The Role of Creative Leadership Manifestations in Creativity and Innovation

ABSTRACT

Creative leadership, leading others toward the attainment of creative outcomes, manifests differently depending on the context. Theory suggests that contextual features can produce three distinct manifestations of leadership – integrative, facilitative, and directive – but that context may vary in strength, potentially allowing for multiple manifestations of creative leadership. This study reports on the development and validation of a scale for measuring the three manifestations and their differential effects on creativity and innovation. The investigation occurred within a relatively weak context wherein multiple manifestations of creative leadership are possible. Examining 646 research unit members nested in 259 research units from nine German universities, this study reveals that the research units’ idea generation was positively related to the facilitating and integrating creative leadership manifestation, while the units’ innovative outcome was positively related to the directing and integrating creative leadership manifestation. This study contributes to understanding and measuring creative leadership and the distribution of creative and supportive tasks between leaders and followers.

Plain Language Summary

This study presents a scale for measuring the creative leadership manifestations currently present within a research unit. Measuring the creative leadership manifestations can help organizations evaluate the distribution of creative and supportive contributions between their leaders and followers. This is important, as different manifestations of creative leadership can be useful in different ways depending on the aim of the unit. This study indicates that the followers should make creative contributions and the leader can either make creative contributions or not if the research unit aims for creativity. If the research unit aims to produce innovative outcomes, this study suggests that it is particularly important for leaders to make creative contributions. Understanding this can help organizations create a context in which the creative tasks are distributed between leaders and followers in a purposeful way. Further, making leaders aware of creative leadership manifestations and their influence on their followers’ work can help them create a work environment that supports the unit in their creative and innovative tasks.

Introduction

Leadership is context-dependent (Osborn, Hunt, & Jauch, 2002). What it takes to engage followers and foster collaboration depends on various contextual features, including social, task and organizational characteristics (Day, Fleenor, Atwater, Sturm, & McKee, 2014; Oc, 2018; Stahl, Filatotchev, Ireland, & Miska, 2023). Context is particularly crucial for leaders to foster creativity and innovation (e.g., Amabile, 2018; Shalley, Zhou, & Oldham, 2004; Zhou & Hoever, 2014, 2023). Creative leadership can be broadly defined as “leading others toward the attainment of a creative outcome” (Mainemelis, Kark, & Epitropaki, 2015, p. 399), but how it is behaviorally manifested varies systematically with context. Depending on the needs, inducements and constraints in the context, leaders can be facilitators, enablers or active participants in the creative process (e.g., Oldham & Cummings, 1996; Somech, 2006; Tierney, Farmer, & Graen, 1999; Zhang & Bartol, 2010). In a comprehensive review of creative leadership, Mainemelis, Kark, and Epitropaki (2015) identified three distinct manifestations of creative leadership that are the product of contextual factors. The first is facilitating, which corresponds to the traditional view of creative leadership as supporting and offering resources and not actively contributing. The second is directing, wherein the leader is the primary generator of creative contributions and the followers primarily undertake supportive tasks. The third is integrating, wherein both the leader and the followers generate creative contributions, which the leader then synthesizes to establish a creative synergy (Mainemelis, Kark, & Epitropaki, 2015).

However, the authors acknowledge that contextual factors do not necessarily align to consistently prompt one of these three manifestations of creative leadership. Contexts comprising mixed demands and constraints can be considered weak contexts (Meyer, Dalal, & Hermida, 2010). Given the potential for mixed or competing demands, cues, and constraints, relatively weak contexts may render creative leadership manifestations that are idiosyncratic and do not produce consistent levels of creativity and innovation. Importantly, the way relatively weak contexts shape the manifestation of creative leadership and the consequences for creativity and innovation are not well understood. The lack of empirical insight into weak creative contexts is noteworthy, considering that mixed and even competing demands are increasingly common in the complex contexts of contemporary organizations.

Another critical ambiguity in theory and research on creative contexts is whether those contexts are stable throughout the creative process. Broadly speaking, the creative process can be delineated by distinct activities of creativity and innovation (Anderson, Potočnik, & Zhou, 2014). Creativity or idea generation is often an initial phase of the creative process in which numerous ideas are created, fostering creativity and exploration. Innovation is the subsequent process of refining and implementing selected ideas to create something valuable. We maintain that the configuration of contextual demands of each of these activities is distinct. We therefore posit that in weakly structured contexts, demands and cues are both multitudinous and dynamic, resulting in different patterns of creative leadership across the process of creativity and innovation.

This investigation addresses this important gap by examining creative leadership in a relatively weak creative context, namely, academic research. While creativity and innovation are a cornerstone of research (Heinze, Shapira, Rogers, & Senker, 2009; Keller, 2006), we argue that academic research occurs in a context of diverse and mixed features that prompt multiple manifestations of creative leadership. For example, the academic research enterprise tends to be organized in flexible, fluid, temporary, and networked forms that are apt to prompt integrative leadership. At the same time, the pedagogical demands of training new scientists and scholars may prompt more directive leadership. We further argue that the multiple manifestations of creative leadership may be necessary to produce creativity and innovation, but that the overall pattern differs between the two activities. To that end, we test our hypotheses in the context of academic research units within universities in Germany. Positing that in the academic research context all three creative leadership manifestations are possible and commonly occurring, we directly compare their effects on creativity and innovation.

The research aims and contribution of this study are threefold. First, we extend the work of Mainemelis, Kark, and Epitropaki (2015), by applying it to a relatively weak context. Given the potential for idiosyncratic behavior in response to weaker contexts, we seek to identify the constellation of leadership manifestations that leads to creativity and innovation. In doing so, we add validation and nuance to existing theory on creative leadership contexts. This may also be beneficial for research on creative leadership in contexts outside academia, given that the rise of temporal and networked forms of organizing is likely to create more contexts with mixed demands regarding the distribution and management of creative work.

Second, while much of the work on creative leadership contexts to date has been conceptual or qualitative (especially concerning directing and integrating creative leadership; cf. Kark, Epitropaki, & Mainemelis, 2020), we seek to provide empirical, quantitative validation of the framework. To that end, we develop and validate a measure of the facilitating, directing, and integrating creative leadership manifestations. Proposing that the facilitating and integrating creative leadership manifestation are fostering factors for creative behavior, while the integrating and directing manifestation focus followers’ efforts on achieving innovative outcomes, the study tests these propositions and examines the combined and unique influence of each. Thereby, it provides a more comprehensive understanding of the role of creative leadership in creativity and innovation.

Third, we address calls for careful differentiation between creativity, the process of idea generation, and innovation, the process of idea implementation (Anderson, Potočnik, & Zhou, 2014; Hughes, Lee, Tian, Newman, & Legood, 2018; Montag, Maertz, & Baer, 2012; Perry-Smith & Mannucci, 2016). In this way, we explore whether creative leadership manifestations have different consequences depending on the phase of the creative process. Fourth, it contributes to the understanding of the organization of research units (Shibayama, Baba, & Walsh, 2015) by shedding light on how different leadership manifestations can foster scientific creativity (cf. Heinze, Shapira, Rogers, & Senker, 2009) and the innovative outcome of research units.

Theory

Workplace creativity is commonly defined as the generation of ideas for products, processes, or services, which are both novel and useful for the organization (Amabile, Conti, Coon, Lazenby, & Herron, 1996; Montag, Maertz, & Baer, 2012; Oldham & Cummings, 1996). The outcome of a creative process is an idea (Hughes, Lee, Tian, Newman, & Legood, 2018) and the creativity of an idea can range from incremental, implying minor modifications of existing practices, to radical, implying substantial differences from the current practices in the organization (Madjar, Greenberg, & Chen, 2011; Mumford & Gustafson, 1988). In contrast, innovation usually refers to intentional implementation of ideas, products, or processes, which are new to the organization (West & Farr, 1990). The outcome of an innovative process is therefore a functioning and implemented idea (Hughes, Lee, Tian, Newman, & Legood, 2018). While idea generation requires intrinsic motivation, a supportive atmosphere for creativity, cognitive flexibility, and (creative thinking) skills, idea implementation requires shared vision, understanding, integration skills, and resources (Amabile, 1988; Hunter, Bedell, & Mumford, 2007; Perry-Smith & Mannucci, 2016; van Knippenberg, 2017; West, 2002).

Even though creativity and innovation are closely related constructs, it is important to distinguish between them: Creativity is defined as the generation of novel and useful ideas (Amabile, 1988; Oldham & Cummings, 1996), whereas innovation refers to the conversion of these ideas into new and improved products, processes, or services (West, 2002). While individuals may share their ideas with the group, only when they are successfully implemented are they considered an innovation (Mumford & Gustafson, 1988; Shalley, Zhou, & Oldham, 2004). Implementation is vital, as there is a wide range of opportunity for failure between uttering an idea and implementing it in a product or process (e.g., not finding support for the idea in the team, not being able to mobilize the necessary resources, losing motivation, or being distracted by other work tasks; cf. Perry-Smith & Mannucci, 2016). Thus, creativity may be suitably conceptualized as a first step in the innovative process and an important antecedent for subsequent innovation (West & Farr, 1990). Empirical research similarly suggests the two are not perfectly related (as shown in a meta-analysis by Sarooghi, Libaers, & Burkemper, 2015) and researchers have called for them to be investigated separately (Montag, Maertz, & Baer, 2012), specifically regarding the effects of leadership on creativity and innovation (Hughes, Lee, Tian, Newman, & Legood, 2018). Therefore, in line with empirical research indicating that the two depend on different processes and lead to different outcomes (Anderson, Potočnik, & Zhou, 2014; Hughes, Lee, Tian, Newman, & Legood, 2018), the two are investigated as separate constructs.

Contexts of creative leadership

Creativity and innovation in organizations have been recognized as being the joint result of the work of individual members, leaders, and the organizational context (Hughes, Lee, Tian, Newman, & Legood, 2018; Zhou & Hoever, 2023). In a large-scale integrative review, Mainemelis, Kark, & Epitropaki (2015) note that creative leadership research so far has been located in distinctly different organizational contexts (including jazz combos, haute cuisine, and R&D teams in large corporations). Thus, understanding creative leadership requires unpacking the concept of context.

Context refers to the characteristics of the environment and the situation in which organizational behavior is located, and affects organizational behavior by setting particular constraints, providing opportunities, and altering the meaning of behaviors (Johns, 2018). According to Johns (2006), context comprises two levels, the broader or omnibus context, which refers to general characteristics like occupation and location, and the discrete context, which includes specific variables such as task and social characteristics that directly affect behavior within a given omnibus context. Thus, the discrete context is nested within the omnibus context and mediates the impact of the omnibus context on behavior. Focusing on the effects of context on creative behavior, Mainemelis, Kark, and Epitropaki (2015) identified three collaborative contexts that pre-structure the distribution of tasks in creative work in certain ways, giving rise to three distinct manifestations of creative leadership. These contexts are mainly distinguished by the distribution of creative and supportive contributions made by leaders and followers, respectively: Facilitative contexts (where leaders foster the creativity of followers), integrative contexts (where leaders integrate the various creative contributions made by followers and themselves), and directive contexts (where leaders implement their own creative ideas with the help of their followers). In delineating these three creative contexts, Mainemelis, Kark, and Epitropaki (2015) are able to explain the creative process in widely differing contexts such as R&D (facilitative), jazz ensembles (integrative), and haute cuisine (directive).

Mainemelis, Kark, and Epitropaki (2015) argued that three creative contexts arise from the configuration of several discrete contextual factors that reflect the specific cultural, industry, organizational, personal, and task characteristics embedded in the omnibus context (Mainemelis, 2018). For example, organizational context includes features such as size, structure, and the strategic role of creativity. Social structure refers to social roles, expectations, and norms (Carmeli & Schaubroeck, 2007). Task characteristics include features such as the degree to which tasks are scripted and whether the work is project-based or ongoing, and personal characteristics include individual differences such as personality (Kirton, 1976), motivation, and skills (Amabile, 1988). Nested within the omnibus context, the configuration of these discrete context features determines the nature of collaborative context and hence, the manifestation of creative leadership. In this way, creative leadership manifestations are located in the contextual configuration, forming a connection between the omnibus and discrete context and the behaviors and attitudes they influence (Mainemelis, 2018).

Creative leadership manifestations

The three creative leadership manifestations (i.e., facilitative, integrative, and directive creative leadership) are different approaches to distributing creative and supportive tasks between leaders and followers to attain an innovative outcome. Importantly, the concept does not focus on leaders’ behaviors but on the patterns in which creative work is structured and organized. As such, creative leadership manifestations denote a feature of the discrete context of creative work, and specifically, of the task and social context (Johns, 2006). While some omnibus contexts strongly prompt one particular creative leadership manifestation, other contexts give no such clear prescription regarding the manifestation of creative leadership (Mainemelis, Kark, & Epitropaki, 2015). Thus, we argue that the three creative leadership manifestations may occur to varying degrees, and also occur in combination within a particular omnibus context. We argue that creative leadership manifestations shape the norms and expectations of the group regarding what the leader and the followers should do with regard to creativity and innovation. Thus, creative leadership manifestations are reflected in the shared perception of followers regarding how creative and supportive tasks are distributed between leaders and followers.

In the facilitating creative leadership manifestation, the generation of new ideas is primarily the task of the followers, while leaders contribute to the creative process in a more supportive and evaluative way (cf. Koseoglu, Liu, & Shalley, 2017; Mainemelis, Kark, & Epitropaki, 2015; Mumford, Connelly, & Gaddis, 2003). This manifestation is theoretically based on a prominent perspective within the organizational creativity literature: It is the task of leaders to supply followers with the required resources and establish an environment conducive to creative work, by providing encouragement, support, or a good team climate, for example (e.g., Amabile, Conti, Coon, Lazenby, & Herron, 1996; Janssen, 2005; Madjar, Oldham, & Pratt, 2002; Mumford, Scott, Gaddis, & Strange, 2002; Zhang & Bartol, 2010). Thus, the facilitating manifestation entails the expectation that followers contribute to the creative process by generating novel ideas, and leaders support them in further developing these ideas and moving them toward implementation.

In the integrating creative leadership manifestation, the development of new ideas is the task of both the leader and the followers. The leader synthesizes the heterogeneous creative contributions from both the leader and the followers to establish creative synergy (cf. Mainemelis, Kark, & Epitropaki, 2015). This manifestation is theoretically related to the knowledge integration perspective introduced by van Knippenberg (2017), which stressed the importance of integrating knowledge, perspectives, ideas, and expertise from diverse individuals. First, exposure to and exchange of the diverse information available within the group fosters creativity and the development of new insights in the creative collaboration between the leader and the followers. Second, the group offers a natural platform for the integration and exploitation of diverse information (van Knippenberg, 2017). In the interaction between the leader and the followers, this integration results in a creative synergy, where the collective creative outcomes are greater than the sum of the individual input. The more the followers engage in information integration, the more innovative the outcomes should be (van Knippenberg, 2017). The integrating manifestation entails the expectation that both the leader and the followers make creative contributions and both sides may similarly generate supportive contributions. Besides a single creative leader integrating the heterogeneous contributions of members, integrative creative leadership can also entail creative leadership rotating among the members of a group, being dual in the form of two persons sharing the leadership responsibility (among either two creative leaders or one creative and one administrative leader), or being collectively shared among more than two leaders (Mainemelis, 2018; Mainemelis, Kark, & Epitropaki, 2015).

In the directing creative leadership manifestation, the development of new ideas is primarily the task of the leaders, who direct their followers to provide mainly supportive contributions in order to implement their own creative ideas (cf. Mainemelis, Kark, & Epitropaki, 2015). This manifestation is theoretically related to the idea of the “creative genius” (Simonton, 2003). First, the creative genius has an extraordinary creative capacity. Second, they realize their own creative potential and productively turn it into original innovations. These unusual creative abilities, combined with the drive to create numerous innovations, put the creative genius in an exceptional position to direct subordinates to implement the creative visions of the genius (Simonton, 1988). The directing manifestation entails the expectation that followers make supportive contributions to implement the creative vision of the leader.

Context strength and the academic research context

In their review, Mainemelis, Kark, and Epitropaki (2015) identify strongly structured contexts that are prototypical for one of the three manifestations of creative leadership. These contexts can be conceptualized as mutually exclusive categories of creative leadership manifestations. For example, Mainemelis (2018) suggests that directive creative leadership is more likely to occur in permanent, highly institutionalized, relatively small organizational contexts, where authentic creativity is key to the organization and its members. Facilitative creative leadership is posited to occur in more traditional corporate settings, where creative identity is not seen as essential to the business and its members. Integrative leadership appears to be more frequent in temporary, highly networked, relatively small, egalitarian organizations, where creativity is central to the organization’s products and leadership (Mainemelis, 2018).

However, Mainemelis, Kark, and Epitropaki (2015) acknowledge that contextual features may more or less strongly shape the manifestation of creative leadership. Consider, for example, a commercial bank that is working on a financial technology innovation and has hired new, young employees from a tech startup to develop the technology. While the task and personal characteristics may point to a facilitative manifestation of creative leadership, the risk-averse culture and regulatory environment may encourage a more directive approach. Thus, Mainemelis, Kark, and Epitropaki (2015) propose that different contexts “fall[…] on a continuum from ‘weakly’ to ‘strongly’ structured” (Mainemelis, Kark, & Epitropaki, 2015, p. 454). In this way, the strength of a context is the extent to which the context drives one of the leadership manifestations to the exclusion of other leadership manifestations. Situational strength has been proposed to encompass four facets (Meyer, Dalal, & Hermida, 2010): clarity (cues regarding work requirements and responsibilities are easily available and understandable), consistency (different cues regarding work requirements and responsibilities are compatible with each other), constraint (the extent of limitation by outside forces), and consequences (decisions or actions have important outcomes).

Strongly structured creative contexts would therefore suggest a particular leadership manifestation as the norm that guides the distribution of creative tasks. In weakly structured contexts, however, there is likely to be more variability and choice in structuring the distribution of creative and supportive tasks between themselves and their members (Gebert, Boerner, & Kearney, 2010; Mainemelis, Kark, & Epitropaki, 2015). Indeed, first qualitative studies have reported observing more than one creative leadership manifestation within the same omnibus context (e.g., directing and integrating in choreographers/modern dance groups, Abecassis-Moedas & Gilson, 2018/Rouse & Harrison, 2022; facilitating and integrating in chefs, Feuls, Stierand, Dörfler, Boje, & Haley, 2021).

Moreover, a high prevalence of one leadership manifestation in a particular (omnibus) context might suggest that this manifestation is functional for achieving the goals and outcomes that are key to an organization’s success. For weakly structured creative contexts, however, there are no clear predictions so far regarding the consequences and functionality of a leader’s choice of facilitative, directive, or integrative manifestation, or a combination thereof.

To address this gap, our research focuses on investigating how creative tasks are structured and distributed between leaders and followers in “weak” creative contexts where the creative work is not clearly pre-structured. We assume that creative leadership manifestations may be more variable and multi-faceted in such contexts and have a more or less positive impact on creativity and innovation. Thus, we propose that these patterns may be more or less functional for attaining different types of creative outputs (e.g., innovation quality or innovation quantity) in a given industry or creative field. In the following section, we will characterize the context in which our study is located, as such a weakly structured creative context.

Academic research as a weakly structured creative context

On Mainemelis, Kark, and Epitropaki's (2015) proposed continuum of strongly to weakly pre-structured creative contexts, we propose that academic research is rather weakly pre-structured, allowing for variability of creative leadership manifestations, as we will argue below. Academic research hosts a plurality of work environments and research communities, with international collaborations and interdisciplinary interactions. We therefore argue that in the academic context, situational cues regarding work requirements and responsibilities are often not very clear, mixed, or even inconsistent with each other, which is typical for contexts of low situational strength (cf. Meyer, Dalal, & Hermida, 2010). Below, we explain in more detail how this results in all three creative leadership manifestations being likely to be present in academic research as a creative context.

On the one hand, academic research has highly institutionalized elements, being embedded in the permanent and formalized structures of large universities. Every research discipline has particular standards regarding the research process, such as what counts as an acceptable and valid methodology, how findings are presented and communicated, and the ethical norms that are considered a standard within the discipline. For example, institutionalized practices for the serial production of creative outcomes have been empirically documented in a STEM research context (Heinze, Shapira, Rogers, & Senker, 2009). At the same time, research is organized in flexible, fluid, temporary, and networked forms that exist parallel to the university institutions. While some research units may work together in relatively stable teams, with the same group of coauthors working on a research topic throughout many projects and papers, each publication project allows for an entirely new team to be built. The size and composition of such project teams can also vary to a great degree, from two authors with similar qualifications belonging to the same research group to large international and interdisciplinary teams.

According to Mainemelis (2018), facilitative and directing creative leadership are prevalent in permanent organizational structures, while integrative creative leadership typically occurs in temporary, networked organizations. The simultaneous presence of permanent and fluid structures renders academic research a relatively weak context with inconsistent demands regarding work structure and organization (cf. Meyer, Dalal, & Hermida, 2010). Thus, we expect all three leadership manifestations to occur in academia.

With exploration and discovery being essential elements of research, creativity is an important aspect of academic work (Lee, Walsh, & Wang, 2015; Simonton, 2004) which has been reported to be a part of professors’ self-perception and goal striving (Shreffler et al., 2021). However, academic research is also a context in which junior researchers are being trained to become independent scholars in the future (Shibayama, 2019), and empirical research supports the view that followers (in this case junior researchers) develop their creative skills (Amabile, 1988) in the long run with the help of the supervisors, in this case professors, acting as their creative mentors during their PhD studies (Wang & Shibayama, 2022). For research group heads, this requires a compromise between granting autonomy to the junior researchers to pursue their own ideas and managing the research lab toward productive outcomes in line with the unit head’s research program (Shibayama, 2019). Thus, while it is clear that creativity is expected from research unit leaders (which might point to directive or integrative leadership being most prevalent, Mainemelis, 2018), research also requires a substantial amount of facilitative creative leadership in the training of junior researchers.

In summary, we expect creative leadership to manifest in all three ways in the context of academic research: facilitative, directive, and integrative. This is mainly due to the multiple and somewhat competing contextual factors in the organization (i.e., institutionalized and permanent) and team (i.e., flexible and fluid) structure of the academic research enterprise. However, the contextual demands are dynamic as the creative process unfolds. Specifically, the task and relevant contextual features involved in creativity, where idea generation is the end goal, are different from the task demands of innovation, where the research product is the end goal. In the following sections, we will develop our hypotheses regarding the consequences of the different leadership manifestations on creativity and innovation as outcomes in academic research.

The relation of creative leadership manifestations to creativity and innovation

In this study, we seek to identify the constellation of leadership manifestations that leads to creativity and innovation: We propose that the facilitating and integrating creative leadership manifestations are fostering factors for creative behavior. Further, we propose that the integrating and directing manifestations are fostering factors for innovative outcomes. In the following paragraphs, the hypotheses explaining how the three manifestations differentially relate to the idea generation behavior and the innovative outcome of research units are derived.

Idea generation behavior

In the facilitating creative leadership manifestation, followers are the primary idea generators, and the leader merely supports the followers in their creative endeavors. Thus, followers are expected to contribute creatively, while the leader is expected to make supportive contributions. This study maintains that the facilitating creative leadership manifestation will positively relate to the idea generation behavior of research units. In line with this proposition, several studies have highlighted the positive effect of social expectations and supervisor support on the creativity of followers (e.g., Gilson & Shalley, 2004; Hunter, Bedell, & Mumford, 2007; Janssen, 2005; Zhang & Bartol, 2010). First, Carmeli and Schaubroeck (2007) showed that expectations for creativity by leaders have a positive influence on the creative self-expectations of followers, which in turn are positively associated with followers’ creative job involvement. If team members perceive that the social norm within their group entails exhibiting creative behavior or making creative contributions, they are more likely to behave creatively and generate ideas (e.g., Scott & Bruce, 1994; Unsworth, Wall, & Carter, 2005). This could also be due to a kind of “Pygmalion effect,” as Tierney and Farmer (2004) theorize: If leaders expect their followers to be creative, they might exhibit higher degrees of supportive behavior. Second, supportive contributions (psychological, social, material support for creativity) are essential for shaping a supportive climate for creativity and “triggering, enabling, and sustaining creative thinking and behavior” in followers (Mainemelis, Kark, & Epitropaki, 2015, p. 400). Supervisor support can enhance follower creativity by facilitating knowledge sharing and creative problem-solving as well as by increasing followers’ intrinsic motivation and psychological safety and removing obstacles to creativity (Amabile, Conti, Coon, Lazenby, & Herron, 1996; Carmeli, Gelbard, & Reiter‐Palmon, 2013; Madjar, Oldham, & Pratt, 2002; Shalley & Lemoine, 2018; Woodman, Sawyer, & Griffin, 1993). These findings indicate a positive effect of expectation for creativity and leaders’ support on research units’ idea generation behavior. Thus, the following is proposed:

Hypothesis 1.

The facilitating creative leadership manifestation is positively related to the research unit’s idea generation behavior.

In an integrating creative leadership manifestation, both the leader and the followers are idea generators. Thus, both the leader and the followers are expected to make creative contributions, which are integrated to realize a creative goal. The study maintains that the integrating creative leadership manifestation will positively relate to the idea generation behavior of the research unit. The involvement of the leader and the followers in creative tasks allows followers to perceive their leader as a creative role model and observe the leader's creative skills (Koseoglu, Liu, & Shalley, 2017; Reiter-Palmon & Illies, 2004). In line with this proposition, Jaussi & Dionne (2003) found that if leaders are perceived as creative role models, the creativity of their followers increases. According to Shalley and Gilson (2004), modeling helps clarify expectations concerning creative and innovative performance and, furthermore, leads to enhanced skill acquisition. As Shalley and Perry-Smith (2001) found, exposure to a creative model leads to higher creative performance. Similarly, the mere observation of leaders’ creative problem-solving skills by followers increased the followers’ creativity (Basadur, 2004; Hemlin & Olsson, 2011; Reiter-Palmon & Illies, 2004).

Further supporting the argument, the integrating creative leadership manifestation entails expectations that both the leader and the followers will make creative contributions. This fosters joint decision-making or at least involvement in decision-making by followers as well as sharing of information and ideas between leaders and followers (Somech, 2006; West & Anderson, 1996). In turn, this may foster an atmosphere in which followers strive to propose new ideas (West, 2002) in line with expectations. In this way, participation is considered to enhance creativity (Andriopoulos, 2001). Thus, the following is proposed:

Hypothesis 2.

The integrating creative leadership manifestation is positively related to the research unit’s idea generation behavior.

In a directing creative leadership manifestation, leaders are the primary idea generators, while followers support their leaders by implementing their creative vision. As followers are not expected to make creative contributions but only supportive contributions, this study maintains that the directing creative leadership manifestation will not be related to the idea generation behavior of the research unit.

Innovative outcomes

In the directing creative leadership manifestation, the leader is the primary idea generator, and followers merely support their leader in the realization of the leader's creative creative vision. Thus, followers are expected to align with and implement their leader’s creative vision. This study maintains that the directing creative leadership manifestation will positively relate to the groups’ innovative outcome. As leaders are expected to make creative contributions, the group benefits from the leader’s expertise in shaping the implementation process successfully. Past research has emphasized the positive effects both of leaders’ creative contribution of providing creative direction and evaluating ideas and of their expertise with respect to idea implementation (for a review, see Amabile, 1997; Mumford, Scott, Gaddis, & Strange, 2002). Given the ill-defined nature of creative work (Mumford, Scott, Gaddis, & Strange, 2002), groups working in a directing creative leadership manifestation further receive essential feedback regarding which ideas are qualified for implementation through the creative involvement, direction, and evaluation of the leader.

Finally, the leader’s creative contribution of directing the group toward a specific creative goal, namely, the implementation of the creative idea generated and evaluated by the leader, enhances the group’s innovative outcome (Mumford, Connelly, & Gaddis, 2003; Mumford, Scott, Gaddis, & Strange, 2002). The directing creative leadership manifestation entails the expectation that followers will align with their supervisor’s vision. The clear goal envisioned by the leader may help the group adapt to the uncertainty and sometimes ill-defined nature of creative work (Mumford, Scott, Gaddis, & Strange, 2002). Consequently, groups who share clear objectives showed better innovative outcomes (for a review and meta-analysis, see Hülsheger, Anderson, & Salgado, 2009). Finally, if the leader alone engages in idea development and evaluation, processes that consume a considerable amount of time and cognitive resources, followers have the capacity to focus completely on implementing innovative outcomes. Thus, the following is proposed:

Hypothesis 3.

The directing creative leadership manifestation is positively related to the research unit's innovative outcome.

In an integrating creative leadership manifestation, both the leader and the followers participate in decision-making on the innovative processes and outcomes. Participation is considered to enhance innovation (Hülsheger, Anderson, & Salgado, 2009). Thus, this study maintains that the integrating creative leadership manifestation will positively relate to the research unit’s innovative outcomes. According to Axtell et al. (2000), participation in decision-making and support for innovation are strong predictors of innovative outcomes. West (1990) argued that this is due to an increase in experienced ownership, which followers feel because of the outcomes these decisions result in. The feeling of ownership is further increased, as the followers are expected to contribute ideas (as in the facilitating creative leadership manifestation) and to develop and implement them (as in the directing creative leadership manifestation). High degrees of ownership for an idea were associated with identification with and commitment to the idea (Basadur, 2004; Rouse, 2013), which may lead to a higher likelihood of idea implementation.

At the same time, the integrating creative leadership manifestation entails a synergistic factor, as the leader collects, combines, and integrates ideas from the followers with her own. The integration of information was assumed to be conducive to innovative outcomes (van Knippenberg, 2017). In addition, as leaders are involved in making creative contributions in the integrating manifestation, they are able to shape the innovation process with their expertise and provide creative direction and idea evaluation (Mumford, Connelly, & Gaddis, 2003; Mumford, Scott, Gaddis, & Strange, 2002). Past research has emphasized the positive effects both of leaders’ expertise and of leaders providing direction and evaluation on innovative outcomes (for a review, see Mumford, Scott, Gaddis, & Strange, 2002). According to Amabile (1988), both technical skills and in-depth organizational expertise are required for successful implementation planning.

In conclusion, in an integrating creative leadership manifestation, leaders enhance a group’s innovative outcome by integrating ideas through expertise-informed evaluation and direction in the implementation process, while followers improve innovative outcomes through high levels of commitment. Thus, the following is proposed:

Hypothesis 4.

The integrating creative leadership manifestation is positively related to the research unit’s innovative outcome.

In a facilitating creative leadership manifestation, leaders are not expected to make creative contributions. Therefore, the lack of an overarching creative vision by the leader as well as the lack of leader involvement, expertise, and direction in shaping and evaluating the implementation process may hinder the achievement of innovative outcomes. This may counteract the positive effect of increased ownership followers experience when implementing the ideas they contributed. As these two effects may cancel each other out, we do not expect the facilitating creative leadership manifestation to be related to innovative outcomes in research units.

Methods

Data collection and sample

Data was collected between September 2018 and January 2019 as part of an online survey of scientific personnel at universities in Germany. Contact details were obtained systematically via the publicly available homepages of academic departments and participants recruited via email. Participation was voluntary, and confidentiality was assured.

The survey was sent to members of the academic supervisors’ research units, which consisted of PhD students or post-doctoral researchers. In total, 1,873 participants completed the survey. For scale development, all participants without missing values on the items of the developed scale were included. The sample for the scale development consisted of 1,334 research unit members. For hypothesis testing, participants with missing values on any of the key measures were excluded. Further, the samples were restricted to research units with at least two participants, less than 40 unit members and an r_wg ≥.5 on the scale for measuring idea generation behavior. The final sample for the hypotheses testing consisted of 646 individuals nested in 259 research units from nine universities. Forty-eight percent of the participants were female; 33.6% held a PhD. The average age was 33.04 (SD = 7.67). The average number of unit members participating in the survey was 2.49 (SD = .82), with a range from two to seven participants per unit. Total unit size ranged from two to 39 (M = 13.75, SD = 8.64). A total of 29.7% of the research units worked in the field of law, social science, or economics, 29.0% in math or science, 19.3% in the humanities, 11.2% in engineering, and 10.8% in other fields.

Scale development

Item generation

Building on the theoretical concept of Mainemelis, Kark, and Epitropaki (2015), an initial pool of 18 items was generated using a theory-driven, deductive approach (Hinkin, 1995) to capture the abovementioned three content domains of creative leadership manifestations: directing, facilitating, and integrating (a complete list of items can be found in Table 1). The items focused on the expected distribution of creative and supportive tasks between leader and followers. To allow for a context-sensitive assessment (as called for by Mainemelis, 2018), the items were adapted to the academic research context (i.e., “professor” instead of “leader,” “scientific staff” instead of “followers”).

Table 1. Rotated factor matrix of exploratory factor analyses on items for measuring creative leadership manifestations.

	EFA on initial 18 items^a			EFA on final 9 items^b
Items	1	2	3	1	2	3
Facilitating
The task of the scientific staff is to develop new research ideas.	0.15	−0.32	0.76	0.15	−0.33	0.69
New research ideas are developed primarily by the academic staff (and not by the professor).	−0.03	−0.31	0.72	0.02	−0.22	0.89
The professor is mainly the “manager” of the group and is hardly involved in the creative work.	−0.18	−0.13	0.50	−0.12	−0.06	0.57
The professor creates the conditions for scientific staff to develop their own research ideas.	0.53	−0.25	0.53
The professor actively encourages scientific staff to develop their own research ideas.	0.53	−0.22	0.58
Leadership means that the professor promotes the creativity of others in generating research ideas.	0.50	−0.21	0.61
Directing
The professor provides research ideas and expects them to be implemented.	−0.17	0.84	−0.15	−0.12	0.85	−0.16
Leadership means that the professor translates his/her own creative research ideas into reality through the work of the scientific staff.	−0.11	0.85	−0.19	−0.07	0.86	−0.21
The task of the scientific staff is to implement the research ideas of the professor.	−0.09	0.87	−0.18	−0.05	0.87	−0.20
New research ideas are developed primarily by the professor (and not by the academic staff).	−0.02	0.66	−0.38
The professor is clearly the “creative mind” of the group that develops the research ideas.	0.06	0.72	−0.32
The professor establishes strong guidelines for research topics and projects.	−0.20	0.79	−0.17
Integrating
Leadership in this team means that the professor brings together the different creative contributions of himself/herself and the academic staff.	0.88	−0.01	0.03	0.89	−0.04	0.02
The professor integrates his/her creative research ideas with those of the scientific staff.	0.91	−0.06	0.02	0.93	−0.11	−0.02
The professor creates synergies between his/her creativity and the creativity of the scientific staff.	0.92	−0.04	0.04	0.94	−0.10	−0.02
New research ideas are developed by both the professor and the research associates.	0.82	−0.17	0.08
The professor and the scientific staff are equally involved in the development of new research ideas.	0.82	−0.16	0.05
The professor works closely with the scientific staff to develop new research ideas.	0.87	−0.01	0.01

Notes. N = 641. Extraction method: Principal axis analysis; Rotation method: Varimax.

^aTotal variance explained: 68.8%.

^bTotal variance explained: 74.9%.

Qualitative feedback

A pilot test of the items was conducted in a sample of 10 PhD students familiar with and experienced in organizational behavior research. They were asked to provide feedback regarding the comprehensibility, clarity, and readability of the items. In addition, three individuals who held M.Sc. degrees in psychology and who were literate in scientific methodology sorted the randomized items and labeled the resulting dimensions. This sorting was a test of the content validity of the items written to reflect each dimension. All individuals came to the same conclusion and made no misclassifications. Therefore, we retained all 18 items.

Factor analysis

As part of a rigorous psychometric analysis of the new measure, exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) were conducted. An EFA was utilized to support the selection of a reduced set of items and to explore the reliability and dimensionality of the scale. A CFA was conducted to confirm and validate the resulting scale.

Following the procedure recommended by Krzystofiak and colleagues (1988), a randomly selected split-half subsample of 641 individuals (N₁) was used to perform the EFA, while the second split-half subsample of 693 individuals (N₂) was used for the CFA. The resulting subsamples still exceeded the minimum sample size of 500 recommended for EFA and CFA (MacCallum, Widaman, Zhang, & Hong, 1999). The initial 18-item scale was analyzed by using the data of the first subsample (N₁) and employing principal axis factor extraction with orthogonal rotation (Varimax) in SPSS. After the first EFA, the scale was reduced by excluding items with loading weights smaller than .50 or cross-loadings on more than one factor larger than .32 (Tabachnick & Fidell, 2001) and balanced the scales. Then, a CFA was performed on the now nine-item scale in AMOS using the second subsample (N₂).

Measures for hypothesis testing

Creative leadership manifestations

Creative leadership manifestations were measured with the newly- developed Creative Leadership Manifestation Scale (CLMS). Three items each were averaged to measure facilitating, directing, and integrating manifestations scale. The final scale can be found in Table 1. Research unit members rated items on a scale ranging from “1” for strongly disagree to “7” for strongly agree.

Idea generation behavior

Creativity was operationalized as idea generation behavior (cf. Montag, Maertz, & Baer, 2012) and was measured via four items by Tierney, Farmer, and Graen (1999) adapted for self-rating. They were back-translated following the procedure recommended by Brislin (1970). The items used were “I generate novel but operable work-related ideas,” “I try new ideas and approaches to problems,” “I seek new ideas and ways to solve problems,” and “I generate ideas revolutionary to the field” (Cronbach’s alpha = .81). Research unit members rated items on a scale ranging from “1” for strongly disagree to “7” for strongly agree.

Innovative outcome

Innovation was operationalized as innovative outcome. To combine the survey results on the creative leadership manifestations with an objective measure (as called for by Kark, Epitropaki, & Mainemelis, 2020), the innovative outcome of the research unit was measured by the number of publications in peer-reviewed journals coauthored by the academic supervisor and at least one follower weighted for journal impact. Academic supervisors in Germany typically publish papers coauthored by one or more members of their research unit (cf. Braun, Peus, Weisweiler, & Frey, 2013). If supervisors in the sample did publish without one or more followers,¹ we excluded these publications from the analysis. Publications from 01.01.2018 to 31.07.2019 were identified via ISI Web of Knowledge and used for subsequent analyses. As the survey in which the creative leadership manifestations were assessed took place at the end of 2018, a concurrent assessment of creative leadership manifestations and innovative outcome was established. The number of publications included per research unit ranged from zero to 56, with an average number of 4.79 (SD = 7.69) publications per team. Between 01.01.2018 and 31.07.2019, 69.9% of the units had five publications or fewer, 16.2% of the units had between six and 10 publications, 8.1% of the units had between 11 and 20 publications, and 5.8% of the units had more than 20 publications. The number of publications per unit was weighted by the Journal Citation Indicator (JCI) for 2021 of these publications, which was also identified for each publication via ISI Web of Knowledge. The JCI is a field-normalized metric that represents the average citation impact over the past three years (Torres-Salinas, Valderrama-Baca, & Arroyo-Machado, 2022).

Control variables

To exclude alternative explanations for the relationships between the variables investigated, the influence of third variables that exert influence on the proposed outcome variables, as indicated by prior research, were controlled for. We controlled for the expertise of followers (Amabile, 1988; by assessing the ratio of research unit members who already had a PhD to those who did not), the average unit tenure (average duration of working with the respective academic supervisor), and the unit size (Hülsheger, Anderson, & Salgado, 2009). Further, we controlled for the field of research (STEM vs. non-STEM field), since the number of publications may vary greatly between different fields of research.

Results

Scale development

Exploratory factor analysis

Using the first split sample, an EFA on the initial 18-item scale was conducted. The magnitude and scree plot of the eigenvalues supported a three-factor structure, which accounted for 68.8% of the total variance. After items that cross-loaded were eliminated, the scales were balanced to be composed of an equal number of items, selecting the nine items with the highest factor loadings, grouped into three dimensions: directing creative leadership manifestation (α = .92), facilitating creative leadership manifestation (α = .77), and integrating creative leadership manifestation (α = .95). In a second principal axis factor analysis of the remaining nine items, the magnitude and scree plot of the eigenvalues again supported a three-factor structure, which now accounted for 74.9% of the total variance. Table 1 shows the factor loadings of both EFAs.

Confirmatory factor analysis (CFA)

Using the second split sample, a CFA was performed on the reduced 9-item scale using maximum likelihood estimates in AMOS. The model fit the data well (χ²(24) = 111.88, p < .001, CFI = .98, RMSEA = .07 and SRMR = .06). All first-order factor loadings were statistically significant at a significance level of p < .001 and reasonably large, ranging from .58 to .95 (M = .85).

Unit-level aggregation

To check whether the intergroup agreement and intergroup reliability were high enough for unit-level aggregation, intraclass correlations (ICC1), the reliability of group mean index (ICC2), and r_wg values using a uniform null distribution were calculated. All ICC values were significant at a level of p < .001, indicating that a significant amount of variance in these measures was associated with between-group differences and that the groups are a valid factor for clustering the data. The ICC1 values were .35, .36, .22, and .14 for the facilitating, directing, and integrating creative leadership manifestations and idea generation behavior, respectively, indicating medium- to large-sized dependence in responses within groups (Biemann & Heidemeier, 2017; Bliese & Hanges, 2004; Kozlowski & Klein, 2000; LeBreton & Senter, 2008). The ICC2 values were .57, .58, .42, and .29. These ICC2 indices were within ranges observed in other research on leadership (e.g., Braun, Peus, Weisweiler, & Frey, 2013; Chen, Farh, Campbell-Bush, Wu, & Wu, 2013) and in team-level constructs (c.f. Woehr, Loignon, Schmidt, Loughry, & Ohland, 2015). Further, relatively low values of these indices are expected when developing new measures and when dealing with groups within similar contexts (Bliese, 2000; LeBreton & Senter, 2008). The mean value of r_wg was r_wg = .67 (SD = .33) for the facilitating creative leadership manifestation, r_wg = .66 (SD = .35) for the directing creative leadership manifestation, r_wg = .60 (SD = .38) for the integrating creative leadership manifestation, and r_wg = .82 (SD = .22) for idea generation behavior, indicating a moderate to strong agreement (LeBreton & Senter, 2008). In sum, these values indicate acceptable within-group agreement, which justifies aggregation of the four scales to the group level (James, Demaree, & Wolf, 1993; LeBreton & Senter, 2008; Woehr, Loignon, Schmidt, Loughry, & Ohland, 2015).

Table 2 shows the means, standard deviations, and correlations of the examined variables aggregated on the unit level.

Table 2. Means, standard deviations, and correlations of the key study variables.

Variables	Mean	SD	1	2	3	4	5	6	7	8
1. PhD^a	0.34	0.33
2. STEM field^b	0.40	0.49	−0.11
3 Group tenure	58.88	36.83	0.57*	−0.06
4. Group size	13.74	8.64	0.00	0.41*	0.03
5. Facilitating	3.96	1.15	−0.02	0.15*	−0.02	0.18*	(0.80)
6. Directing	3.67	1.38	−0.06	0.25*	0.03	0.18*	−0.44*	(0.92)
7. Integrating	4.27	1.33	−0.02	0.07	−0.05	−0.11	0.03	−0.15*	(0.95)
8. Creativity	5.07	0.67	0.00	0.09	−0.06	0.08	0.23*	−0.13*	0.19*	(0.81)
9. Innovation	6.38	11.52	−0.06	0.39*	0.09	0.40*	0.10	0.21*	0.00	0.06

N = 259 research units. Cronbach’s alpha is displayed on the diagonal in parentheses. ^aPhD is coded 1, no PhD is coded 0. ^bSTEM field is coded 1, non-STEM field is coded 0. *p < .05.

Hypothesis testing

Prior to hypothesis testing, the study examined whether there was significant between-organization variance in the outcome variables by running null (intercept-only) models. The results showed that the organization did not account for a significant amount of variance in idea generation behavior (χ²(8) = 2.52, p = .112). Therefore, the hypothesis regarding idea generation behavior was tested using ordinary least squares (OLS) regression. Note that modeling the random effect did not alter the pattern of significance. In contrast, the results for innovative outcome showed that a significant amount of variance in innovative outcome resided at the organization level (χ²(8) = 4.55, p = .049), which necessitated using a random coefficients model to account for the level-2 random effect of organization.

Creative behavior

As shown in Table 3, the addition of the facilitating, directing, and integrating creative leadership manifestations as predictors for research unit’s idea generation behavior significantly increased the model fit (∆F = 7.13, p < .001). The coefficient of the facilitating creative leadership manifestation was positive and significant (b = .11, p < .010), supporting H1. Similarly, the coefficient of the integrating creative leadership manifestation was positive and significant (b = .09, p = .004), supporting H2.

Table 3. OLS regression analyses for creative leadership manifestations predicting creativity.

	Model 1		Model 2
	B (SE)	p	B (SE)	p
Intercept	5.03 (0.10)	<0.001	4.29 (0.30)	<0.001
PhD^a	0.11 (0.15)	0.450	0.09 (0.14)	0.509
STEM field^b	0.09 (0.09)	0.336	0.05 (0.10)	0.634
Group tenure	0.00 (0.00)	0.226	0.00 (0.00)	0.300
Group size	0.00 (0.01)	0.413	0.00 (0.01)	0.362
Facilitating creative leadership			0.11 (0.04)	0.010
Directing creative leadership			−0.02 (0.04)	0.600
Integrating creative leadership			0.09 (0.03)	0.004
R^2	0.02		0.09
∆F			7.13	<0.001

N = 259 research units.

Unstandardized coefficients and standard errors (in parentheses) are reported.

^aPhD is coded 1, no PhD is coded 0.

^bSTEM field is coded 1, non-STEM field is coded 0.

Innovative outcomes

The measure of innovative outcome, the number of publications weighted by citation impact, is a zero-inflated count variable, which does not meet the requirement of equidispersion. Assuming equidispersion in an overdispersed model may lead to incorrect conclusions due to underestimation of standard errors or overestimation of the significance of beta coefficients (cf. Payne et al., 2015). Therefore, the hypotheses were tested via a two-level negative-binomial generalized linear multilevel model. As shown in Table 4, the addition of the facilitating, directing, and integrating creative leadership manifestations as predictors for a unit’s innovative outcome significantly increased the model fit (χ² = 29.07, p < .001). The coefficient of the directing creative leadership manifestation was positive and significant (b = .49, p < .001), supporting H3. Similarly, the coefficient of the integrating creative leadership manifestation was positive and significant (b = .19, p = .018), supporting H4. Unexpectedly, the analysis further displayed a positive and significant coefficient the for facilitating creative leadership manifestation (b = .38, p < .001).

Table 4. Random coefficient analysis for creative leadership manifestations predicting innovation.

	Model 1		Model 2
	Estimate (SE)	p	Estimate (SE)	p
Intercept	−0.06 (0.26)	0.814	−4.16 (0.84)	<0.001
Organization	0.01 (0.10)		0.04 (0.21)
PhD^a	−0.21 (0.35)	0.558	−0.05 (0.35)	0.890
STEM field^b	1.08 (0.23)	<0.001	0.93 (0.22)	<0.001
Group tenure	0.01 (0.00)	0.029	0.00 (0.00)	0.095
Group size	0.05 (0.01)	<0.001	0.05 (0.01)	<0.001
Facilitating creative leadership			0.38 (0.11)	<0.001
Directing creative leadership			0.49 (0.09)	<0.001
Integrating creative leadership			0.19 (0.08)	0.018
Deviance	1316.7		1370.3
AIC	1330.7		1390.3
BIC	1356.2		1426.7
Chi^2			29.07	<0.001

N = 259 research units, 9 organizations.

Unstandardized coefficients and standard errors (in parentheses) are reported.

^aPhD is coded 1, no PhD is coded 0.

^bSTEM field is coded 1, non-STEM field is coded 0.

Post-hoc analysis

While we have asserted that the academic research setting is generally a weak creative context, there are reasons to expect that there are varying degrees of context strength within this setting. For example, research in the STEM field (science, technology, engineering, and mathematics) features normative expectations and task characteristics (e.g., use of prescribed apparatus and metrics, regulatory guidelines in research involving animals) that may point to a more directing creative leadership manifestation than is observed in non-STEM disciplines. To explore potential differences in context strength within the academic research setting, we examined the difference between creative leadership manifestations in STEM versus non-STEM disciplines (e.g., law, social science, humanities). As expected, directing creative leadership was significantly more prevalent in the STEM field (b = .63, p < .001, where STEM = 1, non-STEM = 0). The mean of the directing manifestation was M = 4.21 (SD = 1.26) in the STEM units and M = 3.47 (SD = 1.39) in non-STEM units. No significant differences were observed for facilitating (b = .19, p = .219; M_STEM = 4.09, M_non-STEM = 3.77) and integrating (b = .33, p = .063; M_STEM = 4.32, M_non-STEM = 4.17) creative leadership manifestations.

To further investigate how STEM versus non-STEM disciplines impact the strength of the hypothesized relationship, we tested the interaction of STEM vs. non -STEM (coded 1= STEM., 0 = non-STEM) and creative leadership manifestations on creativity and innovation. The results for creative behavior indicate no significant interaction between STEM and either of the three creative leadership manifestations. This suggests that facilitating, directing, and integrating creative leadership is similarly relevant in STEM and non-STEM fields.

The results for innovative outcomes reveal a significant interaction between STEM and directing creative leadership (b = −.50, p = .009). Subgroup analyses show that while in both cases there was a significant and positive relationship, the relationship was stronger for non-STEM (b = .77, p < .001) versus STEM (b = .28, p = .048). This finding suggests that while directing leadership is relevant regardless of discipline, it is more important to innovation in the non-STEM fields. Additionally, there was a significant interaction between STEM and integrating creative leadership (b = −.40, p = .009). Subgroup analyses indicated that the relationship was not significant in the STEM subgroup (b = .04, p = .634), but was significant in the non-STEM subgroup (b = .46, p = .002). While caution should be exercised in interpreting null effects, these findings suggest that integrating leadership is not relevant to innovation in the STEM fields.

Taken together, these results suggest that, first, the creative leadership manifestations are relevant for creativity and innovation in both STEM and non-STEM fields. Second, the context of STEM-related research arguably seems to be stronger, as we observed a more focused cluster of directive and facilitating leadership than in non-STEM fields.

Discussion

This study set out to extend and empirically test a theory on creative leadership manifestations (facilitating, directing, and integrating) and their relation with creativity and innovation in the relatively weakly pre-structed creative context of academic research units. Drawing on the classifications of Mainemelis, Kark, and Epitropaki's (2015) model of creative leadership, it developed a measure of creative leadership manifestations and validated its tripartite factor structure. Using a sample of 259 academic research units, the relationships between these manifestations and both the units’ self-report of idea generation behavior and the units’ actual output of research publications weighted by citation impact as a measure of innovative outcome were examined.

The results reveal that the creative leadership manifestations contribute to creativity and innovation differently. As hypothesized, the study found that, whereas the facilitating and integrating leadership manifestations were positively and significantly related to idea generation behavior, the directing and integrating leadership manifestations were positively and significantly related to innovative outcome. It is noteworthy that by examining these relationships simultaneously, the study was able to demonstrate the unique impact of each manifestation above and beyond the impact of other manifestations.

Unexpectedly, the facilitating creative leadership manifestation was also related to innovative outcome. This finding may be specific to the context of this study and the innovative outcome measure used, the number of journal publications weighted by citation impact. In the academic context, a main hurdle to the implementation of ideas in the form of publications is that they are rejected by the peer-reviewed journals they were submitted to and thus have to be resubmitted or submitted to other journals. This leads to an iterative process of reworking the original idea. First, the creativity-enhancing environment of the facilitating creative leadership manifestation may lead to a bigger pool of creative ideas concerning the revision and resubmission of papers, which in turn may enhance the likelihood of publication. Second, while in corporate contexts ideas need to be novel and useful in order to be implemented, in the academic context for particularly original work sometimes only the novelty criterion may be decisive for acceptatance by a journal. Innovative outcome may be enhanced in the facilitating manifestation, as in the iterative process of revision and resubmission the gradual enhancement of novelty and creative quality is encouraged, which makes the papers eventually ripe for publication.

Implications for theory and practice

This study advances previous research on creative leadership in several ways. First, it extends the work of Mainemelis, Kark, and Epitropaki (2015) by applying it to a relatively weak context where multiple manifestations of creative leadership may be prompted. In addition, it develops and validates a measurement tool for these creative leadership manifestations, namely, the Creative Leadership Manifestation Scale (CLMS). This scale showed good psychometric properties in terms of internal consistency and construct validity of its three dimensions. Thus, it contributes to creative leadership research by adding validation and nuance to existing theory on creative leadership contexts (Mainemelis, 2018; Mainemelis, Kark, & Epitropaki, 2015) and offering a tool for measuring leaders’ creative characteristics within work groups (Jaussi & Dionne, 2003; Koseoglu, Liu, & Shalley, 2017).

Second, the study examines the combined and unique influence of each creative leadership manifestation, thereby providing a more comprehensive understanding of the role of leadership in creativity and innovation. It shows that the creative leadership manifestations have important implications for the outcomes of research units: The findings suggest that the facilitating and integrating manifestations seem to be especially relevant to fostering creativity and the integrating and directing manifestations are relevant to producing innovation.

Third, it answers the call for careful differentiation between creativity and innovation (Anderson, Potočnik, & Zhou, 2014; Hughes, Lee, Tian, Newman, & Legood, 2018; Perry-Smith & Mannucci, 2016) by separately investigating the relationship of the creative leadership manifestations with both outcomes. Similarly, it follows Montag, Maertz, and Baer's (2012) suggestion to further differentiate between creative behavior and its outcomes. In this vein, this study also speaks to research on the organization and management of research groups (Lee, Walsh, & Wang, 2015; Perry, Hunter, & Currall, 2016) and on the training of future scientists (Shibayama, 2019), pointing to the relevance of facilitating and integrating creative leadership manifestations for nurturing the creative potential of research group members.

This study offers important implications for practice. First, it develops and validates a scale for measuring the creative leadership manifestations present within a group. This measurement tool can help leaders and organizations determine their status concerning the distribution of creative and supportive contributions between leaders and followers. Second, introducing the concept of creative leadership manifestations in leadership guidelines and development programs may help raise awareness in light of current demands and suggest ways of distributing creative and supportive tasks between leaders and followers. Third, the results of this study provide the first insights into which creative leadership manifestation is effective in which phase of the innovation cycle in academic research units. This knowledge may be a straightforward lever for strategically influencing the context of creative and innovative processes in research units. Focusing on the purposeful distribution of creative tasks between leaders and teams can help leaders intentionally shape their own and their followers’ creative context and thereby induce group idea generation behavior or group innovative outcome. Finally, when staffing groups for creative work, attention should be paid to the distribution of creative tasks between leaders and followers, which should also involve thinking about the leaders’ role. Overall, this study provides an opportunity for leaders and organizations to develop a strategy to augment creativity and innovation.

Limitations and future research

When interpreting the findings of this study, it should be noted that it is not free of limitations. As it is the first study to measure the three creative leadership manifestations quantitatively and simultaneously within one context, future research to test, validate, and extend the theoretical model and measurement of creative leadership manifestations is encouraged. The scale we developed can be adapted to a variety of other contexts in and outside of academia to gain insight into the organization of creativity in contexts with mixed demands, as these are becoming increasingly common in the business world and beyond.

Particular attention should be paid to the following issues: First, the analyses and results do not allow strong causal claims. Even though a number of alternative explanations for the relationship of the creative leadership manifestations with the proposed outcome variables were excluded by controlling for third variable influence, the study cannot provide evidence on the direction of the effects observed in the regression analyses or the mechanisms leading to these effects. Future research should employ (field/quasi-) experimental methods and time series analyses to answer questions of causality. In addition, future research investigating mediators and moderators would help illuminate the mechanisms of the influence of creative leadership manifestations on outcomes. For example, followers’ individual creativity within a directing creative leadership manifestation might be dependent on their creative role identity, as highly creative people or people identifying as such might display creative behaviors even if the creative leadership manifestation present in the group does not encourage them to contribute creative ideas. Besides investigating the role of follower identity (cf. Randel & Jaussi, 2019), examining leader personality and their need for control (cf. Flocco, Canterino, Cirella, Coget, & Shani, 2018) as influencing factors may present an interesting avenue for future research. Finally, it would be an interesting avenue for future research to compare and contrast the perceptions of followers and leaders concerning the creative leadership manifested within a group. The congruence or deviation of perspectives of leaders and followers might act as a moderating factor in predicting the effectiveness of creative leadership (cf. Atwater & Yammarino, 1992).

Second, it is important to consider some of the unique features of the setting. First, the academic research environment is a dynamic context characterized by short-term contracts and low group stability. Thus, future research should investigate the impact of leadership manifestations in longer-term groups. Second, an unusual objective metric of innovative outcome was employed: research publications weighted by citation impact. It differs from measures of innovation in other contexts in the definition of the usefulness of the implemented idea. While in corporate contexts novel ideas are only implemented if they are directly useful in a practical way (e.g., by increasing revenue, reducing expenses), research articles are selected and vetted in regard to the usefulness of their contribution, not their commercial profit. In sum, in addition to studying creative leadership manifestations in one specific omnibus context, future research would profit from investigating a vaster array of “weak” omnibus contexts and from examining the existence and consequences of variability in creative leadership modes within these contexts.

Finally, this study did not address the question of temporality. For instance, creative leadership manifestations might be shaped over time in a negotiation process between the leader and the followers. Alternatively, the increasing expertise of followers (and/or leaders) might change the creative leadership manifestation over the lifetime of the group, as suggested by Abecassis-Moedas & Gilson (2018). In addition, different phases of the creative/innovative process might benefit differently from different creative leadership manifestations, and studying this would require a finer-grained assessment of the iterative phases of working together to attain a creative goal. Thus, future research may find interest in taking on a process perspective and investigating the temporal dimension of creative leadership (cf. Kark, Epitropaki, & Mainemelis, 2020).

In addition, future research may investigate whether the integrating creative leadership manifestation, being associated with both creativity and innovation, is also associated with other favorable outcomes, such as high levels of group learning, group member satisfaction, or a higher percentage of junior scholars who decide to pursue a career in academia.

Conclusion

This study extends the concept of creative leadership manifestations by Mainemelis, Kark, and Epitropaki (2015) by applying it to a relatively weak context where multiple manifestations of creative leadership may be prompted. In addition, the study developed and validated a tool for measuring the facilitating, directing, and integrating creative leadership manifestations. Finally, it demonstrated the differential relationships of the creative leadership manifestations with idea generation behavior and innovative outcome of research units. Despite the stated limitations, this study contributes to organizational creativity research by offering initial insights into how to distribute creative and supportive tasks between leaders and followers.

Disclosure statement

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

Notes

1. Less than 24% of publications by supervisors were published without one or several team members.