Shared leadership and organizational ambidexterity : the mediating role of paradoxical cognition and reflection

Shared Leadership and Organizational Ambidexterity:

The Mediating Role of Paradoxical Cognition and Reflection

Author: Jordi Borst Student number: 10004489

Date: June 2014

Qualification: MSc. in Business Studies – Strategy Track

Institution: Amsterdam Business School, University of Amsterdam Supervisor: Pepijn van Neerijnen

Table of Contents

Abstract 03

Introduction 04

Literature Review and Hypotheses 06

Methodology 14

Analysis and Results 19

Discussion and Conclusion 22

References 26

Abstract

Ambidexterity, which is defined as the organizational ability to simultaneously explore and exploit, has been shown to be necessary for organizations to survive in the long run. Although prior scholars found that top management teams play a key role in achieving ambidexterity, it has remained unclear how they influence the ability of top managers to deal with the contradictory cognitive and structural requirements that exploration and exploitation impose. This study proposes that shared leadership within top management teams allows top managers to reconcile their conflicting interests and identify structural configurations that facilitate ambidextrous organizational behavior. It is furthermore proposed that this process is mediated by paradoxical cognition and reflection. Drawing upon data from 282 Dutch and German organizations, it is shown that paradoxical cognition indeed fully mediates the relationship between shared leadership and organizational ambidexterity. Surprisingly, it appears that reflection only partially mediates this relationship. These findings provide various new insights about ambidexterity and its antecedents.

Key words: Ambidexterity, exploration and exploitation, shared leadership, top management teams, paradoxical cognition, reflection.

Introduction

Exploration and exploitation have been defined as two fundamentally different organizational learning activities (March 1991). Whereas exploration aims to achieve variance in organizations’ activities by discovering new possibilities (Katila and Ahuja 2002, Rosenkopf and Nerkar 2001), exploitation aims to achieve efficiencies in existing ones (Beckman et al. 2004, Benner and Tushman 2003). Although exploration and exploitation require distinct organizational structures and processes in order to be effective (Ancona et al. 2001, He and Wong 2004), both learning activities are necessary for organizations to survive in the long run (Burgelman 2002, Gibson and Birkenshaw 2004, Levinthal and March 1993). The organizational ability to simultaneously explore and exploit is referred to as ambidexterity (Duncan 1976).

As strategic decisions are generally made at the apex of the organization, top management teams play a key role in achieving ambidexterity (Eisenhardt and Zbaracki 1992, Hambrick and Mason 1984). Prior scholars found that top managers face significant challenges in balancing exploration and exploitation (Denison et al. 1995, Floyd and Lane 2000). This is because the long-term experimental focus of exploration is often in conflict with the short-term efficiency focus of exploitation (Benner and Tushman 2004, O’Reilly and Tushman 2004). In addition, top managers are often responsible for units that focus on either one of these activities while, at the same time, organizations have limited resources to divide between them (Eisenhardt et al. 1997, Tushman and O’Reilly 1996). Top managers are therefore often in competition with one another (Gupta et al. 2006). This stimulates self-interested behavior, and diminishes top managers’ acceptance of strategic decisions (O’Reilly and Tushman 2004). Top management teams should therefore make top managers aware of the fact that exploration and exploitation are equally important to the organization’s success if they wish to achieve organizational ambidexterity (Smith and Tushman 2005, Weick 1979). However, our understanding of how top management teams are able to do so remains rather limited (Smith and Lewis 2011). In response to the scholarly call for more research in this area (Adner and Helfat 2002, Smith and Lewis 2011), this study examines how shared leadership within top management teams influences the ability of top managers to reconcile their conflicting interests and identify structural configurations that facilitate ambidextrous organizational behavior.

In addition, this study proposes that paradoxical cognition and reflective behavior are necessary for top managers to overcome the tension between exploration and exploitation.

Paradoxical cognition, on the one hand, has been shown to stimulate top managers both to differentiate and integrate their thinking about exploration and exploitation (Smith and Lewis 2011). Top managers are thereby provided a lens through which to identify linkages between them (Smith et al. 2010, Smith and Lewis 2011). The identification of these linkages enhances top managers’ awareness about the importance of both exploration and exploitation, and stimulates them to make more balanced decisions (Smith and Lewis 2011, Lüscher and Lewis 2008). Reflective behavior, on the other hand, has been shown to stimulate dialectic discussion, which allows top managers to share ideas and raise doubts and questions about particular organizational concerns (Schippers et al. 2008). Reflective behavior thereby improves top managers’ understanding of how exploratory and exploitative activities relate to one another (Leung and Kember 2003). This allows top managers to identify structural configurations that facilitate paradoxical organizational behavior (Andriopoulos and Lewis 2009, O’Reilly and Tushman 2008). Figure 1 depicts the study’s conceptual model.

Figure 1. Conceptual Model

This study contributes to the literature in at least three distinct ways. First, this is the first study that empirically examines how shared leadership within top management teams influences the ability of top managers to achieve organizational ambidexterity. Smith and Tushman (2005) suggested that leadercentric and teamcentric structured top management teams are equally able to overcome the tension between exploration and exploitation. The results of this study show that this is not the case. More specifically, it is shown that top managers within teamcentric structured top management teams are better able to balance both learning activities than top managers within leadercentric structured ones.

H2 H1 H3 Organizational Ambidexterity Shared Leadership Paradoxical Cognition Reflection

Second, it has been suggested that paradoxical cognition plays an important role in allowing top managers to overcome the tension between exploration and exploitation (Smith and Tushman 2005). This study shows that paradoxical cognition is in fact a necessary condition to achieve organizational ambidexterity. In addition, it has been suggested that leadercentric and teamcentric structured top management teams are equally able to stimulate top managers to engage in paradoxical behavior (Amabile 1996). This study shows that in reality, teamcentric structured top management teams are better able to stimulate this behavior than leadercentric structured ones.

Finally, the literature emphasizes the importance of reflective behavior is overcoming the tension between exploration and exploitation (Leung and Kember 2003). This study shows that reflective behavior indeed enhances the ability of top managers to achieve a balance between both learning activities. In contrast to the study’s expectations, however, it appears that reflective behavior is not a necessary condition for top managers to successfully do so.

The section that follows presents the literature review and hypotheses underlying this study. Next, and drawing upon data from 282 organizations operating within a wide range of industries, the empirical findings are presented. The study concludes with a discussion of the results, their theoretical and practical implications, and future research suggestions.

Literature Review and Hypotheses

Learning is an important capability for organizations to survive (Barney 1991, Teece et al. 1997). March (1991) identified exploration and exploitation as two fundamentally different learning activities. Exploration entails radical innovation, discovery, and experimentation (Katila and Ahuja 2002, Rosenkopf and Nerkar 2001). In contrast, exploitation entails incremental innovation, refinement, and implementation (Beckman et al. 2004, Benner and Tushman 2002). Both learning activities require contradictory organizational structures and processes in order to be effective (Ancona et al. 2001, He and Wong 2004). Organizations can overcome these contradictions by adopting structural configurations that focus on either exploration or exploitation (Doty et al. 1993, Ghemawat and Costa 1993, Wernerfelt and Montgomery 1988). However, focusing too much on exploration prevents organizations from learning by doing and achieving efficiencies (He and Wong 2004).

Focusing too much on exploitation, on the other hand, leads to dynamic conservatism and inertia (Benner and Tushman 2002, Sull 1999). Hence, survival depends on the organizational ability to simultaneously explore and exploit (Gibson and Birkenshaw 2004, Gupta et al. 2006). This ability is referred to as ambidexterity (Duncan 1976).

Top management teams have the important task to balance the contradictory agendas of exploration and exploitation (Eisenhardt and Zbaracki 1992, Hambrick 1994, Weick 1979). They do so through the allocation of resources between both learning activities, and by making structural design decisions (Edmondson et al. 2003, Hambrick 1994, Lubatkin et al. 2006). Top management teams are said to make balanced decisions when they support both exploratory and exploitative activities over time (Bazerman 1998, Smith and Tushman 2005). However, top managers face cognitive and structural barriers that prevent them from making such decisions (Bazerman and Watkins 2004, Smith and Tushman 2005, O’Reilly and Tushman 2008). Kahneman and Teversky (1979) and Leana and Barry (2002), for instance, respectively found that top managers are risk averse and are looking for consistency. In addition, Levinthal and March (1993) suggested that top managers are rather short-term oriented. It has furthermore been shown that top managers are unwilling to change organizational designs that historically have been proven to be successful (Henderson and Clark 1990). Of course, this imposes considerable risks for the organization, given that environmental circumstances change over time (Levinthal and March 1993). Research furthermore found that top managers develop certain routines over the years (Schwenk 1984). When these routines consistently prove their worth over a certain period of time (commercially or in terms of social stability), top managers are likely to stick to them (Katz and Allen 1982). Situations like these are especially likely to occur in top management teams in which top managers work together for a longer period of time. Katz and Allen (1982) found, for instance, that the level of innovative behavior in teams rises up to one and a half years after their formation, then stays stable for some time, and declines notably after five years.

As these findings show, top managers tend to favor exploitative activities over exploratory ones (Kahneman and Teversky 1979, Leonard-Barton 1992). However, if top management teams wish to balance both types of learning activities, this is not an option. In fact, the literature has suggested that overcoming the tension between exploration and exploitation is only possible when the contradictions of both learning activities are embraced (Smith and Lewis 2011, Smith and Tushman 2005). When top managers fail to do so, they will continue to think about exploration and exploitation as two separate

organizational activities and will be unable to achieve ambidexterity (Smith and Lewis 2011). Top managers should therefore be stimulated to embrace the tension between exploration and exploitation if top management teams wish to successfully act upon it (Lewis 2000, Smith and Berg 1987). Only in this way, top managers are able to develop an awareness of how both learning activities benefit the organization (Smith and Tushman 2005), and identify those structural configurations that facilitate paradoxical organizational behavior (Andriopoulos and Lewis 2009, O’Reilly and Tushman 2008). Somewhat surprisingly, our understanding of how top management teams influence the ability of top managers to overcome the tension between exploration and exploitation is rather limited (Adner and Helfat 2002, Smith and Lewis 2011). This is due to the fact that very few scholars (with the notable exception of Smith and Tushman 2005 and Jansen et al. 2008) have focused on the relationship between top management teams and organizational ambidexterity. This study starts bridging this research gap by explaining how shared leadership within top management teams influences the ability of top managers to achieve organizational ambidexterity.

Shared Leadership and Organizational Ambidexterity

The literature distinguishes between two alternative top management team decision-making structures (Hambrick 1994, Perlow et al. 2004). Within leadercentric structured top management teams, the CEO makes decisions to balance exploration and exploitation primarily on its own (Gilbert 2005, Westerman et al. 2006). In contrast, top managers within teamcentric structured top management teams work collectively to achieve this balance (Benner and Tushman 2003, Tushman and O’Reilly 1996). There are two reasons why top managers within teamcentric structured top management teams are better able to achieve organizational ambidexterity than top managers within leadercentric structured ones. First, top managers within teamcentric structured top management teams frequently interact with each other, as this provides them opportunities for sharing information (Hackman 2002, Weick and Roberts 1993). The higher the quality of these interactions, the more information is being shared (Eisenhardt et al. 1997). Sharing information enhances top managers’ know-how, as well as their ability to make more balanced decisions (Eisenhardt et al. 1997). Sharing information is also necessary for top managers to collectively achieve organizational goals (Weick and Roberts 1993). By integrating their thinking, top managers are furthermore able to avoid interpersonal conflict over scarce resources (Gilbert 2005). In contrast, leadercentric structured top management teams are

characterized by limited interaction between top managers. The reasoning is that the separation of tasks may result in interpersonal conflict, as top managers compete with each other for scarce resources (Deutsch 1973, Sherif 1971). This limited interaction can withhold top managers from embracing the tension between exploration and exploitation (Smith and Tushman 2005).

Second, exploratory and exploitative activities in leadercentric structured top management teams are generally separated (Edmondson 1999, Hackman 2002, Van de Ven et al. 1999). This enables top management teams to balance the organization’s current activities with ones that are not supported by existing structures and processes (Dutton and Jackson 1987). By focusing on either one of these learning activities, top managers are able to avoid the contradictions of the other (Brown and Eisenhardt 1997, Edmondson 1999). The separation of exploration and exploitation furthermore allows the CEO to assign the most appropriate top manager to each learning activity (Leonard-Barton 1992). Although this allows top managers to provide the CEO with more specific information about the requirements of each learning activity (Adler et al. 1999, Benner and Tushman 2002), it limits their ability to overcome the tension between them (O’Reilly and Tushman 2008). This is because the separation of exploratory and exploitative activities does not allow top managers to experience for themselves how both learning activities relate to each other. Achieving ambidexterity within leadercentric structured top management teams therefore highly depends on the cognitive capacity of a single person, namely that of the CEO. Exploratory and exploitative activities within teamcentric structured top management teams are, as within leadercentric structured top management teams, generally separated (Hackman 2002). However, whereas leadercentric structured top management teams use this separation to reinforce differentiation, teamcentric structured ones use it to reinforce integration (Ancona and Nadler 1989). Such integration stimulates top managers to extend their focus beyond their individual task (Hackman 2002), and share information and ideas that would otherwise have been suppressed in order to quickly achieve group consensus (Eisenhardt et al. 1997). Top managers are thereby encouraged to work collectively towards organizational goals (Wageman 2001), which enhances their ability to identify organizational structures that reinforce both learning activities (Andriopoulos and Lewis 2009, O’Reilly and Tushman 2008).

Hypothesis 1. Shared leadership positively relates to organizational ambidexterity.

Paradoxical Cognition and Reflection as Mediators between Shared Leadership and Organizational Ambidexterity

The ability of teamcentric structured top management teams to help top managers overcome the tension between exploration and exploitation is likely to be mediated by paradoxical cognition and reflection. This is because paradoxical cognition and reflective behavior hold the potential to overcome the cognitive and structural barriers that withhold top managers from making balanced decisions (Leung and Kember 2003, Smith and Tushman 2005). Paradoxical cognition, on the one hand, provides top managers a lens through which to identify linkages between exploration and exploitation (Smith et al. 2010, Smith and Lewis 2011). The identification of these linkages enhances top managers’ awareness about the importance of both learning activities, and stimulates them to make more balanced decisions (Smith and Lewis 2011, Lüscher and Lewis 2008). Reflective behavior, on the other hand, stimulates dialectic discussion, which allows top managers to improve their understanding of how exploratory and exploitative activities relate to one another (Leung and Kember 2003). This improved understanding, in turn, enables top managers to identify structural configurations that facilitate paradoxical organizational behavior (Andriopoulos and Lewis 2009, O’Reilly and Tushman 2008). In the subsequent sections, the mediating roles of paradoxical cognition and reflection in the relationship between shared leadership and organizational ambidexterity are explored in greater depth.

Paradoxical Cognition. Defined as “managerial frames and processes that recognize and embrace contradiction” (Smith and Tushman 2005, pp. 523), paradoxical cognition has been shown to play an instrumental role in achieving ambidexterity (Smith et al. 2010, Smith and Lewis 2011). This is because paradoxical cognition provides top managers a lens through which to identify linkages between both learning activities (Smith et al. 2010, Smith and Lewis 2011). The identification of these linkages enhances top managers’ awareness about the importance of both exploration and exploitation, and enables them to make more balanced decisions (Smith and Lewis 2011, Lüscher and Lewis 2008). Paradoxical cognition does so by stimulating top managers both to differentiate and integrate their thinking about exploration and exploitation (Smith and Lewis 2011). Differentiating between both learning activities allows top managers to formulate distinct goals with regard to exploration and exploitation (Daft and Weick 1984, Walsch 1995). The formulation of these distinct goals allows top managers to reframe their mental model which, in turn, enhances their understanding about both learning activities (Smith and

Lewis 2011). Moreover, distinguishing between exploratory and exploitative activities reduces stress and anxiety that are typically associated with exploration (Smith and Berg 1987). By integrating their improved understandings about exploration and exploitation, top managers are able to identify conceptual linkages between both learning activities they previously were unaware of (Suedfeld et al. 1992). The identification of these conceptual linkages enhances the ability of top managers to achieve ambidexterity (Lüscher and Lewis 2008).

There are two reasons why top managers within teamcentric structured top management teams are more likely to engage in paradoxical thinking than top managers within leadercentric structured top management teams. First, top managers within teamcentric structured top management teams are able to assign top managers to either exploratory or exploitative activities, while allowing them to extend their knowledge about the other (Ancona and Nadler 1989, Eisenhardt et al. 1997, Smith and Tushman 2005). This is because teamcentric structured top management teams are focused on the collective achievement of goals (Van de Ven et al. 1999, Van Wijk et al. 2008), and are generally less formalized and centralized than leadercentric ones (Ancona and Nadler 1989). The integration of their knowledge allows top managers to achieve levels of cognition that are greater than the sum of each top manager’s individual understanding (Weick and Roberts 1993). This is because different top managers bring in their own perspectives and experiences (Audia and Goncalo 2007). These different perspectives and experiences allow top managers to look at exploration and exploitation from different angles (Audia and Goncalo 2007, Hargadon and Bechky 2006). This, in turn, improves their understanding of how both learning activities benefit the organization (Dutton and Jackson 1987).

Second, top managers within teamcentric structured top management teams are more likely to engage in paradoxical cognition than top managers within leadercentric structured top management teams because they more frequently interact with each other (Hackman 2002, Weick and Roberts 1993). Frequent interaction has been shown to contribute to the development of top managers’ understanding of how both learning activities enhance the organization’s success (Eisenhardt et al. 1997). Top managers are furthermore able to catch each others’ mistakes when they frequently interact with each other (Hinsz 1990). This allows them to screen out bad ideas, and identify structural configurations that benefit both exploratory and exploitative activities. Within leadercentric structured top management teams, in contrast, frequent interaction between top managers is seen as dysfunctional, as it stimulates interpersonal conflict over organizational resources

(Deutsch 1973, Sherif 1971). Top managers within leadercentric structured top management teams are therefore less likely to develop the cognitive capacity to overcome the tension between exploration and exploitation than top managers within teamcentric structured top management teams.

The contradictory nature of exploration and exploitation imposes considerable role conflicts on top managers (Emsley 2003, Kahneman 1973). Top managers tend to resolve these role conflicts by focusing their efforts on either one of these learning activities (Cao et al. 2009, Floyd and Lane 2000). However, if top managers want to contribute to the achievement of organizational ambidexterity, this is not an option. In fact, top managers should be made aware of how both exploration and exploitation benefit the organization (Smith and Lewis 2011). Such an awareness can only be fully developed through paradoxical cognition (Lewis 2000, Smith and Berg 1987). This study therefore proposes that paradoxical cognition is necessary for top managers to achieve ambidexterity.

Hypothesis 2. Paradoxical cognition mediates the relationship between shared leadership and organizational ambidexterity.

Reflection. Defined as “the extent to which teams collectively reflect upon and adapt their work methods and functioning” (Schippers et al. 2014, pp. 1), reflection has been shown to play an important role in achieving organizational ambidexterity (Lewis 2000, Lüscher and Lewis 2008). This is because reflective behavior stimulates dialectic discussion between top managers (Schippers et al. 2008). These discussions allow top managers to share ideas with one another, and raise doubts and questions about particular organizational concerns (Schippers et al. 2008). Reflective behavior thereby enables top managers to link new insights to existing knowledge bases and past experiences (Smith and Tushman 2005, Tsoukas 2009). This, in turn, allows them to critique assumptions that would otherwise have been taken-for-granted (Raelin 2001), and may result in an improved understanding of how the activities which they are responsible for relate to those of their colleagues (Kemmis 1985, Smith and Tushman 2005). Put differently, reflective behavior makes top managers more aware of how both exploratory and exploitative activities benefit the organization, and provides them insights of how both learning activities can possibly be combined (Schippers et al. 2014).

Becoming more reflective requires that top managers challenge their organization’s existing organizational structures, processes, and strategies (Gray 2007). Because

organizational change may lead to intra-organizational skepticism and conflict, top managers tend to stick to the old way of doing things (Alvesson and Willmott 1992, Reynolds 1999). Top managers should therefore be actively stimulated to engage in reflective behavior. The literature highlights two important ways in which this can be achieved. First, it has been suggested that reflective behavior can be stimulated through storytelling (Gray 2007). Storytelling is a collective activity that enables top managers to improve their understanding about the norms, values, and ideas of their colleagues (Gold and Holman 2001). It encourages top managers to share information with each other, which allows them to establish social cohesion (Bolton 2001). Storytelling furthermore enables top managers to explain complex problems in an understandable way to their colleagues throughout the organization (Gold et al. 2002). This may result in a broader organizational acceptance when top managers have to make decisions that radically change the course of the organization (Eisenhardt et al. 1997). Such an acceptance is especially useful in situations in which environmental circumstances frequently change. Top managers within teamcentric structured structures are more likely to engage in storytelling than top managers within leadercentric ones. This is because top managers within teamcentric structures more frequently interact with their colleagues than top managers within leadercentric structures do (Hackman 2002, Weick and Roberts 1993). In fact, interactions between colleagues are generally limited within leadercentric structures (Hackman 2002). Shared leadership therefore plays an important role to stimulate top managers to engage in reflective behavior.

Second, reflective behavior can be stimulated through constructive conversation (Gray 2007). Constructive conversation allows top managers to reflect upon their own ideas and thoughts, as well as upon their behavior in reaction to the ideas and thoughts of others (Mirvis and Ayas 2003). This critical examination provides top managers new insights about the way in which their colleagues think about particular topics (Isaacs 1993). This, in turn, allows them to develop a common language (Jacobs and Heracleous 2005). However, top managers will only start developing this common language if they believe that their colleagues’ different ways of thinking has the potential to valuably contribute to the achievement of the overall organizational goals (Schein 1996). There are two reasons why top managers within teamcentric structured top management teams are more likely to engage in reflective behavior than top managers within leadercentric structured top management teams. First, top managers within teamcentric structured top management teams more frequently interact with their colleagues than top managers within leadercentric

structured ones do (Hackman 2002, Weick and Roberts 1993). This means that top managers within teamcentric structured top management teams face more situations that allow them to reflect upon their own behavior, as well as upon the behavior of their colleagues. Second, top managers within teamcentric structured top management teams are more likely to develop a common language with their colleagues than top managers within leadercentric structured top management teams do. This is because exploratory and exploitative activities in teamcentric structured top management teams are integrated, which stimulates organizational members to extend their focus beyond their own individual task (Edmondson 1999, Eisenhardt et al. 1997, Gilbert 2005). Top managers are therefore more likely to believe that colleagues are able to valuably contribute to the achievement of the organizational goals.

Prior research found that top managers prefer activities that involve action over ones that involve reflection (Daudelin 1996). Focusing on action alone, however, does not allow top managers to learn (Jarvis 1995). In fact, it has been suggested that reflection is a necessary condition for learning to take place (Gray 2007). Since the ability of top managers to overcome the tension between exploratory and exploitative activities depends on their understanding of how both learning activities benefit the organization (Dutton and Jackson 1987), and this understanding is developed by their ability to learn from each other (Eisenhardt et al. 1997), it seems that reflective behavior is necessary for top management teams to achieve ambidexterity.

Hypothesis 3. Reflection mediates the relationship between shared leadership and organizational ambidexterity.

Methodology

This study is part of an ongoing research project designed to develop an innovation benchmark among Dutch and German organizations. Data were obtained by means of an online survey in 2012 among top managers from small and medium organizations (employing 20 to 500 full-time employees). Although top management teams in small and medium organizations face similar competitive pressures as top management teams in large organizations do, they lack slack resources that enable top management teams in large

organizations to focus entirely on either exploration or exploitation (Nooteboom 1994, Voss and Voss 2013). Top management teams in small and medium organizations therefore rely on their ability to achieve ambidexterity even more than top management teams in large organizations do (Lubatkin et al. 2006, Tushman and O’Reilly 1997).

After screening the data on missing values and the absence of key informant criteria, data from 282 organizations were retained. On average, these organizations employed 124.83 full-time employees (s.d. = 110.73), and were 46.44 years old (s.d. = 8.33). The average top management team consisted of 7.82 team members (s.d. = 10.76). Furthermore, 85.10 per cent of the organizations mainly served business clients, 19.90 per cent of them were family-owned, and 47.52 per cent of the organizations were located in Germany against 52.48 per cent in the Netherlands. The final sample frame covered a wide range of industries including manufacturing (33.7%), professional services (20.2%), wholesale (14.5%), administrative support services (5.7%), construction (4.6%), and other industries (21.3%).

Several tests were performed in order to ensure the usefulness, reliability, and validity of the data for testing the proposed hypotheses. First, the data were tested for non-response bias. In order to do so, respondents and non-respondents were compared to each other on the basis of the number of full-time employees, total assets, and prior financial performance of their respective organization. T-tests showed no significant differences between both groups on any of these factors (p > 0.05). As such, non-response bias seemed not to impose a problem.

Second, the data were tested for individual respondent bias. In order to do so, a second top manager from each organization was invited to participate in the survey. This resulted in 41 responses, representing 14.54 per cent of the final sample. Interrater agreement scores (rwg) between the first and second group of top managers were calculated for their responses on the main variables (James et al. 1993). The median interrater agreement scores were 0.85 for exploration, 0.82 for exploitation, 0.91 for shared leadership, 0.81 for paradoxical cognition, and 0.93 for reflection. All interrater agreement scores are above the cut-off value of 0.7, which indicates adequate agreement amongst respondents (James et al. 1993).

Third, the data were tested for reliability issues associated with single informant data. In order to do so, intra-class correlations were examined. These appeared to be significantly (p < 0.001), indicating a strong level of interrater reliability (Jones et al. 1983).

Fourth, the convergent and discriminant validity of the data was assessed. In order to do so, an integrated confirmatory factor analysis was conducted (Hair et al. 2006). The model fitted the data adequately (χ² = 471.57, DF = 280, CFI = 0.94, NFI = 0.90, RMSEA = 0.06), and all item loadings were as proposed and significant (p < 0.01). The discriminant validity of the data was confirmed by means of a factor correlation matrix. All factor correlations were below the cut-off value of 0.7.

Finally, the risk for common method bias was reduced and assessed. In order to reduce the risk for common method bias, respondents of the survey were assured confidentiality and were offered a report that compared their organization against the industry’s average. The items of the survey had furthermore been pretested, and respondents were clearly explained the purpose of the survey (Podsakoff et al. 2003). Next, the risk for common method bias was assessed by means of a Harman’s one-factor test. All main variables clearly loaded on their intended distinct factor, and together explained 66 per cent of the variance. Individually, the largest factor explained 28 per cent of the variance (rotated loadings). Therefore, common method bias did not seem to impose a problem (Podsakoff and Organ 1986).

Measurement and Validation of Constructs

All of the study’s constructs were measured on a seven-point scale ranging from (1) “strongly disagree” to (7) “strongly agree”. Organizational ambidexterity and reflection were measured by using existing scales from the literature. These scales were verified through various analyses. Because appropriate scales were not available for shared leadership and paradoxical cognition, the literature was reviewed to develop new scales for these constructs. To refine the scale items and enhance the validity of both these scales, five managers and two corporate members were asked to provide suggestions for further refinement. This resulted in a five-item scale for shared leadership and a six-item scale for paradoxical cognition. All items can be found in the appendix.

Organizational Ambidexterity. In accordance with prior scholars, a measure for organizational ambidexterity was developed in two steps (Gibson and Birkinshaw 2004, He and Wong 2004). First, information on exploration and exploitation was captured by means of two separate scales. These scales were adopted from Jansen et al. (2006). The exploration scale measures the extent to which organizations depart from existing knowledge bases in order to pursue radical innovations for emerging customers. Example

items are “We regularly experiment with new products and services in the market” and “We frequently utilize new opportunities in new markets”. The exploitation scale, on the other hand, measures the extent to which organizations build upon existing knowledge bases in order to pursue incremental innovations for existing customers. Example items are “Our organization expands services for existing clients” and “We regularly implement small adaptations to existing products and services”. Exploratory factor analyses indicated that one item from each scale had to be dropped. This resulted in a six-item scale for both exploration (α = .87) and exploitation (α = .83). A joint exploratory factor analysis showed that both factors had eigenvalues greater than one. Each item furthermore clearly loaded on its intended factor (all factor loadings were above 0.74 with cross-loadings below 0.25). Second, the multiplicative interaction between exploration and exploitation was computed. This resulted in a measure for organizational ambidexterity as a whole (Cao et al. 2009, Gupta et al. 2006).

Shared Leadership. An unawareness of any published scale that measures shared leadership led to the development of a new one. Based on Smith and Tushman (2005), a scale was developed that measures the extent to which strategic decisions within top management teams are jointly made and implemented. Five academic experts and two corporate members were consulted for further refinement of the scale’s items. This resulted in a five-item scale. An exploratory factor analysis confirmed the five-item scale to be reliable (α = .92). All items of the scale can be found in the appendix.

Paradoxical Cognition. Because of an unawareness of any published scale that measures paradoxical cognition, a new one was developed. The items on this scale measure the extent to which top management teams cognitively engage in differentiation and integration (Smith and Tushman 2005). Five academic experts and two corporate members were consulted for further refinement of the scale’s items. For differentiation, the scale measures the extent to which top management teams distinguish between existing products and new ones. For integration, it measures the extent to which top management teams identify linkages and synergies between existing products and new ones. An exploratory factor analysis led to the exclusion of one item, resulting in a five-item scale (α = .83). All items can be found in the appendix.

Reflection. Information on reflection was obtained by means of a scale developed by Schippers et al. (2007). This scale measures the extent to which top management teams reflect upon and modify their functioning. Example items are “Within our top management team we talk about different ways in which we can reach our objectives” and “Within our top management team we work out what out what we can earn from past activities”. An exploratory factor analysis pointed at a six-item scale (α = .94).

Control Variables. This study controlled for several alternative explanations. Larger organizations may be better able to achieve ambidexterity than smaller ones (Ahuja and Lampert 2001). Therefore, this study controlled for organizational size, measured as natural logarithm of the number of full-time employees. Gilbert (2005) found that young organizations are more likely to engage in exploratory activities than older ones. This study therefore controlled for organizational age, measured as natural logarithm of the number of years since founding. The size of an organization’s top management team may affect the extent to which leadership is shared. This study therefore controlled for top management team size, measured by the number of top managers who are responsible for strategic decisions making (Siegel and Hambrick 2005). Slack resources allow organizations to focus entirely upon either exploration or exploitation (Voss and Voss 2013). This study therefore controlled for slack resources. Furthermore, the organizations from the sample may have specialized themselves in different markets with different products and services. Different markets may influence the extent to which organizations balance exploratory and exploitative activities. Following Jansen et al. (2006), this study therefore controlled for organizational client focus. Organizations which mainly focused on private clients were coded as 0 and organizations that mainly focused on business clients were coded as 1. Duncan (1972) and Swamidass and Newell (1987) found that managerial decisions are based upon perceptions about the environment, rather than on the actual environmental state itself. Perceived environmental uncertainty therefore seems to play an important role in top management teams’ structural decisions. This study therefore controlled for environmental uncertainty. Ensley and Pearson (2005) showed that there are significant differences between family owned and non-family owned organizations in terms of cohesion, conflict, group potency, and strategic cognition sharing. This study therefore included a dummy variable for organizational ownership. More specifically, non-family owned organizations were coded as 0 and family owned ones were coded as 1. Finally, the data used in this study were obtained from top managers from both Dutch and German organizations. It may be that there are

differences between these countries that influence the results (Schwartz 1999). This study therefore controlled for country by coding Dutch organizations as 0 and German organizations as 1.

Analysis and Results

Table 1 presents the scale means, standard deviations, intercorrelations, and reliabilities for each of the study’s variables. Variance inflation factors (VIFs) were calculated in order to determine the degree of multicollinearity among the different variables. The maximum VIF value was 2.40. This value is well below the cut-off point of 10 (Neter et al. 1990).

Table 2 presents the results of the hierarchical regression analyses for organizational ambidexterity (Models 1-5), paradoxical cognition (Models 6 and 7), and reflection (Models 8 and 9). Models 1, 6, and 8 are baseline models that contain the control variables. Model 2, introduces shared leadership as an antecedent for organizational ambidexterity. Model 3 examines the mediating role of paradoxical cognition in the relationship between shared leadership and organizational ambidexterity. Model 4 does the same for reflection. Models 7 and 8 examine the direct effect of shared leadership on paradoxical cognition and reflection. Mean SD 1 2 3 4 5 6 7 8 9 10 11 12 1. Organizational ambidexterity 22.15008.660( – ) 2. Shared leadership 05.14001.230-.350(.92) 3. Paradoxical cognition 04.80001.080-.440-.480(.83) 4. Reflection 05.07001.170-.350-.620-.580(.94) 5. Organizational size* 04.53000.790-.120-.010-.020-.010( – ) 6. Organizational age* 03.82000.190-.030-.130-.030-.160-.040( – ) 7. Top management team size 07.82010.760-.020-.050-.030-.070-.190-.100( – ) 8. Organizational slack 03.95001.160-.270-.250-.140-.300-.170-.050-.020( – ) 9. Organizational client focus 00.15000.360-.070-.080-.040-.090-.000-.000-.010-.060( – ) 10. Environmental uncertainty 04.35000.960-.250-.450-.480-.530-.110-.020-.080-.160-.060( – ) 11. Family ownership 00.80000.400-.050-.090-.020-.040-.150-.060-.090-.110-.110-.120( – ) 12. Country 00.48000.500-.110-.000-.150-.160-.250-.010-.220-.320-.060-.140-.260( – )

Note: N=282, * natural log, correlations greater than r = |.12| are significant at p < .05.

M od el 1 M od el 2 M od el 3 M od el 4 M od el 5 M od el 6 M od el 7 M od el 8 M od el 9 Co ntr ols O rg an iz at io na l s iz e a -.09 *** -.09 *** -.06 *** -.08 *** -.06 *** -.08 *** -.07 *** -.05 *** -.04 *** O rg ani za tiona l a ge a -.04 *** -.07 *** -.07 *** -.09 *** -.07 *** .02 *** .02 *** .14 ** * .09 * ** To p m an ag em en t t ea m siz e .03 *** .05 *** .05 *** .04 *** .04 *** -.02 *** .00 *** .02 *** .05 *** O rg an iz at io na l s la ck .20* * * .15* ** .14 * ** .11 *** .13 * ** .10 *** .03 *** .27 *** .19 *** O rg an iz at io na l clie nt fo cu s b .07 *** .06 *** .06 *** .05 *** .05 *** .02 *** .00 *** .07 *** .05 *** E nv iro nm en ta l u nc er ta in ty .24 *** .13 * ** .02 *** .08 *** .01 *** .45 *** .31 *** .46 *** .28 *** Fa m ily o w ne rs hip c -.04 *** -.06 *** -.03 *** -.05 *** -.03 *** -.05 *** -.07 *** -.03 *** -.06 *** Cou nt ry d -.05 *** -.05 *** -.11 *** -.10 *** -.12 * ** .16 ** * .16 ** * .20 *** .20 *** M ain ef fec t Sh ar ed le ad er sh ip .26* ** .12 *** .18 * ** .12 *** .34 *** .44 *** M ed ia tor s Pa ra do xic al co gn itio n .36 *** .35 *** Re fle ct io n .18 * ** .06 *** Ad just ed R 2 .11 *** .16 *** .25 *** .18 *** .25 *** .23 *** .32 *** .37 *** .51 *** Δ A dj us te d R 2 .05* ** .08 *** .02 * ** .08 *** .09 *** .14 *** Re fle ct io n Pa ra dox ic al Cog ni tion T ab le 2. H iera rch ica l R eg res si on An al ys es : E ffect s o n O rg an iza tio na l Amb id ex teri ty, P ara do xi ca l C og ni tio n, a nd R ef lect io n O rg an iz at io na l A m bid ex te rit y N ote : St an da rd iz ed re gr es sio n co ef fic ie nt s a re re po rte d, * p < .05, * * p < .01, * ** p < .001, a N at ur al lo g, b D um m y v ar ia ble ( 0= Bu sin es s, 1= Pr iv at e) , c D um m y v ar ia ble ( 0= Fa m ily o w ne d, 1 = N on -fa m ily o w ne d) , d D um m y v ar ia ble ( 0= Th e N et he rla nd s, 1 = G er m an y) .

Hypothesis 1 proposed that shared leadership within top management teams allows organizations to better balance exploration and exploitation. Model 2 shows that the relationship between shared leadership and organizational ambidexterity is positive and significant (β = 0.26, p < 0.001). Hypothesis 1 is therefore supported.

Hypotheses 2 and 3 proposed that paradoxical cognition and reflection mediate the relationship between shared leadership and organizational ambidexterity. Following Baron and Kenny (1986), these hypotheses were tested by means of a four-step procedure. First, the relationship between shared leadership and organizational ambidexterity was examined. Model 2 shows that this relationship is positive and significant (β = 0.26, p < 0.001). Second, it was assessed whether shared leadership significantly relates to the mediating variables. Models 7 and 9 show that this is the case for paradoxical cognition (β = 0.34, p < 0.001) and reflection (β = 0.44, p < 0.001). Third, it was assessed whether paradoxical cognition and reflection significantly relate to organizational ambidexterity. Models 3 and 4 show that this is the case for paradoxical cognition (β = 0.36, p < 0.001) and reflection (β = 0.18, p < 0.05). Fourth, the significant relationship between shared leadership and organizational ambidexterity has to become insignificant when the mediating variables make their entrance into the regression model. Model 4 shows that this is the case for paradoxical cognition (β = 0.14, insignificant). However, Model 3 shows that the relationship between shared leadership and organizational ambidexterity does not become insignificant when reflection is introduced into the regression model (β = 0.18, p < 0.05).

Overall, the results indicate that paradoxical cognition mediates the relationship between shared leadership and organizational ambidexterity. A bootstrapped Sobel test with bias-corrected confidence estimates was performed to confirm this finding (Hayes 2013). The 95 per cent confidence intervals of the proposed mediating effect were obtained with 5000 bootstrap resamples (Preacher and Hayes 2013). The results confirm the mediating role of paradoxical cognition in the relationship between shared leadership and organizational ambidexterity (Sobel Z = 4.35, CI = 0.61 to 1.51, p < 0.001). Hypotheses 2 is therefore supported.

The results furthermore show that although the relationship between shared leadership and organizational ambidexterity does not become insignificant when reflection is introduced, the relationship does decrease in power (from β = 0.26, p < 0.001 to β = 0.18, p < 0.05). This indicates that reflection partially mediates the relationship between shared leadership and organizational ambidexterity (Sobel Z = 3.37, CI = 0.42 to 1.83, p <

0.001). This is however not in accordance with Hypothesis 3. As such, this hypothesis found no support.

Discussion and Conclusion

Strategy scholars have acknowledged the importance of balancing exploratory and exploitative activities in enhancing organizational effectiveness, performance, and survival (Gibson and Birkenshaw 2004, Gupta et al. 2006). Yet, the contradictory requirements of both learning activities have proven this to be a rather difficult task (Raisch and Birkenshaw 2008, Smith and Tushman 2005). Although prior scholars emphasized that top management teams play an important role in managing these contradictions, scholars that focus on how they contribute to achieving ambidexterity remain scarce (Adner and Helfat 2002, Smith and Lewis 2011). This study proposed that shared leadership within top management teams plays an important role in achieving organizational ambidexterity. The results show that leadercentric and teamcentric structured top management teams indeed differentially influence the ability of top managers to balance exploration and exploitation. To extend our understanding of how top managers overcome the tension between these two learning activities, this study furthermore examined the role of paradoxical cognition and reflection in achieving ambidexterity. Paradoxical cognition, on the one hand, is shown to fully mediate the relationship between shared leadership and ambidexterity. Surprisingly, reflection appears only to partially mediate this relationship. Overall, this study not only extends our understanding about ambidexterity and its antecedents, but also provides some practical handholds that might enable organizations to actually become more ambidextrous.

Theoretical and Practical Implications

The literature distinguishes between two alternative top management team decision-making structures. Leadercentric structures, on the one hand, facilitate differentiation between exploratory and exploitative activities, which enables top management teams to balance their current activities with ones that are not supported by existing structures and processes (Dutton and Jackson 1987, Gilbert 2005). Teamcentric structured top management teams, on the other hand, facilitate a balance between exploratory and exploitative activities by integrating the requirements of both learning activities (Van Wijk et al. 2008, Smith and

Tushman 2005, Suedfeld et al. 1992). Whereas Smith and Tushman (2005) have argued that leadercentric and teamcentric structured top management teams are equally able to overcome this tension, this study shows that this is not the case. More specifically, it is shown that teamcentric structured top management teams are better able to achieve ambidexterity than leadercentric structured ones. This finding extends our knowledge of how top management teams are able to develop the ability of top managers to overcome the tension between exploration and exploitation. More specifically, the ability of top management teams to balance exploration and exploitation appears to be associated with the integration of top managers’ tasks, roles, and responsibilities. This implies that top managers should be stimulated to wear multiple hats if top management teams want to enhance their ability to achieve ambidexterity (Ancona and Nadler 1989). The definition of overarching organizational goals might help them to successfully do so (Dutton and Jackson 1987). However, it may be that organizations do not benefit from shared leadership in each and every instance. Therefore, future research is necessary to deepen our understanding about the relationship between shared leadership and organizational ambidexterity. It could be, for instance, that shared leadership within top management teams positively relates to organizational ambidexterity only when organizations operate within certain industries. In that sense, organizations that operate in fast-changing industries like the telecom industry may benefit more from shared leadership than organizations that operate in more traditional industries like the fruit picking industry. Similar, it may be that organizational culture plays an important role in the (dis)functionality of shared leadership within top management teams.

This study furthermore examined the mediating role of paradoxical cognition and reflection in overcoming the tension between exploration and exploitation. Prior studies suggested that top managers should overcome this tension if they wish to meaningfully contribute to the organization’s success (Smith et al. 2010). Making top managers aware of the importance of both exploration and exploitation is a first step in doing so (Smith and Tushman 2005, Weick 1979). This study shows that paradoxical cognition is necessary for top manager to develop such an awareness. By doing so, this study shows that paradoxical cognition plays a key role in achieving ambidexterity, thereby extending the strategic management literature (Poole and Van de Ven 1989). The results of this study furthermore provide valuable contributes to the literature on conflict and contradiction (Lewis 2000, Poole and Van de Ven 1989). This is because, until now, it has been suggested that leadercentric and teamcentric structured top management teams are equally able to

stimulate top managers to engage in paradoxical behavior (Amabile 1996). This study shows that this is not the case. In fact, the results show that shared leadership positively and significantly relates to paradoxical cognition (β = 0.34, p < 0.001). For that reason, teamcentric structured top management teams appear to be better able to stimulate top managers to behave paradoxically than leadercentric structured ones. Of course, this finding provides various interesting avenues for future research. This is especially the case since Eisenhardt (1999) noted that teamcentric structured top management teams should be aware of dysfunctional conflict and political behavior, as both negatively influence the organization’s performance. It has furthermore been suggested that top management teams should be aware of the fact that conflict stimulates groupthink (Janis 1982), and biased information seeking (Schulz-Hardt et al. 2002). This leads to the question under what circumstances conflict positively or negatively contributes to the organization’s success. It may be that conflict appears to be more or less important at different organizational levels. Furthermore, it may be that uncertainty plays an important role here. Future research on managing conflict and contradiction may therefore enhance our understanding about a wide range of organizational issues (Cameron and Quinn 1988, Poole and Van de Ven 1989).

Finally, this study examined the role of reflection in overcoming the tension between exploration and exploitation. Prior research found that reflection involves a purposeful process of discovery that enables organizations to link new insights to existing knowledge bases and past experiences (Gray 2007, Leung and Kember 2003). Because this linking enables top managers to identify organizational configurations that facilitate paradoxical organizational behavior (Andriopoulos and Lewis 2009, O’Reilly and Tushman 2008), it seemed reasonable to propose that reflective behavior of top managers is necessary to achieve ambidexterity. Surprisingly, the results of this study do not support this hypothesis. More specifically, it appears that reflection only partially mediates the relationship between shared leadership and organizational ambidexterity. A possible explanation for this observation lies in the very nature of reflection, which can best be described by comparing the concept to paradoxical cognition. Whereas paradoxical cognition always confronts top managers with the distinct nature of exploration and exploitation, this is not necessarily the case for reflection. In other words, although reflective behavior holds the potential for top managers to learn from each other, this does not mean that it contributes to an improved understanding about the importance of either exploration or exploitation in each and every situation. For instance, in the situation wherein two top managers who are both responsible

for exploratory activities collectively reflect upon their behavior, it is unlikely that any one of them improves its understanding of the organizational necessity of exploitation. It follows that top managers may only be able to overcome the tension between exploration and exploitative through reflective behavior when the reflection group consists of top managers that are responsible for both exploratory and exploitative activities.

Limitations and Future Research

This study has several limitations that open up avenues for future research. First, and because the literature suggested that small and medium organizations do not have the resources to focus entirely on either exploration or exploitation (Nooteboom 1994, Voss and Voss 2013), this study exclusively focused on small and medium organizations. Small and medium organizations rely on their ability to balance exploration and exploitation even more than large organization do, which makes them ideal candidates to study ambidexterity. However, focusing exclusively on small and medium organizations has implications for the external validity of this study’s findings. Future research could therefore expand this study by focusing on large organizations as well.

Second, the data in this study have been obtained in a cross-sectional manner. For that reason, the results of this study should be interpreted with caution. In that sense, it remains unclear how shared leadership within top management teams impacts the ability of top managers to contribute to achieve ambidexterity over time. It may be that over time, top managers develop contradictory ideas about what the organization should look like in the near future. Contradictory ideas may increase the risk of interpersonal conflict within top management teams, and might reduce the ability of top managers to achieve ambidexterity. Future research could therefore replicate the study after the collection of longitudinal data.

Finally, this study examined the impact of shared leadership on organizational ambidexterity. A measure for organizational ambidexterity was developed in two steps. First, information about both exploration and exploitation was captured by means of two separate scales that were adopted from Jansen et al. (2006). Second, the multiplicative interaction between exploration and exploitation was computed, resulting in a measure for organizational ambidexterity as a whole (Cao et al. 2009, Gupta et al. 2006). Future research could focus on how shared leadership within top management teams influences exploration and exploitation separately. Such research would lead to a deeper understanding about the role of top management teams in achieving ambidexterity. In addition, although the scales

of exploitation and exploration were proven to be valid and reliable, they remain proxies. Future research could examine the impact of leadercentric and teamcentric structured top management teams on measurable organizational aspects like financial performance, market share, or survival.

Notwithstanding its shortcomings, this study provides several new insights about organizational ambidexterity and its antecedents. Most importantly, it is shown that shared leadership within top management teams plays an important role achieving ambidexterity, and that paradoxical cognition is necessary for top managers to overcome the tension between exploration and exploitation. This study furthermore shows that reflection enhances the ability of top managers to overcome the tension between both learning activities, yet reflective behavior does not appear to be a necessary condition for top management teams to achieve ambidexterity.

References

Adler, P. S., Goldoftas, B., and Levine, D. I. (1999). Flexibility versus efficiency: A case study of model changeovers in the Toyota production system. Organization Science, 10: 43-68.

Adner, R., and Helfat, C. (2002). Corporate effects and dynamic managerial capabilities. Strategic Management Journal, 24: 1011-1025.

Ahuja, G., and Lampert, C. M. (2001). Entrepreneurship in the large corporation: A longitudinal study of how established firms create breakthrough inventions. Strategic Management Journal, 22: 521-543.

Alvesson, M., and Willmott, H. (1992). On the idea of emancipation in management and organizational studies. Academy of Management Review, 17: 432-64.

Amabile, T. M. (1996). Creativity in context (4th ed.). Boulder: Westview Press.

Ancona, D. G., and Nadler, D. A. (1989). Top hats and executive tales: Designing the senior team. Sloan Management Review, 31: 19-28.

Ancona, D. G., Okhuysen G. A., and Perlow, L. A. (2001). Taking time to integrate temporal research. Academy of Management Review, 26: 512-529.

Andriopoulos, C., and Lewis, M. W. (2009). Exploitation-exploration tensions and organizational ambidexterity: Managing paradoxes of innovation. Organization Science, 20: 696-717.

Audia, P. G., and Goncalo, J. A. (2007). Past success and creativity over time: A study of investors in the hard disk drive industry. Management Science, 53: 1-15.

Barney, J. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17: 99-120.

Baron, R. M., and Kenny, D. A. (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51: 1173-1182.

Bazerman, M. (1998). Judgment in managerial decision making. New York: Wiley.

Bazerman, M., and Watkins, M. (2004). Predictable surprises: The disasters you should have seen coming and how to prevent them. Boston: Harvard Business School Press.

Beckman, C. M., Haunschild, P. R., and Phillips, D. J. (2004). Friends or strangers? Firm-specific uncertainty, market uncertainty, and network partner selection. Organization Science, 15: 259-275.

Benner, M. J., and Tushman, M. L. (2002). Process management and technological innovation: A longitudinal study of the photography and paint industries. Administrative Science Quarterly, 47: 676-706.

Benner, M. J., and Tushman, M. L. (2003). Exploitation, exploration, and process management: The productivity dilemma revisited. Academy of Management Review, 2: 238-256.

Bolton, G. (2001). Reflective practice: Writing and professional development. London: Paul Chapman Publishing Limited.

Brown, S. L., and Eisenhardt, K. M. (1997). The art of continuous change: Linking complexity theory and time-paced evolution in relentlessly shifting organizations. Administrative Science Quarterly, 42: 1-34.

Burgelman, R. A. (2002). Strategy as vector and the inertia of coevolutionary lock-in. Administrative Science Quarterly, 47: 325-357.

Cameron, K., and Quinn, R. (1988). Organizational paradox and transformation. In R. Quinn, and K. Cameron (Eds.), Paradox and transformation (pp. 1-18). Cambridge: Ballinger Publishing Company.

Cao, Q., Gedajlovic, E., and Zhang, H. (2009). Unpacking organizational ambidexterity: Dimensions, contingencies, and synergistic effects. Organization Science, 20: 781-796. Daft, R. L., and Weick, K. E. (1984). Toward a model of organizations as interpretation

Daudelin, M. W. (1996). Learning from experience through reflection. Organizational Dynamics, 24: 36-48.

Denision, D. R., Hooijberg, R., and Quinn, R. E. (1995). Paradox and performance: Toward a theory of behavioral complexity in managerial leadership. Organizational Science, 6: 524-540.

Deutsch, M. (1973). The resolution of conflict: Constructive and destructive processes. New Haven: Yale University Press.

Doty, D. H., Glick, W. H., and Huber, G. P. (1993). Fit, equifinality, and organizational effectiveness: A test of two configurational theories. Academy of Management Journal, 36: 1196-1250.

Duncan, R. B. (1972). Characteristics of organizational environments and perceived environmental uncertainty. Administrative Science Quarterly, 17: 313-327.

Duncan, R. B. (1976). The ambidextrous organization: Designing dual structures for innovation. In R. H. Kilmann, L. R. Pondy, and D. Slevin (Eds.), The management of organization (pp. 167-188). New York: North Holland.

Dutton, J. E., and Jackson, S. B. (1987). Categorizing strategic issues: Links to organizational action. Academy of Management Review, 12: 76-90.

Edmondson, A. C. (1999). Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44: 350-383.

Edmondson, A. C., Roberto, M. R., and Watkins, M. (2003). A dynamic model of top management team effectiveness: Managing unstructured task streams. Leadership Quarterly, 219: 1-29.

Eisenhardt, K. M. (1999). Strategy as strategic decision making. Sloan Management Review, Spring: 65-72.

Eisenhardt, K. M., Kahwajy, J. L., and Bourgeois, L. J. I. (1997). Conflict and strategic choice: How top management teams disagree. California Management Review, 39: 42-62. Eisenhardt, K. M., and Zbaracki, M. J. (1992). Strategic decision making. Strategic

Management Journal, 13: 17-37.

Emsley, D. (2003). Multiple goals and managers’ job-related tension and performance. Journal of Managerial Psychology, 18: 345-356.

Ensley, M. D., and Pearson, A. W. (2005). An exploratory comparison of the behavioral dynamics of top management teams in family and non-family new ventures: Cohesion, conflict, potency, and consensus. Entrepreneurship Theory and Practice, 29: 267-284.

Floyd, S., and Lane, P. (2000). Strategizing throughout the organization: Managing role conflict in strategic renewal. Academy of Management Review, 25: 154-177.

Ghemawat, P., and Costa, J. (1993). The organizational tension between static and dynamic efficiency. Strategic Management Journal, 14: 59-73.

Gibson, C. B., and Birkinshaw, J. (2004). The antecedents, consequences, and mediating role of organizational ambidexterity. Academy of Management Journal, 47: 209-226. Gilbert, C. G. (2005). Unbundling the structure of inertia: Resource vs. routine rigidity.

Academy of Management Journal, 48: 741-763.

Gold, J., and Holman, D. (2001). Let me tell you a story: An evaluation of the use of storytelling and argument analysis in management education. Career Development International, 6: 384-395.

Gold, J., Holman, D., and Thorpe, R. (2002). The role of argument analysis and storytelling in facilitating critical thinking. Management Learning, 33: 371-388.

Gray, D. E. (2007). Facilitating management learning: Developing critical reflection through reflective tools. Management Learning, 38: 495-517.

Gupta, A. K., Smith, K. G., and Shalley, C. E. (2006). The interplay between exploration and exploitation. Academy of Management Journal, 4: 693-706.

Hackman, J. R. (2002). Leading teams: Setting the stage for great performances. Boston: Harvard Business School Press.

Hair, J. F., Black, W. C., Babin, B. J., Anderson, R. E., and Tatham, R. L. (2006). Multivariate data analysis (6th ed.). New Jersey: Prentice Hall.

Hambrick, D. (1994). Top management groups: A conceptual integration and reconsideration of the “team” label. In B. M. Staw, and L. Cummings (Eds.), Research in Organizational Behavior (171-214). Greenwich: JAI Press.

Hambrick, D. C., and Mason, P. A. (1984). Upper echelons: The organization as a reflection of its top managers. Academy of Management Review, 9: 193-206.

Hargadon, A. B., and Bechky, B. A. (2006). When collections of creatives become creative collectives: A field study of problem solving at work. Organization Science, 17: 484-500. Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A

regression-based approach. New York: The Guilford Press.

He, Z.-L., and Wong, P.-K. (2004). Exploration vs. exploitation: An empirical test of the ambidexterity hypothesis. Organization Science, 15: 481-494.

Henderson, R. M., and Clark, K. B. (1990). Architectural innovation: The reconfiguration of existing product technologies and the failure of established firms. Administrative Science Quarterly, 35: 9-30.

Hinsz, V. B. (1990). Cognitive and consensus processes in group recognition memory performance. Journal of Personality and Social Psychology, 59: 705-718.

Isaacs, W. (1993). Taking flight: Dialogue, collective thinking and organizational learning. Organizational Dynamics, 22: 24-40.

Jacobs, C. D., and Heracleous, L. T. (2005). Answers for questions to come: Reflective dialogue as an enabler of strategic innovation. Journal of Organizational Change Management, 18: 338-52.

James, L. R., Demaree, R. G., and Wolf, G. (1993). Rwg: An assessment of within-group interrater agreement. Journal of Applied Psychology, 78: 306-309.

Janis, I. L. (1982). Groupthink (2nd ed.). Boston: Houghton Mifflin.

Jansen, J. J. P., George, G., Van den Bosch, F. A. J., and Volberda, H. W. (2008). Senior team attributes and organizational ambidexterity: The moderating role of transformational leadership. Journal of Management Studies, 45: 982-1007.

Jansen, J. J. P., Van den Bosch, F. A. J., and Volberda, H. W. (2006). Exploratory innovation, exploitative innovation, and performance: Effects of organizational antecedents and environmental moderators. Management Science, 52: 1661-1674. Jarvis, P. (1995). Adult and continuing education (2nd ed.). London: Routledge.

Jones, A. P., Johnson, L. A., Butler, M. C., and Main, D. S. (1983). Apples and oranges: An empirical comparison of commonly used indices of interrater agreement. Academy of Management Journal, 26: 507-519.

Kahneman, D. (1973). Attention and effort. Englewood Cliffs: Prentice-Hall.

Kahneman, D., and Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica, 47: 263-291.

Katila, R., and Ahuja, G. (2002). Something old, something new: A longitudinal study of search behavior and new product introduction. Academy of Management Journal, 45: 1183-1194.

Katz, R., and Allen, T. J. (1982). Investigating the not invented here (NIH) syndrome: A look at the performance, tenure, and communication patterns of 50 R&D project groups. R&D Management, 12: 7-19.