The strengthening effect of organizational transactive memory systems on the relation between shared decision-making and organizational ambidexterity

(1)

The Strengthening Effect of Organizational Transactive

Memory Systems on the Relation Between Shared

Decision-Making and Organizational Ambidexterity.

Name: Eline van Walsem

Student Number: 10002989

Qualification: Msc. Business Administration – Strategy track

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

(2)

Statement of Originality

This document is written by Student Eline van Walsem, who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

Abstract

Current literature showed the importance of the role top management teams (TMT) play in achieving organizational ambidexterity. Until now it remained unclear what is the best way to organize a TMT. This study distinguishes team-centric teams from leader-centric teams on the basis of information-processing capacity within the TMT with regard to shared decision-making. It is argued that team-centric teams, who take decision collectively have a better influence on organizational ambidexterity than leader-centric teams. To contribute to the ambidexterity literature, a model is proposed in which organizational transactive memory systems (TMS) moderate the relationship between TMTs and organizational ambidexterity. Organizational TMS improve the information sharing and retrieving of employees within an organization. Research is conducted among 256 Dutch and German SMEs and provided support for the first hypothesis, the direct effect of shared decision-making on organizational ambidexterity, and for the moderating hypothesis, the strengthening effect of organizational TMS on the relationship between shared decision-making and organizational ambidexterity.

Keywords: organizational ambidexterity; top management team; shared decision-making;

(4)

Introduction

Every organization that strives to be successful should be able to see and react to interesting opportunities in the future, while maintaining their current business. A strategy to remain successful over long periods of time is that an organization must be ambidextrous (Tushman and O’Reilly, 1996; Burgelman, 2002). Originally, ambidexterity is defined in the dictionary as the ability to use the right and left hand equally well and at the same time. In the business environment, the concept of ambidexterity is applied to organizations and is called organizational ambidexterity, in which organizations simultaneously explore and exploit (e.g. Duncan, 1976; Tushman and O’Reilly, 1996; Gibson and Birkinshaw, 2004; Lubatkin et al., 2006; Raisch and Birkinshaw, 2008; Heavey and Simsek, 2014). That organizations need to do both, exploit and explore, for long term results is generally agreed upon (March, 1991), but it is not easy to pursue both strategies simultaneously. Two contradicting strategies executing at the same time leads to a paradoxical situation, because contradicting strategies aim at different goals and need different knowledge and resources to reach their goals. Continuously differentiating and integrating the exploitative and exploratory resources and activities is required to achieve organizational ambidexterity (Lewis, 2000; O’Reilly and Tushman, 2008; Smith and Lewis, 2011).

The top management team (TMT) is regarded as a vital point of integration between on the one hand exploitation and on the other hand exploration (Tushman and O’Reilly, 1996; Smith and Tushman, 2005). The TMT is very important in this process because it gives direction to the strategy with their decision making to become an ambidextrous organization and therefore achieve positive organizational results. The quality of the strategic decisions of the top management team is important, since it influences the organizational performance. Previous literature about the quality of decisions identified two antecedents of decision quality: first, the cognitive capabilities of a TMT and second, the interaction process through which the decision is made within the TMT (Amason, 1996). But, in the current literature it remains unclear if organizations perform better when the decisions are fully made by the leader, or when the decision-making is shared throughout the TMT (Smith and Tushman, 2005; Lubatkin et al., 2006; Cao et al., 2010; Smith, Binns and Tushman, 2010; O’Reilly and Tushman, 2013; Heavey and Simsek, 2014). Central in this debate is the influence these distinct decision-making approaches have on information processing. Top managers play a

(6)

central role in assessing and integrating knowledge requirements for exploitation and exploration (Heavey et al., 2015), this is achieved with the dispersion of knowledge and expertise of the employees throughout the organization. The TMT is ‘pulling’ the knowledge by retrieving necessary knowledge from experts within particular business subjects throughout the organization, and ‘pushing’ the knowledge by triggering the experts to contribute to issues of their expertise (Van Neerijnen, 2016a).

Transactive memory systems are most of the time concerned with information processing in a team, but it can be applied on organization as a whole (Argote and Ren, 2012, Peltokorpi, 2012, Heavey and Simsek, 2014). A transactive memory system (TMS) provides a shared understanding of who knows what (Lewis and Herndon, 2011). More simply, members within an organization know which colleague they need to get particular information. Little research has been done until now into organizational TMS (Ren and Argote, 2011; Peltokorpi, 2012), and because of the fact that the TMT is an important institution within the organization (it is the overarching team of the organization and represents all the functional departments), this study is focused on organizational TMS. Since an organizational TMS provides information about which employee is an expert in which domain, it facilitates a collective filtering on information about new innovative opportunities and provide a flow in which the relevant employees are connected to suitable knowledge issues (Teece, 2007). Organizational TMS is concerned with the available information within the organization as a whole and is a possible solution for information processing between exploration and exploitation (Heavey and Simsek, 2014), and thus might strengthen organizational ambidexterity.

In response to these unresolved issues in the literature, this thesis captures the top management team level as well as the organizational level, and is focused on information processing within the TMT members (internally) on the basis of shared decision-making and information processing within the organization (externally) with regard to organizational TMS. Since organizational ambidexterity and organizational TMS are both at the organizational level, it is relevant to combine these two with shared decision-making at the TMT level, because the TMT determines the strategy of the organization and sets the internal context for the employees. The behavior of the TMT is very decisive for the organization. This study measures the direct relationship between shared decision-making and organizational ambidexterity first. This relationship is concerned with information processing

(7)

effects of organizational TMS (Argote and Ren, 2012; Peltokorpi, 2012), the moderating effect of this organizational information processing system is tested on the direct relationship between shared decision-making and organizational ambidexterity. Organizational TMS is a mechanism that enlarges the information processing within the organization, and the more information available the more deliberately decisions will be made. Therefore it is expected that organizational TMS will positively amplify the relationship between shared decision-making and organizational ambidexterity. The conceptual model of this research is presented in figure 1.

Figure 1: Conceptual Model

Currently the literature on organizational ambidexterity holds some ambiguity about how the management team should be organized. This study makes several contributions to different scientific business areas related to this subject. First of all, it contributes by building on the organizational ambidexterity literature and extending the current thinking on the antecedents of organizational ambidexterity. The first part of this study is focused on how decisions are made in the TMT (Smith and Tushman, 2005) and strives to empirically verify the relationship between shared decision-making and organizational ambidexterity (Cao et al., 2010). Shared decision-making is a feature of team-centric teams (Smith and Tushman, 2005; Lubatkin et al., 2006; Cao et al., 2010) and therefore, it is expected that team-centric teams will be more positively related to organizational ambidexterity than leader-centric teams where the decision-making is a full responsibility of the leader.

The second contribution of this study is the fact that it is dealing with organizational TMS, while most of the current scientific literature is focused on team TMS (Ren and Argote, 2011; Peltokorpi, 2012). Peltokorpi (2012) defined organizational TMS as: “… networks of interdependent work groups that use each other as external cognitive aids to accomplish shared tasks.” The presence of organizational TMS leads to more information processing within the organization, since employees are more aware of the available

(8)

information of their colleagues, it is easier to integrate and differentiate distinct subunits, which results in increasing organizational ambidexterity and therefore is a relevant antecedent in the achievement of ambidexterity.

The last contribution is the fact that this study is concerned with information processing within a TMT as well as within an organization. On the one hand this study is looking at the TMT level, by examining shared decision-making on organizational ambidexterity (Lubatkin et al., 2006; Cao et al., 2010), on the other hand this study is concerned with organizational TMS perceived by the TMT (Peltokorpi, 2012; Heavey and Simsek, 2014). The unique combination of both levels of information processing is contributing to the current literature about the effects of information processing on organizational ambidexterity. Information sharing is one of the pivotal characteristics that distinguish TMT structures from each other (Smith and Tushman, 2005).

This study is addressing the effects of information processing on organizational ambidexterity, taking multiple levels into account. In the next section the study continues with a literature review and the hypotheses with regard to our conceptual model (figure 1). Subsequently, the methodology is described and the results and findings of this study. The last section contains the discussion with regard to this research and a final conclusion.

Literature Review and Hypotheses

Organizational Ambidexterity

As mentioned before, an organization must be ambidextrous for long-term survival (March, 1991; Tushman & O’Reilly, 1996; Burgelman, 2002). O’Reilly and Tushman (2013) reviewed for their research many ambidexterity-related articles and concluded that ambidexterity has been shown to be positively associated with sales growth, subjective ratings of performance, innovation, market valuation and firm survival. Therefore, ambidexterity is very important and interesting for organizations, but there are still some uncertainties with regard to the antecedents of the achievement of organizational ambidexterity as mentioned in the introduction.

Organizational ambidexterity is the ability to simultaneously pursue both incremental (exploitation) and discontinuous (exploration) innovation from hosting multiple contradictory structures, processes and cultures within the same firm (Duncan, 1976; O’Reilly and

(9)

the capability of an organization to adapt to environmental changes while at the same time managing the daily business. This entails that an organization must be able to implement both incremental and revolutionary change (Tushman and O’Reilly, 1996). Incremental change can be realized with exploitation, which is the refinement and extension of existing competences, technologies and paradigms, while revolutionary change is realized with exploration, which consists of the experimentation of new alternatives (March, 1991; Jansen et al., 2009).

Organizations that want to achieve ambidextrous results have to adapt an ambidextrous design to their organization. Adler et al. (1999) describe ambidextrous organizations as organizational firms that build internally inconsistent architectures and cultures into business units so that the firms can both explore and exploit. It is important for a firm to exploit and explore simultaneously, but pursuing contradicting strategies at the same time can result in paradoxical challenges for organizations. Lewis (2000) describes a paradox as: “contradictory yet interrelated elements – elements that seem logical in isolation but absurd and irrational when appearing simultaneously”. In the case of organizational ambidexterity, the interrelated elements are the resources and capabilities that are needed for exploitation as well as exploration, and the contradictions are the total different focus points of these strategies. Examples of a paradoxical situation in organizations are the balance between quality and cost, between stability and change, between local adaptation and global integration, between exploitation and exploration or between cohesion and division (Lewis, 2000; Smith, 2014). As mentioned before, in the case of organizational ambidexterity, it is balancing between on the one hand exploitation and on the other hand exploration. A way to deal with this paradoxical situation is to differentiate and integrate the strategic agendas for exploitation and exploration simultaneously (Heavey and Simsek, 2014).

Paradoxical Situation in Achieving Organizational Ambidexterity

There is a paradox concerning the achievement of organizational ambidexterity because of the different focus points of the contradicting strategies and the need for specific knowledge and available resources. Smith and Tushman (2005) described the paradoxical situation as: ‘where exploration is rooted in variance-increasing activities, learning by doing and trial and error, exploitation is rooted in variance-decreasing activities and disciplined problem solving.’ Besides the different activities within the paradoxical situation, the main goal of these activities is quite different as well. The goal of explorative activities is to meet the needs of new/emerging markets and customers (Benner and Tushman, 2003; He and Wong,

(10)

2004). These activities contain experimentation of new alternatives for which speed and flexibility are very important (March, 1991; Tushman and O’Reilly, 1996; Jansen et al., 2009). But the main goal of exploitative activities is to meet the needs of existing customers (Benner and Tushman, 2003; He and Wong, 2004). In contrast to explorative activities, these activities are concerned with the refinement and extension of existing competences for which cost efficiency and incremental innovation are very important (March, 1991; Tushman and O’Reilly, 1996; Jansen et al., 2009). Exploitation and exploration also differ when looking at their associated returns; exploitative returns are easier to estimate, more certain and closer in time, while explorative returns are more uncertain, variable and distant in time (He and Wong, 2004). Although ambidextrous organizations face paradoxical challenges in which exploitation and exploration are each other’s extremes, it is important that the employees of an ambidextrous organization are able to process the contradictions within their knowledge and capabilities using the existing flow of information. It requires behavioral complexity in which the employees have an expansive collection of suitable actions on which they can rely when demands are rapidly changing or in conflict with each other (Quinn, 1988). To build up a base of potential actions, information processing within the organization is necessary, because this leads to broad knowledge about several domains to respond rapidly to future issues.

These contradicting strategies have to be managed by the top management team, since the TMT is responsible for managing the agenda, setting priorities and allocating the resources (Heavey and Simsek, 2014). Although managing these contradicting strategies sounds appropriate, it appears to be more difficult in practice. A lot of organizations tend to focus more on only one of the two either, so more concerned with exploitation or more concerned with exploration. Engaging only in exploitation will lead to organizations that are trapped in suboptimal stable equilibria and caught in their own success, with hardly any exciting challenges. Only adapting to existing environmental demands might result in structural inertia, which arises when the speed adaptation in an organization is much lower than the rate at which environmental conditions change. Structural inertia obstructs an organization’s ability to respond to environmental changes and relevant opportunities (Hannan and Freeman, 1984). On the other hand, March (1991) and Chen and Katila (2008) argued that engaging only in exploration will lead to high costs of experimentation without gaining many of its benefits; organizations exhibit too many undeveloped new ideas and too little distinctive competence, this will lead to failure traps. As Levinthal and March (1993)

(11)

perfectly describe “… failure leads to search and change which lead to failure which leads to more search and so on. New ideas and technologies fail and are replaced by other new ideas and technologies, which fail in turn.” Studies have shown that organizations most often tend to lean toward too much exploitation and more infrequently toward too much exploration (Chen and Katila, 2008).

Therefore it is important for an organization to engage in exploitation and exploration simultaneously, however there are two overarching challenges for top managers when managing complex business models like the one necessary for organizational ambidexterity. First, it is important to focus on differentiation in the organization’s strategy, agendas and knowledge, and second, the focus on the integration of the differentiated agendas to achieve exploitation as well as exploration (Smith and Tushman, 2005; Smith, Binns and Tushman, 2010; Heavey and Simsek, 2014). In the article of Benner and Tushman (2003) the importance of the simultaneous use of integration and differentiation is stressed as well, and they mention that this occurs at the senior team level. Ambidextrous organizations are designed by the TMT as organizational forms that are not loosely connected, nor do they switch between the contrasting structures, but they are composed of multiple tightly connected subunits that are themselves loosely coupled with each other (Benner and Tushman, 2003). Raisch et al. (2009) conclude that differentiation and integration are complementary mechanisms for achieving organizational ambidexterity and that the tension between these mechanisms requires continuously managerial attention. It is important to determine the right balance between differentiation and integration, which is dependent on the relative importance of exploitative en exploratory activities (Gulati and Puranam, 2009), to deal with the paradoxical challenge of organizational ambidexterity.

Smith and Tushman (2005) define differentiation as: “…clarifying differences in strategy and organizational architectures.” These organizational strategies and architectures are differentiated by their focus on existing products (exploitation) and new innovations (exploration) and foster the generation of more information about these strategies and increase the effectiveness in responding to challenging (paradoxical) situations (Smith and Tushman, 2005). Integration, on the other hand, is necessary to connect the differentiated dimensions with each other. Smith and Tushman (2005) define integration as: “…identifying synergies between strategy and organizational architectures.” Thus, the strategy and organizational architectures for exploitation and exploration are shaped by the differentiation between them, but should at the same time be integrated with each other to create synergies.

(12)

It is very important for the TMT to create these synergies, because it ensures that the resources and capabilities for both architectures are optimally utilized.

Role of TMT in Achieving Organizational Ambidexterity

An ambidextrous organization does not originate from itself, but has to be shaped by the top management team. The TMT consists of senior managers/executives and is led by a chief executive officer (Ancona & Nadler, 1989; Brockmann & Anthony, 2002). They are concerned with balancing the contradicting agendas of, on the one hand; exploitation and on the other; exploration (Hambrick and Finkelstein, 1979; Smith and Tushman, 2005). The TMT creates this balance with their dynamic way of taking strategic decisions (Finkelstein, 1992; Amason, 1996; Smith, Binns and Tushman, 2010), by focusing on differentiation and integration at the same time (Heavey and Simsek, 2014) and the careful allocation of the resources between exploitation and exploration (He and Wong, 2004), which results in appropriate synergies.

Especially SMEs, the type of organizations used in this study, have to rely more on the ability of their TMT to attain ambidexterity (Lubatkin et al., 2006). This is because of the facilitating mechanisms, as slack resources, administrative systems and the possibility for creating separate business units that SMEs often lack, and therefore they need more direction from the TMT, where the members of the TMT are closer connected to the firm’s operating core (Lubatkin et al, 2006). The role of the leader in managing paradoxical strategies with integration and differentiation to achieve organizational ambidexterity is important. As Smith, Binns and Tushman (2010) argue: “…leaders must be capable of communicating an overarching vision, building and maintaining organizational designs that are internally consistent, of managing ongoing conflict and of engaging in long term, integrative thinking.” Top managers need to have an enriched awareness of all the strategic alternatives for exploitation and exploration to build and maintain the most suitable organizational designs (Heavey and Simsek, 2014). To have an enriched awareness, information processing is very crucial within the top management team and within the organization as well. As Smith and Tushman (2005) have argued, processes of differentiation leads to a lot of information generation about all kind of subjects, which fosters the effectiveness of responding to a changing environment. Another important characteristic of top management teams with regard to information management is connectedness, in which the teams aim to build a collective understanding between the team members (Smith and Tushman, 2005).

(13)

Connectedness facilitates knowledge exchange between members. Because it provides a common base of understanding in which the members of a TMT are triggered to have the ability and motivation to integrate and recombine their knowledge with other knowledge sources within the TMT (Jansen et al., 2009). The research of Jansen et al. (2009) showed evidence that connectedness is directly contributing to the achievement of ambidexterity, and therefore it is an important characteristic for TMTs.

In the article of Cao et al. (2010) the blended impact of the CEO and the TMT on ambidexterity is examined. They argue that, besides the extensiveness of the CEOs information network, it is important that the CEO is joining the TMT more as a TMT member than as a hierarchical leader. The reason for this is that, to answer the ambidexterity demands, it is hard for the CEO to fully address all the information and knowledge related to the ambidexterity demands on its own. This blending impact is achieved with three aspects: communication richness, functional complementarity and power decentralization (Cao et al., 2010). Communication richness and power decentralization are both concerned with information processing within the TMT. Communication richness between the CEO and the members of the TMT reduces information asymmetry within the TMT. Communication richness is focused on the quality of information sharing and is build up out of the amount of interaction and the mode of interaction. The more widespread information and knowledge lead to better analyses of the TMT members, higher quality discussions and more strategic understanding within the TMT (Cao et al., 2010). All these advantages of information processing within the TMT foster organizational ambidexterity. Power decentralization, is defined by Cao et al. (2010) as the extent to which responsibilities for strategic decisions are shared within the TMT rather than being dominated by the CEO. Power decentralization will activate other TMT members to participate in assisting the CEO with decision-making, which will lead to more extensive and deliberate decisions, because the TMT members have more information about the current affairs, and thus positively contribute to organizational ambidexterity. Other potential consequences of power decentralization are, first, the positive effect on the sharing of private information (Edmondson et al., 2003), and second, a more open exchange of ideas and more candid discussions (Eisenhardt and Bourgeois, 1988). These consequences are both positively associated with information processing.

The blended impact of the CEO (Cao et al., 2010) can be compared to the behaviorally integrated TMT of Lubatkin et al. (2006). Lubatkin et al. (2006) describe a behaviorally integrated TMT as a team in which every member is able to exchange their

(14)

knowledge and information in a free and open way throughout the team, and where a set of shared perceptions is created that is integrated in the team and where the members act upon to become a more ambidextrous organization.

Characteristics of the TMT

In light of the arguments mentioned in this literature review, it is clear that information processing plays a crucial role in the achievement of ambidexterity, and that the TMT gives direction to this achievement. Previous literature had examined two TMT structures (team-centric vs. leader-(team-centric) or approaches with different characteristics (eg. information processing) in their relation with organizational ambidexterity. Most of these studies describe the TMTs in the most fruitful examples with team-centric characteristics, such as behavioral integration, power decentralization, communication richness and functional complementarity (Lubatkin et al., 2006; Cao et al., 2010). Although, other studies are arguing that leader-centric teams and team-leader-centric teams have equally effective results in achieving organizational ambidexterity (Smith and Tushman, 2005; Smith, Binns and Tushman, 2010).

The study of Smith, Binns and Tushman (2010) is focused on managing strategic paradoxes and they propose two approaches to manage it. First, the leader-centric approach, in which the full responsibility of the management of the contradictions is placed at the leader, and second, the team centric approach, in which the responsibility of managing contradictions is exercised by the top management team collectively. Integration is a central aspect that is distinguishing leader-centric teams from team-centric teams, and information processing within the TMT is contributing to this integration. In leader-centric teams most of the information resides at the leader, while in team-centric teams it is more dispersed across the TMT.

The first, leader-centric teams are teams in which the leader integrates the contradictory agendas by collecting information about each agenda, processes that information and makes the decisions primarily on their own (Smith and Tushman, 2005). The leader is responsible for managing the conflict between the contradicting agendas and the leader is the one possessing the most information about the different aspects within the organization. It is the task of the leader to integrate the divergent activities of on the one hand exploitation and on the other hand exploration, while the other TMT members are solely focused on specific tasks that are assigned to them (Smith and Tushman, 2005). As Smith and Tushman (2005) have argued: “…the behaviors of leader-centric teams reinforce the focus on

(15)

differentiating products.” For example one member is engaged with all the activities with regard to new innovations (exploration), while another member is totally focused on their current portfolio and the development of it (exploitation). The leader of the TMT has interactions with the other members of the TMT, but there is no mutual interaction between the members of the TMT. The quality of the decisions made in leader-centric teams is therefore dependent on the interactions the leader has with the other TMT members (Smith and Tushman, 2005).

The second, team-centric teams are teams who integrate the contradictory agendas themselves. They share paradoxical frames to enable team members to build a collective understanding of the team’s complex goals and a collective acknowledgement of the tensions and conflicts between their contrasting agendas (Smith and Tushman, 2005). The available information within the TMT is common for all members of the team, instead of only for the leader. The team-members collaborate more with other team-members and they have more information about the knowledge and expertise of other members. Because of this, they know better which member knows what about which domain (Smith and Tushman, 2005). The behaviors of the leader in centric teams increase integrative behaviors of the team-members (Smith and Tushman, 2005), which is an important characteristic of managing the paradoxical challenge of organizational ambidexterity (Raisch et al., 2009).

Somech (2006) distinguished the team structures by leadership style and is more focused on decision-making, using different names, but with agreement to the distinction of Smith and Tushman (2005). Both leadership styles are associated with high levels of team outcomes, although they have different characteristics. The first is directive leadership which is defined by Kahai et al. (1997) as: “…leader behaviors that seek followers’ compliance with directions about how to accomplish a problem-solving task.” The directive team leader provides a framework for the team members with regard to decision-making and action in alignment with the team-leader’s vision. The team-leader is the only person who is aware of overlapping aspects between other team members, since the leader possesses the information from the team members. This one is related to the leader-centric team structure and the leader is collecting the information from the individual members he needs to make decisions on his own. The second is the participative leadership and is defined by Kahai et al. (1997) as: “…sharing of problem solving by a leader with followers by consulting them before making a decision.” In this leadership style, the team members join the decision making process with their team leader, or at least have some shared influence on the decision-making. This one is

(16)

related to the team-centric team structure. When team members participate in the decision-making process, they obviously have more information about the current affairs within the organization, which leads to a more purposeful decision-making. In this research the terms leader-centric and team-centric are used.

In the article of Smith and Tushman (2005) the importance of decision-making is stressed and it is argued that balanced strategic decisions are based on two criteria related to the integration and differentiation challenge. The first is that the decisions should have a distributive nature, focused on making balanced trade-offs over time with regard to the division of resources between the existing products and the new innovations (Smith and Tushman, 2005). The second criterion for a balanced strategic decision is the integrative nature, which encompasses the recognition of opportunities, linkages, and synergies between the exploitative and exploratory activities (Smith and Tushman, 2005). The studies of Lubatkin et al. (2006), Jansen et al. (2009), Cao et al. (2010) and Smith, Binns and Tushman (2010) about the role of TMT in achieving organizational ambidexterity are described in the paragraph above. The characteristics they mention will make the differentiation and integration challenge more feasible and are positively associated with organizational ambidexterity. It is remarkable that most of these characteristics have overlap with the characteristics of team-centric teams. For example connectedness (Jansen et al., 2009) is a characteristic of team-centric teams, because it is focused on collective understanding (Smith and Tushman, 2005); and the behavioral integrated team explained by Lubatkin et al. (2006) is focused on collaborative behavior, information exchange and joint decision-making, which is in line with the characteristics of the team-centric teams in the studies of Smith and Tushman (2005) and Somech (2006). Table 1 provides an overview of a couple of articles from the current literature in which the required characteristics for organizational ambidexterity are assigned to the related type of team structure.

Article Leader-centric Team-centric

Amason (1996) Interactive process for

decision-making, consensus among team-members

Kahai et al. (1997) Greater expression of solutions and

(17)

Smith and Tushman (2005) Leader knows exactly who knows what, high-quality interactions with the members, limited coordination requirements

Collective understanding, collaborative decision-making, integrative thinking is reinforced

Lubatkin et al. (2006) Behaviorally integrated team,

connectedness between members, knowledge exchange and joint decision-making

Somech (2006) Framework for decision-making

and alignment with overall vision

Joint decision-making

Jansen et al. (2009) Connectedness, common

understanding

Cao et al. (2010) Power decentralization,

communication richness within TMT, blend CEO with TMT Smith, Binns and Tushman

(2010)

CEO responsible for making decisions while taking multiple businesses into account

Entire team is responsible for managing the tensions, extensive communication with each other and collective decision making Table 1: Leader-centric vs. Team-centric characteristics

It can be concluded that the TMT structures differ on a lot of aspects and that previous research showed different organizational results. In this research, the distinction between team-centric teams and leader-centric teams is made on the basis of decision-making, in which shared decision-making is related to team-centric teams and decision-making only by the leader is related to leader-centric teams. Information processing at the TMT level is an important aspect associated with decision-making, and therefore is considered as internal information processing, while external information processing in this study is focused on the extent of shared information within the whole organization.

Information processing within the TMT

Internal information processing plays an important role within the TMT, because the more information a TMT ‘possesses’, the more deliberately decisions can be made which will have a positive influence on the organizational performance (Smith and Tushman, 2005; Jansen et al., 2009; Cao et al., 2010). Information gathering is increasing the perceptual accuracy of a TMT, which is influencing the performance of an organization that is managed by the TMT (Rau, 2006). In the article of Rau (2006) is mentioned that a couple of studies found support for the relationship between more information gathering and better organizational

(18)

performance, although this was not the case for all the organizations, possibly because of differences in information processing within the TMT.

The overview of the current literature shows that the TMTs with team-centric characteristics outperform in achieving organizational ambidexterity, while Smith and Tushman (2005), Somech (2006) and Smith, Binns and Tushman (2010) argued that both types of TMTs are equally effective in managing contradictions. Shared decision-making is one of the pivotal characteristics that distinguish team-centric teams from leader-centric teams (Edmondson et al., 2003; Lubatkin et al., 2006; Somech, 2006; Cao et al., 2010). Strategic decision-making is very important for organizations, because the quality of these decisions influences organizational performance (Amason, 1996). There are two factors for producing decisions of high quality, the first are the cognitive capabilities of a TMT and the second is the interaction process through which the team produces its decision (Amason, 1996), and that interaction process is a central aspect in team-centric teams. Amason (1996) explains: “…each decision process must build consensus among team members, because without understanding and commitment successful implementation of the decision is unlikely”. Therefore, it is important that all the team members have somewhat information about the trending topics within the organization, which contributes to achieving consensus. Besides required information and consensus it is important that the members of the TMT have affective relationships with each other that makes them willing to participate and allow them to work together effectively (Amason, 1996). Amason (1996) concludes that: “…decision quality is enhanced when multiple perspectives are brought to bear on a particular issue.” Therefore, it should be logical to conclude that shared decision-making, taking multiple perspectives into account, is an antecedent for higher decision quality and thus for better organizational performance.

Cao et al. (2010) found positive support for the fact that power decentralization and communication richness both interacts with the extensiveness of the network of the CEO in affecting organizational ambidexterity. The reason for this is that a CEO in a centralized organization with less communal interaction with the other TMT members, is not able to process and utilize the large quantities of information on its own, and therefore might make less thoughtful decisions, which results in more negative organizational results (Cao et al., 2010). Thus, it can be concluded according to the article of Cao et al. (2010) that it is important that the CEO is ‘blended’ with the rest of the TMT members to

(19)

central position in between all the other TMT members instead of the head of a hierarchical pyramid.

Kahai et al. (1997) distinguish participative leaders from directive leaders, comparable with Somech (2006), and argue that TMT members are likely to have a positive attitude toward a team-centric leader, because they will feel more respected, independent and self-expressed in which their higher-order needs will be fulfilled. This creates a reliable atmosphere in which the members of the TMT are willing to share more information.

In light of these arguments, it is clear that there are differences between team-centric and leader-centric teams with regard to information processing within the TMT and thus in the decision-making process. Shared decision-making is one of the distinguishing factors and plays an important role in the performance of an organization. This study will discover whether shared decision-making in TMTs is positively related with organizational ambidexterity.

Hypothesis 1: Team-centric teams with shared decision-making are positively related to organizational ambidexterity.

Information Processing at the Organizational Level

Information processing within the TMT (internally) is important; because of the positive influence on organizational performance (Rau, 2006; Cao et al., 2010), but outside the TMT (externally) exists much expertise, knowledge and information sources that could become available for the organization as a whole. Rau (2006) explains that organizations will have rich knowledge structures when individual employees possess diverse knowledge structures and know where useful complementary expertise is located within the organization. The use of others’ knowledge is stimulated through transactive memory systems (Hollingshead, 2001; Lewis, 2003; Heavey and Simsek, 2014).

In this thesis, information processing, which is supported through TMS, is regarded as an organizational level construct, because the organizational level of TMS is an underdeveloped research area and therefore interesting to study more in-depth compared to other TMS levels. The frequently used levels for TMS are: organizational level, team/group level and individual level. Most of the studies into TMS are aimed at team/group level TMS and individual level TMS (Wegner, 1987; Liang et al, 1995; Lewis, 2003; Argote and Ren,

(20)

2012; Heavey and Simsek, 2014), while Peltokorpi (2012) aims at the organizational level TMS. As Lewis (2003) argues: “…although TMS is a group-level phenomenon, it exists as a function of the structure, content, and credibility of a members’ individual knowledge.” And the research of Yuan et al. (2007) tested a transactive memory model to discover how individuals allocate and retrieve task-related information in work teams. This means that a TMS will occur at the team level, but should be measured at the individual level, because that is where the knowledge is used and the TMS development starts. This concept can be extended to the organizational level and is defined by Peltokorpi (2012) as: “…networks of interdependent work groups that use each other as external cognitive aids to accomplish shared tasks.”

Hollingshead (2001) defines TMS as: ‘’…the organizing schemes that provide the connection between knowledge held by individuals with knowledge held by others throughout the whole organization”. It can be compared with a knowledge management system, which is a network inside an organization that enables the acquisition, storage, distribution, and retrieval of organizational knowledge and information (Adams and Lamont, 2003). Or in other words: ‘…a shared system that people in relationships develop for encoding, storing, and retrieving information about different substantive domains’ (Ren and Argote, 2011). Additionally, TMS provides that team members are external knowledge sources who can rely on each other because of the responsibility they have for different but complementary areas of learning and expertise (Heavey and Simsek, 2014). Unique features of TMS are contributing to the information processing within the organization. The first unique feature is that it facilitates the exchange of unique knowledge and the development of it, and not target possibly redundant knowledge, and the second unique feature is that TMS provides a team with efficient knowledge exchange and the coordination of this knowledge among the team members (Heavey and Simsek, 2014).

Lewis (2003) described three fundamental elements of TMS that cover the whole TMS process: specialization, credibility and coordination. The transaction of information between group members takes place in the best possible way when combining all the dimensions together. The coordinating processes exist to nurture the processes of credibility of others’ knowledge and localization of others’ specialized knowledge, and increase the capacity to on the one hand ‘pull’ the processed information and on the other hand ‘push’ the processed information through the organization between the exploitative and exploratory

(21)

possess, and ‘pushing’ information by triggering the employees with the most appropriate knowledge or expertise to react to occurring issues. In the context of exploitation the problem is the diffusion of the knowledge, which is limited to individuals, and has to be located, accessed and utilized by the TMT. And in the context of exploration the knowledge problem is the search for new knowledge, which the TMT has to explore divergent from their existing knowledge, perspectives and interpretations (Heavey and Simsek, 2014).

Heavey and Simsek (2014) propose that transactive memory systems (TMS) within the TMT serve as a cognitive mechanism that amplify information processing and foster how TMTs meet the differentiation-integration challenge of ambidexterity. This continuously information processing is important because firms can sustain positive performance by exploiting existing products and processes and exploring new ideas for innovations (He & Wong, 2004) and it makes the transition between the knowledge of exploitative and exploratory activities easier (Kogut & Zander, 1992; Im & Rai, 2008; Raisch et al., 2009).

Organizational TMS provides the possibility for employees to find, use and develop the knowledge and expertise of local and distant others, and therefore the employees make use of the ‘collective mind’ of an organization (Wegner, 1987). The reason for this is that the coordination process results in shorter search chains between the employees’ knowledge and there is more focus on the awareness of ‘who knows what’ and ‘who knows whom’. The total process of organizational TMS contributes to the processing of information within the organization and is in this study meant as external information processing, because it is outside the TMT, but within the whole organization.

Organizational TMS connects employees with each other through strong network ties via which relevant information is transferred, this network becomes more complex and the information is complementing (Peltokorpi, 2012). In the article of Peltokorpi (2012) is mentioned that organizational TMS improves knowledge differentiation, transfer and retention, which supports organizational performance as well as individual and team performance. Organizational TMS develops better when organizations have decentralized decision-making, provide flexible job assignments and where work has to be done within interrelated work groups. Besides this it is important that the leader is motivating the team members to share information and actively utilize other’s information.

(22)

Moderating Effect of Organizational Transactive Memory Systems

According to the above mentioned arguments it is expected that organizational TMS has a positive effect on organizational performance, but this is due to organizational characteristics. In this study organizational TMS is measured as a moderator, because this study will discover whether the effect of shared decision-making on organizational ambidexterity is influenced by organizational TMS, because shared decision-making is an important TMT characteristic.

The strength of the effect of the organizational TMS might be different between the two TMT structures, team-centric teams and leader-centric teams, because, as mentioned before, these TMT structures have different characteristics with regard to information processing. Organizational TMS contributes to organizational performance, because it makes it easier for employees to identify and contact experts within their organization, which leads to more deliberate activities with regard to current market signals, market developments or innovative ideas. The exploitation side of the processed information of the TMS is concerned with the use and refinement of existing knowledge and the exploration side is concerned with the pursuit of new knowledge and opportunities (March, 1991). Therefore it is interesting to discover to what extent this relevant information processing within the organization differs between team-centric teams and leader-centric teams and the effect on organizational ambidexterity.

Heavey and Simsek (2014) argue that when the members of the TMT are intrapersonal functionally diverse; TMS can foster the contradicting agendas of the differentiation and integration issue. In team-centric teams, the team members integrate the contradictory agenda by themselves and have more interactions with each other, hereby the members contribute to the team and learn from the contributions of others (Smith and Tushman, 2005). Smith and Tushman (2005) also argue that: “… a real team allows team members to allocate clear tasks to one who else has information, and work with the other members to create shared mental models and shared processes.” The members in team-centric teams have more information about multiple domains that is why organizational TMS might be more extensive and broader in organizations with team-centric leaders. In leader-centric teams, the leader is the person within the TMT who knows the best which employee possesses which knowledge (Smith and Tushman, 2005). Because of this the leader will have more in-depth information about every business unit, while in team-centric teams this information might be broader but more superficially. Therefore the organizational TMS in organizations with a leader-centric leader might be more specific and profound, although

(23)

there are less ‘active participants’ in this organizational TMS, because the direction of the information processing is more one-sided (merely to the leader) than two-sided (interactive between employees).

Another aspect resulting from organizational TMS and contributing to organizational performance is the creation of synergies, because coordination across functions and activities between a team ensure that the performance of the whole is greater than the sum of its parts (Ancona & Nadler, 1989). Leader-centric teams focus much less on aspects like this, while team-centric teams are more concerned with collaboration between team members. The organizational TMS with its interrelated processes with regard to sharing knowledge and expertise is contributing to these synergetic results.

Organizations are much larger than just a team; therefore the use of organizational TMS is different than team TMS. Moreland (1999) explained two broad categories in which the way of executing organizational TMS can be divided. The first is interpersonal in which the members within an organization know who the specific person is from whom they want to obtain particular knowledge, and the second is technological, in which knowledge is shared and retrieved via supportive tools such as computers (Moreland, 1999). Especially the interpersonal category is a category that might be better developed for organizations in which the leader is more integrated with the rest of the management team, and therefore with the organization as a whole.

In the article of Heavey and Simsek (2014) is stated that when knowledge, skills, and connections available to the team are extended, diversity in organizational experience and functional background is likely to positively influence the association between TMS at the TMT level and ambidexterity. This study expects that organizational TMS has a positive moderating effect on the relationship between shared decision-making in top management teams and organizational ambidexterity.

Hypothesis 2: Organizational transactive memory systems positively moderate the relation between team-centric teams and organizational ambidexterity.

(24)

Methodology

Data and Sample

To answer the research question and underlying hypotheses of this thesis, a database research is conducted. The project around the database is called the Innovation Benchmark, it is used to collect the relevant data and is set up by four researchers of the Rotterdam School of Management (Tempelaar, M., Van de Vrande, V., van Neerijnen, P. and Jansen, J). It is a questionnaire for companies to gain insight into their own performance as well as compared to other (anonymous) firms in the same sector, thus a benchmark. Since the data has been used for previous research, the data had to be refined to be suitable for this study. To start with the refinement of the data, a selection is made for small-to-medium sized enterprises with a headcount of 25-249 employees; this amount is set in alignment with the European Commision (2003) who set 250 employees as the maximum amount for the SME range. In most nations’ economies SMEs represent a vital component, because a large share of all the organizations consists of SMEs (Lubatkin et al., 2006). And as mentioned before, since SMEs have fewer hierarchical levels, the TMT is more related to strategic as well as operational roles, and therefore more concerned with internal and external pressures (Lubatkin et al., 2006). Thus, SMEs are very appropriate for this study.

The Innovation Benchmark Database consists of the data of about 700 Dutch and German SMEs in 2012/2013 and is submitted by members of the top management team. After refining the dataset, and excluding the companies that fall outside the SME range, as well as removing the items with missing values linked to the key concepts (shared decision-making, transactive memory systems and organizational ambidexterity), this study retained sample data from 256 organizations. The average number of fulltime employees of the suitable SMEs was 89.09 (SD=54.51). These firms operate in a wide variety of sectors: metal and wood manufacturing; retail and wholesale trade; administrative support, management of companies, professional services, scientific and technical services; utilities and construction; arts, entertainment and food industries; health care, social assistance and educational services; and other services. The large amount of firms and the wide variety of the different sectors contribute to the generalizability of the results of this study.

To demonstrate whether there is any non-response bias, several t-tests are done and showed no significant differences (p < 0.05), which means that nonresponse bias is not an issue in this study.

(25)

Measurements

In this research three constructs are used. The first; organizational ambidexterity, which contains the questionnaire items about exploitation and exploration (March, 1991). The second; shared decision-making in top management teams, in which a distinction is made between team-centric teams and leader-centric teams (Smith and Tushman, 2005). And the last one; organizational TMS, consisting of the questionnaire items about knowledge management with regard to coordination, credibility and localization (Van Neerijnen, 2016b). Most of the items of each measurement are based on a 7-point Likert scale (strongly disagree – strongly agree). If another scale is used it is mentioned for that particular measurement. The number of items varies per measurement and an overview of the items of the questionnaire is presented in the appendix (Survey Items).

Organizational Ambidexterity. Organizations having an ambidextrous design are organizations that can exploit and explore simultaneously (March, 1991; Gibson and Birkinshaw, 2004; He and Wong, 2004; Smith and Tushman, 2005; Lubatkin et al., 2006), which leads to long-term performance and firm survival. The measurements of exploitation and exploration from the article of Jansen et al. (2006) are adopted in this study. The exploitation scale measures to what extent organizations build on existing knowledge and are able to meet the needs of existing customers and improve the fulfillment of their future needs. While the exploration scale measures the organization’s use and expansion of existing knowledge and development of (radical) innovations for emerging customers and markets (Jansen et al., 2006). Both components of organizational ambidexterity consist of seven items. Exploratory factor analysis showed in this study that for the measurement of exploitation the second item had to be removed (α=.79), resulting in a scale of six items, and for the measurement of exploration none of the items had to be removed, thus this scale remains with seven items (α=.87). To show whether each item is loading clearly on the desired factor, a subsequent joint exploratory factor analysis shows the structure of the two components, which showed that each item loaded clearly on the intended factor. All factor loadings were above .59 with cross loadings under .31. Both factors have eigenvalues greater than one.

There are multiple ways of measuring ambidexterity, since it consists of two constructs. First, the additive way in which the measurement of exploitation and the measurement of exploration are added together (Lubatkin et al., 2006; Jansen et al., 2009). Second, the multiplicative way, in this way the measurements of exploitation en exploration

(26)

are combined with each other through multiplying (Gibson & Birkinshaw, 2004; He & Wong, 2004; Cao et al., 2010). And the last way is the balancing or subtractive way, in which the exploitation measurement and exploration measurement are subtracted from each other (He & Wong, 2004; Cao et al., 2009). Ambidexterity is measured in this study with the sum of exploitation and exploration, as it is suggested by Lubatkin et al. (2006) and Cao et al. (2010), who followed the Edwards test (1994) in composing a construct of two components. The exploitation and exploration items used in this study, as mentioned before, are adopted from Jansen et al. (2006), who used the additive method in their study as well. To check the convergent validity of the operationalization of ambidexterity in this study, confirmatory factor analysis compared three different additive models on the basis of their fit with this study (AMOS 20). The first model is a second-order factor structure (χ²=282.4, DF=64 , CFI=0.86, RMSEA=0.092), in the second model all items were loading on the same underlying factor (χ²=607.3, DF=65, CFI=0.653, RMSEA=0.143), and the third model treats exploitation and exploration as independent variables (χ²=1654.2, DF=91, CFI=0.000, RMSEA=0.206). The first model clearly showed the best fit to the data, therefore, in this study ambidexterity is operationalized as a second-order construct.

Shared Decision-Making in TMT. Smith and Tushman (2005) emphasize the importance of the top management team: “…the top management team serves as the point of integration between these contradicting agendas,” and found support that both team structures had equally effective results with regard to organizational ambidexterity. Their article provided a base for the scale of shared decision-making in the Innovation Benchmark questionnaire. This variable is measured by the difference between leader-centric versus team-centric decision-making and consists of five items. Since shared decision-making is related to team-centric teams, the higher the score on shared decision-making, the more ‘team-centric’ the team is. The shared decision-making scale has a high reliability (α=.91), principal component analysis showed that all the items loaded well enough on the scale and because of the high factor loadings, none of the items have to be removed. This scale measures whether the decisions about the organizational strategy within the top management team are solely at the level of the CEO (leader-centric teams) or whether they are at the team wide level (team-centric team).

Organizational Transactive Memory Systems. Organizational TMS is build up out of multiple components, with regard to the structures defined in the articles of Liang et al. (1995), Moreland et al. (1998), Lewis (2003) and Ren and Argote (2011). The TMS is

(27)

measuring the active use of transactive memory by two or more people to cooperatively store, retrieve, and exchange information (Lewis, 2003; Argote and Ren, 2012) and creates an atmosphere in which colleagues are aware of who knows what and make use of this knowledge (Wegner, 1987; Lewis, 2003; Ren and Argote, 2011). The measurement in this research is concerned with organizational transactive memory systems in the way it is perceived by the TMT, and therefore is concerned with the whole organization. The measurement for organizational TMS is based on three components: localization, coordination and credibility (Van Neerijnen, 2016b). Faraj and Sproull (2000) defined localization as, the situation in which employees know how the knowledge is distributed among the organization. Thus, they know where to find the knowledge about specific subjects, which might be complex in large organization. Credibility is the belief individuals have about the reliability of the knowledge of other team members, and the trust they have in the other members’ expertise (Lewis, 2003). The last one is coordination, which is the process of structuring the knowledge across the team members.

The localization dimension consists of five item, of which the third had to be removed resulting in a reliability score of α=.84. From the credibility dimension, consisting of four items, the second item had to be removed resulting in a reliability score of α=.86, and the last dimension, coordination, consists of five items and none of them had to be removed (α=.90). In the appendix (Survey Items) is showed which items were deleted. To show whether each item is loading clearly on the desired factor, a subsequent joint exploratory factor analysis shows the structure of the three components, which showed that each item loaded clearly on the intended factor. All factor loadings were above .67 with cross loadings under .45. The overall reliability of the organizational TMS scale resulted in α=.91.

Control Variables. It might be that there are possible alternative explanations having an effect on organizational ambidexterity. To provide a more critical view on our hypotheses, six relevant control variables at the team, firm, and environmental level of analysis are included in the analysis. The first control variable, at the organizational level, is organizational size; organizational size is measured with regard to the number of fulltime employees in an organization (Cao et al., 2009; Jansen et al., 2009; Heavey and Simsek, 2014). Ahuja and Lampert (2001) argue that large organizations lack the motivation and ability to create breakthrough inventions and therefore it is harder for them to become an ambidextrous organization. The second control variable, at the team level, is TMT size, and entails the number of members in the top management team (Cao et al., 2009; Jansen et al.,

(28)

2009; Heavey and Simsek, 2014). This control variable is used, because Amason and Sapienza (1997) discovered that the larger the TMT, the more cognitive and affective conflict the team experiences, which could implicate that it is harder to achieve organizational ambidexterity. Attentional engagement (α=.88) is the third control variable and it comprises the extent to which TMT members discuss strategic issues with the lower levels in the organization. This organizational level variable is expected to have a positive influence on organizational ambidexterity, because when the TMT is attentional engaged, it is aware of the issues at all levels of the organization. The fourth variable is reflexivity (α=.94) within the top management team. Team reflexivity is defined as the extent to which teams collectively reflect upon and adapt their working methods and functioning (West, 2000). Non-reflexive teams show little awareness of the team objectives, strategies and environment in which they operate (Widmer et al., 2009), and thus can be concluded that these teams are less able to achieve organizational ambidexterity. State uncertainty (α=.88) is the fifth control variable and occurs at the environmental level, because Ashill and Jobber (2010) argue that in uncertain environments it is more challenging to sustain a careful balance between exploitation and exploration. The last control variable, at the organizational level, is organizational slack (α = .74, after removing one item). Organizational slack has different and sometimes conflicting influences on product exploitation and exploration (Voss et al., 2008). This means that if an organization has slack resources it is very hard to be exploitative and exploratory at the same time.

Aggregation and Measurement Analysis

For some companies, more than one TMT member submitted the questionnaire. Since these answers concern the same company, it is verified whether these ‘doubles’ are allowed to be aggregated. Calculating the intraclass correlations and interrater agreement (rwg) estimates will show whether aggregation is applicable for doubles in this research (LeBreton and Senter, 2008). The inter-rater agreement scores are .72 for exploitation, .70 for exploration, .49 for TMT decision-making and .83 for organizational TMS. Since these values are almost all above 0.5, it shows a reasonable fit (James et al., 1993). Besides the interrater agreement scores, the intraclass correlation showed .77 for exploitation, .86 for exploration, .90 for TMT decision-making and 0.84 for organizational TMS. According to the article of LeBreton and Senter (2008) this means that those are the percentages of individuals that are unique for each variable. The results of the intraclass correlation and interrater agreement showed that it is allowed to aggregate in this database. All the measurements in this study were individually

(29)

tested for construct validity with reliability testing according to Cronbach’s alpha and principal component analysis and showed reliable results.

Common Method Variance and Analysis

Since this database is already used for previous research, Van Neerijnen (2016b) took several steps to control the potential effects of common method variance within the database of the Innovation Benchmark. The rules of Podsakoff et al. (2003) have been followed, by taking precautions to limit the risk of common method variance. Pretesting the survey items, ensuring anonymity, and an introduction added to the survey, to explain the purpose of the research, were done to ensure this.

This study tests the moderating effect of organizational TMS on the relationship between TMT decision-making and organizational ambidexterity using hierarchical multiple regression. Besides this, according to Hayes (2013), the significance of the moderation is demonstrated with the SPSS add-in PROCESS and the Johnson-Neyman technique is used to show the interplay between the constructs of this thesis.

Results

Table 2: Descriptive statistics, correlations and reliabilities.

The descriptive statistics, correlations between the variables and reliabilities in this study are presented in table 2. The variance inflation factor scores are measured to examine potential multicollinearity. In this study the maximum VIF score is 2.43, which is below the cut-off

Mean SD 1 2 3 4 5 6 7 8 9 1 Ambidexterity 9.48 1.72 (0.868) 2 TMT Decision Making 5.18 1.16 0.427** (0.908) 3 Organizational TMS 4.93 0.87 0.485** 0.535** (0.911) 4 Organizational Size 87.71 54.52 -0.078 0.009 -0.66 (-) 5 TMT Size 8.32 16.48 0.03 0.055 0.049 0.111 (-) 6 Attentional Engagement 4.56 1.22 0.413** 0.518** 0.497** 0.004 0.027 (0.879) 7 Reflexivity 5.09 1.15 0.411** 0.627** 0.656** 0.063 0.107 0.492** (0.939) 8 State Uncertainty 4.46 1.08 0.266** 0.438** 0.457** 0.129* 0.095 0.372** 0.498** (0.875) 9 Organizational Slack 4.2 1.02 0.317** 0.369** 0.486** -0.008 -0.007 0.250** 0.382** 0.289** (0.743)

**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is sifnificant at the 0.05 level (2-tailed). Reliabilities on the diagonal.

The strengthening effect of organizational transactive memory systems on the relation between shared decision-making and organizational ambidexterity