ORIGINAL RESEARCH published: 16 December 2019 doi: 10.3389/fpsyg.2019.02788 How and When Do Leaders Influence Employees’ Well-Being? Moderated Mediation Models for Job Demands and Resources Rita Berger 1* , Jan Philipp Czakert 1 , Jan-Paul Leuteritz 2 and David Leiva 1 1 Department of Social Psychology and Quantitative Psychology, Facultat de Psicologia, Universitat de Barcelona, Barcelona, Spain, 2 Human Factors Engineering, Fraunhofer Institute for Industrial Engineering (IAO), Stuttgart, Germany Following the call of recent reviews on leadership and well-being, the purpose of this study is to examine how and when two contrasting leadership styles, transformational leadership (TFL) and passive-avoidant leadership (PAL), are related to employees’ anxiety and thereby either promote or inhibit employees’ well-being. Using the prominent job demands-resources (JD-R) model as a theoretical framework, we propose that the relationship between leadership behavior and anxiety is mediated by Edited by: organizational job demands, namely, role ambiguity (RA), and job resources, namely, Radha R. Sharma, team climate for learning (TCL), as well as moderated by autonomy as important job Management Development Institute, India characteristic. A sample of 501 knowledge workers, working in teams in a German Reviewed by: research and development (R&D) organization, answered an online survey. We tested Thomas Rigotti, moderated multiple mediation models using structural equation modeling (SEM). Results Johannes Gutenberg University Mainz, Germany demonstrated that the relationships between TFL as well as PAL on the one hand Guiquan Li, and anxiety on the other hand were fully mediated by RA and TCL. Job autonomy Rennes School of Business, France moderated the quality of the leadership–job demand relationship for TFL and PAL. This *Correspondence: paper contributes to understanding the complex relationship between leadership and Rita Berger ritaberger@ub.edu followers’ well-being taking into account a combination of mediating and moderating job demands and resources. This is the first study that examines the effects of TFL and Specialty section: PAL on well-being taking into account the job demand RA and team processes and This article was submitted to Organizational Psychology, autonomy as resources. a section of the journal Frontiers in Psychology Keywords: job demands-resources model, transformational leadership, passive-avoidant leadership, role ambiguity, team climate for learning, job autonomy, employee well-being Received: 25 July 2019 Accepted: 27 November 2019 Published: 16 December 2019 INTRODUCTION Citation: Berger R, Czakert JP, Modern business organizations are increasingly aware of the importance to sustain and promote Leuteritz J-P and Leiva D (2019) How employees’ well-being in order to gain and maintain competitive advantage (Nielsen et al., 2017). and When Do Leaders Influence Simultaneously, the fast-paced and increasingly complex environments in which organizations Employees’ Well-Being? Moderated Mediation Models for Job Demands operate demand for constant innovation (Reuveni and Vashdi, 2015) and performance (Nielsen and Resources. et al., 2017). Consequently, teamwork and collaborative learning are becoming more important, Front. Psychol. 10:2788. since teams pool different skills and expertise to deal with new and diverse tasks (Edmondson and doi: 10.3389/fpsyg.2019.02788 Nembhard, 2009; Ramírez Heller et al., 2014). Additionally, promoting teamwork in organizations Frontiers in Psychology | www.frontiersin.org 1 December 2019 | Volume 10 | Article 2788 Berger et al. Leaders Influence Employees’ Well-Being is often seen as a means to attain organizational goals and build (Heinitz et al., 2005), a passive leadership style with a negative competitive advantage (Day et al., 2004). Teams also have many relationship to well-being (Frooman et al., 2012); the latter has resources that have been associated with higher levels of well- been researched less (Inceoglu et al., 2018). Furthermore, the being and lower strain (Avanzi et al., 2015), explaining why selection of our mediators was based on Job design (Hackman working in teams has also been linked to employee well-being and Oldham, 1976) and included the job demand and work (e.g., Di Fabio, 2017). characteristic role ambiguity (RA) and therefore a motivational Albeit organizational and psychological benefits of working mediator. Additionally, our mediator selection was based on the in a team, also psychosocial risk factors may arise which may social learning theory (Bandura, 1977), including the job resource lead to employee stress (Navarro et al., 2011). For example, it is of team climate for learning (TCL) (Brodbeck et al., 2010), a a basic tenet in psychological science that stress originates from relational and social-cognitive mediator, since both pathways dealing with ambiguity and uncertainty (Peters et al., 2017). This are relevant for both positive and negative leadership styles is especially applicable to knowledge workers that are exposed (Inceoglu et al., 2018). We selected the concept of RA as one of to uncertain tasks, tough deadlines, and conflicting requirements the most commonly investigated occupational stressors (Bowling (e.g., Savelsbergh et al., 2012; Leuteritz et al., 2017). Consequently, and Beehr, 2006), related to negative well-being outcomes, and organizations need to prepare their employees successfully to deal TCL, which is a relevant resource for well-being (Lases et al., with ambiguity and provide the necessary resources for well- 2019). In this sense, TFL is known to positively influence being. In this regard, collaborative learning is thought to help employee outcomes via, e.g., inspirational motivation, which is in dealing with those challenges and acts as an interpersonal in line with the motivational path of the JD-R and which has resource (Lases et al., 2019). also been shown to increase learning resources of employees Another way of dealing with these ambiguity-related work via, e.g., intellectual stimulation (Schaufeli, 2015). Contrastingly, demands is transformational leadership (TFL) behavior, as it PAL does not increase resources (Barling and Frone, 2017). offers support resources via, e.g., individual consideration and We chose state anxiety to be an important negative aspect of has been recognized since several years ago as important for subjective well-being (Rajgopal, 2010; Skakon et al., 2010; Cheng employees’ well-being (Kelloway and Barling, 2010; Arnold, and McCarthy, 2018). It describes the proximal and internally 2017). Over the last few years, researchers’ interest in the directed response to perceived stressors (Glazer and Kruse, 2008; relationship between TFL (Bass, 1985) on employee well-being Pyc et al., 2016) and is known to be positively related to RA has grown, and the amount of available literature has increased (Glazer and Beehr, 2005). Thus, we contributed to research on substantially (Skakon et al., 2010; Arnold, 2017). Despite the negative indicators of well-being which until now have been less fact that the relationship between TFL and employee well- researched in relation with leadership behaviors (Inceoglu et al., being has been well-established in research, the exact underlying 2018) but which are necessary to understand the mechanisms complex mechanisms through which transformational leaders included in the negative pathway. Until now, few studies (e.g., exert their health-promoting effects still remain unclear (Skakon Cole et al., 2009; Franke and Felfe, 2011; Holstad et al., 2014; et al., 2010; Arnold, 2017). Specifically, the questions of how Diebig et al., 2017) investigated the role of moderators for the and when leadership affects followers’ well-being, and even more TFL–followers’ well-being relationship showing the importance interestingly how these mechanisms can be combined in models, of the conditions when TFL predicts employees’ well-being. Less are yet to be addressed (Arnold, 2017). studies also investigating moderating effects in mediation models Our study aimed to invoke answers to these questions (Holstad et al., 2014; Diebig et al., 2017) with mixed results. following the suggestions of the recent reviews (Arnold, 2017; Hence, we followed a recent call (Arnold, 2017) that requires Inceoglu et al., 2018) by exploring possible both, mediating further research on moderated mediation to answer how and and moderating mechanisms between TFL and employees’ well- when leadership influences followers’ well-being. Therefore, we being. Furthermore, previous research lacked a clear theoretical integrated with job autonomy one well-known moderator, a approach regarding the inclusion or not of specific mediators core job resource in line with the JD-R model (Bakker and in the TFL–well-being relationship (Diebig et al., 2017). Our Demerouti, 2007) that describes a motivational aspect (Inceoglu study, in line with recent relevant work (e.g., Schaufeli, 2015; et al., 2018). More specifically, to answer the question of how Diebig et al., 2017), tried to overcome this limitation by using the and when leadership influences employee well-being, we propose job demands-resources (JD-R) model (Bakker and Demerouti, that the relationship between TFL and anxiety is mediated by 2007) as a prominent theoretical framework, which allowed RA and TCL, and that the quality of the TFL -RA relationship is us to combine job demands and job resources. As leadership influenced by job autonomy. As research on negative leadership behavior does not always affect followers’ well-being in the same style and employees’ well-being and the underlying mechanism way (Inceoglu et al., 2018), we decided to analyze differential is scare and needed (Inceoglu et al., 2018), we include PAL as mediator pathways including two different leadership behaviors “contrasting” to TFL (Hetland et al., 2011, p. 163) to uncover and and two different mediators, specifically job demands and evidence the linkages of both sides of the leadership spectrum resources that should facilitate information about how leadership with job demands and resources, and ultimately with followers’ behavior increases resources and decreases demands and impacts well-being. In line with this, we propose that the relationship in followers’ well-being. We therefore used TFL, a change- for TFL’s negative counterpart, namely, PAL, is also mediated oriented leadership style with a positive relationship to well-being and moderated by the same factors. This is in line with previous (Inceoglu et al., 2018), and passive-avoidant leadership (PAL) studies that mirrored the effects of PAL and TFL in one study Frontiers in Psychology | www.frontiersin.org 2 December 2019 | Volume 10 | Article 2788 Berger et al. Leaders Influence Employees’ Well-Being to highlight the importance of TFL and the destructiveness (Bandura, 1977) for social-cognitive and relational pathways of PAL (e.g., Hetland et al., 2011; Frooman et al., 2012) via using this category of mediators and resources allow to analyze the same mechanism. developmental aspects that can lead to development. With this study, we contribute to existing research in several ways. First, we further built up on the literature combining The JD-R Model and Leadership the JD-R Model and the TFL employee well-being relationship. There is a growing number of studies that show the link To our knowledge, relevant research combining both, job between leadership style and employee well-being (e.g., Kelloway demands and job resources in the same model has been limited and Barling, 2010) but that often lack a clear conceptual (Hentrich et al., 2017). Thus, by analyzing the role of leadership approach (Inceoglu et al., 2018). The JD-R model offers a together with motivational and social-cognitive and relational broad theoretical framework for explaining the indirect positive mechanisms (Inceoglu et al., 2018), as well as influencing core relationship between TFL and well-being (Diebig et al., 2017) job characteristics, we tried to overcome limitations of previous and a negative one with PAL (Molero Alonso et al., 2010). Being research. Specifically, we broadened the scope of mediators of the initially included as a job resource (Bakker and Demerouti, 2007), TFL–well-being relationship by including TCL and RA. Second, recent research integrated leadership as a standalone predictive we highlighted influencing job characteristics as moderators in factor to the JD-R that affects both, job demands and resources the TFL–well-being relationship. Third, we mirrored the effect (e.g., Schaufeli, 2015; Diebig et al., 2017). This is in line with of TFL with the antagonist PAL, so that the contrast of the the social environment model (Katz and Kahn, 1978), which leader’s effect becomes even more apparent and, in this manner, suggests that perceptions of a leader frame a person’s perceptions we increase the scarce research on this topic. Lastly, considering of experienced work characteristics, which in turn relate to that research in TFL and well-being has shown inconsistencies outcomes and strains. Thus, it is widely regarded as a contextual in regard of whether the relationship between TFL and well- variable (Skakon et al., 2010; Oc, 2018). Therefore, we contributed being is fully or partially mediated (Arnold, 2017), we explored with our study in the frame of the prominent JD-R with a both, direct and indirect effects and thus, contribute to fill concept-based analysis approach. To be able to get differential this research gap. information on the mechanisms working in the leadership–well- being relationship, we included two different leadership styles in the analysis: the positive and change-oriented TFL and the THEORETICAL BACKGROUND AND negative and passive PAL that until now has been researched less RESEARCH MODEL (Arnold, 2017; Inceoglu et al., 2018). The JD-R model is the most prominent framework to study the effects of variables on different levels (structural, interpersonal, Transformational Leadership and etc.) on employees’ mental health (Nielsen et al., 2017). Employee Well-Being Compared to conservation of resources theory (Hobfoll and Transformational leadership, based on the full range leadership Lilly, 1993; Halbersleben et al., 2014), which also underlines model (Bass, 1985), is one of the most widely used leadership the importance for the individual to gain and preservation of theories (Gawke et al., 2017) and contributes to the personal resources, the JD-R model offers an appropriate consideration growth of followers by providing them with idealized influence, of job demands with respect to the available resources, without individualized consideration, intellectual stimulation, and narrowing its focus on resources acquisition and conservation. inspirational motivation (Bass, 1990, 2000). It is related to According to the JD-R model, every working environment multiple positive outcomes at individual and team level has its own specific factors linked to job related stress. These (Kelloway et al., 2012; Harms et al., 2017) and considered factors can generally be classified into two categories: job as change-oriented leadership (Inceoglu et al., 2018). Since demands and job resources. Job demands “refer to those physical, recently, TFL is recognized as a leadership style that as an psychological, social, or organizational aspects of the job that antecedent positively influences employee well-being (Arnold, require sustained physical and/or psychological (cognitive and 2017), such as decreased risk of sick absence (Kuoppala et al., emotional) effort or skills and are therefore associated with 2008; Skakon et al., 2010), and that reduces stress (Harms certain physiological and/or psychological costs” (Bakker and et al., 2017). This has been well documented in both, profit and Demerouti, 2007, 312), while job resources “refer to those non-profit organizations (e.g., McMurray et al., 2010) as well physical, psychological, social, or organizational aspects of the as in cross-sectional (e.g., Arnold et al., 2007) and longitudinal job that are either/or functional in achieving work goals, reduce studies (e.g., Tafvelin et al., 2011). According to the JD-R model, job demands and the associated physiological and psychological leaders manage the allocation of job demands and job resources costs, and stimulate personal growth, learning, and development” on their followers (Schaufeli, 2015). In other words, TFL (Bakker and Demerouti, 2007, 312). Within the JD-R as a leaders provide job resources, strengthen subordinates’ personal broad framework, diverse pathways (Inceoglu et al., 2018) can resources and reduce job demands (Diebig et al., 2017; Hentrich be combined. Building on job design (Hackman and Oldham, et al., 2017). For example, leaders may provide social support 1976), motivational pathways using motivational mediators and (Holstad et al., 2014) through individualized consideration or demands allow to analyze organizational information that can reduce job demands if one cannot adequately carry out the lead to redesign. Grounding on the social learning theory responsibilities (Holstad et al., 2014; Hentrich et al., 2017). Also, Frontiers in Psychology | www.frontiersin.org 3 December 2019 | Volume 10 | Article 2788 Berger et al. Leaders Influence Employees’ Well-Being transformational leaders who encourage subordinates to use authors (Skogstad et al., 2007) have found a positive relationship resources effectively for coping are a beneficial resource for them between laissez-faire leadership (embedded in PAL) on the one (Harms et al., 2017). Thus, they promote workers’ well-being hand, and role conflict and RA on the other hand. Furthermore, through healthy practices at work (Inceoglu et al., 2018). they have also found the effect of laissez-faire on distress However, the underlying mechanism behind the relationship (as well as on workplace bullying) to be mediated by these between TFL and various indicators of well-being is still stressors. Moreover, such a relationship has also been found unclear (Skakon et al., 2010; Arnold, 2017). Research on TFL between passive-avoidant leaders and the following variables: and employee well-being has thus proposed models where role overload, role conflict, and RA (Chênevert et al., 2013; job demands or job resources mediate the TFL well-being Barling and Frone, 2017). Furthermore, we follow the idea relationship (Holstad et al., 2014; Schaufeli, 2015; Diebig of Frooman et al. (2012), who argued earlier that PAL could et al., 2017; Hentrich et al., 2017). Nevertheless, research on be seen as the very opposite of TFL. Specifically, they state how transformational leaders promote follower well-being that that “by their very definitions, it does not seem possible for includes both, job demands, and job resources at the same time someone to be both transformational and passive avoidant” has clearly received less attention (Hentrich et al., 2017). For (Frooman et al., 2012, p. 450). In this sense, it seems reasonable example, Hentrich et al. (2017) found that TFL is associated to argue that when PAL prevails, all the positive effects of TFL with lower job demands, such as work intensity, but higher on employees’ well-being are lacking. For instance, within the occupational self-efficacy (personal resource), which in turn same research, it was shown that TFL positively predicts job contributes to lower levels of strain (here irritation). Similarly, satisfaction of the follower, whereas PAL negatively predicts the it has been found that daily TFL behavior is related to followers’ same factor of well-being (Frooman et al., 2012). Following the daily level of stress on a day-to-day basis through role conflict conclusion by Inceoglu et al. (2018), this study contributed to a (job demand) and team cooperation (job resource) (Diebig et al., better understanding of how the positive change-oriented TFL 2017). To broaden the existing evidence base (Diebig et al., 2017; influences followers’ well-being by including the negative passive Hentrich et al., 2017), the combined analysis of job demand and PAL into the analysis. resources will be one of the contributions of our paper. Anxiety as a Negative Indicator of Passive-Avoidant Leadership and Well-Being Employee Well-Being We chose state anxiety as an important negative well-being Based on the review done by Inceoglu et al. (2018), additional indicator as outcome variable (Rajgopal, 2010) for several information on how leadership influences employee well-being reasons. First, rather few studies analyze the relationship is needed. To underscore the importance of TFL, to better between leadership and negative indicators of employees’ well- understand its relationship with well-being and to contribute being, e.g., anxiety (Montano et al., 2017; Inceoglu et al., to closing the apparent research gap regarding passive forms of 2018). Second, it is defined as “a state physio-psychological destructive leadership styles (Inceoglu et al., 2018), this study also sensation, addressing people’s perceptions of psychological included PAL as an independent variable (Frooman et al., 2012; and physiological states” (Glazer and Beehr, 2005, 469) and Barling and Frone, 2017). includes both, physiological and psychological outcomes of In the contrast to the active form of TFL, other authors have stress. Third, anxiety is especially related to RA across different linked management by exception passive, previously embedded countries (Glazer and Beehr, 2005) in working environments in the transactional style (Bass, 1998; Avolio, 1999; as cited in that are characterized by uncertainty and ambiguity (Leuteritz Judge and Piccolo, 2004) of the full range leadership model (Bass, et al., 2017; Peters et al., 2017). Fourth, it has also been 1985), and laissez-faire style under the label of PAL (Heinitz et al., conceptualized as the most proximal response in supervisor– 2005). PAL represents both the passive mode of reaction and the subordinate relationships that is directed internally (Pyc et al., lack of reaction at all (Molero Alonso et al., 2010). 2016). Fifth, we analyzed state anxiety as a crucial outcome Passive-avoidant leadership-related behaviors do not because it is a great indicator for one of the most frequent accomplish any of the tasks associated with TFL. Working with mental disorders, namely, anxiety disorder, in the European leaders that do not provide subordinates with relevant feedback, Union (Wittchen et al., 2011). Sixth, when dealing with work- employees are less likely to know how they have to orientate related stress, state anxiety is seen as an initial response to their efforts to complete their task. Not only is it argued that stress (Glazer and Kruse, 2008) and significantly related to the employees under PAL lack social support by the supervisor, it is outcome of chronic work-related stress (Koutsimani et al., 2019), also assumed resources are not strengthened, and job demands organizational commitment, and turnover intention (Glazer and such as role stressors can even increase, consequently leading to Beehr, 2005). So, it makes sense to attack the predictor of increased health impairment (Barling and Frone, 2017). Research those later consequences in an early stage to avoid severely on PAL is scarce (Chênevert et al., 2013; Barling and Frone, 2017). negative outcomes. To the authors’ knowledge, no comprehensive meta-analyses on the specific issue exist until today. Nonetheless, some articles Mediators provided strong empirical support to link passive leadership To get information on and contribute to answering the styles to decreased employees’ well-being. For instance, some question how leadership behavior influences followers’ Frontiers in Psychology | www.frontiersin.org 4 December 2019 | Volume 10 | Article 2788 Berger et al. Leaders Influence Employees’ Well-Being well-being, we included two different mediators in the the well-being of followers. This contributes to the scarce leadership well-being relationship. Within the broad JDR, existing research. we combine and base on job design (Hackman and Oldham, H1a: RA mediates the negative relationship between 1976) for demands and social learning theory (Bandura, TFL and anxiety. 1977) for resources two theoretical approaches that can allow following Inceoglu et al. (2018) to analyze two H1b: RA mediates the positive relationship between important mediator pathways: a motivational one via the PAL and anxiety. demand RA and a social-cognitive and relational one via resource TCL. Team Climate for Learning as a Job Resource Grounding on social learning theory (Bandura, 1977), the Role Ambiguity as a Job Demand TFL (Inceoglu et al., 2018) has also been related to diverse Some researchers have analyzed work characteristics as mediators team processes as resources such as team cooperation (Diebig in the relationship between TFL and followers’ well-being. For et al., 2017). This resource that, e.g., offers social support example, Nielsen et al. (2008) found in a longitudinal study that (Bakker and Demerouti, 2007), therefore, may be also linked work characteristics partially mediated the relationship between to reduced strain (Schaufeli, 2017). Some researchers found TFL and followers’ well-being at time 1 and fully mediated team self-efficacy (Nielsen et al., 2009) or a positive climate the same relationship at time 2. Even so, there is not yet for innovation (Tafvelin et al., 2011) mediating the relationship enough research suggesting that leaders may reduce strain via job between TFL and employees’ well-being. As a recent study demands (Hentrich et al., 2017). could show, the aspect of having a positive learning climate Building on Job design (Hackman and Oldham, 1976), we is crucial for the individual well-being of the team members have chosen with RA, which is one of the most investigated (Lases et al., 2019). occupational stressors (Bowling and Beehr, 2006), an important Nevertheless, there is not yet much research that analyzes and well-known work characteristic and demand. RA, the role of mediating team processes in the leadership–employee which has been linked with multiple indicators of strain, well-being relationship (Jung and Sosik, 2002), and even less e.g., the frequency of symptoms of depression, generalized applying TCL as a job resource. TCL is defined as “shared anxiety, hyper-alertness (Dobreva-Martinova et al., 2002), and individual perceptions of work settings among members of a tension/anxiety (Parker and DeCotiis, 1983; Glazer and Beehr, team or an organization that promote or hinder learning in the 2005), is prevalent in research organizations (Schulz, 2013), workplace” (Ramírez Heller et al., 2014, p. 544). Team members and thus, relevant to our sample that consisted of employees with positive TCL (Brodbeck et al., 2010): (1) offer support to the of a research organization. Those knowledge workers are other members and show understanding and mutual trust; (2) often confronted with uncertain tasks, conflicting priorities, have regular contact with formal or informal communication; (3) multiple stakeholders, and requirements that may exceed their work toward common goals with clear, realistic, and achievable available resources (Savelsbergh et al., 2012; Leuteritz et al., objectives while a notion of democracy prevails with no team 2017). At the same time, teams often suffer compositional member dominating over others; and (4) perceive a kind of changes that provoke a change in responsibilities which individual development as a result of the group enhancing their may, in turn, result as unclear or incompatible (Alcover creativity and providing them with various resources such as new et al., 2011). RA constitutes a demotivational job design ideas and support. (Hackman and Oldham, 1976; Inceoglu et al., 2018). RA Based on the JD-R, in the present study, TCL is considered refers to feelings of uncertainty or to a lack of necessary therefore as a job resource and relational and social-cognitive information concerning role functions and responsibilities mediator (Inceoglu et al., 2018) in the leadership–well-being (Glazer and Beehr, 2005; Örtqvist and Wincent, 2006). relationship for various reasons: it may reduce job demands Previous research has linked role stressors to the JD-R (e.g., Lases et al., 2019), it promotes individual development (Bakker and Demerouti, 2007; Hentrich et al., 2017) and and resources, since it stimulates learning, and it is functional showed that transformational leaders may actively play a role in achieving work goals, since it is linked to increased positive in decreasing RA and consequently, reducing anxiety (e.g., performance indicators, team functioning, and support for Diebig et al., 2016). This may be done by clarifying roles and innovation (Ramírez Heller et al., 2014). Moreover, according motivating team members by providing a vision. Moreover, to Bakker and Demerouti (2007), job resources may include transformational leaders may train and stimulate employees opportunities for development, feedback, and role clarity, while to be able to conduct special non-routine decision-making daily team cooperation is a job resource that is linked to reduced (Harms et al., 2017). In contrast, PAL has been associated daily levels of stress (Diebig et al., 2017). As described above, with perceptions of unclear guidelines at work (Hetland et al., a positive TCL facilitates individual development while mutual 2011), demonstrating that leaders who neglect their duties trust and face-to-face frequent contact may help team members create uncertain and ambiguous environments. Other studies in clarifying role expectations (Marín Puchades, 2015). It can link PAL particularly to RA (Skogstad et al., 2007; Chênevert be concluded that working in a team that is prevailed by a et al., 2013; Barling and Frone, 2017). In the present study, we positive climate for learning can make employees less vulnerable included RA as a job demand and motivational mediator to to resource loss and more capable and likely to gain other be able to analyze how both leadership styles are influencing resources, which, in turn makes them more resilient to stress Frontiers in Psychology | www.frontiersin.org 5 December 2019 | Volume 10 | Article 2788 Berger et al. Leaders Influence Employees’ Well-Being (Hobfoll and Lilly, 1993; Xanthopoulou et al., 2009). This possibly working situation (Volmer et al., 2012; Ju et al., 2019). For promotes the growth of personal competences of employees, instance, when employees are receiving motivational triggers by making them more efficient in dealing with demands, and may the transformational leader to schedule their own action plans influence in the reduction of stress reactions (Dackert, 2010). and to clarify their roles, they are likely to do so if the job context Previous research linked TFL and PAL with the learning climate allows it (Ju et al., 2019). In contrast, when job autonomy is in a team setting and emphasized both the constructive impact low, the leader might stimulate employees to enact clarifying of TFL and destructive impact of PAL on the learning climate activities, albeit the role stressor will continue to prevail because (e.g., Hetland et al., 2011). According to the input-mediator- the specific working context is determined (Volmer et al., 2012). output (IMO) model (Ilgen et al., 2005), team climate has been Hence, we expect that the TFL–RA relationship is influenced by widely regarded as an emergent state (Mathieu et al., 2008). job autonomy, and specifically that the higher the job autonomy TFL may be considered as an input variable which affects level, the greater the link of TFL. TCL which, in turn, may affect team members’ well-being, Second, the reinforcing effect of PAL on RA is arguably e.g., anxiety levels. influenced by the degree of autonomy in the way that PAL paired with a low level of autonomy strengthens the effect of a H2a: TCL mediates the negative relationship between leader’s passive-avoidant behavior on RA. When employees are TFL and anxiety. not able to reduce ambiguity by themselves, or autonomously, H2b: TCL mediates the positive relationship between the perception of ambiguity due to leader passivity might be even PAL and anxiety. higher (Hetland et al., 2011). Job Autonomy as an Interacting Factor H3a: Job autonomy moderates the relationship between in the Leadership–Job Demands TFL and anxiety via RA. The higher the levels of job autonomy, the higher the influence of TFL on RA. Relationship To get information about when leadership behavior influences followers’ well-being, we wish to add to the present lack of H3b: Job autonomy moderates the relationship between research (2017) and included with job autonomy an important PAL and anxiety via RA. The lower the levels of job motivational moderator, since it seems to be one of the autonomy, the higher the influence of PAL on RA. most powerful job resources (Bakker and Demerouti, 2007) and simultaneously an influencing factor for TFL leadership In conclusion, we hypothesize moderated multiple mediation effectiveness (Den Hartog and Belschak, 2012). models where RA and TCL mediate the relationship between TFL Job autonomy is defined as “the degree to which the job and strain. We argue that TFL is positively related to TCL and in provides substantial freedom, independence, and discretion to turn negatively related to anxiety. Similarly, we hypothesize that the individual in scheduling the work and in determining the TFL is negatively related to RA and in turn, negatively related procedures to be used in carrying it out” (Hackman and Oldham, to anxiety. Job autonomy moderates the effect (Figure 1). In 1976, p. 258). It constitutes a contextual variable, where high contrast, we further hypothesize that PAL is negatively related levels of autonomy describe contexts in which employees can to TCL and positively related to RA, which eventually increases take full advantage of their abilities at the job (Morgeson et al., anxiety. Job autonomy is hypothesized to moderate the effect 2005), and where environments are more dynamic and uncertain between PAL and RA (Figure 2). (Den Hartog and Belschak, 2012). There is strong evidence that our sample of knowledge workers in research and development (R&D) organizations work in such described environments (Schulz, 2013). Former research suggested that certain leadership theories, such as leader–member exchange (Volmer et al., 2012), empowering leadership (Ju et al., 2019), and also TFL (Den Hartog and Belschak, 2012), work better in less-prescribed environments by modeling successfully the moderating function of job autonomy in their works. It was found that leaders through work characteristics such as, e.g., autonomy might enable resources (Halbersleben et al., 2014). In line with this, we argue that the impact of leadership on the job demand RA is influenced by this core job characteristic (Hackman and Oldham, 1976), and that this applies for both contrasting leadership styles, TFL and PAL. These assumptions seem legitimate for the following reasons: First, the link of TFL on RA arguably depends strongly on FIGURE 1 | Hypothesized model TFL. the actual capabilities of the employee to clarify his or her Frontiers in Psychology | www.frontiersin.org 6 December 2019 | Volume 10 | Article 2788 Berger et al. Leaders Influence Employees’ Well-Being Measures Transformational Leadership We measured TFL, based on the full range leadership model (Bass, 1985), using the German version of the Human System Audit Short-Scale of TFL shown to be unidimensional (Berger and Antonioli, 2019), since it is shorter than the Multifactor Leadership Questionnaire MLQ-5X-Short (Avolio and Bass, 2004) and an easy to apply instrument (Berger et al., 2011). The scale consists of eight items (e.g., “My leader promotes the use of intelligence a means of overcoming obstacles”) that measure TFL using a five-point Likert scale (from 1 = strongly disagree to 5 = strongly agree). Previous research has provided empirical evidence for the construct validity of this measure (Berger et al., 2011, 2012) in diverse languages. Cronbach’s alpha for the eight items for the German sample was α = 0.93. FIGURE 2 | Hypothesized model PAL. Passive-Avoidant Leadership For PAL (Heinitz et al., 2005), we used the Multifactor Leadership Questionnaire MLQ-5X-Short (Avolio and Bass, 2004). The MATERIALS AND METHODS answers to the items are measured with a five-point Likert scale with responses regarding frequency (from 1 = never to 5 = nearly Participants and Procedure always). We only used the eight items that measure PAL. Four of A total of 501 employees of an R&D organization of these items belong to the management by exception passive sub- approximately 15,000 employees participated in an online survey scale (e.g., “Avoids intervening until the problems get serious”) in Germany. As Table 1 shows, out of the 501 participants, 343 and four represent laissez-faire style (e.g., “Has avoided telling (68.5%) were males and 158 (31.5%) were females with ages me how to perform my job”). Cronbach’s alpha for the PAL was ranging from 18 to 70 years (M = 35.6, SD = 10). Moreover, α = 0.88, which implies a high reliability. This is in line with all participants were part of teams, mainly R&D teams (85.8%). previous studies (e.g., α = 0.84 in Molero Alonso et al., 2010). A lower percentage of individuals participated in administration (9.6%), facility (2%), and other tasks (2.6%). Role Ambiguity In order to measure the constructs involved in this research, Three items of The Role Stressors Questionnaire (Glazer and four questionnaires were applied with a total of 48 items, Beehr, 2005) that refer to RA (e.g., “I know exactly what is as well as questions regarding demographic data. All scales expected of me”) were translated into German and used to used where translated into German following the guidelines measure RA following the ITC Guidelines (2017). The items are of the International Test Commission (2017) for translating rated with a seven-point Likert scale (from 1 = strongly disagree and adapting tests. All team leaders were excluded from the to 7 = strongly agree). The Cronbach’s alpha in our sample for the data collection, and all participants were informed about data three items used was α = 0.81. confidentiality policy in regard of the administration of the questionnaire by a written text. Moreover, participation was Team Climate for Learning voluntary; the declaration of consent of the individuals was asked The German version of the TCL Questionnaire (Brodbeck et al., before completion of the questionnaire and found to be accepted 2010) was used for measuring TCL. Thirty-three items were when submitted. In order to promote participation, a lottery of used that include nine subscales: (1) Mutual Trust, (2) Goal vouchers was used. Alignment, (3) Attendance, (4) Regular Contact, (5) Democracy, There were no empty cells in the final dataset, as participants (6) Team Management, (7) Individual Development, (8) Open could only return completely answered questionnaires. Hence, Exchange, and (9) Motivation and Interest (Brodbeck et al., missing values were not an issue in this study. 2010; Ramírez Heller et al., 2014). The items (e.g., “My team provides me with useful ideas and practical support”) were rated using a seven-point Likert scale (from 1 = strongly disagree to TABLE 1 | Sample description (N = 501). 7 = strongly agree). n Percent (%) Job Autonomy Male participants 343 68.5 Three items of the German version of the well-established Job Female participants 158 31.5 Diagnostics Survey (Hackman and Oldham, 1976; Schmidt et al., Job: R&D 430 85.8 1985) were used to assess job autonomy. One item was “How Job: administration 48 9.6 much autonomy is there in your job? That is, to what extent does Job: facility 10 2 your job permit you to decide on your own how to go about Job: other 13 2.6 doing the work?” and is rated with a seven-point Likert scale Frontiers in Psychology | www.frontiersin.org 7 December 2019 | Volume 10 | Article 2788 Berger et al. Leaders Influence Employees’ Well-Being (from 1 = very little; the job gives me almost no personal “say” formal and informal topics”). Factors 1–3 showed a very good about how and when the work is done to 7 = very much; the job internal consistency with alphas ranging from α = 0.89 to α = 0.92. gives me almost complete responsibility for deciding how and when Cronbach’s alpha for factor 4 was α = 0.64. the work is done). The other two items are “The job gives me This four-factor solution was supported by using a second- considerable opportunity for independence and freedom in how I order CFA. The CFA revealed a satisfactory model fit [χ2 (185) do the work” and “The job denies any chance to use my personal (N = 247) = 532.61, p < 0.001; χ2 /df = 2.879; Tucker–Lewis initiative or judgment in carrying out the work.” These items are Index (TLI) = 0.89; comparative–fit index (CFI) = 0.91; root- rated with a seven-point Likert scale (from 1 = strongly disagree to mean-square error of approximation (RMSEA) = 0.087]. Values 7 = strongly agree). Cronbach’s alpha in our sample for the three larger than 0.90 for TLI and CFI and lower than 0.10 for RMSEA items used was α = 0.77. indicate an acceptable model fit (Byrne, 2009). Regarding the χ2 /df, various rules of thumb have been recommended; a value Strain up to 2.0 is considered very good and between 2.0 and 5.0 We translated the four items out of the Anxiety Scale of Parker acceptable (Hair et al., 2006). All of the standardized parameter and DeCotiis (1983) that refer to anxiety (e.g., “I have felt fidgety estimates (i.e., factor loadings) were significant at p < 0.01 and or nervous as a result of my job”) into German. Similar to ranged from 0.51 to 0.97 (mean standardized loading = 0.78). previous research, the remaining 11 items of the scale were Maximum likelihood estimation was applied and the omitted because they were referring to time stress (Glazer and goodness-of-fit of the tested models was evaluated using several Beehr, 2005). The participants were asked to answer the questions indices: (1) the chi-square test statistic, (2) the normed chi-square using a seven-point Likert scale (from 1 = strongly disagree to (χ2 /df ), (3) the RMSEA, (4) the TLI, and (5) the CFI. Based 7 = strongly agree). Cronbach’s alpha in our sample for the four on recommendations by Hu and Bentler (1998), the following items used was α = 0.88. cut-offs were used to indicate adequate model fit: TLI and CFI > 0.95 and RMSEA < 0.06. Data Analysis The tests of indirect effects (i.e., Hypotheses 1 and 2) were Our hypothesized models, displayed in Figures 1, 2, are analyzed simultaneously, and a two-step approach was adopted moderated multiple mediation models where TFL and PAL are (Kline, 2016). First, a measurement model was tested. This model the independent variables, anxiety the dependent variable, RA determines if the indicators well reflect the latent constructs. and TCL two mediators, and job autonomy the moderator in Finally, the second step (structural part) allows us to test the the leadership–job demand relationship. In order to test the proposed relationships among the latent variables. As proposed hypothesized models, we applied structural equation modeling by Preacher and Hayes (2008), we applied a bootstrap procedure (SEM) using IBM SPSS AMOS 25.0 (Arbuckle, 2017). We for testing statistical significance of mediations. More specifically, assessed the hypothesized moderating effects separately, because 1000 bootstrap samples were produced which were used to we were interested in the specific interaction effects for the estimate 95% bias-corrected confidence intervals (CIs) to test for individual leadership style (Dawson, 2014). significant indirect effects. To assess the impact of common source variance, we followed To test the interaction effect of leadership and job autonomy the recommendations of experts in the field (Podsakoff et al., on RA, we adopted the three-step procedure of residual mean 2003, 2012) and applied the so-called “single-common-method- centering for latent constructs interaction effects introduced factor approach” (Podsakoff et al., 2003, 895) with CFA. Hence, by Little et al. (2006), since – unlike mean centering – we included an additional latent variable to our CFA model it completely alters item variances and covariances and is and allowed all items to load on their theoretical constructs as therefore a recommended orthogonalizing tool for addressing well as on a common method variance factor. Subsequently, the interaction effects in SEM (Geldhof et al., 2013). In a first common method variance factor was removed, and the difference step, all first-order variables of the two latent variables were of the loadings of the indicators on their constructs between the multiplied. In a second step, each product term was regressed models was compared. Since standardized cut-off criteria for this on all effect indicators. Finally, the resulting residuals were method are yet to be agreed on, we selected differences higher integrated in the model and the covariances between the than 0.2 as our criterion based on Cohen’s (1988) effect size two latent variables and the interaction effect were set to categorization. For hypothesis testing, we set the significance level zero. Likewise, 1000 bootstrap samples were produced, and at α = 0.05. 95% CI analyzed to test the interaction effect for significance In order to validate the questionnaire for TCL, we split the (Little et al., 2006). sample in half and carried out an exploratory factor analysis (EFA) with one half, followed by a confirmatory factor analysis (CFA) (Cabrera-Nguyen, 2010). EFA revealed a four-factor RESULTS structure for 33 items. The four factors were labeled as: (1) Group Interaction and Group Support (e.g., “My team provides me Preliminary Analysis with useful ideas and practical support”), (2) Goal Alignment Table 2 displays the means, standard deviations, Cronbach’s (e.g., “The goals of my group are useful and suitable”), (3) Open α, and intercorrelations of all variables included in this study. Exchange (e.g., “All opinions are respected”), and (4) Regular All measures relevant to our proposed mediation model Contact (e.g., “The group members meet frequently to discuss were significantly correlated; in other words, measures of Frontiers in Psychology | www.frontiersin.org 8 December 2019 | Volume 10 | Article 2788 Berger et al. Leaders Influence Employees’ Well-Being TABLE 2 | Means, standard deviations, correlations, and alphas of the modeled latent variables. Variable Mean SD 1 2 3 4 5 6 1. TFL 3.62 0.80 (0.91) 2. PAL 2.19 0.77 −0.64 (0.88) 3. RA 2.68 1.18 −0.46 0.36 (0.81) 4. TCL 5.42 0.90 0.63 −0.53 −0.51 (0.85) 5. Job Autonomy 5.88 0.97 0.29 −0.21 −0.25 0.35 (0.77) 6. Anxiety 3.61 1.58 −0.34 0.35 0.41 −0.39 −0.15 (0.88) N = 501. Cronbach’s alpha in parentheses. All correlations significant at the 0.01-level (two-tailed). TFL, PAL, the hypothesized mediating variables (TCL, RA), TFL to anxiety was not significant, therefore revealing that the hypothesized moderator job autonomy, and anxiety were the relationship between TFL and anxiety is fully mediated by correlated as expected. role RA and TCL. H3a was also supported: the model showed a positive Confirmatory Factor Analysis significant interaction effect between TFL and job autonomy with The measurement model comprised the six latent factors (TFL, 0.19 being significant at the 0.05-level and 95% CI = [0.042, PAL TCL, RA, job autonomy, anxiety) and their indicators. TCL 0.43]. The positive interaction shows: the higher the degree of job was included as a latent variable with four indicators [mean scores autonomy, the greater the effect of TFL on RA. Figure 3 shows of factors: (1) group interaction and group support, (2) goal the identified interactions. alignment, (3) open exchange, and (4) regular contact] as the previously conducted CFA revealed. Testing the Hypothesized Model for The results of the measurement model revealed a reasonable Passive-Avoidant Leadership fit with the data [χ2 (881) (N = 501) = 2364.798, p < 0.001; We further tested the moderated mediation model where χ2 /df = 2.684; TLI = 0.90; CFI = 0.92; RMSEA = 0.058]. All items perceived job autonomy moderates the link between PAL and loaded in the expected direction on their corresponding latent RA. The estimation of the proposed model yielded acceptable fit factors, and all standardized factor loadings were significant indices [χ2 (1258) (N = 501) = 2558.62, p < 0.001; χ2 /df = 2.034; at p < 0.001, ranging between 0.481 and 0.915, supporting TLI = 0.93; CFI = 0.93; RMSEA = 0.05]. Hypotheses 1b and 2b convergent validity. Moreover, all correlations were below 0.80, were supported. More specifically, the standardized indirect effect supporting discriminant validity. Therefore, the hypothesized from PAL through RA and TCL to anxiety was 0.15 (95% CI factor structure of the measurements was supported. [0.064, 0.28]). The direct path linking PAL to anxiety was not significant, therefore revealing that the relationship between PAL Assessment of Common Method Bias and anxiety is fully mediated by RA and TCL (Table 4). To assess common method bias, we applied the single-common- Hypothesis 3b was also supported, since there was a significant method-factor approach (Podsakoff et al., 2012). The differences interaction effect between PAL and RA. The moderation effect only ranged between −0.084 and 0.11, i.e., the effect sizes of was −0.71 (95% CI [−5.72; −0.077]). That is, perceived job the indicators on their theorized latent factor: (a) were not autonomy had a buffering effect on the negative link between PAL substantially different when controlling for common method and RA. Figure 4 illustrates the identified relationships. variance and (b) were mostly inflated, but also deflated in some cases. Since all values scored lower than our cut-off criterion of <0.2, we did not impute the factor loadings of the model with DISCUSSION single common method factor into our hypothesized structural equation models. Findings This research made four contributions to leadership research. Testing the Model for Transformational First, it addressed the limitations of previous research on Leadership TFL and employee well-being by bringing together job We first tested the moderated mediation model where perceived demands and job resources to explain how positive and job autonomy moderates the link between TFL and RA. The negative leadership influences follower well-being. Second, it estimation of the proposed model yielded adequate fit indices expanded the complexity of this research field’s commonly used [χ2 (1258) (N = 501) = 2791.93, p < 0.001; χ2 /df = 2219.34; mediator models by modeling moderated mediation models, TLI = 0.91; CFI = 0.92; RMSEA = 0.047]. H1a and H2a were thereby finding additional answers to when leadership behaviors supported. TFL had a significant indirect effect on follower’s level influence job demands. Third, it emphasized the importance of on anxiety via TCL and RA. More specifically, the standardized bringing together resources and demands from motivational and indirect effect from TFL through RA and TCL to anxiety was relational and social-cognitive categories when analyzing well- −0.30 (95% CI [−0.46, −0.16]). As can be seen in Table 3, being at the workplace. Fourth, TFL was analyzed in contrast to all mediating paths were significant. The direct path linking PAL to better understand the role of the positive TFL. Frontiers in Psychology | www.frontiersin.org 9 December 2019 | Volume 10 | Article 2788 Berger et al. Leaders Influence Employees’ Well-Being TABLE 3 | B coefficients, SE, p-values, and 95% B CI of the transformational leadership model (TFL × Autonomy moderation). 95% B CI Predictor Outcome B SE p Lower Upper TFL TCL 0.50 0.06 ∗∗∗ 0.37 0.63 PAL TCL −0.19 0.06 ∗∗∗ −0.33 −0.083 TFL RA −0.34 0.07 ∗∗∗ −0.51 −0.19 PAL RA 0.20 0.07 ∗∗ 0.054 0.36 Autonomy RA −0.15 0.05 ∗∗ −0.29 −0.017 TFL × Autonomy RA 0.19 0.05 ∗∗∗ 0.042 0.43 RA ANX 0.50 0.09 ∗∗∗ 0.26 0.66 TCL ANX −0.28 0.11 ∗∗ −0.53 −0.047 PAL ANX 0.16 0.11 0.18 −0.072 0.38 TFL ANX −0.04 0.13 0.81 −0.31 0.25 N = 501. ∗∗∗ indicates a p-value <0.01. ∗∗ indicates a p-value <0.05. ANX = anxiety. that positive leadership indirectly affects follower’s well-being by primarily increasing resources, whereas negative leadership does so by primarily impacting job demands (Schaufeli, 2015; Inceoglu et al., 2018). The results suggest that transformational leaders have a slightly stronger effect via relational social- cognitive than motivational paths, but motivational paths are closer linked to anxiety. Specifically, transformational leaders may reduce employee anxiety by fostering a positive TCL. Such a climate may be induced by promoting open exchange, knowledge sharing, mutual trust, communication, and goal setting between team members. Another mechanism through which transformational leaders may contribute to the reduction of their followers’ anxiety is by giving them guidance for efforts, by encouraging their self-confidence to pursue new pathways for growth, and by reducing ambiguity (Harms et al., FIGURE 3 | Empirical model TFL. ∗∗∗ indicates a p-value < 0.01. ∗∗ indicates 2017). The absence of a direct effect of TFL on employee a p-value < 0.05. anxiety is noteworthy and in line with existing research that has demonstrated that the relationship between TFL and well- being is a fully mediated relationship (e.g., Arnold et al., 2007; The principal aims of the current study were to examine Tafvelin et al., 2011; Kelloway et al., 2012). In turn, the results how and when transformational leaders may contribute in showed that PAL not only increases motivation-related job promoting their followers’ well-being, and what happens when demands but also decreases important job resources (relational leaders show opposing behaviors. Based on the JD-R model, two and social-cognitive TCL) and is thus by far not a form of moderated multiple mediation models were hypothesized where zero-sum leadership, but a severe factor for employee stress. job demands and job resources mediate the relationship between This is in line with previous research (Skogstad et al., 2007; TFL – respectively, PAL – and anxiety, and where job autonomy Hetland et al., 2011). as an important job characteristic determines the quality of the Another finding is that job autonomy acts as a moderator for leadership–job demand relationship. the relationship between TFL and RA. Our results suggest that The main findings are that the relationships between TFL TFL is generally able to decrease the demand of RA, but that TFL and PAL on the one hand and anxiety on the other hand works better when the situation provides the employee enough are fully mediated by the motivational RA (Hypothesis 1) and freedom to act. In other words, when job design constraints the relational and social-cognitive TCL (Hypothesis 2) paths, prevail, positive leadership is less effective to reduce job demands, confirming our two hypotheses regarding the question how which is in line with previous research (Volmer et al., 2012). leadership influences employee well-being. In line with previous In contrast to this amplifying moderating effect and in line research, the health-impairment process via the job demand was with previous research (Hetland et al., 2011), job autonomy stronger than the link via job resource on anxiety (Schaufeli buffered the connection between PAL and RA, suggesting as it and Taris, 2014; Schaufeli, 2015). TFL’s relationship with TCL is to be an impactful resource when dealing with leader passivity. was stronger than its relationship with RA. In contrast, PAL’s Furthermore, the considerably high negative covariation between relationship with RA was higher than its relationship with TFL and PAL strengthened the view of PAL as a direct TCL. This is in line with previous research, that suggested antagonist of TFL. Frontiers in Psychology | www.frontiersin.org 10 December 2019 | Volume 10 | Article 2788 Berger et al. Leaders Influence Employees’ Well-Being TABLE 4 | B coefficients, SE, p-values, and 95% B CI of passive-avoidant leadership model (PAL × Autonomy moderation). 95% B CI Predictor Outcome B SE p Lower Upper PAL TCL −0.19 0.05 ∗∗∗ −0.32 −0.074 TFL TCL 0.50 0.07 ∗∗∗ 0.37 0.63 PAL RA 0.22 0.07 ∗∗∗ 0.065 0.38 TFL RA −0.34 0.07 ∗∗∗ −0.51 −0.19 Autonomy RA −0.15 0.05 ∗∗ −0.29 −0.023 PAL × Autonomy RA −0.71 0.47 ∗∗ −5.72 −0.077 RA ANX 0.46 0.09 ∗∗∗ 0.27 0.65 TCL ANX −0.29 0.11 ∗∗ −0.53 −0.056 TFL ANX −0.03 0.13 0.85 −0.31 0.26 PAL ANX 0.17 0.11 0.17 −0.062 0.40 N = 501. ∗∗∗ indicates a p-value <0.01. ∗∗ indicates a p-value <0.05. ANX = anxiety. Third, this research could highlight the importance of positive leadership for follower’s well-being by contrasting the effects of TFL and PAL, and also the interplay of resources (TFL and job autonomy). Fourth, TFL and PAL were not directly associated with follower’s well-being, suggesting full indirect relationships. Practical Implications With the aim of maintaining a mentally healthy workforce, leadership training may be a productive occupational health intervention. Leaders should be aware of their potentially significant role in helping employees cope with anxiety – and that passivity can have severe consequences – and so be trained to that they can adopt more transformational behaviors. FIGURE 4 | Empirical model PAL. ∗∗∗ indicates a p-value < 0.01. ∗∗ indicates Training programs may be designed not only for leaders but a p-value < 0.05. for teams as well. Employees need to be aware that their team may consist of a valuable resource when it comes to learning, dealing with ambiguity and, most importantly, their well-being. Theoretical Implications Therefore, training programs on motivational, relational, and Apart from the described findings, the study further revealed social-cognitive aspects that promote communication within five theoretical implications. We expanded the ongoing research the team, knowledge transfer, and planning skills may result work on the JD-R model and leadership both conceptually and effective. Moreover, through team building activities, employees methodologically. First, conceptually, we included job resources may build trust with their fellow team members and improve and job demands, on both motivational and relational and social- their communication. cognitive level. To our knowledge, no research up until now Ultimately, and next to HR practices, job design measures that has examined TCL and RA as possible mediators between the grant employees more autonomy in the way they carry out their TFL well-being relationship. Specifically, although there is plenty tasks should be considered. of evidence that transformational leaders increase job resources (e.g., Arnold et al., 2007; Nielsen et al., 2008) and in turn promote Limitations and Future Research employee well-being, only a small amount of research has Although our study contributes to the further understanding proposed models in which transformational leaders exert their of the relationship between TFL/PAL and employee well-being, health promoting effects by reducing job demands (Hentrich it has certain limitations. First, we applied a cross-sectional et al., 2017). Second, methodologically, we modeled not only design, which clearly limits the degree to which we could mediator or moderator models, but created moderated multiple make causal inferences regarding the relationships proposed mediation models, thereby integrating different interactions into between the variables. However, following Hetland et al. (2011), one frame. In doing so, we followed the call of previous meta- only experiments overcome cause and effect issues. For the analyses to combine moderators and mediators in the same purpose of this manuscript, studying the phenomena linked to model (Arnold, 2017). a real-life context seemed appropriate. Consistent with Cerin’s Frontiers in Psychology | www.frontiersin.org 11 December 2019 | Volume 10 | Article 2788 Berger et al. Leaders Influence Employees’ Well-Being (2010) recommendations, we mitigated the limitations associated group level and examine whether TFL may affect followers’ with cross-sectional mediation analysis by relying on a large collective well-being. sample, which helps reduce bias in regression estimates due to Sixth, the sample was unbalanced toward researchers and measurement error (Kline, 2016). male participants, which was due to the true distribution of Additionally, the proposed models are supported by previous genders (34% were women) and jobs (55% were researchers) research and current theory, while the relationship between in the organization (based on HR data from the year 2014). TFL and well-being has also been demonstrated by longitudinal However, our data correctly represent today’s R&D sector studies (e.g., Tafvelin et al., 2011). Furthermore, the JD-R with its limited gender diversity. Future research with more model has been attested longitudinal evidence in reviews, balanced samples would be interesting to confirm or refute the which is why other research regarding this relationship in obtained results. the context of the JD-R model has also applied cross- sectional designs (e.g., Schaufeli, 2015; Hentrich et al., 2017). Explicitly, Schaufeli and Taris (2014) state in their critical CONCLUSION review of the JD-R model that “job demand and job resources have an impact over time on burnout” (Schaufeli In summary, leadership behavior affects the well-being level, and Taris, 2014, p.48). In any case, the findings of the namely, anxiety, of employees by influencing perceived demands present study would need to be replicated by longitudinal or and resources to cope with demands. TFL is not only associated experimental designs, so that the mediation effects causality can with a positive TCL, but also with reducing RA. TFL works be adequately tested. better in high-autonomous contexts. In contrast, leader passivity Second, our study used only one source of information increases job demands and reduces job resources. Autonomy- for data gathering (self-reported questionnaire), which raises promoting job designs buffer the link between PAL and RA. concerns about common method bias (Podsakoff et al., 2003). Hence, estimates of the effects of the predictors on the criterion variable and other constructs might be artificially inflated, DATA AVAILABILITY STATEMENT deflated, or non-significant (Podsakoff et al., 2012). To mitigate this effect, we followed state of the art recommendations and The datasets generated for this study are available on request to applied procedural (e.g., avoiding ambiguous items by using the corresponding author. simple and concise statements and syntax, reverse coding, and counterbalancing question order) and statistical remedies (single-common-method-factor approach) to statistically control ETHICS STATEMENT for a common latent variable (Podsakoff et al., 2012). Although we designed our questionnaire carefully and statistical testing The studies involving human participants were reviewed and suggested that common method bias was not an issue in our approved by two ethics committees: by the workers’ council study, we encourage future researchers to collect data from (German: “Gesamtbetriebsrat”) of the participating organization, multiple sources in order to reduce Type-I error probability and by the academic commission of the Ph.D. program HDK 14 (Podsakoff et al., 2012). (“Comissió de Acadèmica del programa de Doctorat HDK 14 - Third, we built our research study on the JDR as it is the Psicologia Social i de les Organitzacions”) of the Universitat de most prominent and broad framework that helps explaining Barcelona (UB). The patients/participants provided their written the indirect relationship between TFL and well-being. Other informed consent to participate in this study. approaches, e.g., leader–member exchange theory (Graen and Uhl-Bien, 1995), can add further information and should be used in future research. AUTHOR CONTRIBUTIONS Fourth, we included only one job resource and job demand as mediators, and the list of potential mediators RB, JC, and DL contributed to definition of research objectives, is unquestionably longer than this (see Arnold, 2017 for a models, and hypotheses. RB and J-PL contributed to provision meta-analysis on mediators and moderators). When it comes of materials (i.e., questionnaires). J-PL contributed to data to role stressors, due to the particularities of working in a collection. RB and DL contributed to data analysis plan. JC research institution, the present study only included RA as a and DL contributed to data analysis. RB and JC contributed to possible mediator. Therefore, future research should examine principal article writing. RB, DL, and J-PL contributed to article the possible mediating effects of role conflict and role overload, revision and proofreading. RB, DL, JC, and J-PL contributed for instance. Regarding the detrimental effect of leader passivity, to final approval. future research should further investigate which resources work for which particularly increased job demand, since it is likely that only specific resources work for specific demands FUNDING (Schaufeli and Taris, 2014). Fifth, the interest of the present study focused on individual The Universitat de Barcelona will pay the open access health. However, future studies should consider analysis at a publishing fee. Frontiers in Psychology | www.frontiersin.org 12 December 2019 | Volume 10 | Article 2788 Berger et al. Leaders Influence Employees’ Well-Being REFERENCES Cohen, J. (1988). Statistical Power Analysis for the Behavioral Sciences, 2nd Edn. Hillsdale, N.J: Lawrence Erlbaum. Alcover, C. M., Rico, R., and Gil, F. (2011). Equipos de trabajo en contextos Cole, M. S., Bruch, H., and Shamir, B. (2009). Social distance as a moderator of organizacionales: dinámicas de cambio, adaptación y aprendizaje en entornos the effects of transformational leadership: both neutralizer and enhancer. Hum. flexibles. Papeles Psicol. 32, 7–16. Relat. 62, 1697–1733. doi: 10.1177/0018726709346377 Arbuckle, J. L. (2017). IBM AMOS 25 User’s Guide. Chicago: SPSS. Dackert, I. (2010). The impact of team climate for innovation on well-being and Arnold, K. (2017). Transformational leadership and employee psychological stress in elderly care. J. Nurs. Manag. 18, 302–310. doi: 10.1111/j.1365-2834. wellbeing: a review and directions for future research. J. Occup. Organ. Psychol. 2010.01079.x 22, 381–393. doi: 10.1037/ocp0000062 Dawson, J. F. (2014). Moderation in management research: what, why, when, and Arnold, K. A., Turner, N., Barling, J., Kelloway, E. K., and McKee, M. C. (2007). how. J. Bus. Psychol. 29, 1–19. doi: 10.1007/s10869-013-9308-7 Transformational leadership and psychological well-being: the mediating role Day, D. V., Gronn, P., and Salas, E. (2004). Leadership capacity in teams. Leadersh. of meaningful work. J. Occup. Health Psychol. 12, 193–203. doi: 10.1037/1076- Q. 15, 857–880. doi: 10.1016/j.leaqua.2004.09.001 8998.12 3.193 Den Hartog, D. N., and Belschak, F. D. (2012). When does transformational Avanzi, L., Schuh, S. C., Fraccaroli, F., and van Dick, R. (2015). Why leadership enhance employee proactive behavior? The role of autonomy and does organizational identification relate to reduced employee burnout? The role breadth self-efficacy. J. Appl. Psychol. 97, 194–202. doi: 10.1037/a0024903 mediating influence of social support and collective efficacy. Work Stress 29, Di Fabio, A. (2017). The psychology of sustainability and sustainable development 1–10. doi: 10.1080/02678373.2015.1004225 for well-being in organizations. Front. Psychol. 8:1534. doi: 10.3389/fpsyg.2017. Avolio, B. J. (1999). Full Leadership Development: Building the Vital Forces in 01534 Organizations. Thousand Oaks, CA: Sage. Diebig, M., Bormann, K. C., and Rowold, J. (2016). A double-edged sword: Avolio, B. J., and Bass, B. M. (2004). Multifactor Leadership Questionnaire. Manual relationship between full-range leadership behaviors and followers’ hair cortisol and Sampler Set, 3rd Edn. Redwood City, CA: Mindgarden. level. Leadersh. Q. 27, 684–696. doi: 10.1016/j.leaqua.2016.04.001 Bakker, A. B., and Demerouti, E. (2007). The job demands−resources model: state Diebig, M., Bormann, K. C., and Rowold, J. (2017). Day-level transformational of the art. J. Manag. Psychol. 22, 309–328. doi: 10.1108/02683940710733115 leadership and followers ’ daily level of stress: a moderated mediation model Bandura, A. (1977). Social Learning Theory. Oxford: Prentice-Hall. of team cooperation, role conflict, and type of communication. Eur. J. Work Barling, J., and Frone, M. R. (2017). If only my leader would just do something! Organ. Psy. 26, 1–16. doi: 10.1080/1359432X.2016.1250741 Passive leadership undermines employee well−being through role stressors and Dobreva-Martinova, T., Villeneuve, M., Strickland, L., and Matheson, K. (2002). psychological resource depletion. Stress Health J. Int. Soc. Investig. Stress 33, Occupational role stress in the canadian forces: its association with individual 211–222. doi: 10.1002/smi.2697 and organizational well-being. Can. J. Behav. Sci. 34, 111–121. doi: 10.1037/ Bass, B. M. (1985). Leadership and Performance Beyond Expectations. New York, h0087161 NY: Free Press. Edmondson, A. C., and Nembhard, I. M. (2009). Product development and Bass, B. M. (1990). Transformational leadership: learning to share the vision. learning in project teams: the challenges are the benefits. J. Prod. Innovat. Organ. Dyn. 18, 19–31. doi: 10.1016/0090-2616(90)90061-S Manag. 26, 123–138. doi: 10.1111/j.1540-5885.2009.00341.x Bass, B. M. (1998). Transformational Leadership: Industrial, Military, and Franke, F., and Felfe, J. (2011). How does transformational leadership impact Educational Impact. Mahwah, NJ: Lawrence Erlbaum Associates. employees’ psychological strain? Examining differentiated effects and the Bass, B. M. (2000). The future of leadership in learning organizations. J. Leadersh. moderating role of affective organizational commitment. Leadership 7, 295–316. Organ. Stud. 7, 18–40. doi: 10.1177/107179190000700302 doi: 10.1177/1742715011407387 Berger, R., and Antonioli, T. C. (2019). Psychometric properties of the human Frooman, J., Mendelson, M. B., and Murphy, J. K. (2012). Transformational and system audit transformational leadership short scale in Germany and passive avoidant leadership as determinants of absenteeism. Leadersh. Org. Dev. Philippines: a cross-cultural study. Psychologica 62, 165–185. doi: 10.14195/ J. 33, 447–463. doi: 10.1108/01437731211241247 1647-8606_62-1_9 Gawke, J., Gorgievski, M., and Bakker, A. (2017). Personal costs and benefits of Berger, R., Romeo, M., Guardia, J., Yepes, M., and Soria, M. A. (2012). employee intrapreneurship: disentangling the employee intrapreneurship, well- Psychometric properties of the spanish human system audit short-scale of being, and job performance relationship. J. Occup. Health Psychol. 23, 508–519. transformational leadership. Spanish J. Psychol. 15, 367–375. doi: 10.5209/ doi: 10.1037/ocp0000105 revSJOP.2012.v15.n1.37343 Geldhof, G. J., Pornprasertmanit, S., Schoemann, A. M., and Little, T. D. (2013). Berger, R., Yepes, M., Gómez Benito, J., Quijano, S., and Brodbeck, F. C. (2011). Orthogonalizing through residual centering: extended applications and caveats. Validity of the human system audit transformational leadership short scale Educ. Psychol. Meas. 73, 27–46. doi: 10.1177/0013164412445473 (HSA-TFL) in four european countries. Univ. Psychol. 10, 657–668. doi: 10. Glazer, S., and Beehr, T. A. (2005). Consistency of implications of three role 11144/Javeriana.upsy10-3.vhsa stressors across four countries. J. Organ. Behav. 26, 467–487. doi: 10.1002/ Bowling, N. A., and Beehr, T. A. (2006). Workplace harassment from the victim’s job.326 perspective: a theoretical model and meta− analysis. J. Appl. Psychol. 91, Glazer, S., and Kruse, B. (2008). The role of organizational commitment in 998–1012. occupational stress models. Int. J. Stress Manag. 15, 329–344. doi: 10.1037/ Brodbeck, F. C., Guillaume, Y. R. F., and Winkler, M. (2010). “Team climate for a0013135 learning in higher education,” in In WOP Working Paper No. 2010/1, (Munich: Graen, G. B., and Uhl-Bien, M. (1995). Relationship-based approach to leadership: LMU). development of leader-member exchange (LMX) theory of leadership over 25 Byrne, B. M. (2009). Structural Equation Modeling with AMOS. New York, NY: years: applying multi-level multi-domain perspective. Leadersh. Q. 6, 219–247. Routledge. doi: 10.1016/1048-9843(95)90036-5 Cabrera-Nguyen, E. P. (2010). Author guidelines for reporting scale development Hackman, J. R., and Oldham, G. R. (1976). Motivation through the design of and validation results in the journal of the society for social work and research. work: test of a theory. Organ. Behav. Hum. Decis. Process. 16, 250–279. doi: J. Soc. Soc. Work Res. 1, 99–103. doi: 10.5243/jsswr.2010.8 10.1016/0030-5073(76)90016-7 Cerin, E. (2010). Ways of unraveling how and why physical activity influences Hair, J. F., Black, W. C., Babin, B. J., Anderson, R. E., and Tatham, R. L. (2006). mental health through statistical mediation analyses. Mental. Health Phys. Act. Multivariate Data Analysis. Upper Saddle River, NJ: Prentice Hall Pearson 3, 51–60. doi: 10.1016/j.mhpa.2010.06.002 Education. Chênevert, D., Jourdain, G., Cole, N., and Banville, B. (2013). The role of Halbersleben, J. R. B., Neveu, J.-P., Paustian-Underdahl, S. C., and Westman, organizational justice, burnout and commitment in the understanding of M. (2014). Getting to the “COR”: understanding the role of resources absenteeism in the canadian healthcare sector. J. Health Organ. Manag. 27, in conservation resources theory. J. Manag. 40, 1334–1364. doi: 10.1177/ 350–367. doi: 10.1108/JHOM-06-2012-0116 0149206314527130 Cheng, B. H., and McCarthy, J. M. (2018). Understanding the dark and bright Harms, P. D., Credé, M., Tynan, M., Leon, M., and Jeung, W. (2017). Leadership sides of anxiety: a theory of workplace anxiety. J. Appl. Psychol. 103, 537–560. and stress: a meta-analytic review. Leadersh. Q. 28, 178–194. doi: 10.1016/j. doi: 10.1037/apl0000266 leaqua.2016.10.006 Frontiers in Psychology | www.frontiersin.org 13 December 2019 | Volume 10 | Article 2788 Berger et al. Leaders Influence Employees’ Well-Being Heinitz, K., Liepmann, D., and Felfe, J. (2005). Examining the factor structure of Mathieu, J., Maynard, M. T., Rapp, T., and Gilson, L. (2008). Team effectiveness the MLQ: recommendation for a reduced set of factors. Eur. J. Psychol. Assess. 1997-2007: a review of recent advancements and a glimpse into the future. 21, 182–190. doi: 10.1027/1015-5759.21.3.182 J. Manag. 34, 410–476. doi: 10.1177/0149206308316061 Hentrich, S., Zimber, A., Garbade, S. F., Gregersen, S., Nienhaus, A., and McMurray, A. J., Pirola−Merlo, A., Sarros, J. C., and Islam, M. M. (2010). Petermann, F. (2017). Relationships between transformational leadership and Leadership, climate, psychological capital, commitment, and wellbeing in a health: the mediating role of perceived job demands and occupational self- non−profit organization. Leader. Organ. Dev. J. 31, 436–457. doi: 10.1108/ efficacy. Int. J. Stress Manag. 24, 34–61. doi: 10.1037/str0000027 01437731011056452 Hetland, H., Skogstad, A., Hetland, J., and Mikkelsen, A. (2011). Leadership and Molero Alonso, F., Saboya, P., and Cuadrado Guirado, I. (2010). Transformational learning climate in a work setting. Eur. Psychol. 16, 163–173. doi: 10.1027/1016- and transactional leadership: an analysis of the factor structure of the 9040/a000037 multifactor leadership questionnaire (MLQ) in a spanish sample. Psicothema Hobfoll, S. E., and Lilly, R. S. (1993). Resource conservation as a strategy for 22, 495–501. community psychology. J. Community Psychol. 21, 128–148. doi: 10.1002/1520- Montano, D., Reeske, A., Franke, F., and Hüffmeier, J. (2017). Leadership, followers 6629(199304)21:2<128::aid-jcop2290210206>3.0.co;2-5 mental health and job performance in organizations: a comprehensive meta- Holstad, T. J., Korek, S., Rigotti, T., and Mohr, G. (2014). The relation analysis from an occupational health perspective. J. Org. Behav. 38, 327–350. between transformational leadership and follower emotional strain: the doi: 10.10002/job2124 moderating role of professional ambition. Leadership 10, 269–288. doi: 10.1177/ Morgeson, F., Delaney−Klinger, K., and Hemingway, A. M. (2005). The 1742715013476083 importance of job autonomy, cognitive ability, and job-related skill for Hu, L.-T., and Bentler, P. M. (1998). Fit indices in covariance structure modeling: predicting role breadth and job performance. J. Appl. Psychol. 90, 399–406. sensitivity to underparameterized model misspecification. Psychol. Methods 3, doi: 10.1037/0021-9010.90.2.399 424–453. doi: 10.1037/1082-989X.3.4.424 Navarro, J., Quijano, S. D., Berger, R., and Meneses, R. (2011). Grupos en Ilgen, D. R., Hollenbeck, J. R., Johnson, M., and Jundt, D. (2005). Teams in las organizaciones: herramienta básica para gestionar la incertidumbre y organizations: from input-process-output models to IMOI models. Annu. Rev. ambigüedad crecientes. Papeles Psicol. 32, 17–28. Psychol. 56, 517–543. doi: 10.1146/annurev.psych.56.091103.070250 Nielsen, K., Brenner, S. O., Randall, R., and Yarker, J. (2008). The effects of Inceoglu, I., Thomas, G., Chu, C., Plans, D., and Gerbasi, A. (2018). Leadership transformational leadership on followers’ perceived work characteristics and behavior and employee well-being: an integrated review and a future research psychological well-being: a longitudinal study. Work Stress 22, 16–32. doi: 10. agenda. Leadersh. Q. 29, 179–202. doi: 10.1016/j.leaqua.2017.12.006 1080/02678370801979430 International Test Commission, (2017). The ITC Guidelines for Translating and Nielsen, K., Nielsen, M. B., Ogbonnaya, C., Känsälä, M., Saari, E., and Isaksson, Adapting Tests, 2ed Edn. Kolkata: ITC. K. (2017). Workplace resources to improve both employee well-being and Ju, D., Ma, L., Ren, R., and Zhang, Y. (2019). Empowered to break the silence: performance: a systematic review and meta-analysis. Work Stress 31, 101–120. applying self-determination theory to employee silence. Front. Psychol. 10:485. doi: 10.1080/02678373.2017.1304463 doi: 10.3389/fpsyg.2019.00485 Nielsen, K., Yarker, J., Randall, R., and Munir, F. (2009). The mediating effects of Judge, T. A., and Piccolo, R. F. (2004). Transformational and transactional team and self-efficacy on the relationship between transformational leadership, leadership: a meta-analytic test of their relative validity. J. Appl. Psychol. 89, and job satisfaction and psychological well-being in healthcare professionals: 755–768. doi: 10.1037/0021-9010.89.5.755 a cross-sectional questionnaire survey. Int. J. Nurs. Stud. 46, 1236–1244. doi: Jung, D. I., and Sosik, J. J. (2002). Transformational leadership in work groups: the 10.1016/j.ijnurstu.2009.03.001 role of empowerment, cohesiveness, and collective-efficacy on perceived group Oc, B. (2018). Contextual leadership: a systematic review of how contextual factors performance. Small Group Res. 33, 313–336. doi: 10.1177/10496402033003002 shape leadership and its outcomes. Leadersh. Q. 29, 218–235. doi: 10.1016/j. Katz, D., and Kahn, R. L. (1978). The Social Psychology of Organizations. New York: leaqua.2017.12.004 Wiley. Örtqvist, D., and Wincent, J. (2006). Prominent consequences of role stress: a meta- Kelloway, E. K., and Barling, J. (2010). Leadership development as an intervention analytic review. Int. J. Stress Manag. 13, 399–422. doi: 10.1037/1072-5245.13. in occupational health psychology. Work Stress 24, 260–279. doi: 10.1080/ 4.399 02678373.2010.518441 Parker, D. F., and DeCotiis, T. A. (1983). Organizational determinants of job Kelloway, K., Turner, N., Barling, J., and Loughlin, C. (2012). Transformational stress. Organ. Behav. Hum. Perform. 32, 160–177. doi: 10.1016/0030-5073(83)90 leadership and employee psychological well-being: the mediating role of 145-9 employee trust in leadership. Work Stress 26, 39–55. doi: 10.1080/02678373. Peters, A., McEwen, B. S., and Friston, K. J. (2017). Uncertainty and stress: why 2012.660774 it causes diseases and how it is mastered by the brain. Prog. Neurobiol. 156, Kline, R. B. (2016). Principles and Practice of Structural Equation Modeling, 4th 164–188. doi: 10.1016/j.pneurobio.2017.05.004 Edn. New York, NY: The Guilford Press. Podsakoff, P., Mackenzie, S. B., and Podsakoff, N. P. (2012). Sources of method bias Koutsimani, G., Montgomery, A., and Georganta, K. (2019). The relationship in social science research and recommendations on how to control It. Annu. between burnout, depression, and anxiety: a systematic review and meta- Rev. Psychol. 63, 539–569. doi: 10.1146/annurev-psych-120710-10045 analysis. Front. Psychol. 10:238. doi: 10.3389/fpsyg.2019.00284 Podsakoff, P. M., MacKenzie, S. B., Lee, J.-Y., and Podsakoff, N. P. (2003). Common Kuoppala, J., Lamminpää, A., Liira, J., and Vainio, H. (2008). Leadership, job well- method biases in behavioral research: a critical review of the literature and being, and health effects: a systematic review and a meta-analysis. J. Occup. recommended remedies. J. Appl. Psychol. 88, 879–903. doi: 10.1037/0021-9010. Environ. Med. 50, 904–915. doi: 10.1097/JOM.0b013e31817e918d 88.5.879 Lases, L., Arah, O. A., Busch, O., Heineman, M. J., and Lombarts, K. (2019). Preacher, K. J., and Hayes, A. F. (2008). Asymptotic and resampling strategies for Learning climate positively influences residents’ work-related well-being. Adv. assessing and comparing indirect effects in multiple mediator models. Behav. Health Sci. Educ. Theory Pract. 24, 317–330. doi: 10.1007/s10459-018-9868-4 Res. Methods 40, 879–891. doi: 10.3758/BRM.40.3.879 Leuteritz, J.-P., Navarro, J., and Berger, R. (2017). How knowledge worker teams Pyc, L. S., Meltzer, D. P., and Liu, C. (2016). ). Ineffective leadership and employees’ deal effectively with task uncertainty: the impact of transformational leadership negative outcomes: the mediating effect of anxiety and depression. Int. J. Stress and group development. Front. Psychol. 8:1339. doi: 10.3389/fpsyg.2017. Manag. 24, 196–215. doi: 10.1037/str0000030 01339 Rajgopal, T. (2010). Mental well-being at the workplace. Indian J. Occup. Environ. Little, T. D., Bovaird, J. A., and Widaman, K. F. (2006). On the merits Med. 14, 63–65. doi: 10.4103/0019-5278.75691 of orthogonalizing powered and product terms: implications for modeling Ramírez Heller, B., Berger, R., and Brodbeck, F. C. (2014). Does an adequate team interactions among latent variables. Struct. Equ. Model. 13, 497–519. doi: 10. climate for learning predict team effectiveness and innovation potential? A 1207/s15328007sem1304_1 psychometric validation of the team climate questionnaire for learning in an Marín Puchades, V. (2015). Leadership and Stress in Team Settings. master thesis, organizational context. Procedia Soc. Behav. Sci. 114, 543–550. doi: 10.1016/j. Universitat de Barcelona, Barcelona sbspro.2013.12.744 Frontiers in Psychology | www.frontiersin.org 14 December 2019 | Volume 10 | Article 2788 Berger et al. Leaders Influence Employees’ Well-Being Reuveni, Y., and Vashdi, D. R. (2015). Innovation in multidisciplinary teams: the Skogstad, A., Einarsen, S., Torsheim, T., Aasland, M. S., and Hetland, H. (2007). moderating role of transformational leadership in the relationship between The destructiveness of laissez-faire leadership behavior. J. Occup. Health professional heterogeneity and shared mental models. Eur. J. Work and Organ. Psychol. 12, 80–92. doi: 10.1037/1076-8998.12.1.80 Psy. 24, 678–692. doi: 10.1080/1359432X.2014.1001377 Tafvelin, S., Armelius, K., and Westerberg, K. (2011). Toward understanding Savelsbergh, C., Gevers, J. M. P., van der Heijden, B. I. J. M., and Poell, R. F. (2012). the direct and indirect effects of transformational leadership on well-being: Team role stress: relationships with team learning and performance in project a longitudinal study. J. Leadersh. Organ. Stud. 18, 480–492. doi: 10.1177/ teams. ıGroup Organ. Manag. 37, 67–100. doi: 10.1177/1059601111431977 1548051811418342 Schaufeli, W. B. (2015). Engaging leadership in the job demands-resources model. Volmer, J., Spurk, D., and Niessen, C. (2012). Leader–member exchange (LMX), Career Dev. Int. 20, 446–463. doi: 10.1108/CDI-02-2015-0025 job autonomy, and creative work involvement. Leadersh. Q. 23, 456–465. doi: Schaufeli, W. B. (2017). Applying the job demands-resources model: a ‘how to’ 10.1016/j.leaqua.2011.10.005 guide to measuring and tackling work engagement and burnout. Organ. Dyn. Wittchen, H. U., Jacobi, F., Rehm, J., Gustavsson, A., Svensson, M., Jonsson, B., 46, 120–132. doi: 10.1016/j.orgdyn.2017.04.008 et al. (2011). The size and burden of mental disorders and other disorders Schaufeli, W. B., and Taris, T. W. (2014). “A critical review of the job demands- of the brain in Europe 2010. Eur. Neuropsychopharmacol. 21, 655–679. doi: resources model: implications for improving work and health,” in Bridging 10.1016/j.euroneuro.2011.07.018 Occupational, Organizational and Public Health: A Transdisciplinary Approach, Xanthopoulou, D., Bakker, A., Demerouti, E., and Schaufeli, W. (2009). Reciprocal eds G. F. Bauer, and O. Hämmig (New York, NY: Springer Science + Business relationships between job resources, personal resources, and work engagement. Media), 43–68. doi: 10.1007/978-94-007-5640-3_4 J. Vocat. Behav. 74, 235–244. doi: 10.1016/j.jvb.2008.11.003 Schmidt, K.-H., Kleinbeck, U., Ottmann, W., and Seidel, B. (1985). Ein Verfahren zur diagnose von arbeitsinhalten: der job diagnostic survey (JDS). Z. Arbeits Conflict of Interest: The authors declare that the research was conducted in the Organisationspsychol. 29, 162–172. absence of any commercial or financial relationships that could be construed as a Schulz, J. (2013). The impact of role conflict, role ambiguity and organizational potential conflict of interest. climate on the job satisfaction of academic staff in research-intensive universities in the UK. Higher Educ. Res. Dev. 32, 464–478. doi: 10.1080/ Copyright © 2019 Berger, Czakert, Leuteritz and Leiva. This is an open-access article 07294360.2012.680209 distributed under the terms of the Creative Commons Attribution License (CC BY). Skakon, J., Nielsen, K., Borg, V., and Guzman, J. (2010). Are leaders’ well-being, The use, distribution or reproduction in other forums is permitted, provided the behaviours and style associated with the affective well-being of their employees? original author(s) and the copyright owner(s) are credited and that the original A systematic review of three decades of research. Work Stress 24, 107–139. publication in this journal is cited, in accordance with accepted academic practice. No doi: 10.1080/02678373.2010.495262 use, distribution or reproduction is permitted which does not comply with these terms. Frontiers in Psychology | www.frontiersin.org 15 December 2019 | Volume 10 | Article 2788