attention to such unpleasant events and aversive feelings, and subsequently motivates them to engage in deviant behavior. At the same time, the experience of abusive supervision is associated with decreased PA among subordinates. When they lose hope and happiness in this situation, our findings indicate that they may see no other option than to look for alternative job opportunities.

In addition, based on the idea of secondary appraisal, we examined if coworker emotional support functions as a coping resource for subordinates in a way that it weakens or strengthens the relation between subordinates’ affective states and their fight or flight responses. By doing so, we also addressed conflicting results for the moderating effect of coworker support (Fischer et al., 2021) and highlight that coworker emotional support is not always a beneficial resource, but that it also has the potential to deteriorate a negative situation. In line with the latter, our results showed that high coworker emotional support strengthens the positive relation between abusive supervision and CWB-O via NA. This finding aligns with prior studies which have found a similar moderating effect of coworker support for the relation of abusive supervision with subordinates’ emotional exhaustion and job satisfaction (Caesens et al., 2019; Wu & Hu, 2009), showing that subordinates felt more emotionally exhausted and less satisfied when they received more support from their colleagues after experiencing abusive supervision.

As for the reason why coworker emotional support did not alleviate the positive mediating effect of PA for the relation between abusive supervision and turnover intentions, we draw upon previous literature (e.g., Chiaburu & Harrison, 2008; Duffy, 2009; Tews et al., 2019) and assume that interpersonal relationships with the supervisor may be more impactful compared to those with coworkers in determining whether employees consider to leave the organization. In other words, abusive supervision has a more potent influence on subordinates than positive interactions with coworkers. Even experiencing coworker emotional support cannot help employees to reappraise the abusive behavior from supervisors, which ultimately makes them contemplate leaving the organization as the last remedy. All of this can explain the non-significant moderation of coworker emotional support for the relation of PA with turnover intentions and for the indirect relation of abusive supervision with turnover intentions via PA.

Lastly, the current study deepens our understanding of CWB-O. First, abused subordinates engage in behavioral responses that are not directly targeted at the supervisor, but transfer their antagonism tendency to the organizations (Peng et al., 2019). It is possible that subordinates consider their supervisors as the embodiment of the organization and thus take revenge by taking actions that are harmful to the organization (Shoss et al., 2013). Further, in the second study we tested the model by measuring CWB-O from multiple sources (i.e., subordinates and their supervisors) and found that the results were no longer significant when using supervisor-rated CWB-O. One explanation could be that supervisors may not always be aware of their subordinates’ deviant behaviors directed toward the organization, such as whether they are daydreaming or stealing from the organization, especially because subordinates usually try to hide their own counterproductive behavior from their supervisors (Fox et al., 2007). In addition, previous literature has shown that abused subordinates may react with surface acting by hiding how they truly feel Wu & Hu, 2013. It is therefore possible that subordinates also engage in more covert forms of CWB-O when experiencing abusive supervision, making it more difficult for supervisors to detect, and subsequently rate, subordinates’ CWB-O.

Practical Implications

Our study has several practical implications for organizations, managers, and employees. Results indicate that abusive supervision has a lot of detrimental consequences for organizations and employees, and that organizations should therefore do everything within their power to prevent the occurrence of abusive supervision. As suggested by Duffy (2009), organizations should define acceptable and unacceptable behaviors for employees, including supervisors, and create strategies to enforce those policies and standards. Abusive supervision occurs more easily in an unhealthy work environment surrounded by fear, mutual mistrust, and lack of transparency and morale (Tepper, 2007). Therefore, organizations should encourage teams to create a positive culture where subordinates and supervisors can openly discuss problems in a supportive and task-focused manner (Mitchell & Ambrose, 2007). Also, in order to reduce supervisors’ abusive behaviors, organizations can offer trainings to supervisors that aim to enhance their sensitivity to and awareness of the negative effects of abusive supervision on subordinates (Peng et al., 2014).

Second, our findings highlight the importance of subordinates’ NA and PA, which implies that organizations attempting to reduce talent loss and deviant behaviors should pay attention to employees’ psychological affective states and wellbeing. To smooth subordinates’ NA, HR managers can proactively initiate conversations and inquire for subordinates’ difficulties or worries, and further provide or take possible solutions without hurting subordinates’ benefits (Harvey et al., 2007). If possible, organizations can also introduce a EAP (i.e., Employee Assistance Program) from a professional third party to help subordinates and simultaneously guarantee subordinates’ anonymity. Also, HR managers or organizations should be aware of the positive power of subordinates’ PA, and thus create more opportunities for positive experiences (Neves & Cunha, 2018). For examples, from the top leader, the organizations should encourage and try to set up a more harmonious, humorous, and trusting atmosphere in the interpersonal relationship. Taking subordinates’ affective states and wellbeing into consideration can help to prevent abused employees from engaging in counterproductive behavior at work or from contemplating to leave the organization.

Further, based on the moderation results, we suggest that when subordinates experience abusive supervision, HR managers should train coworkers to provide appropriate and constructive support for the abused employees. We found that emotional support from coworkers is not always a promising remedy to prevent detrimental consequences of subordinates’ affective states or of abusive supervision. Instead of simply offering emotional support or ruminating together with the abused subordinates, coworkers can offer more constructively supportive help. For example, based on the premise of the bystander intervention framework, Arman (2020) stated that coworkers could inform someone from management about the behaviors of the supervisor towards the abused subordinates, and encourage the abused subordinates to talk to someone from management. Further, organizations can also set up practical access (i.e., anonymous open mailbox) for subordinates to reflect anonymously on their situation of experiencing abusive supervision, train the abused subordinates to sometimes neglect supervisors’ aggression to protect their own mental health, and provide subordinates with effective up-management skills. In any case, our findings suggest that focusing on supervisors to reduce and further eliminate abusive supervision is the fundamental and key solution.

Limitations and Future Research

This study has some limitations that could be overcome in future studies. First, the current study operationalized turnover intentions as a “flight” response, however, as Oh and Farh (2017) mentioned, “flight” responses include a variety of forms, including actual turnover, turnover intentions, and withdrawal. Despite the fact that turnover intentions and actual turnover are positively correlated (Podsakoff et al., 2007), they are not exactly the same. Although one can have the intention to leave an abusive supervisor, subordinates may not actually quit their jobs because other factors prevent them from doing so (e.g., dependence on their jobs, status of their employability; Breevaart et al., 2021). Related to this, our study found that PA mediated the relations between abusive supervision and turnover intentions, but future research is needed to see whether followers actually turn these intentions into actions and if not, what prevents them from doing so.

Second, the current study only measured subordinates’ affective states without including discrete and episodic emotions. We posited that subordinates’ emotions and state affect are both relevant, but that they emerge at different stages after experiencing abusive supervision. As pointed out by Quigley et al. (2014), emotions comprise a short-term mental state when individuals have invoked conceptual knowledge to make sense of their current internal state, whereas state affect is more of a long-term state that can be described as pleasant or unpleasant with some degree of arousal. However, we did not distinguish specific stages and explored a more comprehensive affective development of abusive supervision. Future studies may design a combination of diary and longitudinal studies to track both subordinates’ emotional and affective responses to abusive supervision, and reveal a more complex affective pattern.

Another limitation of the current study is the use of subordinate-ratings of abusive supervision, which are known to be affected by individual characteristics such as subordinates’ affectivity, attribution styles, and attitudes towards their supervisor (e.g., Fischer et al., 2021). For example, Martinko et al. (2018) found that subordinates’ affect influences their evaluations of leadership. Future studies may conduct experiments or use multi-source ratings of abusive supervision to ensure the objective measurement of abusive supervision. Although we used a time-lagged design in both studies in which the measurement of abusive supervision preceded the measurement of the other variables, This, however, does not entirely exclude the possibility that reverse causal relations exist. For example, it is possible that high levels of negative affect or CWB-O among subordinates cause higher levels of abusive supervision. To exclude these possibilities, future longitudinal studies may use random intercepts cross-lagged panel models (Hamaker et al., 2015) to study the bidirectionality of abusive leadership perceptions and subordinates’ affective states (see e.g., Rudolph et al., in press).

Lastly, the mixed moderating effect of coworker emotional support deserves further investigation. Although most of the time coworker emotional support is conceptualized as a buffering factor of employees’ strain and work attitudes (e.g., Chiaburu & Harrison, 2008), we did not find such a buffering effect in the current study. In fact, we only found evidence for the opposite effect: Coworker emotional support exacerbates the negative relation of abusive supervision with CWB-O via NA. However, we still lack a more comprehensive and nuanced theory to make sense of when coworker emotional support would strengthen, weaken, or exert no effect on the relations of abusive supervision with subordinates’ “fight” or “flight” responses. one possibility is that different types of coworker emotional support, such as ruminating about the negative emotions together, criticizing and diminishing the supervisors, or helping the abused subordinates to build confidence, may exert a different influence on the relations. For example, Neves and Cunha (2018) found that coworkers’ humor alleviates the negative effect of abusive supervision on work tension. Future studies should distinguish different kinds of coworker (emotional) support and develop a comprehensive theoretical framework about to disentangle which kinds of emotional support strategies are helpful under which circumstances when dealing with abusive supervision.

Conclusion

Our research shows that investigating state affective responses to abusive supervision provides unique insights for two important subordinate outcomes. More specifically, we demonstrate that NA mediates the positive relation of abusive supervision with CWB-O (but not with supervisor-rated CWB-O), whereas PA mediates the positive relation of abusive supervision with turnover intentions. Moreover, we find that subordinates who receive higher coworker emotional support engage in more CWB-O in response to experiencing NA and abusive supervision, whereas coworker emotional support does not moderate the indirect relation of abusive with turnover intentions via PA.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Kimberley Breevaart

Xichao Zhang

Author Biographies

Wen Zhang is a PhD candidate of psychology department at Beijing Normal University. Her research focuses on leadership and organizational health. She has conducted a series of researches and interventions about how leaderships influence organizational climate as well as employees’ mental health and performance.

Shumin Zheng achieved her master's degree at Beijing Normal University, and now is pursing her PhD at The University of Edinburgh. Her research mainly focuses on recovery, personality and workers’ psychological health.

Jan Luca Pletzer is an Assistant Professor of Organizational Psychology at the Erasmus University Rotterdam in The Netherlands. His research focuses on the relations of broad and narrow personality traits with a wide variety of different outcomes in an organizational context, such as contextual performance, well-being, or leadership, often using meta-analytic methods. For example, he has published several meta-analyses that examine the extent to which personality traits predict outcomes such as workplace deviance or organizational citizenship behavior. Jan is on the editorial boards of Organizational Psychology Review, Stress & Health, and the international Journal of Selection and Assessment.

Daantje Derks is an Associate Professor of Work and Organizational Psychology at the Erasmus University Rotterdam (EUR) in The Netherlands. She is the chair of the Work and Organizational Psychology group. Her main research topics are: work-life interaction, the impact of digital media (mainly smartphones) on employee well-being and proactive behavior. Her research is published in journals such as Computers in Human Behavior, Journal of Occupational Health Psychology, Journal of Vocational Behavior and Human Relations. She is a frequently asked reviewer for multiple high end journals in the field and on the editorial board of Journal of Vocational Behavior.

Kimberley Breevaart is an Associate Professor of Organizational Psychology at the Erasmus University Rotterdam (EUR) in The Netherlands. Her main research topic is leadership: she studies destructive leadership, day-to-day leadership, and the relations between personality, leadership, work engagement, and burnout. In 2018 she received a 2-year fellowship from the EUR for her research on abusive supervision, leading to the publication of The Barriers Model of Abusive Supervision in Academy of Management Perspectives. In this model, she and her co-authors outline the most important barriers that prevent victims of abusive supervision to escape. In addition, she studies proactivity with a specific focus on self-management and playful work design. Her research is published in journals such as The Leadership Quarterly, Journal of Organizational Behavior, Journal of Occupational Health Psychology, Journal of Vocational Behavior, Journal of Research in Personality, and Journal of Business and Psychology. She is section editor at Stress &Health, and she is on the editorial board of Organizational Psychology Review and European Journal of Work and Organizational Psychology.

Xichao Zhang is a professor of Psychology Department at Beijing Normal University in China. His main research focuses on field of occupational health psychology, and he pays close attention to the real-life challenges of Chinese employees. He also introduces Employee Assistance Program to China, and have conducted many theoretical and practical explorations in global renowned companies. He has published papers in journals such as Journal of Business and Psychology, International Journal of Stress Management, Stress and Health, and Journal of Occupational and Organizational Psychology. He is on the editorial board of Journal of International Journal of Stress Management.

References

Afthanorhan

Awang

Abd Majid

Foziah

Ismail

Al Halbusi

Tehseen

(2021). Gain more insight from common latent factor in structural equation modeling. Journal of Physics: Conference Series, 1793(1), 012030. https://doi.org/10.1088/1742-6596/1793/1/012030

Arman

(2020). Bystander intervention in the context of abusive supervision: Effects of power distance orientation and gender. Revista de Psicología del Trabajo y de las Organizaciones, 36(2), 111-124. https://doi.org/10.5093/jwop2020a11

Bamberger

P. A.

Bacharach

S. B.

(2016). Abusive supervision and subordinate problem drinking: taking resistance, stress and subordinate personality into account. Human Relations, 59(6), 723-752. https://doi.org/10.1177/0018726706066852

Barsade

S. G.

Gibson

D. E.

(2007). Why does affect matter in organizations? Academy of Management Perspectives, 21(1), 36-59. https://doi.org/10.5465/amp.2007.24286163

Barsade

S. G.

Knight

A. P.

(2015). Group affect. Annual Review of Organizational Psychology and Organizational Behavior, 2(1), 21-46. https://doi.org/10.1146/annurev-orgpsych-032414-111316

Bennett

R. J.

Robinson

S. L.

(2000). Development of a measure of workplace deviance. Journal of Applied Psychology, 85(3), 349-360. https://doi.org/10.1037/0021-9010.85.3.349

Berkowitz

Harmon-Jones

(2004). Toward an understanding of the determinants of anger. Emotion (Washington, D.C.), 4(2), 107-130. https://doi.org/10.1037/1528-3542.4.2.107

Bollen

K. A.

(1989). Structural equations with latent variables. Wiley.

Boren

J. P.

(2014). The relationships between co-rumination, social support, stress, and burnout among working adults. Management Communication Quarterly, 28(1), 3-25. https://doi.org/10.1177/0893318913509283

10.

Breckler

S. J.

Wiggins

E. C.

(1989). Affect versus evaluation in the structure of attitudes. Journal of Experimental Social Psychology, 25(3), 253-271. https://doi.org/10.1016/0022-1031(89)90022-X

11.

Breevaart

Wisse

Schyns

. (2021). Trapped at work: The barriers model of abusive supervision. Academy of Management Perspectives. https://doi.org/10.5465/amp.2021.0007

12.

Brislin

R. W.

(1970). Back-translation for cross-cultural research. Journal of Cross-Cultural Psychology, 1(3), 185-216. https://doi.org/10.1177/135910457000100301

13.

Britt

C. L.

Weisburd

(2010). Statistical power. In Piquero

Weisburd

(Eds.), Handbook of quantitative criminology (pp. 313-332). Springer.

14.

Burleson

B. R.

(2008). What counts as effective emotional support? Explorations of individual and situational differences. In Motley

M. T.

(Ed.), Studies in applied interpersonal communication (pp. 207-227). Sage.

15.

Byrne

B. M.

(2013). Structural equation modeling with Mplus: Basic concepts, applications, and programming. Routledge, Taylor & Francis Group.

16.

Caesens

Nguyen

Stinglhamber

(2019). Abusive supervision and organizational dehumanization. Journal of Business and Psychology, 34(5), 709-728. https://doi.org/10.1007/s10869-018-9592-3

17.

Chiaburu

D. S.

Harrison

D. A.

(2008). Do peers make the place? Conceptual synthesis and meta-analysis of coworker effects on perceptions, attitudes, OCBs, and performance. Journal of Applied Psychology, 93(5), 1082-1103. https://doi.org/10.1037/0021-9010.93.5.1082

18.

Cohen

Wills

T. A.

(1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98(2), 310-357. https://doi.org/10.1037/0033-2909.98.2.310

19.

Dalal

R. S.

(2005). A meta-analysis of the relationship between organizational citizenship behavior and counterproductive work behavior. Journal of Applied Psychology, 90(6), 1241-1255. https://doi.org/10.1037/0021-9010.90.6.1241

20.

Diener

Emmons

R. A.

(1984). The independence of positive and negative affect. Journal of Personality and Social Psychology, 47(5), 1105-1117. https://doi.org/10.1037/0022-3514.47.5.1105

21.

Duffy

(2009). Preventing workplace mobbing and bullying with effective organizational consultation, policies, and legislation. Consulting Psychology Journal: Practice and Research, 61(3), 242-262. https://doi.org/10.1037/a0016578

22.

Duffy

M. K.

Ganster

D. C.

Pagon

(2002). Social undermining in the workplace. Academy of Management Journal, 45(2), 331-351. https://doi.org/10.5465/3069350

23.

Eichhorn

B. R.

(2014). Common method variance techniques. In Cleveland State university, department of operations & supply chain management (pp. 1-11). SAS Institute Inc.

24.

Ellsworth

P. C.

Scherer

K. R.

(2003). Appraisal processes in emotion. In Davidson

R. J.

Scherer

K. R.

Goldsmith

H. H.

(Eds.), Handbook of affective sciences (pp. 572-595). Oxford University Press.

25.

Fenlason

K. J.

Beehr

T. A.

(1994). Social support and occupational stress: Effects of talking to others. Journal of Organizational Behavior, 15(2), 157-175. https://doi.org/10.1002/job.4030150205

26.

Fischer

Tian

A. W.

Lee

Hughes

D. J.

(2021). Abusive supervision: A systematic review and fundamental rethink. The Leadership Quarterly, 32(6), 101540. https://doi.org/10.1016/j.leaqua.2021.101540

27.

Fox

Spector

Goh

Bruursema

(2007). Does your coworker know what you're doing? convergence of self-and peer-reports of counterproductive work behavior. International Journal of Stress Management, 14(1), 41-60. https://doi.org/10.1037/1072-5245.14.1.41

28.

Fredrickson

B. L.

Losada

M. F.

(2005). Positive affect and the complex dynamics of human flourishing. The American Psychologist, 60(7), 678-686. https://doi.org/10.1037/0003-066X.60.7.678

29.

Frijda

N. H.

(2009). Emotions, individual differences and time course: Reflections. Cognition and Emotion, 23(7), 1444-1461. https://doi.org/10.1080/02699930903093276

30.

George

J. M.

Jones

G. R.

(1996). The experience of work and turnover intentions: Interactive effects of value attainment, job satisfaction, and positive mood. Journal of Applied Psychology, 81(3), 318-325. https://doi.org/10.1037/0021-9010.81.3.318

31.

Hamaker

E. L.

Kuiper

R. M.

Grasman

R. P

. (2015). A critique of the cross-lagged panel model. Psychological Methods, 20(1), 102-116. https://doi.org/10.1037/a0038889

32.

Hannah

S. T.

Schaubroeck

J. M.

Peng

A. C.

(2016). Transforming followers’ value internalization and role self-efficacy: Dual processes promoting performance and peer norm-enforcement. Journal of Applied Psychology, 101(2), 252-266. https://doi.org/10.1037/apl0000038

33.

Harvey

Stoner

Hochwarter

Kacmar

(2007). Coping with abusive supervision: The neutralizing effects of ingratiation and positive affect on negative employee outcomes. The Leadership Quarterly, 18(3), 264-280. https://doi.org/10.1016/j.leaqua.2007.03.008

34.

Hayes

A. F.

(2018). Introduction to mediation, moderation, and conditional process analysis. In A regression-based approach (2nd ed. pp. 417-423). The Guilford Press.

35.

High

A. C.

Dillard

J. P.

(2012). A review and meta-analysis of personcentered messages and social support outcomes. Communication Studies, 63(1), 99-118. https://doi.org/10.1080/10510974.2011.598208

36.

Hom

P. W.

Caranikas-Walker

Prussia

G. E.

Griffeth

R. W.

(1992). A meta-analytic structural equations analysis of a model of employee turnover. Journal of Applied Psychology, 77(6), 890-909. https://doi.org/10.1037/0021-9010.77.6.890

37.

Judge

T. A.

Hulin

C. L.

Dalal

R. S.

(2012). Job satisfaction and job affect. In The Oxford handbook of organizational psychology (Vol. 1, pp. 496-525). Oxford University Press.

38.

Kaplan

Bradley

J. C.

Luchman

J. N.

Haynes

(2009). On the role of positive and negative affectivity in job performance: A meta-analytic investigation. Journal of Applied Psychology, 94(1), 162-176. https://doi.org/10.1037/a0013115

39.

Klein

Moosbrugger

(2000). Maximum likelihood estimation of latent interaction effects with the LMS method. Psychometrika, 65(4), 457-474. https://doi.org/10.1007/BF02296338

40.

Lazarus

Folkman

(1984). Stress, appraisal, and coping. Springer Pub.

41.

Lazarus

R. S.

(1966). Psychological stress and the coping process. McGraw-Hill.

42.

Lazarus

R. S.

(1991). Cognition and motivation in emotion. American Psychologist, 46(4), 352-367. https://doi.org/10.1037/0003-066X.46.4.352

43.

Lazarus

R. S.

Folkman

(1987). Transactional theory and research on emotions and coping. European Journal of Personality, 1(3), 141-169. https://doi.org/10.1002/per.2410010304

44.

Lerner

J. S.

Keltner

(2001). Fear, anger, and risk. Journal of Personality and Social Psychology, 81(1), 146-159. https://doi.org/10.1037/0022-3514.81.1.146

45.

Chen

Zhang

Luo

(2017). How do authoritarian leadership and abusive supervision jointly thwart follower proactivity? a social control perspective. Journal of Management, 47(4), 930-956. https://doi.org/10.1177/0149206319878261

46.

Little

T. D.

Rhemtulla

Gibson

Schoemann

A. M.

(2013). Why the items versus parcels controversy needn’t be one. Psychological Methods, 18(3), 285-300. https://doi.org/10.1037/a0033266

47.

Mackey

J. D.

Frieder

R. E.

Brees

J. R.

Martinko

M. J

. (2017). Abusive supervision: A meta-analysis and empirical review. Journal of Management, 43(6), 1940–1965. https://doi.org/10.1177/0149206315573997

48.

Martinko

M. J.

Mackey

J. D.

Moss

S. E.

Harvey

McAllister

C. P.

Brees

J. R.

(2018). An exploration of the role of subordinate affect in leader evaluations. Journal of Applied Psychology, 103(7), 738-752. https://doi.org/10.1037/apl0000302

49.

Martinko

M. J.

Harvey

Brees

J. R.

Mackey

(2013). A review of abusive supervision research. Journal of Organizational Behavior, 34(S1), S120-S137. https://doi.org/10.1002/job.1888

50.

Maslyn

J. M.

Uhl-Bien

. (2001). Leader–member exchange and its dimensions: Effects of self-effort and other's effort on relationship quality. Journal of Applied Psychology, 86(4), 697–708. https://doi.org/10.1037/0021-9010.86.4.697

51.

Mitchell

M. S.

Ambrose

M. L.

(2007). Abusive supervision and workplace deviance and the moderating effects of negative reciprocity beliefs. Journal of Applied Psychology, 92(4), 1159-1168. https://doi.org/10.1037/0021-9010.92.4.1159

52.

Mobley

W. H.

Horner

S. O.

Hollingsworth

A. T.

(1978). An evaluation of precursors of hospital employee turnover. Journal of Applied Psychology, 63(4), 408-414. https://doi.org/10.1037/0021-9010.63.4.408

53.

Moors

Ellsworth

P. C.

Scherer

K. R.

Frijda

N. H.

(2013). Appraisal theories of emotion: State of the art and future development. Emotion Review, 5(2), 119-124. https://doi.org/10.1177/1754073912468165

54.

Neves

Cunha

M. P. E.

(2018). Exploring a model of workplace ostracism: The value of coworker humor. International Journal of Stress Management, 25(4), 330-347. https://doi.org/10.1037/str0000069

55.

T. W. H.

Sorensen

K. L.

(2008). Toward a further understanding of the relationships between perceptions of support and work attitudes:A meta-analysis. Group & Organization Management, 33(3), 243-268. https://doi.org/10.1177/1059601107313307

56.

J. K.

Farh

C. I. C.

(2017). An emotional process theory of how subordinates appraise, experience, and respond to abusive supervision over time. Academy of Management Review, 42(2), 207-232. https://doi.org/10.5465/amr.2014.0347

57.

Patrizia

Garofalo

Bottazzi

Caretti

(2017). Faces of shame: implications for self-esteem, emotion regulation, aggression, and well-being. The Journal of Psychology, 151(2), 171-184. https://doi.org/10.1080/00223980.2016.1248809

58.

Peng

A. C.

Schaubroeck

J. M.

Chong

(2019). Discrete emotions linking abusive supervision to employee intention and behavior. Personnel Psychology, 72(3), 393-419. https://doi.org/10.1111/peps.12310

59.

Peng

A. C.

Schaubroeck

J. M.

(2014). Social exchange implications of own and CoWorkers’ experiences of supervisory abuse. Academy of Management Journal, 57(5), 1385-1405. https://doi.org/10.5465/amj.2012.0080

60.

Podsakoff

N. P.

LePine

J. A.

LePine

M. A.

(2007). Differential challenge stressor-hindrance stressor relationships with job attitudes, turnover intentions, turnover, and withdrawal behavior: A meta-analysis. Journal of Applied Psychology, 92(2), 438-454. https://doi.org/10.1037/0021-9010.92.2.438

61.

Podsakoff

P. M.

MacKenzie

S. B.

Lee

J. Y.

Podsakoff

N. P.

(2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5), 879-903. https://doi.org/10.1037/0021-9010.88.5.879

62.

Polk

D. E.

Cohen

Doyle

W. J.

Skoner

D. P.

Kirschbaum

(2005). State and trait affect as predictors of salivary cortisol in healthy adults. Psychoneuroendocrinology, 30(3), 261-272. https://doi.org/10.1016/j.psyneuen.2004.08.004

63.

Poon

J. M.

(2011). Effects of abusive supervision and coworker support on work engagement. International Proceedings of Economics Development and Research, 22, 64-70.

64.

Preacher

K. J.

Rucker

D. D.

Hayes

A. F.

(2007). Addressing moderated mediation hypotheses: Theory, methods, and prescriptions. Multivariate Behavioral Research, 42(1), 185-227. https://doi.org/10.1080/00273170701341316

65.

Qian

J. J

. (2001). Perceived social support scale. Chinese Journal of Behavior and Medicine Science, 10(10), 41-43. ISSN : 1005-8559

66.

Qiu

Zheng

Wang

Y. F

. (2008). Revision of the positive emotion and negative emotion scale (PANAS). Chinese Applied Psychology, 14(3), 249-254. ISSN: 1006-6020

67.

Quigley

K. S.

Lindquist

K. A.

Barrett

L. F.

(2014). Inducing and measuring emotion and affect: Tips, tricks, and secrets. In Reis

H. T.

Judd

C. M.

(Eds.), Handbook of research methods in social and personality psychology (pp. 220-252). Cambridge University Press.

68.

Rogers

W. M.

Schmitt

(2004). Parameter recovery and model fit using multidimensional composites: A comparison of four empirical parceling algorithms. Multivariate Behavioral Research, 39(3), 379-412. https://doi.org/10.1207/S15327906MBR3903_1

69.

Rozin

Royzman

E. B

. (2001). Negativity Bias, Negativity Dominance, and Contagion. Personality and Social Psychology Review, 5(4), 296-320. https://doi.org/10.1207/S15327957PSPR0504_2

70.

Russell

J. A.

Carroll

J. M.

(1999). On the bipolarity of positive and negative affect. Psychological Bulletin, 125(1), 3-30. https://doi.org/10.1037/0033-2909.125.1.3

71.

Scherer

K. R.

(1999). Appraisal theory. In Dalgleish

Power

M. J.

(Eds.), Handbook of cognition and emotion (pp. 637-663). John Wiley & Sons Ltd.

72.

Shoss

M. K.

Eisenberger

Restubog

S. L. D.

Zagenczyk

T. J.

(2013). Blaming the organization for abusive supervision: The roles of perceived organizational support and supervisor's organizational embodiment. Journal of Applied Psychology, 98(1), 158-168. https://doi.org/10.1037/a0030687

73.

Smith

C. A.

Ellsworth

P. C

. (1985). Patterns of cognitive appraisal in emotion. Journal of personality and social psychology, 48(4), 813–838.

74.

Spector

P. E.

Fox

(2002). An emotion-centered model of voluntary work behavior: Some parallels between counterproductive work behavior and organizational citizenship behavior. Human Resource Management Review, 12(2), 269-292. https://doi.org/S1053-4822(02)00049-9

75.

Spector

P. E.

Fox

(2005). The stressor-emotion model of counterproductive work behavior. In Fox

Spector

P. E.

(Eds.), Counterproductive behavior. Investigations of actors and targets (pp. 151-174). Washington, DC: American Psychological Association.

76.

Stine

(1989). An introduction to bootstrap methods. Sociological Methods & Research, 18(2), 243-291. https://doi.org/10.1177/0049124189018002003

77.

Tepper

B. J.

(2000). Consequences of abusive supervision. Academy of Management Journal, 43(2), 178-190. https://doi.org/10.5465/1556375

78.

Tepper

B. J

. (2007). Abusive supervision in work organizations: Review, synthesis, and research agenda. Journal of Management, 33(3), 261-289. https://doi.org/10.1177/0149206307300812

79.

Tepper

B. J.

Simon

Park

H. M.

(2017). Abusive supervision. Annual Review of Organizational Psychology and Organizational Behavior, 4(1), 123-152. https://doi.org/10.1146/annurev-orgpsych-041015-062539

80.

Tett

R. P.

Meyer

J. P.

(2006). Job satisfaction, organizational commitment, turnover intention, and turnover: path analyses based on meta-analytic findings. Personnel Psychology, 46(2), 259-293. https://doi.org/10.1111/j.1744-6570.1993.tb00874.x

81.

Tews

M. J.

Michel

J. W.

Stafford

(2019). Abusive coworker treatment, coworker support, and employee turnover. Journal of Leadership & Organizational Studies, 26(4), 413-423. https://doi.org/10.1177/1548051818781812

82.

Thoresen

C. J.

Kaplan

S. A.

Barsky

A. P.

Warren

C. R.

de Chermont

(2003). The affective underpinnings of job perceptions and attitudes: a meta-analytic review and integration. Psychological Bulletin, 129(6), 914-945. https://doi.org/10.1037/0033-2909.129.6.914

83.

Van Emmerik

I. H.

Euwema

M. C.

Bakker

A. B.

(2007). Threats of workplace violence and the buffering effect of social support. Group & Organization Management, 32(2), 152-175. https://doi.org/10.1177/1059601106286784

84.

Vogel

R. M.

Bolino

M. C.

(2020). Recurring nightmares and silver linings: Understanding how past abusive supervision may lead to posttraumatic stress and posttraumatic growth. Academy of Management Review, 45(3), 549-569. https://doi.org/10.5465/amr.2017.0350

85.

Watson

Clark

L. A.

Tellegen

(1988). Development and validation of brief measures of positive and negative affect: the PANAS scales. Journal of Personality and Social Psychology, 54(6), 1063-1070. https://doi.org/10.1037/0022-3514.54.6.1063

86.

Watson

Wiese

Vaidya

Tellegen

. (1999). The two general activation systems of affect: Structural findings, evolutionary considerations, and psychobiological evidence. Journal of Personality and Social Psychology, 76(5), 820-838.

87.

Watson

Spence

M. T.

(2007). Causes and consequences of emotions on consumer behaviour: A review and integrative cognitive appraisal theory. European Journal of Marketing, 41(5), 487-511. https://doi.org/10.1108/03090560710737570

88.

T.-Y.

(2009). Abusive supervision and employee emotional exhaustion:dispositional antecedents and boundaries. Group & Organization Management, 34(2), 143-169. https://doi.org/10.1177/1059601108331217

89.

T.-Y.

(2013). Abusive supervision and subordinate emotional labor: The moderating role of openness personality: Abusive supervision and emotional labor. Journal of Applied Social Psychology, 43(5), 956-970. https://doi.org/10.1111/jasp.12060

90.

Zhang

Liu

Yang

L.-Q.

Bednall

T. C.

(2019). Why abusive supervision impacts employee OCB and CWB: A meta-analytic review of competing mediating mechanisms. Journal of Management, 45(6), 2474-2497. https://doi.org/10.1177/0149206318823935

91.

Zimet

G. D.

Dahlem

N. W.

Zimet

S. G.

Farley

G. K.

(1988). The multidimensional scale of perceived social support. Journal of Personality Assessment, 52(1), 30-41. https://doi.org/10.1207/s15327752jpa5201_2