Safety Nudged or Empowerment Fueled? Leadership Mechanisms and Boundary Condition for Follower’s Adaptive Communication Behaviors During the COVID-19 Pandemic Crisis in China

ABSTRACT

Incorporating social information processing theory and literature on organizational learning, this study explores the process approach to internal communication by theorizing and empirically testing a framework of employees’ adaptive communication amidst the COVID-19 pandemic crisis. A cross-sectional survey on full-time employees (n = 983) in Mainland China demonstrated how followers’ perception of leadership inclusiveness was related to the two dimensions of adaptive communication behaviors (i.e. internal voicing and external scouting). Results also showed the two mediating appraisals of inclusive leadership, i.e. psychological empowerment and psychological safety, functioned differently for followers with lower-level positions versus with higher-level ones. Theoretical and practical implications on internal communication, crisis communication, and public relations were elaborated in detail.

KEYWORDS:

• Adaptive communication behavior
• inclusive leadership
• internal communication
• pandemic crisis

Introduction

Unlike other pandemics (e.g., H1N1, Ebola), the COVID-19 posed threats not only to public health on a societal level but also to the functioning and existence of organizations from different walks of life around the globe (Ruck & Men, 2021). Organizational crisis responses have far exceeded the scopes of governments and hospitals (Diers-Lawson et al., 2021). Virtually every type of organization had to face issues and challenges on adjusting teamwork processes if they had to maintain operations (J. Li et al., 2021). These challenges were especially difficult for the internal stakeholders (i.e., employees), for their organizations’ ability to provide instructing and adjusting information to them was hindered due to the “sticky” nature of the pandemic (Coombs et al., 2020). Under contexts like the workplace in China, pandemic-related repercussions were prevalent to Chinese workers, partly due to the nationwide zero-COVID control measures and policies, such as lockdowns, social distancing, community PCR tests, mask and quarantine mandates, and information censorship (e.g., T. Li et al., 2023; Qian & Fan, 2020). Stress, anxieties, and burnouts appeared frequently during lockdowns, as physical movements were restricted for indefinite periods (Zhou et al., 2023). Long-existing issues got amplified such as loss of vitality in performing tasks (Binyamin & Brender-Ilan, 2018), low resource appraisals (Semmer et al., 2021), and blurred work – life boundaries (Paustian-Underdahl et al., 2016).

In academia, public relations and internal communication scholars have looked for solutions to these dilemmas. On the one hand, the conventional approach prescribed what the organization should do-two-way symmetrical communication and its role in enhancing employee outcomes such as engagement (e.g., Sun et al., 2023) and well-being (e.g., Qin & Men, 2023). On the other hand, a less functionalist approach recognized that virtually no managers could disseminate the “one best answer” due to the unpredictable crisis environment, not to mention ushering a transition back into a normal non-crisis state (Doshi et al., 2021; Heide & Simonsson, 2021). Admittedly, the senior management is still responsible for providing crisis responses to external stakeholders, but all individual coworkers inside the organization – leaders and followers alike – had to adapt to the pandemic crisis via their own communication, interaction, and socialization (Einwiller et al., 2021; Heide & Simonsson, 2015; Strandberg & Vigsø, 2016).

This study echoes with the latter stream of research on adapting and learning. We examine employee followers’ adaptive communication and how inclusive leaders promote such behavior amidst the COVID-19 pandemic crisis in China. To construct the theoretical framework, this study draws from organizational learning literature (e.g., Gnyawali & Stewart, 2003) and social information processing theory (Salancik & Pfeffer, 1978). While the former overarches our conceptualization of adaptive communication, the latter allows us to pinpoint the antecedent and the mechanisms through which to induce adaptive communication. As a synthesis of these two perspectives, we propose that inclusive leadership fosters followers’ adaptive communication via two cognitive pathways, i.e., psychological empowerment and psychological safety. Using cross-sectional survey data collected in China, we substantiate relationships among perceived leadership inclusiveness, psychological empowerment, psychological safety, and the two dimensions of adaptive communication behavior (i.e., internal voicing and external scouting) during the COVID-19 pandemic crisis. We also identify employee’s status in the organizational hierarchy as a boundary condition for the proposed model.

Overall, our study contributes to theory building of internal communication and public relations in three ways. First, it adds to the growing research on internal crisis communication, which advocates explanations of why and how employees communicate upon the omnipresent threat of a crisis (Heide & Simonsson, 2015, 2021; Johansen et al., 2012). By examining employees’ internal voicing and external scouting during the COVID-19 pandemic crisis, this study regards employees as micro-level actors that socialize with their immediate work environment with some degrees of agency (Salancik & Pfeffer, 1978). As such, we look beyond the conventional sense of internal communication, i.e., centralized administration of information, which regards employees as a “public” separate from management processes (e.g., Kang & Sung, 2017; Lee, 2021). Second, our study advances an internal/employee angle for crisis research in the broader field of public relations (e.g., Einwiller et al., 2021; Y. Kim, 2018; Y. Kim & Lim, 2020). We achieve this by unraveling the psychosocial mechanisms through which supervisors’ inclusive leadership fosters employee followers’ adaptive communication behaviors during the COVID-19 pandemic. Whereas past crisis management research focused on reputational recovery from an episodic crisis event in the eyes of external stakeholders (see Coombs, 2010; cf.; Heide & Simonsson, 2015), our study delineates how internal members enacted on-going, adjustive learning during a chronic, exogenous crisis period (Gnyawali & Stewart, 2003; Kersten, 2005; Roux-Dufort, 2007). Third, our study answers the recent call for diverse cultural perspectives in public relations (Sriramesh et al., 2013; Verčič, 2020). By illustrating how employee followers’ status in the organizational structure alters their cognitive processing of leader inclusiveness, we reveal the unique socio-cultural contexts of employee communication in the Chinese organizations.

Literature review and hypotheses development

The process approach toward internal communication during crises

Internal communication is also known as “employee relations” or “internal public relations” within public relations scholarship (Lee & Yue, 2020). Men and Bowen (2017) defined internal relations as “managing interdependence and building mutually beneficial relationships between the organization and its employees.” (p. 12). Because internal communication scholars were mostly influenced by the Excellence Theory of Public Relations (Grunig & Hunt, 1984), constructs like employee – organization relationship (EOR) championed by Grunig have become prevalent in internal communication research (Lee & Yue, 2020). However, one limitation is the treatment of employees as one organizational “public,” a separate entity from management processes. The presumptions are that organization equals senior management and that an ideal mode of internal communication equals information exchanges without the strains of status asymmetry (e.g., Kang & Sung, 2017).

An alternative stream of research, however, revolves around the complexity of employees’ day-to-day communication processes, i.e., the specific mechanisms reflected in the interactions between leaders and followers (e.g., G. Fairhurst & Connaughton, 2014; Heide & Simonsson, 2015; Ruck et al., 2017; Tourish, 2014). Heide and Simonsson (2021) summarized this alternative as a process approach to internal communication. During the COVID-19 pandemic crisis, this theoretical approach is crucial, because it is not adequate anymore to treat all employees as quiescent recipients of information after a crisis event (Heide & Simonsson, 2015, 2020). Instead, different communication flows emerged. Leaders/managers may become facilitators to overcome cognitive barriers by inviting inputs and facilitating collaborations (Edmondson & Moingeon, 1998). Followers/subordinates can act as producers of solutions and progress under the guidance of a leader (Grint, 2005; Tourish, 2014).

Employee’s adaptive communication behavior during the pandemic crisis

As the process approach theorizes, internal communication during turbulent times is not merely top-down information delivery. Upon a chaotic and yet high-stake reality, employees do not just wait for the organizations’ remedy. Rather, they try to provide, collect, and/or exchange information (Heide & Simonsson, 2021; Wall et al., 2002). In this regard, employees’ own communication behavior is pivotal because failure to do so could threaten not only the enactment of professional identity but also the wellbeing of organization (Cheney et al., 2014; K. Weick et al., 2005).

Aligning with this train of thoughts, this study examines adaptive communication. It refers to employees’ proactive conveyance and retrieval of information, which intends to help themselves, others, and the organization to navigate the confusion, ambiguity, and disorientation induced by a chronic and exogenous crisis. Part of theoretical root of adaptive communication behavior is in the organizational learning literature (e.g., Edmondson, 2004; Edmondson & Moingeon, 1998; Gnyawali & Stewart, 2003; Schein, 1999). This literature contends that employees contribute to the organization’s knowledge pool by acquiring, storing, sensemaking, distributing, and exchanging relevant information about difficult situations like crises (Simonsson & Heide, 2018; van Zoonen et al., 2022). This way, they help themselves to accumulate practical knowledge and the organization to sketch a viable work-in-progress manual for the challenges brought by crises (Gnyawali & Stewart, 2003; K. E. Weick, 1988).

Two dimensions of adaptive communication: internal voicing and external scouting

We theorize adaptive communication behavior as having two dimensions: internal voicing and external scouting. Internal voicing refers to an employee’s proactive conveyance of information (such as raising concerns, expressing opinions, and sharing tips) inside the work team. External scouting refers to an employee’s proactive retrieval of information (such as news, feedback, and complaints) from external sources to the organization.

The conceptualization of internal voicing takes inspirations from prior research on proactive communication. There were two primary perspectives. First, subordinate employee’s communication inside the organization was often considered extra-role and citizenry in nature (e.g., Detert & Burris, 2007; Grant & Ashford, 2008; Morrison et al., 2011). Contrary to silent employees, those proactive ones engage voluntarily as they anticipate and strive to alter the status quo appearing stagnant (Milliken et al., 2003; Parker et al., 2010). Second, previous research on proactive communication emphasized different tenors of the content, i.e., what is communicated can either be promotive or prohibitive (Liang et al., 2012). Promotive messages are constructive and generative forms of content, whereas prohibitive messages are dissenting and oppositional to the management (Kassing, 1997, 1998; Kassing & Armstrong, 2002; Y. Kim & Lim, 2020). However, internal voicing does not intend to specify an in-/extra-role typology nor to discriminate between message contents. Internal voicing is centered around how employees generate insights inside their organization. Message of internal voicing may be in-role or extra-role; its content may be promotive or prohibitive, but the purpose is to help the focal employees, others, and/or the organization adapt to difficulties.

J. Kim and Rhee’s (2011) research was the main reference for external scouting. J. Kim and Rhee (2011) coined the internal – external differentiation in employee’s everyday communications. That is, employees either bring information into the organization from outside (i.e., scouting) or send information out from within the organization (i.e., megaphoning). While they included both inward and outward information flows under normal, non-crisis circumstances, this study picks just an employee’s action of gathering insights from external sources (external scouting). Megaphoning is of little relevance to organizational adaptiveness under the context of a chronic, exogenous pandemic crisis, because employees’ role as “information agents” during the turbulent times primarily lies in retrieving useful information for their organization rather than publicizing information on behalf of their organization (Park et al., 2014; Stephens et al., 2020; Verhoeven & Madsen, 2022).

The two dimensions of adaptive communication behavior are related. Here, we argue that more frequent internal voicing increases the likelihood of external scouting. This is mainly because prior research has demonstrated that an organization’s ability to communicate well internally is a prerequisite for cohesive communication with external constituencies (e.g., C. Chen et al., 2018; Mazzei et al., 2012; Simonsson & Heide, 2018; Sommerfeldt & Taylor, 2011). We speculate that the same logic applies to individual employees’ adaptive communication. Employees need to first be able to verbalize crisis-related problems and/or opinions within the organizational boundary, i.e., with coworkers and supervisors, before collecting further information from friends, family, clients, competitors, or regulators for the organization’s improvement (Y. Kim, 2018; Y. Kim & Lim, 2020; Park et al., 2014).

Now that the adaptive communication concept has been introduced, in the following sections, we will discuss inclusive leadership as an antecedent to adaptive communication and two psychological mechanisms through which this effect occurs by applying the Social Information Processing Theory.

Perceived leadership inclusiveness as a key antecedent to adaptive communication

Around the recent times of uncertainties and chaos in the COVID-19 pandemic, a growing amount of research contended that employee followers derive appraisals of the internal climate and subsequent communication behaviors from their interactions with and the impacts of leaders (e.g., Santoso et al., 2022; Sun et al., 2023; Wu et al., 2021). We select the main antecedent to employees’ adaptive communication – inclusive leadership – based on Social Information Processing Theory (SIP; Salancik & Pfeffer, 1978). According to SIP, a focal individual relies heavily on the social information available to them to form perceptions and make behavioral decisions in the workplace (Salancik & Pfeffer, 1978). Leadership provides one of the most prevalent social contexts (G. Fairhurst & Connaughton, 2014; Morrison, 2011; Tourish, 2014). In the rest of this section, we elaborate what inclusive leadership is and why it is the key to employee’s adaptive communication during the COVID-19 pandemic crisis.

Inclusive leadership refers to the words and actions by leaders that indicate invitation and appreciation for others’ contributions (Nembhard & Edmondson, 2006). In organizational processes facilitated by inclusive leaders, followers are likely to be valued for their uniqueness while feeling part of the group (Randel et al., 2018; L. Shore et al., 2011). It is this perception of a balanced treatment of both uniqueness and belongingness that embodies inclusion in the internal climate (Roberson & Perry, 2022). Because the conceptual emphasis appears on the experiences of the focal employee followers, we use the two terms, “perceived leadership inclusiveness” and “inclusive leadership” interchangeably (Carmeli et al., 2010).

Inclusive leadership as a construct is related to transformational leadership in that the leader is willing to delegate authority to followers (Aldoory & Toth, 2004). However, inclusive leadership focuses on facilitating processes with proactive feedback (Edmondson, 1999), whereas transformational leadership emphasizes a leader’s visionary side that inspires followers to create their own solutions (Men, 2014). Another related construct to inclusive leadership is servant leadership, as both underline the humbling choices made by a leader to renounce a superior status (Sendjaya et al., 2008). However, a hallmark of servant leadership is rather the willingness to help followers in need vis-à-vis the concern for self-interests (van Dierendonck, 2011). This quality is different from an inclusive leader, who solicits followers’ input into the group decision-making. Inclusive leadership is also related to management openness (Detert & Burris, 2007), as both highlight subordinates’ perceptions that the supervisor listens to concerns, but perceived leadership inclusiveness implies an individual employee’s awareness that the supervisor values their contributions by adopting a participative way of decision making rather than seeking unanimity in an authoritative manner (Nembhard & Edmondson, 2006).

As evidenced in empirical research, inclusive leadership is positively associated with both dimensions of adaptive communication (i.e., internal voicing and external scouting). First, employee followers perceiving a higher degree of leadership inclusiveness would be more likely to engage in voicing within their work teams, because the idea of inclusion fundamentally addresses the difficulty of a lower social position and to overcome communication barriers brought by it (Nembhard & Edmondson, 2006; Ng & Feldman, 2012; L. Shore et al., 2011). Under inclusive leaders’ supervision, employee followers would become motivated to take risks in showing their concerns about the organization (Carnevale et al., 2017; Morrison, 2011). Thus, pointing out uncertainties, doubts, concerns, and even inefficiencies related to crisis internally would be regarded as precursors – rather than impediments – to critical adjustments (Grint, 2005; Morrison, 2011; van den Heuvel et al., 2013). Second, under the influence of inclusive leadership, employee followers would also expect to increase in external scouting (Lee, 2022; Lee et al., 2018). As an informal, nonspecialist form of environment scanning, external scouting represents employees’ discretionary effort to help their organization decrease the cost of information gathering (Y. J. Kim & Rhee, 2011; Stoffels, 1994; Verhoeven & Madsen, 2022). Upon turbulent times like the COVID pandemic crisis, inclusive leaders acknowledge that a higher status in the organization does not necessarily presume an absolute superiority in possessing knowledge (Doshi et al., 2021). Because of such humble mind-set, inclusive leaders would stay open to any relevant data, trends, policies, or practical wisdom outside the scope of the organization’s existing knowledge protocol (Lee, 2022; Park et al., 2014). In this sense, inclusive leadership enables employee followers to creatively leverage their personal sources external to the organization (Park et al., 2014; L. M. Shore & Chung, 2022). In a word, organizations tend to manage challenging times resiliently if members perceive that leadership invites input from them (Einwiller et al., 2021; Neill & Bowen, 2021).

So far, we have established the direct effect of perceived leadership inclusiveness on the two dimensions of adaptive communication respectively. In the previous section, we have also elaborated the direct effect of internal voicing on external scouting. Now we further propose an indirect effect of perceived leadership inclusiveness on external scouting via internal voicing. This proposition is also based on social information processing’s thesis (Salancik & Pfeffer, 1978). For employee followers to act as “scouts” for the organization’s improvement, inclusive leaders first familiarize their followers with an internal climate where lateral or upward influences are viable (L. M. Shore & Chung, 2022; Simonsson & Heide, 2018). Once employee followers get used to identifying and discussing potential pitfalls within their immediate work environment, they will then consider expanding their efforts to generating informational assets for the organization’s adaptiveness (Lee & Kim, 2017; Park et al., 2014). Based on the above arguments, the following hypotheses are proposed.

H₁: Perceived leadership inclusiveness is positively associated with follower’s (a) internal voicing and (b) external scouting.

H₂: Follower’s internal vocalizing positively mediates the association of perceived leadership inclusiveness with external scouting.

Psychological mechanisms for employees’ adaptive communication

Social information processing theory (Salancik & Pfeffer, 1978) also suggests that it is crucial to understand how a worker decodes meanings out of the social environment as cognitive antecedents to his/her corresponding behaviors. We propose two possible ways through which employees decode the meanings of inclusive leadership during the COVID-19 pandemic crisis. One is a heightened control appraisal, and the other is low perceived costs (Parker et al., 2010). The former is psychological empowerment and the latter psychological safety. We will explain the two mechanisms one by one.

Psychological empowerment: the “heightened control” mechanism. G. M. Spreitzer (2008) defined psychological empowerment as “a set of psychological states that are necessary for individuals to feel a sense of control in relation to their work” (p. 56). There are four subcategories pertinent to a worker’s empowerment: competence, impact, meaning, and self-determination (G. Spreitzer, 1995). Competence refers to beliefs about self-efficacy to successfully perform a task (Bandura, 1989). Impact refers to the degree to which an individual believes their work can make a difference in fulfilling the outcomes of the work unit (Thomas & Velthouse, 1990). Meaning refers to the centrality of an occupation or task in their own cognitive schema of values (Hackman & Oldham, 1980). Self-determination refers to a sense of autonomy in the initiation, regulation, and continuation of work behaviors (Deci et al., 1989). The four dimensions jointly account for the unitary psychological empowerment construct (Seibert et al., 2011).

From a SIP perspective, employee followers who experienced inclusive leadership should increase appraisals of empowerment (Edwards & Collinson, 2002; Maynard et al., 2012). Since inclusive leaders openly invite followers’ critiques and suggestions, such actions send a signal to followers that their contributions can be impactful toward workgroup processes (Venkataramani & Tangirala, 2010). As all team members see the viability of mutual persuasion rather than blind conformity, the internal climate cultivated by an inclusive leader thus helps diminish status difference (Z. Chen & Aryee, 2007; G. T. Fairhurst & Chandler, 1989). This enhances the followers’ sense of control over their work activities (Liden et al., 2000; Nishii & Mayer, 2009). Through ongoing interactions with inclusive leadership, followers foster their own meaning, confidence, and purpose about their jobs (Frazier & Fainshmidt, 2012; Grant, 2008). In turn, those psychologically empowered employees are more likely to speak up at work and make upward influences (Cunningham et al., 2002; Tangirala & Ramanujam, 2012; Younas et al., 2022)

Since the essence of empowerment lies in its role as a bridge between inclusive leadership and employee involvement, this study proposes the hypotheses about follower’s psychological empowerment both as a simple mediator and a serial mediator in the framework:

H₃a: Follower’s psychological empowerment positively mediates the association of perceived leadership inclusiveness with external scouting.

H₃b: Follower’s psychological empowerment and internal vocalizing function as positive serial mediators between perceived leadership inclusiveness and external scouting.

Psychological safety: the “lowered cost” mechanism. Edmondson (1999) defined psychological safety as a belief that the workplace is “safe for interpersonal risk taking” (p. 354). She contended that the conceptual focus was on perception of minimized cost for candor; or freedom from concerns over potential embarrassment on self-image, status, or career (Edmondson, 1999; Edmondson & Lei, 2014; Newman et al., 2017). It can appear unwelcome for a follower to voice issues or concerns, but an inclusive team leader helps mitigate it in such a way that teammates foresee fewer costs from raising these ideas (Edmondson, 1999, 2003). To employees, such cost reduction means a mental switch from being nervous about upsetting others to the perception of a non-threatening climate (Morrison & Rothman, 2009; West, 1990). Employees thereby begin communicating willingly with coworkers or supervisors about work issues (Krone, 1992).

Prior literature has established this position of psychological safety as a positive mediator to the association between leadership inclusiveness and follower’s communication behaviors. For example, Maitlis et al. (2013) concurred that safe employees are more likely to become idea generators with less fear of public admission of confusion, uncertainty, or previous errors. Carmeli and colleagues (2009; 2014) confirmed that inclusion – induced psychological safety significantly contributed to the candor about what needs improving and how to do it within the organization. Based on the above arguments, this study proposes the hypotheses about follower’s psychological safety both as a simple mediator and a serial mediator in the framework:

H₄a: Follower’s psychological safety positively mediates the association of perceived leadership inclusiveness with external scouting.

H₄b: Follower’s psychological safety and internal vocalizing function as positive serial mediators between perceived leadership inclusiveness and external scouting.

In addition, it is noteworthy that psychological safety is distinct from psychological empowerment. They are discrete appraisals of inclusive leadership because empowerment pertains to the cognition about one’s tasks or jobs (Frazier et al., 2017), whereas safety is about the interpersonal environment (Carmeli & Gittell, 2009). Since this study theorizes psychological empowerment and psychological safety as two separate appraisals of inclusive leadership (see Figure 1 for the theoretical framework proposed in this study), one research question arises about comparing their mediating effects:

Figure 1. Proposed theoretical framework.

RQ1: Which of the two parallel psychological appraisals, empowerment or safety, has a larger effect size in mediating the associations between perceived leadership inclusiveness and follower’s adaptive communication behaviors (i.e., internal vocalizing and external scouting)?

Follower’s status in Chinese organizations

Chinese organizations are in general characterized by high power distance, i.e., employees presume the existence of hierarchy and accept unequally distributed power (Hofstede, 1980). At the organizational level, Chinese organizations, as a member of the Chinese society, supposedly respect and obey the government regulations and moral obligations (Y-H. C. Huang et al., 2016; J. Jiang & Wei, 2013). At the individual level, subordinate employees largely anticipate their superiors to make decisions and follow them up without much consultation (X. Huang & van de Vliert, 2003; Kirkman et al., 2009). However, since inclusive leadership to some extent runs contrary to the logic of such high power distance, it begs the question: which group of Chinese employees would be more/less susceptible to leaders’ inclusiveness than others? This distinction is especially important during the pandemic crisis, when vocalizing localized issues and assembling adjusting information are urgently needed for employees to cope with the strict COVID measures in China (Sun et al., 2023).

In prior studies, professional status level was found significant in shaping employees’ cognitive and behavioral processes in response to inclusive leadership (e.g., Ergeneli et al., 2007; Nembhard & Edmondson, 2006; Sanner & Bunderson, 2015). However, we still know little about whether and how employee followers’ status (lower- vs. higher-level positions) could alter the impact of inclusive leadership on adaptive communication in Chinese organizations. For the above reasons, the second research question is raised:

RQ2: Do the inclusive leadership – induced psychological mechanisms differ according to an employee follower’s status in the organizational hierarchy? If so, how?

Method

Procedure and sampling

This study used an online survey to collect data of Mainland Chinese employees, both those with a full-time position and working in a team/group setting (filtered for these conditions). A pilot test was first conducted with 154 survey respondents over three days in early April 2022 to check the wording of the questionnaire. The main survey data collection was completed in May 2022 for six days after the Omicron variant outbreak of COVID-19 in Mainland China. Participants were recruited using the online panel Credamo (a Chinese equivalent to Amazon MTurk). It has over 1.5 million panel registrants. All employee participants worked in major industries (see Appendix 1) were categorized by the National Bureau of Statistics (2021).

The total number of participants was 989. In this sample, the average age was 36.74 (SD = 7.49) and those who were 31–40 years old made up 51.7% (n = 511). Female accounted for 57% (n = 564). The majority (96.4%, n = 953) were college graduates or had completed graduate work. For status, 62.4% (n = 613) identified themselves in lower-level positions (i.e., entry-level, experienced, and supervisory) and 37.6% (n = 370) in higher-level positions (i.e., executives and senior executives). Nearly half (44.7%, n = 442) worked in medium-sized organizations with 300–1,000 employees, followed by those (40.1%, n = 397) in small-sized organizations with 20–300 employees. For geographical regions, 33.6% (n = 334) were based in East China, 19.0% (n = 188) in South China, 13.3% (n = 132) in North China, 13.2% (n = 131) in Central China, 10.7% (n = 106) in Northeast China, 2.8% (n = 27) in Northwest China, and 7.1% (n = 71) Southwest China. For work experience, respondents varied from 0–3 years (6.6%, n = 65), 3–7 years (38.9%, n = 385), 7–10 years (31.5%, n = 312), 10–15 years (14.6%, n = 144), to over 15 years (8.4%, n = 83). Compared with the entire workforce in Mainland China, this sample had more females and appeared much more educated, but it was representative about geographical region and industry sector (National Bureau of Statistics, 2021).

Survey instruments

Since all items for core variables were adapted from existing scales written in English, they were translated into Chinese. To ensure an accurate adaptation of meaning and cultural relevance, a back-translation method was employed (retranslated Chinese into English to check and revise inconsistencies; see Brislin, 1970). Chronbach’s alphas for all scales exceeded the 0.70 threshold (Nunnally, 1978).

Perceived leadership inclusiveness (α = .85) was measured by assessing the extent to which individual employees perceive their direct supervisor in their team as open to others’ contributions, available for problem consultation, and accessible for understanding emerging issues. This was done using nine items adapted from Carmeli et al. (2010). Psychological empowerment (α = .87) captured individual employees’ perceived control over their own work tasks, including meaning, competence, self-determination, and impact. This was measured using eight items adapted from G. Spreitzer (1995). Psychological safety (α = .80) captured individual employees’ sentiment of minimal worry about negative consequences for being straightforward in the interpersonal environments of their team. This was measured using seven items adapted from Edmondson (1999). Internal voicing (α = .83) was measured by assessing how frequently individual employees proactively convey information inside their work team, such as raising concerns, expressing opinions, and sharing knowledge. This was done using six items adapted from van Dyne and LePine (1998). External scouting (α = .85) was measured by assessing how frequently individual employees proactively retrieve information from external sources to the organization, such as news, feedback, and complaints. This was done using seven items adapted from J. Kim and Rhee (2011).

Control variables

Proactive personality refers to a person with a “relatively stable behavioral tendency” to initiate change (Bateman & Crant, 1993, p. 105). Past empirical research demonstrated strong associations between the proactive personality trait and employees’ proactive communication such as voice, issue selling, and feedback seeking (Fuller & Marler, 2009; Wijaya, 2019), especially under stressful environments (Hung et al., 2015). To avoid the potential confounding effect of this variable, employee’s proactive personality was assessed using Seibert et al. (1999) scale (α = .86) and included as a covariate in all analyses.

Perceived severity of the COVID-19 pandemic crisis refers to the extent to which an employee believes the COVID-19 pandemic posed serious threats to the normal functioning of their work life. This perception was measured by a 7-point semantic differential scale (from 1 = not serious at all to 7 = extremely serious; M = 4.53; SD = 1.17). This was controlled in all analyses in this study because crisis communication literature suggested that individuals upon crises tend to develop more behavioral intentions to mitigate harms when they perceive strong threats compared to weak threats (Bakker et al., 2018; Jin, 2010).

Years of work experience and organization size were also treated as covariates in all analyses, for prior research suggested that these occupation-related variables were potentially associated with employees’ proactive communication behaviors at work (e.g., Allen & Griffeth, 1997; Baker, 1991; H. Jiang & Men, 2017; Johansen et al., 2012). Demographics (i.e., gender, age, and education) were also controlled in all analyses.

Hypotheses testing and results

All hypotheses were tested using Model 80 in PROCESS, which can test both parallel and serial mediations (Hayes, 2018). H₁a, and H₁b posited positive direct associations between perceived leadership inclusiveness (X) and internal voicing (M₃) and between X and external scouting (Y). As shown in Table 1, inclusive leadership was a significant and positive predictor of internal voicing (b = .185, SE = .044, CIs = [.098, .272]), supporting H₁a. Besides, perceived severity of pandemic crisis (b = .029, SE = .014, CIs = [.003, .056]), proactive personality (b = .317, SE = .043, CIs = [.234, .401]), age (b = .061, SE = .024, CIs = [.014, .108]), and level in organization (b = .058, SE = .014, CIs = [.030, .086]) all significantly and positively predicted internal voicing. However, inclusive leadership was not significantly related to external scouting (see Table 1). H₁b was not supported. External scouting was instead positively predicted by perceived severity of pandemic crisis (b = .044, SE = .017, CIs = [.011, .077]), proactive personality (b = .192, SE = .055, CIs = [.084, .299]), and age (b = .105, SE = .030, CIs = [.046, .164]). See Appendix 2 for descriptive statistics and correlations among all variables.

Table 1. Unstandardized Ordinary Least Squares (OLS) regression coefficients with confidence intervals estimating external scouting among different employee subgroups.

	Mediators									Outcome Variable
	Psychological Empowerment (M1)			Psychological Safety (M2)			Internal Voicing (M3)			External Scouting (Y)
Entire Sample (N = 983)	Coefficient	95% CI		Coefficient	95% CI		Coefficient	95% CI		Coefficient	95% CI
Inclusive Leadership (X)	0.37^*** (0.03)	0.301	0.436	0.50^*** (0.03)	0.440	0.568	0.19^*** (0.04)	0.098	0.272	0.09 (0.06)	−0.125	0.195
Psychological Empowerment (M1)	–	–	–	–	–	–	0.31^*** (0.04)	0.232	0.378	0.12^* (0.05)	0.023	0.212
Psychological Safety (M2)	–	–	–	–	–	–	0.16^*** (0.04)	0.078	0.231	0.12^* (0.05)	0.024	0.218
Internal Voicing (M3)	–	–	–	–	–	–	–	–	–	0.55^*** (0.04)	0.475	0.632
Constant	−0.17 (0.15)	−0.455	0.122	0.77^*** (0.14)	0.494	1.044	−0.33 (0.17)	−0.665	0.002	–1.28^*** (0.21)	−1.698	−0.861
							R^2 = .64			R^2 = .61
							F(7, 975) = 249.95			F(8, 974) = 186.16
Lower-level Subgroup (n = 613)
Inclusive Leadership (X)	0.39^*** (0.05)	0.299	0.474	0.48^*** (0.04)	0.395	0.556	0.14^* (0.06)	0.028	0.254	0.08 (0.07)	−0.062	0.214
Psychological Empowerment (M1)	–	–	–	–	–	–	0.34^*** (0.05)	0.244	0.430	0.16^** (0.06)	0.044	0.280
Psychological Safety (M2)	–	–	–	–	–	–	0.17^** (0.05)	0.069	0.272	0.05 (0.06)	−0.078	0.170
Internal Voicing (M3)	–	–	–	–	–	–	–	–	–	0.55^*** (0.05)	0.452	0.646
Constant	−0.50^* (0.20)	−0.895	−0.095	1.06^***(0.19)	0.691	1.429	–0.78^** (0.24)	−1.254	−0.307	–1.15^*** (0.30)	−1.730	−0.570
							R^2 = .65			R^2 = .61
							F(7, 605) = 157.75			F(8, 604) = 116.56
Higher-level Subgroup (n = 370)
Inclusive Leadership (X)	0.38^*** (0.05)	0.277	0.490	0.55^*** (0.06)	0.435	0.654	0.38^*** (0.07)	0.248	0.511	0.08 (0.10)	−0.113	0.281
Psychological Empowerment (M1)	–	–	–	–	–	–	0.15^* (0.06)	0.033	0.257	0.02 (0.08)	−0.138	0.184
Psychological Safety (M2)	–	–	–	–	–	–	0.14^** (0.06)	0.036	0.253	0.29^*** (0.08)	0.129	0.444
Internal Voicing (M3)	–	–	–	–	–	–	–	–	–	0.53^*** (0.08)	0.382	0.679
Constant	0.53 (0.29)	−0.039	1.104	0.63^* (0.30)	0.042	1.216	1.08^** (0.31)	0.461	1.690	–1.20^** (0.46)	−2.097	−0.307
							R^2 = .62			R^2 = .58
							F(7, 362) = 84.77			F(8, 361) = 61.07

Note. CI: confidence interval; Standard errors in parentheses; *p < .05, **p < .01, ***p < .001.

H₂ posited a simple mediation of internal voicing (M₃) between perceived leadership inclusiveness (X) and external scouting (Y). Results showed a significant indirect effect of leadership inclusiveness on external scouting through internal voicing (Index Effect = .103, SE = .034, CIs = [.039, .171]). Thus, H₂ was supported. H₃a posited a simple mediation of psychological empowerment (M₁) between perceived leadership inclusiveness (X) and external scouting (Y). Empowerment positively mediated the impact of leadership inclusiveness on external scouting (Index Effect = .043, SE = .022, CIs = [.001, .087]), supporting H₃a. H₄a posited the mediation of psychological safety (M₂) between perceived leadership inclusiveness (X) and external scouting (Y). The indirect effect of leadership inclusiveness on external scouting through psychological safety was also significant (Index Effect = .061, SE = .030, CIs = [.003, .122]). Thus, H₄a was supported. Moreover, as the direct relationship between leadership inclusiveness and external scouting became insignificant when mediators were included (b = .09, SE = .06, CIs = [−.125, .195]), the three parallel mediators––empowerment (M₁), safety (M₂), and internal voicing (M₃) ––all fully mediated the relationship between leadership inclusiveness and external scouting.

H₃b posited a serial mediation of psychological empowerment (M₁) and internal voicing (M₃) on the relationship between perceived leadership inclusiveness (X) and external scouting (Y). The indirect effect was significant (Index Effect = .062, SE = .014, CIs = [.037, .093]), supporting H₃b. H₄b posited a serial mediation of psychological safety (M₂) and internal voicing (M₃) between leadership inclusiveness and external scouting. The indirect effect was significant (Index Effect = .043, SE = .013, CIs = [.020, .072]). H₄b was thus supported.

RQ₁ asked to compare the indirect effect sizes of the two serial paths (X ➔ M₁ ➔ M₃ ➔ Y vs. X ➔ M₂ ➔ M₃ ➔ Y). Because both serial mediations were significant, partially standardized indirect effects were compared (Hayes, 2018). Results showed that a 1-point increase on leadership inclusiveness produced an increase of .064 standard deviations on external scouting through the indirect mechanism featured by empowerment (M₁) and internal voicing (M₃) (SE = .014, CIs = [.038, .094]); yet produced an increase of .044 standard deviations on external scouting through the indirect mechanism featured by safety (M₂) and internal voicing (M₃) (SE = .013, CIs = [.021, .073]). For employee respondents across all levels, empowerment (M₁) thus slightly outperformed psychological safety (M₂) in translating follower’s perceived leadership inclusiveness into first internal voicing and then external scouting.

RQ₂ asked how the entire conceptual model (i.e., inclusive leadership – induced mechanisms behind internal voicing and external scouting) may differ among different employee statuses in the organizational hierarchy. Model 80 of PROCESS was run for each employee subsample (high vs. low level) separately, with a focus on comparing indirect effect coefficients.¹ As shown in Table 2, for lower-level employees, the indirect effect of leadership inclusiveness on external scouting through psychological empowerment (M₁) was significant (Index Effect = .063); also significant through internal voicing (M₃) (Index Effect = .077); but insignificant through psychological safety (M₃). The serial mediation of inclusiveness (X) ➔ empowerment (M₁) ➔ internal voicing (M₃) ➔ external scouting (Y) was significant: Index Effect = .072 as well as the serial mediation of X ➔ safety (M₂) ➔ internal voicing (M₃) ➔ Y: Index Effect = .044 (see Table 2 for details). Given the results of both simple and serial mediations, psychological empowerment (M₁) –– rather than safety (M₂) –– was the prevalent appraisal of inclusive leadership among lower-level employees.

Table 2. Comparison of indirect effects of perceived leadership inclusiveness between the two employee subgroups.

Indirect effect of inclusive leadership for employees positioned at lower levels (entry-level, experienced, and supervisory)				95% CI
		Effect	SE	LL	UL
IndH2	Inclusive leadership ➔ Internal voicing ➔ External scouting	.077	.040	.002	.162
IndH3a	Inclusive leadership ➔ Psychological empowerment ➔ External scouting	.063	.031	.007	.013
IndH4a	Inclusive leadership ➔ Psychological safety ➔ External scouting	.022	.037	−.051	.095
IndH3b	Inclusive leadership ➔ Psychological empowerment ➔ Internal voicing ➔ External scouting	.072	.020	.038	.116
IndH4b	Inclusive leadership ➔ Psychological safety ➔ Internal voicing ➔ External scouting	.044	.016	.016	.081
Indirect effect of inclusive leadership for employees positioned at higher levels (executive and senior executive)
IndH2	Inclusive leadership ➔ Internal voicing ➔ External scouting	.201	.053	.110	.318
IndH3a	Inclusive leadership ➔ Psychological empowerment ➔ External scouting	.009	.036	−.066	.077
IndH4a	Inclusive leadership ➔ Psychological safety ➔ External scouting	.156	.054	.055	.266
IndH3b	Inclusive leadership ➔ Psychological empowerment ➔ Internal voicing ➔ External scouting	.029	.016	.002	.066
IndH4b	Inclusive leadership ➔ Psychological safety ➔ Internal voicing ➔ External scouting	.042	.020	.007	.085

Note. The effect is statistically significant when CI does not include zero.

For higher-management employees, the results were quite different. The indirect effect of leadership inclusiveness on external scouting was significant through psychological safety (M₂): Index Effect = .156; significant through internal voicing (M₃): Index Effect = .201; but insignificant through psychological empowerment (M₁): Index Effect = .009 (see Table 2). The serial mediation of inclusiveness (X) ➔ empowerment (M₁) ➔ internal voicing (M₃) ➔ external scouting (Y) was significant: Index Effect = .029 as well as the serial mediation of X ➔ safety (M₂) ➔ internal voicing (M₃) ➔ Y: Index Effect = .042 (see Table 2). The results suggest that for employees at higher management levels, psychological safety (M₂) was a more prevalent appraisal of inclusive leadership than empowerment (M₁).

Apart from identifying different patterns in indirect effects, we used a z-test on the equality of regression coefficients to compare the differences in direct effects between two employee subsamples (lower n = 613; higher n = 370). This has the advantage of minimal bias when the two subgroups of interest have unequal sample sizes (Paternoster et al., 1998, p. 862).² Results suggest that the direct effect of leadership inclusiveness (X) on internal voicing (M₃) was significantly larger for higher status than for lower status (z = 2.60, p < .01), while the effect of leadership inclusiveness did not differ on external scouting, psychological empowerment, or psychological safety. The direct effect of psychological empowerment (M₁) on internal voicing (M₃) was significantly smaller for higher status than for lower status (z = −2.44, p < .05), but little difference on external scouting. The direct effect of psychological safety (M₂) on external scouting (Y) was significantly larger for higher status than for lower status (z = 2.40, p < .05), but little difference on internal voicing (M₃). The direct effect of internal voicing on external scouting also did not differ significantly. Combining the z-test with the above comparison of indirect effects, the findings demonstrated the empirical soundness to divide the sample into two subgroups with different sample sizes. In other words, the boundary condition of follower’s status existed, suggesting the proposed model worked differentially by employees’ status level.

Discussion, implications, and future research

General discussion of findings

The findings overall reveal the leadership mechanisms behind Chinese employees’ two types of adaptive communication behavior, i.e., internal voicing and external scouting, upon the changes induced by the COVID-19 pandemic. When perceiving a greater degree of leadership inclusiveness, followers in Chinese organizations become more engaged in internal voicing, but not so in external scouting directly. In fact, the full mediation result of internal voicing suggests that before employee followers to proactively retrieve external information (i.e., external scouting), they need to first become active communicators about relevant issues within the work team (i.e., internal voicing). This sequence provides empirical support to prior crisis communication research contending that good information circulation internally facilitates resolutions of external challenges (e.g., Heide & Simonsson, 2021; Mazzei et al., 2012; Park et al., 2014; Simonsson & Heide, 2018). Parallel to internal voicing, both psychological empowerment and psychological safety also fully mediate the relationship between inclusive leadership and external scouting. Thus, besides inducing more internal voicing, the impacts of inclusive leadership are manifested by increased work control and decreased interpersonal worries. In turn, the one behavioral (internal voicing) and two cognitive (empowerment and safety) variations contributed to external scouting respectively. Moreover, an older age, a higher level of proactive temperament, and a more serious perception of the pandemic crisis would also predict an increase in the likelihood of external scouting.

Findings also demonstrated that employee status is a boundary condition for the proposed internal communication model. This was clarified as this study compared the functioning of two serial mediations (leadership ➔ empowerment ➔ voicing ➔ scouting vs. leadership ➔ safety ➔ voicing ➔ scouting). To followers at lower-level positions, experiencing an inclusive supervisor would only induce empowerment, not safety; in turn, the empowered state would translate to first internal voicing and then external scouting. Whereas to followers at higher levels, inclusive leadership would only lead to safety, not empowerment; sequentially, the safe sentiment would elicit more engagements in first internal voicing and then external scouting. This relative prevalence of empowerment and safety indicates that employee followers’ susceptibility to inclusive leadership is contingent upon their status in the organization. In other words, an inclusive leader in a prototypical Chinese organization can create a relatively fearless environment for adaptive communication behaviors during the pandemic crisis, but such effect is limited to scenarios of higher statuses.

We speculate that such boundary condition existed because of the normative leader – follower differentials perceived by employee followers (Antonakis & Atwater, 2002). More specifically, the lower an employee follower’s position is in the organizational structure, the more likely he/she expects an absolute difference in status and influence from his/her immediate supervisor (Liu & Liao, 2013). This is often the case in work cultures characterized by high power distance, like organizations in China (Farh et al., 2007; Zhang et al., 2023). For example, compare a sales representative reporting to a sales manager with a CFO reporting to a CEO in our sample of Chinese workers. The sales rep (lower-level follower) appeared to experience a much greater decrease in power distance if her sales manager demonstrated inclusive leadership behaviors to her, whereas the CFO (higher-level follower) would not experience so much gap reduction with her CEO. As such, a sense of increased job control (i.e., psychological empowerment) appeared essential for lower-level followers to engage in agential, role-expanding activities such as adaptive communication. Meanwhile, senior personnel with a peaceful mind (i.e., psychological safety) would probably consider adaptive communication as taking on the strategic responsibility of “sensegiving” in crisis (Christianson & Barton, 2021; Y. Kim, 2018).

Theoretical and practical implications

Our study made three unique theoretical contributions. First, we corroborated the process approach to internal crisis communication (Heide & Simonsson, 2020, 2021). This approach argues that internal communication during crises should not be reduced to information dissemination during the acute phrase of crises (Heide & Simonsson, 2015, 2021). Rather, employees’ discrete communication behaviors need to be dissected as a way of observing how they adapt to and master the challenges during crises (Gnyawali & Stewart, 2003; Kersten, 2005; Roux-Dufort, 2007). In our case, we demonstrated how employee followers became proactive providers and retrievers of insights under the influence of inclusive leaders during the COVID-19 pandemic. Second, we bridged employee communication research with crisis communication literature in the broader field of public relations (Lee & Yue, 2020). Our study showed that both psychological empowerment and psychological safety were important social information processing mechanisms for employee followers to cope with turbulent times like the COVID-19 pandemic crisis. In this sense, we not only provided new evidence to social information processing theory (Salancik & Pfeffer, 1978), but also highlighted the feasibility for researchers to study internal members’ crisis-related perceptions – a different route than responsibility attribution and response strategy preached by the Situational Crisis Communication Theory (SCCT; Coombs, 2007). Moreover, because our results highlighted the day-to-day contributions made by both leadership and followership over an enduring crisis period (Macpherson et al., 2021; Tourish, 2014; Uhl-Bien et al., 2014), we believe that crisis research should also examine how internal and external stakeholders cognitively and/or affectively experience crisis as “a complex, integrated, and emergent phenomenon” (Heide & Simonsson, 2015, p. 237), rather than an episodic event (see also Jaques, 2007, 2009; Ji & Kim, 2020). Third, our findings also echoed with public relations scholars who called for more understanding and explanation of cultural intricacies in theory development (Sriramesh et al., 2013; Verčič, 2020). In our entire research model, Chinese employee followers’ status in the organizational structure existed as a boundary condition. Most direct and indirect effects of inclusive leadership varied based on it. The results thus implied the relative prevalence of power distance orientations in Chinese organizations (Liu & Liao, 2013). In this regard, we provided insights into how employee/internal communication research could integrate cultural contexts outside Anglo countries into their models and frameworks.

This study also provides practical implications to managers, employees, public relations practitioners, and organizations in China as well as around the globe. First, results of this study remind organizational leaders of the importance of inclusion in day-to-day internal communications. Because the findings suggest that empowerment is a more important appraisal for lower-level followers (the majority), leaders – especially middle-level managers – should consider offering more task-related coaching, verbal encouragement, and appreciations to them in the face of difficulties (Lee, 2017; Sharbrough et al., 2006). Second, consistent with the first implication, organizations – regardless of their cultural backgrounds – need to support team supervisors with resources and training about the strategies, techniques, timing, and ethics of inclusive leadership behaviors to facilitate an optimal level of adaptive communication among their subordinate employees (see Wasserman, 2020). Third, Chinese employees need more awareness of how they can engage psychologically in adaptive communication to better navigate the challenges induced by crises. Even if they are well used to predefining “status-appropriate” roles in internal communication, they can contribute to their work teams by being more vocal internally or seeking useful information from outside (Hui et al., 2004; Kirkman et al., 2009). Fourth, this study’s attention to employee followers’ agency does not mean to abandon the top-down cascade of instructing and adjusting information (S. Kim & Sung, 2014); nor does it downplay the conventional sense of leadership, i.e., senior management. To better advise them, however, public relations practitioners can reference this study to improve the system of internal information circulation to fit the cognitive and behavioral needs of different employee segmentations.

Limitations and future research

Conclusions drawn in this study, however, need to be interpreted with caution due to the following limitations. First, this study used one online survey to explore hypotheses and research questions. The cross-sectional nature of this method can only present one slice of the reality encountered by Chinese organizations during the pandemic crisis, while causal inference cannot be made among constructs. An experiment or a time-lagged research design in future research can make up for such weakness. Second, as the sample within this study is female-skewed and more educated compared to the entire population in Mainland China, a sampling bias might exist. However, the impact of such divergence from the population is minimal, for neither gender nor education significantly explained the variances of any core variables in the analyses. Third, though empowerment and safety were regarded as two parallel mediators in this framework, later research can further examine how the two psychological states may or may not affect each other by comparing employee samples from different cultural contexts (see Simonet et al., 2015).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. As suggested by a reviewer, we did a robustness check for the model’s boundary condition. That is, we ran Model 83 with employee status as the moderator “W” and the results were no different. The robustness check results are available from the authors upon request.

2. The z-test equation: $\mathit{z}=\frac{b_1-b_2}{\sqrt{\mathit{SE}{b_1}^2+\mathit{SE}{b_2}^2}}$

Appendix 1

Descriptive Statistics of Respondents’ Industry Sectors

Table

Industry sector	n	%
Agriculture, forestry, fishing, and hunting	12	1.2
Manufacture, mining, energy, and construction	281	28.4
Wholesale and retail	52	5.3
Transportation, warehousing, and postal service	27	2.7
Leisure, hospitality, and catering	43	4.3
Information, telecommunication, and Internet	241	24.3
Real estate	25	2.5
Financial activities	98	9.9
Scientific research and polytechnic services	40	4.0
Education	49	4.9
Health care and social assistance	24	2.4
Culture, entertainment, and sports	32	3.2
Public administration, welfare, and community	38	3.8
Other services (except public sector)	27	2.7
Total	990	100

Appendix 2.

Summary Statistics and Correlations between Variables

Table

Variable	M	SD	1	2	3	4	5	6	7	8	9	10	11	12	13
1. Gender (male = 0; female = 1)	0.570	0.50
2. Age group	2.72	0.69	−.105**
3. Education level	4.11	0.43	−.016	−.023
4. Level in organization	2.83	1.21	−.157**	.300**	.164**
5. Work experience	2.80	1.05	−.160**	.752**	−.087**	.357**
6. Size of organization	2.70	0.71	−.051	.091**	.084**	.121**	.165**
7. Perceived crisis severity	4.54	1.17	.006	.061	.022	.052	.065*	−.063*
8. Proactive personality	5.59	0.71	−.018	.072*	.057	.207**	.107**	.180**	−.031
9. Perceived leadership inclusiveness	5.83	0.64	−.013	.088**	.038	.155**	.130**	.141**	−.037	.769**
10. Psychological safety	5.67	0.70	−.003	.050	.049	.168**	.084**	.137**	−.079*	.748**	.763**
11. Psychological empowerment	5.54	0.79	−.030	.116**	.049	.297**	.172**	.173**	−.056	.803**	.739**	.727**
12. Internal voicing	5.55	0.82	−.059	.153**	.055	.291**	.194**	.140**	.009	.739**	.688**	.673**	.745**
13. External scouting	5.17	0.98	−.012	.178**	.015	.238**	.183**	.149**	.043	.669**	.622**	.613**	.661**	.746**

Note. N = 983; *p < .05; **p < .01; ***p < .001.