Abstract
Working people differentiate between their relationships with their organizations and their relationships with their supervisors. To better understand how these simultaneous relationships influence each other, we examined the effects of organizational unfairness on subordinates’ supervisor-directed destructive voice behaviors, as well as the effects of supervisor unfairness on subordinates’ organization-directed destructive voice behaviors. We examined supervisor organizational embodiment as a boundary condition and perceived obstruction as a mediating mechanism within these relationships. We tested the hypothesized model with an online survey study and an experimental vignette study. Study 1 indicated that organizational unfairness was positively and indirectly related to supervisor-directed destructive voice through perceived supervisor obstruction when supervisor organizational embodiment was higher rather than lower. Study 2 indicated that supervisor unfairness was positively and indirectly related to organization-directed destructive voice through perceived organizational obstruction when supervisor organizational embodiment was higher rather than lower. Although study results differed in the observed significance of organization-to-supervisor and supervisor-to-organization effects, together they demonstrate the importance of controlling for source-target aligned relationships when examining source-target misaligned effects, and extend source-target misalignment research by examining the effect of supervisor organizational embodiment as a moderator of misaligned supervisor- and organization-directed workplace unfairness perceptions and destructive voice behaviors.
Employees maintain ongoing, distinct social exchange relationships with their organizations and supervisors (Lavelle, Rupp, & Brockner, 2007). For example, employees respond to abusive supervision by retaliating toward their supervisors (Mackey, Frieder, Brees, & Martinko, 2017) and respond to organizational unfairness by performing fewer organizational citizenship behaviors (Lavelle et al., 2009). We refer to this consistency between employees’ perceived social exchange sources and subsequent behavioral targets as source-target alignment. Still, inconsistency has been observed between employees’ perceived sources of negative exchange and the target to which they direct subsequent behaviors. For example, experienced abusive supervision is related to retaliatory employee behaviors directed toward not only the abusive supervisor but also the organization (Shoss, Eisenberger, Restubog, & Zagenczyk, 2013). In addition, meta-analytic findings suggest employees may respond to organizational wrongdoing by lashing out interpersonally (Rupp, Shao, Jones, & Liao, 2014). We refer to this inconsistency between employees’ perceived social exchange sources and subsequent behavioral targets as source-target misalignment. The purpose of this research is to shed additional light on the nature of social exchange in the workplace by examining misalignment between employees’ perceived sources of unfairness and the targets of their subsequent destructive voice behaviors while controlling for source-target alignment, with a specific focus on organizations and supervisors as both sources and targets.
Research directly examining misalignment between sources and targets of social exchange has thus far focused on examining how negative supervisor interactions affect employees’ organization-directed behavioral responses (i.e., supervisor-to-organization effects; Mackey, McAllister, Brees, Huang, & Carson, 2018; Shoss et al., 2013). However, employees’ supervisor-directed responses to negative organizational exchanges (i.e., organization-to-supervisor effects) remain to be explored in similar depth. It is important to directly examine organization-to-supervisor effects to understand their impact on outcomes that stem from organizational perceptions and to provide a more complete picture of the determinants of workplace interpersonal mistreatment. The present research extends prior work by examining both supervisor-to-organization effects as well as organization-to-supervisor effects, which contributes to theory, research, and practice by demonstrating that both forms of source-target misalignment shed light on how social exchange occurs in the workplace.
We examine a social exchange model assessing employee relationships with their supervisors and organizations, as well as their behaviors directed toward their supervisors and organizations (please see Figure 1). We first consider employees’ perceptions of supervisor and organizational unfairness as indicators of general social exchange quality perceptions from each of these sources. In doing so, we respond to meta-analyses which indicate that additional studies of specific source-target misalignment are necessary to better understand employee relationship differentiation in the fairness context; this is especially important because much extant research does not adequately control for differentiated sources of unfairness in the same way that it is concerned with specifying different types of unfairness (e.g., Colquitt et al., 2013; Hershcovis et al., 2007; Rupp et al., 2014). Research model. Note. Dotted lines indicate the control variables.
We assess employees’ behavioral responses to unfairness in the form of supervisor- and organization-directed destructive voice, an emerging construct which captures employees’ overly harsh criticism of workplace procedures and policies (Maynes & Podsakoff, 2014). Our research model remains consistent with prior social exchange approaches by examining the mediating role of perceived organizational and supervisor obstruction, which is an employee’s perception of the extent to which their organization [supervisor] gets in the way of his or her goals or well-being (Gibney, Zagenczyk, & Masters, 2009, p. 670). Further, our model suggests employees’ destructive voice targets may differ from the aligned source of unfairness when employees perceive their supervisor to represent their organization (i.e., supervisor organizational embodiment [SOE]; Eisenberger et al., 2010).
Our study was motivated by two goals. First, we sought to demonstrate a more comprehensive multifoci approach to understanding SOE’s role in explaining source-target misalignment by examining both organization-to-supervisor and supervisor-to-organization effects while controlling for source-target alignment. This extends prior research which has focused on understanding employees’ “one-way” (i.e., supervisor-to-organization) social exchange misalignment (Mackey et al., 2018; Shoss et al., 2013) and responds to theoretical and meta-analytic suggestions that misalignment may occur “both ways” (Lavelle et al., 2007; Rupp et al., 2014). This is important because supervisors may face social exchange backlash from employees responding to organizational unfairness, a phenomenon that is in accordance with Lavelle et al.’s (2007) target similarity model, but which has not been considered to be affected by SOE. Second, we sought to examine source-target misalignment while controlling for source-target alignment in order to provide a clearer picture of the extent to which source-target misalignment occurs. Our model allows for this control by examining pairs of social exchange constructs which are substantively similar yet allow for distinction in source (e.g., supervisor unfairness and organizational unfairness).
Overall, we draw upon social exchange theory and research related to Lavelle and colleagues’ (2007) target similarity model to present our hypotheses. We then present two studies, including a survey study and experimental vignette study, so we can test for source-target misalignment while controlling for source-target alignment.
Theoretical Foundations and Hypothesis Development
According to social exchange theory, the relationship between the employee and his or her supervisor or organization comprises ongoing, interdependent exchanges that create and fulfill obligations (Blau, 1964; Cropanzano & Mitchell, 2005). Gouldner (1960) termed the felt obligation to reciprocate the receipt of a social resource the norm of reciprocity. When an organization or supervisor performs an initiating behavior, employees tend to adhere to the norm of reciprocity in order to bring balance to the relationship (Berry, Ones, & Sackett, 2007; Cropanzano & Mitchell, 2005; Lavelle et al., 2007). Social exchange research indicates that the norm of reciprocity is not limited to positive exchanges, but extends to negative exchanges as well (Cropanzano, Anthony, Daniels, & Hall, 2017; Liden, Sparrowe, & Wayne, 1997). This transaction of harmful behaviors characterizes low quality social exchange relationships. It is reasonable to expect that employees’ negatively valenced perceptions of unfairness will be distinguishable in their effects on employees’ subsequent social exchange relationship quality with their respective focal actors (i.e., organizations and supervisors) due to the power of negative perceptions in prompting negative reciprocation (Baumeister, Bratslavsky, Finkenauer, & Vohs, 2001; Colquitt, Long, Rodell, & Halvorsen-Ganepola, 2015; Gouldner, 1960).
The quality of a social exchange relationship is based on the presence and reciprocation of social benefits such as fair treatment. We define unfairness as “a negatively valenced evaluation that the expectation of fair treatment has been violated” (Baer et al., 2018, p. 6) and can thus be seen as indicative of poor-quality social exchange. Employee responses to perceived workplace unfairness include counterproductive work behaviors (Jones, 2009), absenteeism (de Boer, Bakker, Syroit, & Schaufeli, 2002), and retaliatory bargaining (Brebels, De Cremer, & Sedikides, 2008), all of which detrimentally impact effective organizational functioning. We maintain a broad view of unfairness throughout our paper with a focus on how the presence of unfairness operates in a social exchange relationship.
Perceptions of unfairness arise when an individual experiences or observes a disadvantageous or injurious condition that is attributable to an actor and which is deemed transgressive (Folger & Cropanzano, 2001). Accountability, or identifying who to blame, occurs through a process in which the individual considers whether the potentially responsible actor(s) should have and could have done something differently, and whether the outcome would have been fairer had they performed an alternative action (Folger & Cropanzano, 2001). Organizations, however, are complex systems that affect employees in personal ways. Social exchange theory acknowledges that employees may anthropomorphize the organization to view it as an acting entity when determining responsibility for their workplace experiences (Blau, 1964; Cropanzano et al., 2017). In fairness terms, “the ‘who’ that ‘could’ have acted differently may be either an individual (such as the victim’s supervisor) or an organization (i.e., a ‘juristic person’)” (Cropanzano et al., 2001, p. 169).
Perceptions of a given party’s unfairness are more strongly related to low quality social exchange relationship perceptions with that focal party than with other parties. In other words, source-target alignment effects are generally stronger than source-target misalignment effects (Lavelle et al., 2007). Because source-target aligned effects are generally stronger, Lavelle et al. use the term “spillover effects” to describe source-target misalignment effects with the apparent assumption that employees perceive the source of unfairness to also be the responsible party. We challenge this assumption by suggesting that employees may sometimes perceive a third party to be responsible for a given party’s unfairness. We further assert that this attribution of responsibility manifests in employees’ perceived source of their experienced workplace obstruction. We explain these assertions in the following section.
Perceived Obstruction and Embodiment
When a social exchange partner, such as a supervisor or employing organization, violates expectations for fair treatment, employees may feel this lack of reciprocity hinders their ability to attain the goals and well-being that come with a high-quality social exchange relationship. Perceived organizational obstruction describes the presence of an active, undesirable hindrance of employee goals, making it distinct from perceived organizational support, which is the extent to which employees feel their organization is committed to enabling their goal attainment and well-being (Eisenberger et al., 1986). Gibney and colleagues (2009) established empirical support for this theoretical distinction by showing that perceived organizational obstruction explained additional variance beyond perceived organizational support, among other similar constructs. Indeed, perceived organizational obstruction is related to organizational cynicism and feelings of psychological contract breach (Gibney, Zagenczyk, Fuller, Hester, & Caner, 2011), as well as subsequent decreases in organizational citizenship behaviors (Mackey et al., 2018).
We hypothesize that a misaligned link between unfairness and obstruction is possible due to supervisor organizational embodiment (SOE), which is the extent to which employees see their supervisor as a representative of the organization (Eisenberger et al., 2010). SOE is a function of the supervisor’s identification with the organization, which is expressed to employees through the supervisor’s comments and actions. Employees perceive high SOE when they believe their supervisor shares organizational values and goals. When employees perceive high SOE, a generalization effect is theorized to occur through which employees relate positive and negative elements of their relationships with their supervisor to their relationships with their organization (Eisenberger et al., 2010). Indeed, research indicates SOE strengthens the negative relationship between abusive supervision and perceived organizational support (Shoss et al., 2013) as well as the positive relationship between leader-member exchange and perceived organizational support (Eisenberger et al., 2014) because employees perceive supervisors’ actions as representative of the organization’s support for them.
We build upon this established generalization effect of SOE and expand it to consider particularization as well. The notion that employees view high SOE supervisors as organizational agents serves as the foundation for the generalizing effects of SOE (Eisenberger et al., 2010). It is this same foundation that supports our assertion that a particularization effect may occur through which employees specify high SOE supervisors as actors responsible for obstructions created by organizational unfairness. We therefore hypothesize that when employees believe their supervisor strongly embodies the organization, they may attribute responsibility for workplace obstruction they have experienced as a result of unfair organizational practices to their supervisor.
For example, consider a relatively new employee who feels they were unfairly passed over for promotion because the company traditionally bases promotion heavily on employee tenure. If the employee perceives their supervisor to be strongly representative of the company based on their prior actions, then the supervisor may be seen as responsible for obstructing the employee’s professional goal attainment. Perhaps the employee believes the supervisor agrees with the company’s unofficial policy and is happy to enforce the status quo.
We maintain that generalization effects will also occur consistent with prior research. Specifically, high SOE creates a condition in which employees may generalize their supervisor’s unfairness to the organization due to their supervisor’s representation of the organization’s values and goals. In such a situation, employees may feel that their organization is responsible for their experienced workplace obstruction because they perceive their supervisor’s unfair treatment is indicative of their organization’s lack of concern for their well-being or goal attainment.
In a similar example to the previous, consider a relatively new employee who feels they were unfairly passed over for promotion because their supervisor traditionally prioritizes employee tenure when making promotion decisions. If the employee believes their supervisor generally represents company views, then the employee may see the company as being responsible for thwarting their promotion goals. Perhaps the employee believes their company lacks concern for employee goal attainment because it is comprised of policies and actors which systematically enable the status quo.
To summarize, we hypothesize that employee perceptions of organizational unfairness will be positively related to perceived supervisor obstruction when SOE is high because employees will view organizationally representative supervisors as responsible for obstruction that occurs as the result of organizational unfairness. Further, we posit that employee perceptions of supervisor unfairness will be positively associated with subsequent perceptions of organizational obstruction when SOE is high because employees will see their supervisors’ unfairness as an indication of the organization’s exchange intentions regarding goal attainment and well-being.
Hypothesis 1a: SOE will moderate the relationship between organizational unfairness and supervisor obstruction such that the positive relationship will be stronger for employees who perceive higher levels of SOE than employees who perceive lower levels of SOE.
Hypothesis 1b: SOE will moderate the relationship between supervisor unfairness and perceived organizational obstruction such that the positive relationship will be stronger for employees who perceive higher levels of SOE than employees who perceive lower levels of SOE.
Perceived Obstruction and Destructive Voice
Perceived obstruction and support research show that these perceptions are indicative of exchange quality between the employee and their organization or supervisor (Allen, Shore, & Griffeth, 2003), and this exchange quality drives subsequent exchange behaviors. For example, obstruction is positively related to other negative social exchange behaviors, including exit and neglect of work duties (Gibney et al., 2009). Additionally, perceived organizational obstruction is negatively related to constructive voice (Gibney et al., 2009), indicating employees withhold positive exchange behaviors when their goals and well-being are obstructed.
Employee voice is an exchange behavior with potentially wide-reaching effects (Budd et al., 2010; Morrison, 2011; Van Dyne & LePine, 1998). Maynes and Podsakoff (2014) provided evidence that voice varies along two dimensions: preservation versus opposition and promotion versus prohibition. Destructive voice is prohibitive and challenging because it questions the workplace status quo with critical comments. Examples of destructive voice include criticizing, insulting, and making overly critical comments about the target’s workplace policies, procedures, and practices. When employees exhibit destructive voice, others perceive the employee who uses voice to care less about the organization and see the behavior as negatively impacting organizational performance (Maynes & Podsakoff, 2014).
Maynes and Podsakoff’s (2014) categorization of the voice domain made a significant contribution to the voice research literature (Organ, 2018), although little extant research makes use of Maynes and Podsakoff’s measure of destructive voice. It is important to examine the presence of destructive voice rather than a lack of constructive voice because constructive workplace criticism (i.e., constructive voice) may be a required element of employees’ job roles (Maynes & Podsakoff, 2014; Van Dyne & LePine, 1998). Destructive voice, on the other hand, is not representative of typical required job role behaviors, so it more accurately exemplifies a negatively valenced social exchange behavior than does a lack of positive voice.
Destructive voice is similar to but different from other behavioral constructs which are active and undesirable according to Cropanzano et al.’s (2017) classification of social exchange construct dimensions. Specifically, destructive voice could be considered to be a type of counterproductive work behavior (CWB), deviance, or aggression to the extent that these are viewed as broad “umbrella” concepts (Cropanzano et al., 2017). However, destructive voice is distinct from other types of CWB, deviance, and aggression in its explicitly defined mode of delivery (verbal), and its focus on overly harsh criticism of workplace policies and practices. This makes it a strong candidate for examining responses to obstruction because it communicates debasing opinions regarding the very policies and practices that can obstruct employees’ work goals and well-being. Further, we assert that goal and well-being obstruction are perceptions that may be more proximal to employees’ tendency to speak destructively than general feelings of unfairness and are thus more likely to elicit target-similar responses (Lavelle et al., 2007).
We draw from extant social exchange theory and voice research to posit that when employees perceive organizational obstruction, they reciprocate by responding with organization-directed destructive voice behaviors in order to bring balance to their social exchange relationship. These voice behaviors will be prohibitive and challenging as employees criticize the organization for hindering their ability to attain goals and achieve well-being. Likewise, we posit that employees who perceive supervisor obstruction will reciprocate with supervisor-directed destructive voice behaviors.
Hypothesis 2a: Perceived supervisor obstruction will be positively related to destructive voice toward the supervisor.
Hypothesis 2b: Perceived organizational obstruction will be positively related to destructive voice toward the organization.
Conditional Indirect Effects
We combine the prior hypotheses to propose a conditional indirect effect model whereby supervisor (organizational) obstruction mediates the misaligned relationships between organizational (supervisor) unfairness and supervisor-directed (organization-directed) destructive voice. Consider the prior example in which a new employee feels they were unfairly denied a promotion because their company traditionally bases promotion on tenure. Such an unfair denial of promotion may prompt this employee to speak out against their company (an aligned effect). We hypothesize an alternative pathway such that, due to particularization effects, this employee may believe their high-SOE supervisor is responsible for obstructing promotion and responds in accordance with social exchange by speaking harshly toward their supervisor (a misaligned effect).
On the other hand, consider the converse example in which a new employee feels unfairly denied a promotion because their supervisor bases such decisions on tenure. The employee may speak out against their supervisor in such a situation (an aligned effect). However, our hypothesized framework recognizes that, due to generalization effects, this employee may believe their high-SOE supervisor is representative of organizational responsibility for promotion obstruction. In this framework, the employee would respond in accordance with social exchange by verbally criticizing their company (a misaligned effect).
To summarize, we posit that employees’ perceived organizational unfairness will be indirectly and positively associated with subsequent supervisor-directed destructive voice behaviors due to the employees’ blame and responsibility for organizational unfairness captured via perceived supervisor obstruction. Additionally, we suggest this indirect effect will be moderated by SOE such that the relationship will be stronger for employees who view their supervisors as more representative of the organization than others. We also hypothesize that employees’ perceived supervisor unfairness will be indirectly and positively associated with subsequent organization-directed destructive voice through perceived organizational obstruction, such that higher rather than lower SOE will strengthen this positive relationship.
Hypothesis 3a: Organizational unfairness will be positively and indirectly associated with destructive voice toward supervisors through perceived supervisor obstruction, conditional upon SOE such that the conditional indirect effect will be stronger for higher levels of SOE.
Hypothesis 3b: Supervisor unfairness will be positively and indirectly associated with destructive voice toward organizations through perceived organizational obstruction, conditional upon SOE such that the conditional indirect effect will be stronger for higher levels of SOE.
Construct Choice
Construct proliferation is an ever-present concern when examining social exchange. Thus, two overarching goals drove the choice of constructs examined in our research model. First, we chose to build a model with pairs of social exchange constructs which are similar in substance but allow for distinction in source in order to examine source-target misalignment while controlling for source-target alignment. This meant avoiding constructs which are inherently (yet ostensibly) tied to a particular source. For example, assessing abusive supervision would prove difficult for our purpose of source-target alignment control because there is no equivalent construct allowing for the assessment of organizational abuse. On the other hand, unfairness and obstruction can each be perceived as stemming from the supervisor and/or organization and thus allow for a construct-similar comparison of source-target misalignment while controlling for source-target aligned effects.
The second overarching aim was to examine a model that represents broadly accepted social exchange processes (Lavelle et al., 2007) while maintaining similarity in the valence (desirable or undesirable nature) and activity level (active or passive) of the constructs. This similarity in valence and activity level is called homeomorphic reciprocity, which is a tenet of social exchange theory which suggests that, in general, social exchange involves the trading of behaviors that are similarly desirable (or undesirable) and active (or passive). Cropanzano et al. (2017) reviewed prior research to identify matched homeomorphic reciprocity effects as being stronger (empirically and theoretically) than unmatched effects and thus called for future social exchange research to examine models in which exchanged behaviors are similar in valence and activity. Our research model does so by investigating employee perceptions of the presence of unfairness, the presence of obstruction, and the extent to which they engage in destructive voice, rather than constructs that oppose each other in valence or activity (i.e., the presence of fairness, lack of support, or use of supportive voice).
Overview of Hypothesis Testing
We present a two-study approach to examining these hypotheses. In the first study, we administered surveys in which respondents reported unfairness perceptions, obstruction perceptions, SOE, and destructive voice behaviors related to their place of work. After the first study, we conducted a second study in which we employed experimental vignette methodology with a 2×2×2 between-subjects design to better assess the causal relationship between unfairness, obstruction, and likelihood of organization-directed and supervisor-directed destructive voice.
In both studies, we tested misalignment effects while controlling for aligned effects. We found it important to control for aligned effects specifically because the field accepts aligned effects to be stronger (Lavelle et al., 2007). However, extant research on misaligned effects fails to control for these aligned effects, possibly resulting in inflated observed misaligned effects. Our inclusion of aligned controls thus provides a more complete picture of the extent to which misaligned effects occur.
Study 1 Method
Sample
We recruited participants through Amazon’s Mechanical Turk (MTurk), a crowdsourcing marketplace through which individuals can volunteer for paid tasks such as completing surveys. MTurk enables the recruitment of demographically diverse working adults who are as attentive to instructions as participants from other sources (Hauser & Schwarz, 2016). Participants were compensated $1USD per survey, with the potential to complete three surveys separated by approximately 1 month each. There were 1345 initial respondents at Time 1, with 727 returning for the Time 2 survey. A total of 445 respondents completed all three surveys and met inclusion criteria outlined below.
We required participants to be at least 18 years old, speak English, live in the United States, and work at least 20 hours per week at a job with a direct supervisor. Data were retained only for respondents who had supervisor continuity between Time 1 and Time 3. Surveys included one attention check item per time period (DeSimone, Harms, & DeSimone, 2015). Each attention check item was an instructed item that read “please mark ‘agree’ for this response.” If respondents failed an attention check item (350 at Time 1, 69 at Time 2, and 69 at Time 3), they were allowed to complete the survey and were compensated, but their responses were not retained and they were not invited to take part in future time periods. To minimize the impact of common method bias, we included survey items for construct measures we did not use in the study, used both agreement and frequency response types, and collected data at three time periods (Podsakoff, MacKenzie, & Podsakoff, 2012).
The sampling technique resulted in the desired heterogeneous sample of working adults. Participants worked in a variety of industries, with the two most common categories being educational, health, and social services (21.80%) and finance, insurance, real estate, rental, and leasing (13.93%). Participants were an average of 38.29 years old, and 50.34% of the sample was female. Most respondents worked at the staff/associate level (47.19%) in their organizations, and they worked an average of 39.50 hours per week. Participants had spent an average of 6.69 years employed at their current organization and had an average of 5.04 years working with their current supervisor. Over half of respondents (58.43%) held at least a bachelor’s degree.
Measures
Unfairness
We measured unfairness with Colquitt et al.’s (2015) three-item supervisor unfairness scale. To measure organizational unfairness, the referent for each item was changed from “supervisor” to “organization.” An example scale item is “Does your supervisor [organization] do things that are unfair?” Responses were reported along a seven-point scale (1 = “to an extremely small extent” and 7 = “to an extremely large extent”). Unfairness was measured at Time 1 (organizational unfairness: α = .98; supervisor unfairness: α = .97).
Supervisor Organizational Embodiment
We used Eisenberger et al.’s (2010) seven-item scale to measure SOE. Respondents reported their agreement with statements along a seven-point scale (1 = “strongly disagree” and 7 = “strongly agree”). An example item reads “My supervisor and my organization have a lot in common.” SOE was measured at Time 2 (α = .95).
Perceived Obstruction
We measured perceived organizational obstruction with Gibney et al.’s (2009) five-item scale. To measure perceived supervisor obstruction, the referent was changed from “organization” to “supervisor.” An example item is “My organization [supervisor] is a detriment to my well-being.” Respondents reported their agreement with items along a seven-point scale (1 = “strongly disagree” and 7 = “strongly agree”). Perceived obstruction was measured at Time 2 (organizational obstruction: α = .96; supervisor obstruction: α = .96).
Destructive Voice
To measure organization-directed destructive voice, we used the five-item scale developed by Maynes and Podsakoff (2014). We changed the referent for each item from “organization” to “supervisor” to measure supervisor-directed destructive voice. An example item is “I often make overly critical comments about my organization’s [supervisor’s] work practices or methods.” Respondents reported their destructive voice behaviors by rating their agreement with the statements on a seven-point scale (1 = “strongly disagree” and 7 = “strongly agree”). Destructive voice behaviors were measured at Time 3 (destructive voice toward organization: α = .93; destructive voice toward supervisor: α = .94).
Control Variables
For each of our hypotheses, we controlled for the aligned source of unfairness while testing the effects of the misaligned source of unfairness and also controlled for the alternate source of obstruction. Doing so provides a strong test of misaligned effects while accounting for aligned effects. We conducted the analyses reported below while controlling for age and gender, because prior research has indicated that younger people and men appear to perform more workplace deviance than women or older people (Berry et al., 2007; Mackey et al., 2021). However, including gender and age did not change the regression coefficients of our model variables, so they are not included in the presented results.
Study 1 Analysis
We used SPSS 23.0 and AMOS to perform all statistical analyses. We used hierarchical ordinary least squares (OLS) regression to examine Hypotheses 1 and 2, and followed Hayes’ (2013) PROCESS approach to examine the moderated mediation models described in Hypothesis 3. Hayes’ PROCESS procedure is a regression-based approach that uses bootstrapping to analyze conditional indirect effects. Bootstrapping overcomes the need for the assumption of a normal distribution of regression coefficients by generating confidence intervals based upon a bootstrapped sampling distribution that accounts for the coefficients’ actual distribution (Preacher, Rucker, & Hayes, 2007). We used Model 7 in PROCESS, and variables were mean centered before creating interaction terms to more clearly interpret effects (Cohen, Cohen, West, & Aiken, 2003; Hayes, 2013).
Study 1 Measurement Model and Comparison.
Note. N = 445. χ2 = chi-square. df = degrees of freedom. CFI = comparative fit index. TLI = Tucker-Lewis index. RMSEA = root mean square error of approximation. SRMR = standardized root mean square residual. Sup. = supervisor. Org. = organizational.
* p < .05. ** p < .01. ***p < .001.
Modification indices from CFA analyses indicated that specifying certain within-factor and between-factor error terms to be correlated would result in increased model fit. It is likely that correlation between error terms would be due to their proximity in the self-report survey or some similarity in wording (e.g., Goeddeke Jr. & Kammeyer-Mueller, 2010). However, we elected to present results for a model with uncorrelated error terms to aid in parameter value replication, to provide the most conservative estimate of model fit, and to avoid impacting our results with speculation about the true source of error correlation in the model (Hermida, 2015). All factor loadings were significant and ranged from .65 to .98. We were comfortable retaining the two items with factor loadings below .70 to fully represent constructs per their established measures.
We compared the seven-factor measurement model to a series of models with fewer factors which reflected potential construct type similarity (i.e., unfairness and obstruction), source similarity (i.e., supervisors and organizations), and target similarity (i.e., supervisors and organizations). We compared RMSEA 90% confidence intervals to assess the fit of the seven-factor model relative to that of each more parsimonious model (Colquitt, 2001). The 90% confidence interval of the RMSEA for the seven-factor measurement model (90% CI [.087, .094]) was lower than that of each of the comparison models, and none of the comparison models’ confidence intervals overlapped (see Table 1). This indicates the seven-factor measurement model was a significantly better fit to the data than each comparison model. Chi-square difference tests (see Table 1) echoed this finding.
We assessed discriminant validity by comparing the square of the correlation between construct pairs with the average of each pair’s average variance extracted (AVE). Evidence for discriminant validity is present when the average AVE between two constructs is greater than their squared correlation (Fornell & Larcker, 1981). Construct pairs’ average AVE ranged from 0.68 to 0.93. Squared correlations between construct pairs ranged from 0.07 to 0.52. Thus, this comparison yielded evidence of discriminant validity, with all pairs of constructs exhibiting an average AVE that was greater than their squared correlation. We continued with the analysis confident that the measurement model employed yielded acceptable discriminant validity and represented the best potential fit to the data, especially in the context of a study examining potentially representative sources and targets (i.e., supervisors and organizations).
Study 1 Results
Descriptive Statistics and Correlations
Means, Standard Deviations, and Zero-Order Bivariate Correlations among Study 1 Variables.
Note. Statistical tests are based on two-tailed tests (α = .05). Cronbach’s α presented on diagonal. M = mean. SD = standard deviation.
N = 445. * p < .05. ** p < .01. *** p < .001.
Hypothesis Testing
Study 1 Regression Results Predicting Perceived Supervisor Obstruction and Perceived Organizational Obstruction.
Note. Variables were centered about their means prior to creating interaction terms.
n = 445. * p < .05. ** p < .01. *** p < .001. b = unstandardized coefficient. B = standardized coefficient. SE = standard error. SOE = supervisor organizational embodiment.
Regions of significance for the effect of organizational unfairness on supervisor obstruction at values of supervisor organizational embodiment (H1a).
Study 1 Regression Results Predicting Supervisor-Directed Destructive Voice and Organization-Directed Destructive Voice.
Note. Variables were centered about their means prior to creating interaction terms.
n = 445. * p < .05. ** p < .01. *** p < .001. b = unstandardized coefficient. B = standardized coefficient. SE = standard error. SU = supervisor unfairness. OU = organizational unfairness. SOE = supervisor organizational embodiment. PSO = perceived supervisor obstruction. POO = perceived organizational obstruction.
Study 1 Conditional Indirect Effect of Organizational Unfairness on Supervisor-Directed Destructive Voice and Conditional Indirect Effect of Supervisor Unfairness on Organization-Directed Destructive Voice.
Note. n = 445. SOE = supervisor organizational embodiment. B = standardized coefficient. SE = standard error. Interaction terms were centered about their mean to aid in interpretation.
* p < .05. ** p < .01. *** p < .001.
Regions of significance for the indirect effect of organizational unfairness on supervisor-directed destructive voice through supervisor obstruction at values of supervisor organizational embodiment (H3a).
Hypothesis 3b asserted that the indirect effect of supervisor unfairness on organization-directed destructive voice through perceived organizational obstruction was dependent on SOE. We tested this hypothesis in the same manner as Hypothesis 3a, and results are presented in Table 5. The 95% confidence interval for the index of moderated mediation included zero (index = .01, 95% CI [-.012, .034]). Thus, Study 1 failed to provide support for Hypothesis 3b.
Study 2
The results from Study 1 provided some support for our hypothesized model when tested in a field study with a heterogeneous sample. We conducted a second study using experimental vignette methodology (Aguinis & Bradley, 2014) in which we manipulated organizational (supervisor) unfairness and SOE in order to extend our findings and better assess causal relationships. A vignette is “a short, carefully constructed description of a person, object, or situation, representing a systematic combination of characteristics” (Atzmüller & Steiner, 2010, p. 128). Experimental vignette methodology allows for strong internal validity and has been used in recent management and organizational behavior research (e.g., Tse, Lam, Gu, & Lin, 2018). We followed Aguinis and Bradley’s (2014) best practice recommendations for study design and analysis, with a focus on providing contextual immersion while avoiding manipulation of potentially confounding factors.
Study 2 Method
Vignette Study
We employed a 2 × 2 × 2 factorial design which included high and low organizational unfairness, supervisor unfairness, and SOE. Each respondent was randomly assigned to one of the eight conditions, then presented with survey questions assessing manipulation checks, perceived organizational and supervisor obstruction, and intent to perform destructive voice directed toward the organization and the supervisor. Each vignette began with a standard opening statement and visual presentation of their hypothetical place of work, and each vignette ended with a standard closing statement meant to prompt reflection. A visual flowchart representation of each of the eight presented vignettes can be found in Figure 4. Flowchart presentation of vignette. Vignette flow is indicated with numbers preceding the bullet points. There are 5 total statements. Statements 2, 3, and 4 are differentiated (2×2×2), resulting in 8 different conditions.
A sample separate from Study 1 was recruited from MTurk (n = 845). Respondents averaged 36.34 years of age and 58% were male. Respondents were paid $1 USD and were fairly evenly distributed across conditions (ranging from n = 103–109).
Organizational Unfairness Manipulation
We considered the key fairness elements of could have and should have (Folger & Cropanzano, 2001), as well as items found in Colquitt et al.’s (2015) fairness measurement to manipulate organizational (un)fairness in the vignettes. We chose to describe a collective perception of organizational (un)fairness while avoiding specific examples, which would threaten the validity of the study by unnecessarily expanding its scope. We also standardized respondents’ tenure with the company (i.e., 1 year) to provide the perception that adequate time had transpired for them to assess the organization’s (un)fairness. A neutral, fictitious organization name was used (i.e., MidFirm).
Supervisor Unfairness Manipulation
The manipulation for supervisor (un)fairness was similar to that of organizational unfairness. Again, the fairness accountability elements of could have and should have were presented, along with coworkers’ collective perception of supervisor (un)fairness. A gender-neutral supervisor name was used (i.e., “Pat”; Tse et al., 2018).
Supervisor Organizational Embodiment Manipulation
Supervisor organizational embodiment was manipulated with a brief description of the supervisor’s representativeness of MidFirm’s values and goals. We used a visual representation to strengthen this manipulation, with an illustrated gender-neutral supervisor’s text bubble stating “I do things the MidFirm way!” in the high SOE condition, and “I do things my own way!” in the low SOE condition. Additionally, the illustrated supervisor’s nametag in the high SOE condition included the name of the firm, whereas the illustrated supervisor’s nametag in the low SOE condition did not.
Measures
Organizational and Supervisor Obstruction
Respondents reported their perceived organizational and supervisor obstruction with the same five-item measure used in Study 1 (Gibney et al., 2009). The items were preceded with the statement “If you were in the workplace scenario you just read, to what extent would you feel the following?” The items were followed with a seven-point Likert response scale (1 = “not at all” and 7 = “to a very great extent”). The items exhibited strong reliability (organizational α = .97; supervisor α = .97).
Destructive Voice Toward Organization and Supervisor
Respondents reported their likelihood to perform destructive voice toward their organization and supervisor with the same five-item measure used in Study 1 (Maynes & Podsakoff, 2014). The items were preceded with the statement “If you were in the workplace scenario you just read, how likely would you be to perform the following behavior?” The items were followed by a seven-point Likert response scale (1 = “extremely likely” and 7 = “extremely unlikely”). The items exhibited strong reliability (organizational α = .97; supervisor α = .97). The items were thus averaged and recoded such that higher scores indicated a stronger likelihood to exhibit the destructive voice behavior in order to facilitate interpretation.
Study 2 Analysis
Means, Standard Deviations, and Zero-Order Bivariate Correlations among Study 2 Variables.
Note. Statistical tests are based on two-tailed tests (α = .05). Cronbach’s α presented on diagonal. Cronbach’s α not provided for single-item manipulation checks. M = mean. SD = standard deviation. SOE = supervisor organizational embodiment.
N = 845. * p < .05. ** p < .01. *** p < .001.
Factor Analysis
Study 2 Measurement Model and Comparisons.
Note. N = 845. df = degrees of freedom. CFI = comparative fit index. TLI = Tucker-Lewis index. RMSEA = root mean square error of approximation. * p < .05. ** p < .01. *** p < .001.
Manipulation Checks
We assessed the organizational and supervisor unfairness manipulations with a single item each, “In the workplace scenario that you have just read, how would you describe MidFirm [Pat]?” (1 = “fair” and 7 = “unfair”). The SOE manipulation was assessed with the item, “In the workplace scenario that you have just read, to what extent would you say Pat embodies the organization?” (1 = “not at all” and 7 = “to a great extent”). We assessed manipulation checks with a series of ANOVAs for each manipulation check, and found that respondents in the high organizational unfairness condition reported a significantly higher mean level of organizational unfairness (M = 5.13, SD = 1.87) compared to respondents in the low organizational unfairness condition (M = 2.36, SD = 1.68), F(1, 843) = 512.93 p < .001. Likewise, respondents in the high supervisor unfairness condition reported a significantly higher mean level of supervisor unfairness (M = 5.31, SD = 1.87) compared to those in the low supervisor unfairness condition (M = 2.17, SD = 1.54), F(1, 843) = 709.01, p < .001. Finally, those in the high SOE condition reported a significantly higher mean level of SOE (M = 5.24, SD = 1.66) compared to those in the low SOE condition (M = 2.95, SD = 1.89), F(1, 843) = 350.35 p < .001. Thus, manipulation checks indicated all three vignette manipulations were successful.
Study 2 Results
Hypothesis Testing
Study 2 Multivariate Analysis of Variance and Estimated Marginal Means of Supervisor and Organizational Obstruction among Respondents in High and Low Unfairness and SOE Conditions.
Note. n = 845. SOE = supervisor organizational embodiment. Sup. = supervisor. Org. = organizational.
Estimated marginal means for organizational obstruction at conditions of supervisor unfairness and supervisor organizational embodiment (SOE).
The MANOVA also indicated that the organizational unfairness condition was associated with higher ratings of organizational obstruction (F[1837] = 148.13 p < .001), but not with higher ratings of supervisor obstruction (F[1837] = 2.06 p = .151). On the other hand, the supervisor unfairness condition yielded higher ratings of supervisor obstruction (F[1837] = 269.70 p < .001) as well as higher ratings of organizational obstruction (F[1837] = 5.13 p = .024). The organizational and supervisor unfairness conditions did not combine to predict either source of obstruction, and the SOE condition did not have a significant direct effect on either source of obstruction. To summarize, Study 2 yielded support for Hypothesis 1b but not Hypothesis 1a, whereas Study 1 yielded support for Hypothesis 1a but not Hypothesis 1b.
Study 2 Regression Predicting Supervisor-Directed Destructive Voice and Organization-Directed Destructive Voice.
Note. N = 845. b = unstandardized coefficient. B = standardized coefficient. SE = standard error. SU = supervisor unfairness. OU = organizational unfairness. SOE = supervisor organizational embodiment. POO = perceived organizational obstruction. PSO = perceived supervisor obstruction. * p < .05. ** p < .01. *** p < .001.
Study 2 Conditional Indirect Effect of Organizational Unfairness on Supervisor-Directed Destructive Voice and Conditional Indirect Effect of Supervisor Unfairness on Organization-Directed Destructive Voice.
Note. n = 845. B = standardized coefficient. SE = standard error. SOE = supervisor organizational embodiment. Interaction terms were centered about their mean to aid in interpretation.
* p < .05. ** p < .01. *** p < .001.
The test of Hypothesis 3b (see Table 10) indicated that when controlling for the effects of organizational unfairness and supervisor obstruction, supervisor unfairness did not have a significant indirect effect on organization-directed destructive voice through perceived organizational obstruction in the low SOE condition (B = −.09, 95% CI [-.28, .09]) but did in the high SOE condition (B = .40, 95% CI [.22, .58]). In addition, the bootstrapped confidence intervals for the index of moderated mediation did not include zero (index = .49, 95% CI [.23, .76]), indicating that the indirect effect of supervisor unfairness on organization-directed voice through organizational obstruction was significantly different in high and low SOE conditions. Thus, Hypothesis 3b was supported by the results from Study 2. To summarize, Study 2 yielded significant supervisor-to-organization effects, whereas Study 1 yielded significant results for organization-to-supervisor effects.
Discussion
Our studies build upon prior misalignment research in a number of ways. First, the results of CFA and discriminant validity tests lend additional credence to the idea that employees view their supervisors and organizations as distinct sources of unfairness and obstruction. Additionally, we found that the extent to which the supervisor embodies the organization appears to play a role in the subsequent perceived source of work obstruction and target of employee destructive voice, even while controlling for source-target aligned effects. In other words, our studies provide preliminary evidence that high SOE may be a boundary condition which moderates employees’ supervisor-directed responses to organizational unfairness (Study 1) as well as organization-directed responses to supervisor unfairness (Study 2). Overall, our results provide evidence of explainable relationship differentiation and source-target misalignment relationships between organizational unfairness and supervisor-directed destructive voice, and vice versa.
However, our studies provide differing evidence of the direction in which social exchange misalignment might flow; Study 1 supports organization-to-supervisor effects, whereas Study 2 supports supervisor-to-organization effects. This is consistent with Lavelle et al.’s (2007) identification of the relative strength of source-target alignment and speaks to the difficulty of detecting misaligned effects when controlling for source-target aligned effects. These discrepant findings may be due to study design differences, as Study 1 surveyed employees about personal work experiences whereas Study 2 surveyed employees regarding their hypothetical responses to brief vignettes. Perhaps the impact of personally experienced unfairness and obstruction manifests differently from salient hypothetical experiences when it comes to supervisor versus organizational effects. Future research could ask respondents to recall personal experiences which are similar to vignette descriptions in order to increase validity. In addition, experiential indicators of SOE may be broader than those captured in a vignette. Our study was the first to our knowledge which has manipulated SOE with experimental vignette methodology. Future qualitative and longitudinal research regarding the development of SOE perceptions would aid in strengthening vignette representation of SOE.
Additionally, our study results suggest different study methodologies may yield differences in evidenced misaligned social exchange relationships. We encourage future research to build upon our findings by examining misaligned social exchange relationships from multiple methodological angles. In summary, we find our discrepant findings to be more inspiring than limiting, as it suggests that social exchange misalignment research can do more to better understand these effects and the conditions under which they occur. Below, we reflect on what our studies can tell us about social exchange source-target misalignment, consider study limitations, and suggest how future research might add to the present contribution.
Research Implications
This research contributes to the literature in four key ways. First, our studies indicate the importance of including source-target aligned controls when studying how, when, and why employees direct their destructive voice toward different organizational parties. Extant research theoretically recognizes that source-target misalignment is possible (Lavelle et al., 2007), and meta-analysis shows that source-target misalignment exists and is worthy of more research attention (Colquitt et al., 2013; Rupp et al., 2014). However, empirical studies that examine source-target misalignment with a social exchange lens have largely failed to control for the effects of aligned social exchange variables (i.e., Eisenberger et al., 2014; Mackey et al., 2018; Shoss et al., 2013). Although our source-target misaligned indirect effect results varied between studies, the effects we found were significant in the presence of source-target aligned control variables. Thus, our studies communicate that previously reported aligned and misaligned source-target effects may not be as strong as previously thought when considered in a more fully controlled model.
Second, our results contribute to the social exchange literature by providing evidence of the extent to which organizational (supervisor) unfairness yields subsequent supervisor (organizational) obstruction. By focusing on these relationships while controlling for the aligned effects, this research is well suited to identify fairness blame and representation (Folger & Cropanzano, 2001) as key components in explaining misaligned social exchange processes. In addition, we considered the negatively valenced construct of obstruction as potentially stemming from the supervisor as well as the organization (Gibney et al., 2009). In doing so, we hope our studies provide a starting point for future research to consider employees’ misaligned social exchange processes both from organization-to-supervisor and from supervisor-to-organization.
Third, our studies expand upon how SOE might impact misalignment. The generalization effect of SOE has been established (Eisenberg et al., 2010) and has received support in empirical social exchange models (i.e., Mackey et al., 2018; Shoss et al., 2013). However, the idea that high SOE perceptions may be a condition for supervisor-to-organization misalignment (i.e., generalization effects) as well as organization-to-supervisor misalignment (i.e., particularization effects) broadens SOE’s potential role in social exchange models because it suggests employees not only view their supervisor’s actions as representative of the organization, but they may also view their organization’s actions as representative of their supervisor. Overall, we anticipate that increased attention to both the generalization and particularization effects of SOE will result in a better understanding of why employees may appear to misdirect their responses to (un)fair treatment.
Finally, we contribute to an emerging destructive voice literature begun by Maynes and Podsakoff’s (2014) initial work to help “facilitate research on the types of voice that have been relatively ignored in past research” (p. 105). Even though Maynes and Podsakoff’s article has received significant attention (i.e., 444 citations as of December 2021), only a handful of studies have investigated destructive voice. These studies examine destructive voice as either specifically supervisor-directed (Mackey et al., 2020; Puyenbroeck, Stouten, & Vande Broek, 2019) or organization-directed (Unler & Caliskan, 2017). In the present research, we provide a unique point of view that considers potentially harmful outcomes for supervisors that strongly embody an unfair organization as opposed to focusing solely on supervisor-to-organization effects. Additionally, we provide for obstruction’s role in translating perceived unfairness into destructive voice, even across sources.
Practical Implications
The potential for source-target misalignment is practically relevant. Extant practitioner literature communicates the effectiveness of various forms of supervisor organizational embodiment, such as extreme ownership (Willink & Babin, 2017) and ambassadorship (Studer, 2020). On the other hand, the title of an online article from the Society for Human Resource Management (SHRM) advises managers to help their company to “avoid employee lawsuits by being fair” (Nagele-Piazza, 2017). It appears that the popular press recognizes the impact supervisors can have on employee perceptions of their organization, but less attention is given to the potential managerial downsides of embodying an organization’s values when employees perceive the organization itself to be unfair.
The present research findings highlight the potential managerial downsides of representing an unfair organization, as doing so could yield supervisor-directed destructive voice that is not only interpersonally hurtful but which could also extinguish the organizational benefits typically yielded by ownership or ambassadorship. By reviewing employee grievances with a focus on descriptions of when the organization could have or should have done something differently, managers can gain an awareness of their employees’ perceptions of organizational unfairness. Our research suggests that managers should beware of strong organizational embodiment in situations where employees perceive the organization to be unfair. Similarly, managers of organizations which are widely accepted to be fair would be wise to adopt the organization’s values in a way that is visible to their employees.
Likewise, we echo Nagele-Piazza’s (2017) suggestion that upper leadership should encourage fair practices among subordinate leaders to avoid negative employee behaviors directed toward the company. Onboarding and integration for new leaders represents a unique opportunity to emphasize managerial fairness as an organizational priority. This offers the dual potential benefit of influencing perceptions of both managerial fairness and organizational fairness. This could be especially crucial in small organizations or situations in which employee perceptions of the organization are otherwise strongly reliant on managers’ actions, such as remote or isolated work groups.
Limitations and Future Research Directions
We contextualize the impact of our limitations on our findings below. First, all of the variables in both studies were self-reported. Self-report data was necessary because each of the constructs used measured employee perception. However, self-report data can be an issue when investigating sensitive constructs, such as those dealing with negative behaviors at work, because employees may deliberately distort responses (Dalal & Hakel, 2016). For example, destructive voice reports are at risk of being either minimized by the self-reporter or distorted by the perceiver (Maynes & Podsakoff, 2014). Although we took precautions in Study 1 to alleviate self-report concerns (e.g., collecting data across multiple time periods, using agreement and frequency response scales), future research may consider implementing other reports to further validate the findings presented here. Additionally, we failed to include an explicit representation of the “would” element of counterfactual thinking that occurs in assessments of unfairness (Folger & Cropanzano, 2001), although we believe this element of counterfactual thought is included implicitly in the vignettes as they are worded. Future vignette studies may consider including conditional statements describing “would” conclusions when manipulating perceptions of unfairness.
Second, both samples were restricted to people living in the United States. Future studies might examine differences attributable to other cultural contexts. Such differences have been examined with constructive voice (Chelminski & Coulter, 2007), but not destructive voice. Third, our theorizing and measurement of unfairness did not specify distinct types of unfairness. We rather sought to consider broad unfairness and the role of this perception in driving subsequent social exchange processes. Different types of unfairness or injustice such as distributive, procedural, informational, and interactional unfairness may yield important insights about the extent to which these different perceptions affect social exchange processes within and across employees’ relationships with their organizations and supervisors (Colquitt et al., 2013). For example, some employees may perceive their supervisors to be more responsible for interactional unfairness and less responsible for informational unfairness. Future research could expand upon our findings by identifying specific types of unfairness and considering the extent to which employee respondents’ work contexts create boundaries in which supervisors and organizations “do” unfairness in different ways.
In addition, our study did not account for potential differences in SOE that may be related to differences in organizational level. It may be the case that higher level management yields higher SOE among their subordinates to the extent that these managers have more influence and authority (Eisenberger et al., 2010). Our study also did not directly examine the potential bidirectional link between perceived supervisor and organizational obstruction, although we controlled for the presence of both in our full analytical models. Future research could consider examining the extent to which obstruction from one source predicts the perceived presence of obstruction from other sources, as well as the potential role of SOE in these relationships.
Finally, we call for future research to further investigate potential source-target misalignment in the organizational cynicism research space. Organizational cynicism is a negative general attitude (with cognitive, affective, and behavioral elements) that employees feel toward their organizations which reflect perceptions of deceitful treatment (Wilkerson et al., 2008). Because organizational cynicism shares some conceptual space with our focal research model, we suggest future research can better understand its effects on supervisor-directed attitudes and behaviors (and vice versa) by considering the potential moderating role of SOE and controlling for source-target alignment effects. Doing so may reveal similar generalization and particularization effects found in our current studies.
Conclusion
Although employees differentiate between their relationships with supervisors and organizations, misalignment effects can occur (Colquitt et al., 2013; Rupp et al., 2014). This source-target misalignment is especially concerning when it leads to overly hurtful or critical comments directed toward the other party. The present research advances the conversation about employees’ source-target misaligned destructive voice reactions to workplace unfairness. The results from our investigation into source-target misalignment effects provide nuanced insights that are necessary to understand before the relationship differentiation literature can move forward. We hope our findings inspire an active research dialog about how and why employees differentiate between their supervisors and organizations when responding to different sources of workplace unfairness.
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.
Associate Editor: Luke Zhu