Using transformational
leadership behavior to
enhance contextual
ambidexterity
Executive Program in Management Studies, Strategy track
Date: 26-01-2018
Ronald H. Nijenbanning
11144904
E-mail: ronaldnijenbanning@outlook.com
Statement of Originality
This document is written by Ronald Nijenbanning, 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
ABSTRACT
Today’s turbulent environment makes it a challenge for companies to survive since they need to be cost-effective and efficient in their current activities, but they also need the ability to adapt and innovate. This study tries to find a solution to this challenge by investigating the achievement of long-term business-unit performance through the creation of an intra-organizational context where the individual is encouraged to manage the tension between being efficient in the current activities (i.e. exploitation) and being innovative (i.e. exploration), with influence of
transformational leadership behavior. Data from a quantitative study, by means of an online survey taken from people working either on middle-management- or operational level in business-units (N = 114), resulted in the finding that transformational leadership behavior can enhance the positive significant effect of intra-organizational context on contextual
ambidexterity. This study also provides some interesting suggestions for further research.
FOREWORD
My current employment is at the Quality Assurance department of two divisions of CECO Environmental (NASDAQ: CECE), which operate in the gas industry making parts for gas-energy plants. In order to achieve a certain level of quality within the organization, our department analyzes the processes and activities of the organization, standardizes them, and measures effectiveness and efficiency after a certain period of time, resulting in opportunity for improvements. A discussion we currently have within the organization is that standardization can lead to inflexibility, and can suppress creativity and innovation. Since the environment becomes more dynamic and competitive, our organization needs to be able to adapt to changing
circumstances (i.e. be more flexible and organic). Because of this challenge, my interests started to develop in how to balance efficiency with flexibility and innovation within an organization.
I would like to take this opportunity to thank my supervisor prof dr. Ed Peelen for his guidance during my master thesis and for the interesting discussions we had. Besides, I would like to thank all participants of this study for their input.
Finally I would like to dedicate this master thesis to both my grandmothers who passed away during the two-and-a-half years of my master study. They were proud when I started my master study and supported me greatly, and still do till this day.
TABLE OF CONTENTS
1 Introduction ... 6
1.1 Thesis structure ... 8
2 Literature review ... 10
2.1 What is meant with organizational ambidexterity and how is it created? ... 10
2.1.1 In a ‘sequential’ manner ... 12
2.1.2 In a ‘structural’ manner ... 12
2.1.3 In a ‘contextual’ manner ... 13
2.1.4 Sequential and structural ambidexterity versus contextual ambidexterity ... 13
2.2 What is meant with the context of an organization and how is it related to organizational ambidexterity? ... 15
2.2.1 Context as antecedent of organizational ambidexterity as antecedent of performance . 16 2.3 What is meant with leadership and what is meant with leadership behavior? ... 17
2.3.1 Leadership and organizational ambidexterity ... 18
2.3.2 Leadership behaviors ... 19
2.3.3 Transformational leadership and organizational ambidexterity... 20
3. Theoretical framework ... 21
4. Research design ... 24
4.1 Data collection ... 24
4.3 Measures ... 26
4.3.1 Dependent variable – performance ... 27
4.3.2 Independent variable – intra-organizational context ... 27
4.3.3 Moderator variable – transformational leadership behaviors ... 28
4.3.4 Mediator variable – contextual ambidexterity ... 30
4.3.5 Control variables ... 30
5. Results ... 32
5.1 Data preparation ... 32
5.2 Hypotheses testing ... 35
6. Discussion ... 46
6.1 Findings & theory ... 46
6.2 Limitations (general) and further research ... 49
6.3 Implications ... 51
7. Conclusion ... 55
1 INTRODUCTION
Because of disruptive technologies and political turmoil, today’s environment is quite dynamic and turbulent. It makes it a challenge for companies to survive, since they need the ability to adapt and to be innovative and flexible. However, companies need not only being able to adapt. Companies need also to be able to be successful in their current activities; being efficient and competitive on low cost. To conclude, there is a tension which needs to be addressed by companies in order to be successful on the long-term; innovation, flexibility and adaptability, versus efficiency, stability, standardization.
The concept of organizational ambidexterity; ‘an organization’s ability to be aligned and efficient in its management of today’s business demands (also called ‘exploitation’) while simultaneously being adaptive to changes in the environment’ (also called ‘exploration’) (March, 1991), is key to address the tension as mentioned earlier. Dividing resources between exploitation and exploration can be quite a challenge, but managing the tension between them is necessary to achieve long-term performance (March, 1991; He & Wong, 2004; Raisch & Birkinshaw, 2008). A lot of research is done regarding organizational ambidexterity, including antecedents, moderators, outcomes and possible ways to achieve organizational ambidexterity. One of those possible ways to achieve organizational ambidexterity is by influencing contextual factors, also called
‘contextual ambidexterity’. Birkinshaw & Gibson (2004) describe contextual ambidexterity as a way to achieve ambidexterity by creating an intra-organizational context where the individual manages the tension between alignment (i.e. exploitation) and adaptability (i.e. exploration). The collective behavior of individuals would result in contextual ambidexterity at organizational level, which in turn leads to long-term performance. However, they point out that the relation between ‘context’ and ‘contextual ambidexterity’ still needs research. They suggest to further
delineate the fundamental aspects of intra-organizational context and the relationship with contextual ambidexterity. Raisch & Birkinshaw (2008) conduct research on three antecedents (among others) of organizational ambidexterity, with contextual factors as one of them. They also suggest to further investigate the relationship between contextual factors (and the other
antecedents) and organizational ambidexterity, by focusing on possible relationships between the antecedents.
Another topic which needs more research regards the role of leadership in achieving
organizational ambidexterity, such as the role of senior team and leadership behaviors and styles, on organizational ambidexterity (Jansen et al., 2009; Nemanich & Vera, 2009; Rosing et al., 2011; O’Reilly III & Tushman, 2013). There is research done regarding the relationship of leadership on organizational ambidexterity. In their research, Jansen et al. (2008) found that transformational leadership behaviors (hereinafter referred to as ‘TLBs’) have a positive effect as a moderator on the relationship of senior team attributes on organizational ambidexterity. TLBs can be described as a creative exchange between leaders and subordinates, in order to bring about vision driven change in people and context (Bass, 1985; Burke et al., 2006). Jansen et al. (2008) suggest though, to further investigate the effect of leadership behaviors on lower levels within organizations to take the notion of multiple levels within an organization into account. They found a positive effect of leadership on senior team dynamics, and are therefore curious how this would work on lower levels. They also refer to Gibson & Birkinshaw (2004) and their statement that leaders may serve as examples for lower levels regarding management of the tension
between alignment and adaptability. As mentioned earlier, Raisch & Birkinshaw (2008) describe contextual factors as antecedent of organizational ambidexterity, but structural factors and
leadership-based factors as well. Since the possible relationships between the antecedents are not investigated, they make it one of the suggestions for further research.
Summarized, there is still a lot of research to be done within the field of organizational
ambidexterity regarding contextual factors as antecedent, and the role of leadership (behaviors) in achieving it. Combined, research needs to be conducted on how (top) management can enhance the effect of an intra-organizational context where the tension between exploitation and
exploration is managed by individuals throughout an organization. Individuals on top
management level play an important role by setting out the frame-work to manage the tension between adaptability and alignment, but individuals at middle management and operational levels play an important role too. As Bordia et al. (2004) describe, participation leads to awareness and understanding of change events, and that the involvement of lower levels is necessary since they often have more (current) knowledge of operations and related developments.
Therefore, I would like to contribute to the field of organizational ambidexterity by investigating the moderating effect of TLBs on the relationship between intra-organizational context and performance, through contextual ambidexterity. This brings me to the following research question:
What transformational leadership behaviors enhance the effect of intra-organizational context on performance, through contextual ambidexterity?
1.1 Thesis structure
Chapter two provides an overview of the main theoretical concepts organizational ambidexterity, intra-organizational context and transformational leadership behaviors. This overview will result in a better understanding of the research topic and provides insight in what is already researched.
Chapter three provides an overview of all hypotheses that will be tested in order to answer the research question. Argumentation is provided as well for each hypothesis. The chapter concludes with the conceptual model which shows the hypotheses in a graphical way.
Chapter four describes the research design, the variables, sample data, and describes how the variables will be measured.
Chapter five will describe the results of the research and the main findings.
Chapter six will provide finding, limitations of the research, recommendations for future research and implications.
2 LITERATURE REVIEW
This chapter will provide an overview of the main theoretical concepts organizational
ambidexterity, intra-organizational context and transformational leadership behaviors. It will result in a better understanding of the research topic and provides insight in what is already researched.
2.1 What is meant with organizational ambidexterity and how is it created?
“Aligned and efficient in their management of today’s business demands while simultaneously adaptive to changes in the environment”, is how Duncan (1976) describes organizational ambidexterity. It regards the balancing of learning activities exploitation and exploration. Exploitation is described by March (1991) as engaging in activities such as standardization and refinement and as an organization striving for efficiency, and exploration is described as engaging in activities such as innovation and experimentation and striving for flexibility. The challenge for an organization and its management is to balance exploitation and exploration, and to allocate resources between the two, since they are both crucial elements for achieving long-term performance (March, 1991; He & Wong, 2004; Raisch & Birkinshaw, 2008). One of the reasons why this challenge is often present is that it takes longer to see the return on the
investment in exploration than in exploitation (March, 1991). Another reason is that exploitation comes with less uncertainty than exploration, the latter having a higher risk of failure (Levinthal & March, 1993). Therefore, it often happens that there lies too much focus on exploitation. Even though it can lead to short-term performance, it can result in organizations that have trouble to appropriately react to changes in the environment due to a low focus on innovation (Raisch & Birkinshaw, 2008; Ahuja & Lampert, 2001; Leonard-Barton, 1992; Smith & Tushman, 2005).
Kortmann et al. (2014) makes the same point, by stating that too much focus on operational efficiency (similar to exploitation) can result in a negative influence on strategic flexibility. On the other hand, too much focus on exploration can result in a vicious circle of search and
unrewarding change (Raisch & Birkinshaw, 2008; Smith & Tushman, 2005; Volberda & Lewin, 2003). Similar, Kortmann et al. (2014) describe that too much focus on exploration can result in a lack of economies of scale (Grewal and Tansuhaj, 2001; Smith & Tushman, 2005).
Rather, companies should see exploitation and exploration as complements. Cao et al. (2009) describe that synergies can develop between exploitation and exploration. Exploitative activities can result in higher effectiveness regarding the development of new possibilities. The knowledge of current activities and resources can form the foundation for new ideas and possibilities. As mentioned earlier, in order to achieve long-term performance, a balance must be established between exploitation and exploration (Tushman & O’Reilly, 1996; Smith & Tushman, 2005). Multiple researchers have found evidence that organizational ambidexterity has a positive effect on performance, with these studies on various levels such as organizational, project and business-unit level (O’Reilly III & Tushman, 2013). Results of He & Wong’s (2004) research for example, indicated that ambidexterity is positively associated with sales growth. Lubatkin et al. (2006) also found that performance (relative to that of other major competitors on profitability and growth) is affected by a balance between exploitation and exploration, but through behavioral integration (i.e. a team’s wholeness and unity of effort).
The question remains, how do firms establish organizational ambidexterity in their firm? To answer this question, the forms to achieve ambidexterity as defined by O’Reilly III & Tushman (2013) will be followed.
2.1.1 In a ‘sequential’ manner
A first manner to establish organizational ambidexterity is to shift in structures through time, according to the dynamics in the environment of the organization (Duncan, 1976; Kaupilla, 2010). In their article, Siggelkow & Levinthal (2003) describe three forms of organizational structure; the centralized structure, decentralized structure and temporary decentralized structure which later reintegrates. The latter holds that in order to start exploration, a decentralized
structure is handled to provide new ideas and innovations space and resources. Eventually it is refined and formalized in order to achieve efficiency. As their results show, Siggelkow &
Levinthal (2003) describe that in cases when interactions between the organization’s activities are inescapable, this form can lead to higher performance than the pure centralized or decentralized structure. In opposite to the structural manner (as described later on), this manner holds that exploitation and exploration are pursued within the same business-unit, which results in exploration activities staying close to the core business (Gibson & Birkinshaw, 2004).
This manner of is also labeled as ‘sequential ambidexterity’ (O’Reilly III & Tushman, 2013).
2.1.2 In a ‘structural’ manner
A second manner to establish organizational ambidexterity is to separate exploitation and
exploration in the organization, by, for example, creating separate business-units or divisions for both of them (Tushman & O’Reilly III, 2008). The various subdivisions are then organized to meet their different environmental demands (Tushman & O’Reilly III, 2008; Gilbert, 2005). For example, in case radical innovation is necessary because of disruptiveness in the environment, exploratory units need to be separated from the exploitative ones (Christensen, 1998). These separate units are often differently organized from each other, regarding incentive systems and
team structures for example (O’Reilly III & Tushman, 2004). This however, results in the need for coordination. Top management needs to establish a vision along with a shared corporate culture to integrate the different units. This creates coordination costs (Lawrence & Lorsch, 1967).
This manner is also labeled as ‘structural ambidexterity’ (O’Reilly III & Tushman, 2013).
2.1.3 In a ‘contextual’ manner
A third manner to establish organizational ambidexterity is to give the floor to the individuals within an organization to pursue both exploration and exploitation, giving them the opportunity to divide their time between the two (Gibson & Birkinshaw, 2004). For example, should
employees focus on current customer accounts to meet quota, or should they nurture new customers with slightly different needs? (Birkinshaw & Gibson, 2004). Giving the floor to the individual should result in individuals efficiently performing their tasks in order to pursue customer satisfaction, but simultaneously acting in a flexible manner by looking for potential disruptions in the environment and new ways for innovation. This requires strategic direction and a certain context within the organization which is necessary to motivate individuals to act and let them act in a certain behavior to achieve this.
This manner is also labeled as ‘contextual ambidexterity’ (Gibson & Birkinshaw, 2004; O’Reilly III & Tushman, 2013).
2.1.4 Sequential and structural ambidexterity versus contextual ambidexterity
The main difference between sequential and contextual ambidexterity regards the process of ambidexterity. Whereas sequential ambidexterity is more static, contextual ambidexterity is more
organic and more viable to continuously reconfigure a company’s activities to meet the demands of the environment (Raisch et al. 2009).
The main difference between structural and contextual ambidexterity is the structure of
organizations; structural ambidextrous organizations are separated in units pursuing exploitation and units pursuing exploration, whereas contextual ambidextrous organizations will act more as a whole with ambidexterity pursued within the same unit. Gibson & Birkinshaw (2004) describe that in structural ambidextrous organizations it happens that the explorative units (e.g. R&D department) of structural ambidextrous organizations often lack in connections with the core business and operations, which is the main cause of coordination costs which are often present in these type of organizations (Lawrence & Lorsch, 1967). It even happens that the separation leads to isolation of explorative units. In opposition, pursuing ambidexterity within the same unit makes it easier for contextual ambidextrous organization to adapt as a whole.
However, according to Gibson & Birkinshaw (2004) there are successful samples of sequential and structural ambidexterity as well. Therefore, they suggest to use these forms of ambidexterity as complements, using structural ambidexterity as a kind of starting point. This way new ideas and innovations are provided with resources and space to develop. Eventually, the new ideas and innovations should be reintegrated within the organization, so switching back to contextual ambidexterity in a sequential manner. This holds however, that structural ambidexterity should only be temporal.
2.2 What is meant with the context of an organization and how is it related to organizational ambidexterity?
Bartlett & Ghosal (1994) describe context as the invisible set of stimuli that encourage and stresses individuals to act in a certain way, having effect on performance. To configure the context, they refer to four attributes which can be influenced by actions of management through systems, processes, and beliefs that shape individual behaviors in an organization. The four attributes of context are discipline, stretch, trust and support.
Bartlett & Ghosal (1994) describe the ‘discipline’ feature of context as: “members voluntarily strive to meet all expectations generated by their explicit or implicit commitments. Establishment of clear standards of performance and behavior, a system of open, candid, and rapid feedback, and consistency in the application of sanctions contribute to the establishment of discipline.” Bartlett & Ghosal (1994) describe the ‘stretch’ feature of context as: “induces members to voluntarily strive for more, rather than less, ambitious objectives. Establishment of a shared ambition, the development of a collective identity, and the ability to give personal meaning to the way in which individuals contribute to the overall purpose of an organization contribute to the establishment of stretch.”
Bartlett & Ghosal (1994) describe the ‘support’ feature of context as: “an attribute of an organization's context that induces its members to lend assistance and countenance to others. Mechanisms that allow actors to access the resources available to other actors, freedom of initiative at lower levels and personal orientation of senior functionaries that gives priority to providing guidance and help over exercising authority contribute to the establishment of support.”
Bartlett & Ghosal (1994) describe the ‘trust’ feature of context as: “an attribute of context that induces members to rely on the commitments of each other. Fairness and equity in a business-unit’s decision processes, involvement of individuals in decisions and activities affecting them, and staffing positions with people who possess and are seen to possess required capabilities contribute to the establishment of trust.”
The four attributes described by Bartlett & Ghoshal (1994) together should foster an environment where individuals pursue ambitious goals through discipline and stretch, and should also foster a cooperative environment through support and trust.
Individual behavior is affected by these organizational context features, as they result in initiative, cooperation, and learning.
2.2.1 Context as antecedent of organizational ambidexterity as antecedent of performance
Research shows that context is an antecedent of organizational ambidexterity on business-unit level, which is an antecedent of performance (Gibson & Birkinshaw, 2004; Raisch & Birkinshaw, 2008).
Gibson & Birkinshaw (2004) extend the framework of Bartlett & Ghosal (1994) by stating that context is an antecedent for organizational ambidexterity with individuals pursuing both exploration and exploitation. The qualitative interviews that Gibson & Birkshaw (2004) conducted, resulted in the conclusion that the four attributes by Bartlett & Ghoshal (1994) are complements that all have to be available for an organization to become ambidextrous. For their research they could not use the four attributes as described by Bartlett & Ghoshal (1994), and therefore they used two other concepts for the items of the four attributes discipline, stretch, trust and support:
Performance management for ‘discipline’ and ‘stretch’, which is related to motivating people to achieve high performance and making them responsible for their actions
Social context for ‘trust’ and ‘support’, which is related to see to people’s needs and provide care
These should result in a context that can enable organizational ambidexterity. It is of importance to mention, that both concepts need to be balanced, since an overemphasis on one of them can lead to low performance (Gibson & Birkinshaw, 2004).
In describing contextual ambidexterity, Gibson & Birkinshaw (2004) state that to achieve superior performance as an organization, it is essential to build a context that allows capabilities of alignment and adaptability emerge simultaneously.
2.3 What is meant with leadership and what is meant with leadership behavior?
Jago (1982) stated that leadership has two dimensions; a ‘process’ and a ‘property’ dimension. The ‘process’ dimension holds that leadership is the use of a non-forcible power in order to steer the organization, its activities and employees to follow a future developmental path. The
‘property dimension holds the set of competences and characteristics of a leader which executes this non-forcible power in a successful manner. The leadership construct can be divided in two conceptualizations;
Leadership traits; which can be seen as general acknowledged characteristics that can be measured of an individual
Leadership behavior; which can be seen as the patterns in actions individuals can take In general the key research question in the leadership literature is how leaders can influence the organizational context that determines a firm’s financial performance and long-term survival
(Yukl, 2008; Boal & Hooijberg, 2000; Finkelstein & Hambrick, 1996). Put in different words from the perspective of ambidexterity, the key question of leadership is, how can leaders influence organizational context that facilitate short-term alignment for financial performance and simultaneously facilitate adaption for long-term survival.
2.3.1 Leadership and organizational ambidexterity
Multiple authors have predicted that leadership has effect on organizational ambidexterity, mostly in an indirect way. Smith & Tushman (2005) for example, proposed that leaders can play a major role in managing the tension between exploration and exploitation, through decisions with an integrative nature such as the recognition of opportunities, linkages, and synergies. These decisions can lead to creating value. Gibson & Birkinshaw (2004) also propose that the role of leadership in creating ambidexterity is of importance, as creating the context within the
organization in which individuals can manage the tension of exploration and exploitation. However, there are authors who define leadership as a direct effect on organizational
ambidexterity as well. As mentioned earlier, Lubatkin et al. (2006) that ambidexterity is affected through behavioral integration (i.e. a team’s wholeness and unity of effort). Beckman (2006) proposes the affiliations and previous experiences of founding team members and the interactions among them, have direct effect on organizational ambidexterity. Jansen et al. (2008) describe that senior teams and leadership are important antecedents to balance the tension between exploitation and exploration.
2.3.2 Leadership behaviors
Schein (1990) describes that leadership behaviors are important to set an example to the organization and its employees, and therefore influence the context of the organization. In the article of Burke et al. (2006), multiple types of leadership behaviors are mentioned:
Task-focused leadership o Transactional o Initiating structure o Boundary spanning Person-focused leadership o Transformational o Consideration o Empowerment o Motivational
Transformational leadership is described as a creative exchange between leaders and
subordinates, in order to bring about vision driven change in people and context (Bass, 1985; Burke et al., 2006). This suits the research question, since it regards the bottom-up involvement of the operational layer of an organization in the tension of exploration and exploitation, and how leaders can establish a context within the organization to make this happen. Therefore, focus will lie on this kind of leadership.
In the article of Bommer and Rubin (2005), six transformational leadership behaviors are described:
1. Identifying and articulating a vision. A vision should describe a better future for its followers, and it impacts an employee’s perception that current conditions will improve
2. Fostering the acceptance of group goals. Leaders encourage employees to work together to achieve group goals
3. High performance expectations. Leaders communicate high expectations in order to instill confidence in followers to that they can achieve these goals
4. Providing intellectual stimulation. Leaders encourage followers to re-examine some of their assumptions and to find new ways of improving their performance
5. Role modeling. Leaders provide a behavioral example of subordinates to work according to the organization’s beliefs and values
6. Providing individualized support. By respecting followers and oversee their development related to personal needs
2.3.3 Transformational leadership and organizational ambidexterity
Research is executed on the relationship between transformational leadership and organizational ambidexterity. Jansen et al. (2008) found that transformational leadership has a positive effect as a moderator on the relationship between senior team attributes and organizational ambidexterity. Also Nemanich & Vera (2009) determined that a relationship between transformational
leadership and organizational ambidexterity exists. They described some transformational leadership behaviors having impact on organizational ambidexterity, such as providing a clear vision, the encouragement of employees to work for the greater good as a common goal and promote adaptive behavior.
3. THEORETICAL FRAMEWORK
This chapter provides an overview of all hypotheses that will be tested in order to answer the research question.
To ensure that there is (1) a relationship between intra-organizational context and contextual ambidexterity and (2) a relationship between contextual ambidexterity and performance, as is mentioned in the literature, the first hypothesis of this research is:
H1: Contextual ambidexterity mediates the positive relationship between intra-organizational context and performance
H1A: Intra-organizational context has a positive effect on contextual ambidexterity
H1B: Contextual ambidexterity has a positive effect on performance
More specifically, it is expected that the more an intra-organizational context is characterized by an interaction of performance management and social support, the more it will relate to a higher level of contextual ambidexterity, which will in turn lead to a higher level of performance. As mentioned earlier, Raisch & Birkinshaw (2008) suggest for further research to search for possible interrelationships between the antecedents of organizational ambidexterity of their model; structural-, contextual- and leadership-based factors. The different factors could possibly complement each other in the achievement of organizational ambidexterity. Therefore the moderation of leadership (by means of TLBs) on the relationship between intra-organizational context and contextual ambidexterity will be tested. Note: due to the time span of the master thesis, the focus will lie on three TLBs. These three TLBs are not chosen without reason, as is elaborated on below.
As Gibson & Birkinshaw (2004) described with the example of Oracle, one of the key successes of Oracle was that objectives were clear among the organization. In contrary, another example showed that no clear and overarching vision resulted on business-units each devising their own goals and objectives. Hambrick (1994) found this as well; describing that a vision expresses the future advancing direction and can prevent organizations from breaking into pieces. Tushman & O’Reilly (1996) also described the importance of vision as a source for ambidexterity. These articles indicate the importance of a clear vision in relation to organizational ambidexterity which is communicated by management. Therefore, the TLB ‘identifying and articulating a vision by a leader’ is included.
As Gibson & Birkinshaw (2004) described with the example of Oracle, one of the key successes of Oracle was that high expectations and goals were set for the employees. A balance needs to be maintained, since overemphasis on high expectations can lead to burn-outs (Bartlett & Ghoshal, 1994; Gibson & Birkinshaw, 2004), but still, this context is desirable for achieving exploitation and exploration. Therefore, the TLB ‘high performance expectation by a leader’ is included. Tushman & O’Reilly (1996) described supportive leaders as a source of ambidexterity. As Bommer & Rubin (2005) mentioned, individuals are more likely to support ideas of leaders who support them. Therefore, the TLB ‘providing individual support by a leader’ is included.
The moderation effect of TLBs on the relationship between intra-organizational and contextual ambidexterity is tested through the second hypothesis:
H2: The relationship between intra-organizational context and contextual
ambidexterity is positively moderated by transformational leadership behaviors (TLBs). More specifically, we expect that the stronger leadership behaviors are characterized by an interaction of identifying and articulating a vision, high performance expectations, and providing
individual support, the stronger the positive effect of intra-organizational context (characterized by an interaction of performance management and social support) on contextual ambidexterity will be.
The main and final test of this research is to determine whether TLBs (characterized by an interaction of leadership behaviors identifying and articulating a vision, high performance expectations and providing social support) enhance the positive effect of intra-organizational context (characterized by an interaction of performance management and social support) on performance, through the mediating role of contextual ambidexterity. Therefore, the third hypothesis is:
H3: The indirect effect of intra-organizational context on performance through
contextual ambidexterity will be moderated by transformational leadership behaviors
The hypotheses for this research are to be found in the conceptual model (see figure 1).
4. RESEARCH DESIGN
This chapter describes the research design of this study in order to answer the following research question;What transformational leadership behaviors enhance the effect of intra-organizational context on performance, through contextual ambidexterity?
4.1 Data collection
In order to answer the hypotheses and research question, quantitative research by means of a cross-sectional survey was executed. As described by Fowler Jr. (2013), one of the purposes of survey designs can be to ‘tap’ in the subjective feelings of people, in order to get a feeling of their needs or opinions. This is the main reason that the survey method was chosen for this study, since its purpose is to gather opinions and perspectives to answer the hypotheses and ultimately, the research question. Besides, there is a practical reason for the research design, which holds that the thesis’ period of time is quite short and the goal was to reach out to many respondents with consistent questions to have a valid conclusion based on comparable information (Fowler Jr., 2013).
The survey was a self-completed survey (Saunders et al., 2012), electronically distributed and set-up with Qualtrics. See the appendix for the questionnaire and its content. Since the aim lied on the opinions of respondents, Likert scales with a maximum of five possible answers were used (ranging from ‘Completely disagree’ to ‘Completely agree’). Besides, a sixth possible answer was included; “No answer”. This way, respondents were not pushed to answer to questions they were not able to answer. The personal network of the author was used to get respondents (via e-mail and social media). The respondents were also asked to distribute the survey in their personal networks in order to get more respondents. The survey included an introduction containing
information regarding the content of the survey so that respondents had access to the appropriate information for completing the survey. Both can be found in the appendix.
4.2 Sample
As mentioned in the literature review, relationships between intra-organizational context and contextual ambidexterity and between contextual ambidexterity and performance, are supported at business-unit level. Therefore the aim regarding respondents regarded individuals working in units. To ensure this perspective was maintained, the questions were set on business-unit level. For the respondents working in single-business-unit companies, the explanation was given that they could consider ‘business-unit’ as a synonym for ‘single-unit’.
Within business-units, the aim lied at individuals working at middle management and operational levels (i.e. every possible position but top management) to get their perspective of certain
transformational leadership behaviors’ effect on the enhancement of the effect of intra-organizational context on their ambidextrous behavior/activities.
The sample finally consisted of 133 respondents, of which 19 respondents were in top management layer. These were excluded since the aim lied on individuals working at middle management and operational levels, as mentioned earlier. 58 respondents belonged to the
operational layer (50.9%), 50 respondents belonged to the middle management layer (43.9%) and 6 respondents did not indicate which layer they were working at (5.5%).
Of the 114 remaining responses, 85 were completely answered. Most of the respondents were working at companies with more than 250 employees (45.6%) or companies with 0-50 employees (30.7%). Most of the respondents were working in the ‘Industry’ sector (47.4%) and the
the companies respondents were working for was 46.42 years. The average years of working experience of the respondents was 14.95.
Graph 4.1: Respondents’ distribution of industries
4.3 Measures
In this section, all variables and their measures will be explained. All were derived from previous research.
4.3.1 Dependent variable – performance
The concept performance is used differently in multiple studies (Kirby, 2005). According to Richard et al. (2009) this is due to the fact that performance is measured in different ways
because of differences in stakeholders, heterogeneous product market circumstances, and time. In general there are two ways to measure performance (Richard et al., 2009, Junni et al., 2013), which are in an ‘objective’ manner (e.g. based on financial indicators) or in a ‘perceptual’ manner. The latter has in general two ‘sub-manners’, which are ‘relative’ measurement (i.e. benchmarked with competitors) and ‘absolute’ measurement (i.e. not benchmarked with
competitors). Most studies measure performance through requesting/looking up financial data or interviewing/surveying top management levels to provide insight on relative performance (Junni et al., 2013). Due to the fact that the level of analysis of this research is on business-unit level, and the time-span of the thesis is quite short, the decision was made to (1) not gather financial information of each business-unit which will be surveyed and (2) not survey top-management levels to provide insight regarding performance. Gibson & Birkinshaw (2004) measure performance in an absolute perceptual manner, by surveying top-management and middle-management levels with four items. Their way of measurement was partly used for this research, with the difference that middle-management and operational levels were surveyed. An example of a question regarding the opinion of a respondent on performance: “This business-unit is achieving its full potential”.
4.3.2 Independent variable – intra-organizational context
The independent variable intra-organizational context was measured as an interaction of the factors performance management (with four items) and social context (with four items), as partly
depicted by Gibson & Birkinshaw (2004). The difference with the latter is that three and five (respectively) items were deleted in order to shorten the survey time. As described in chapter two, performance management counts for ‘discipline’ and ‘stretch’, which is related to motivating people to achieve high performance and making them responsible for their actions. Social context counts for ‘trust’ and ‘support’, which is related to see to people’s needs and provide care. An example of a question regarding the opinion of a respondent on performance context: “Systems in this organization encourage people to set challenging/aggressive goals”. An example of a
question regarding the opinion of a respondent on social support: “Systems in this organization devote considerable effort to developing their subordinates”.
4.3.3 Moderator variable – transformational leadership behaviors
As mentioned in chapter two, transformational leaders are capable of giving more meaning to the work, which results in employees who are more committed and perform beyond expectations (De Hoogh et al., 2004). Most studies with transformational leadership behavior as a variable,
measure by means of the ‘Multifactor Leadership Questionnaire (MLQ)’ of Bass and Aviolo (1993) (Eisenbeiß & Boerner, 2013, De Hoogh et al., 2004, Jansen et al., 2008, Nemanich & Vera, 2009). However, there has been criticism on the MLQ. In their article, Muenjohn & Armstrong (2008) provide a summary of this criticism. They state that the four transformational leadership behaviors mentioned by Bass (1985), which were each used as a single factor, may be best used as a single factor since they were highly correlated with each other. Another point of criticism Muenjohn & Armstrong (2008) provide regards the unclear conceptual distinction between the transformational leadership behaviors ‘idealized influence’ and ‘inspirational motivation’. The conclusion of Muenjohn & Armstrong (2008) is that the MLQ has an unclear
factor structure, and that it is therefore doubtful to use it for measurement of transformational leadership behaviors. That is why the ‘Charismatisch leiderschap in Organisaties (CLIO)’ questionnaire of De Hoogh et al. (2004) was used, since it deals with the criticism above as a response to the MLQ. In their article and questionnaire, De Hoogh et al. (2004) test the transformational leadership behaviors as mentioned in chapter two somewhat different. They equal identifying and articulating a vision as ‘charismatic leadership’ (Dutch: ‘charismatisch leiderschap’) and measure it by means of six items. However, for this research five items were used in order to shorten the survey time. An example of a question regarding the opinion of a respondent on identifying and articulating a vision: “Management of this business-unit has a vision and direction for the future”. De Hoogh et al. (2004) equal high performance expectations with a part of transactional leadership, named ‘contingent reward’ (Dutch: ‘contingent belonen’) and measure it by means of three items. However, they point out that ‘contingent reward’
correlates stronger with transformational leadership than transactional leadership. An example of a question regarding the opinion of a respondent on high performance expectations:
“Management of this business-unit delegates challenging responsibilities to subordinates”. De Hoogh et al. (2004) equal individual support by a leader as ‘empowerment’, which is a combination of individual consideration and intellectual stimulation (Dutch: ‘individuele
consideratie en intellectuele stimulatie’) and measure it by means of four items. An example of a question regarding the opinion of a respondent on individual support: “Management of this business-unit stimulates employees to develop their skills as much as possible”.
Since the goal is to measure the moderating effect of transformational leadership behaviors as a whole, the three transformational leadership behaviors will be multiplied with each other to have one variable tested.
4.3.4 Mediator variable – contextual ambidexterity
The mediator variable contextual ambidexterity was measured in line with the construct as presented by Gibson & Birkinshaw (2004), by asking questions regarding the concepts
alignment/exploitation and adaptability/exploration. As mentioned in chapter two, exploitation can be defined by key words standardization, refinement and efficiency, whereas exploration can be defined by key words innovation, experimentation and flexibility. Due to the fact that the time-span of the thesis was quite short, the decision was made to not survey top-management levels to provide insight regarding contextual ambidexterity. Gibson & Birkinshaw (2004) measure contextual ambidexterity by interviewing top-management and middle-management levels. Their way of measurement was partly used for this research, with the difference that middle-management and operational levels were surveyed. An example of a question regarding the opinion of a respondent on alignment/exploitation: “The systems in this business-unit work coherently to support the overall objectives of this organization”. An example of a question regarding the opinion of a respondent on adaptability/exploration: “The systems in this business-unit are flexible enough to allow us to respond quickly to changes in our markets”.
4.3.5 Control variables
The control variables for this research were company size, (business) unit size, company age, total years of work experience, years with the company, operating layer within the (business) unit and industry.
Company and business-unit size were chosen to determine flexibility of the organization. Larger companies are in general less flexible with lots of reporting levels, whereas smaller companies are in general more able to adjust quickly to opportunities and treats.
Company age was chosen to determine the conservativeness within the organization. Older companies tend to be more conservative and stick to the current working methods and products (i.e. exploitation), in comparison with newer companies.
The total amount of years of work experience was chosen since there is difference in opinion and view between someone who just graduated and someone who has ten years of work experience. Similar, years with the company was chosen to differentiate the newbies from the oldies, where the latter often have a more knowledge of the company and its methods than someone who has been working at the company for less than a year.
Again similar, operating layer within the company was chosen to differentiate operational level from middle management level, since the different kind of responsibilities of both often results in different viewpoints.
Industry was chosen to differentiate the general levels of exploration and exploitation. A common steel manufacturing company is likely to be more focused on efficiency (i.e. exploitation) than a high-tech developing company which will be more focused on innovation (i.e. exploration).
5. RESULTS
This chapter describes the results of the research and the main findings.
5.1 Data preparation
To prepare the data for hypotheses testing, some steps were taken.
First the data was screened on odd and missing answers. Odd answers included typos and misunderstood questions. These answers were all corrected, for example: an answer on the question what year the company the respondent works for was founded; “After the merger: seven years ago” was changed to “2010”. Also, all cells indicating the answer possibility ‘No answer’ were deleted and indicated as ‘missing values’ since they do not contribute the analyses (list wise deletion). One of the questions was how large the business-unit is the respondent is working at. One of the answer possibilities was ‘Not applicable (single unit)’. Of the 133 filled answers, only 10 were answered with ‘Not applicable (single unit)’. These answers were deleted and indicated as ‘missing values’ (list wise deletion).
Second, normality checks for skewness and kurtosis were executed. Only the item “The
organization in this BU…works coherently to support the overall objectives” (AL1) and control variables company size, business-unit size, and operating level in the organization had skewness and/or kurtosis issues. However, for item AL1 the decision was made to not correct it, since the skewness and kurtosis were minor (-1.139 and 1.046 respectively). Also for the control variables no corrections were made, but reference is made to the ‘Discussion’ chapter of this report. Third, outlier checks were executed. Only the item “My manager…shows to be convinced of his/her ideals, conceptions and values” showed a Z-score of >3 (-3.40297). When checking the
distribution via histogram, it showed little negative skew and a little kurtosis, and therefore no corrections were made.
Fourth, the reversed item “The organization in this BU…causes us to use resources on unproductive activities” (rAL4), was recoded into a new variable (AL4).
Fifth, a principal axis factoring analysis (PAF) was conducted on the scales of the factors (1) ‘identifying and articulating a vision’ (IAV), (2) ‘high performance expectations’ (HPE), and (3) ‘providing individual support’ (IS) to see whether the variable transformational leadership behaviors is valid. The Kaiser-Meyer-Olkin measure verified the sampling adequacy for the analysis, KMO = 0.907. Bartlett’s test of sphericity χ² (66) = 727,233, p < 0.001, indicated that correlations between items were sufficiently large for PAF. An initial analysis was run to obtain eigenvalues for each component in the data. Two components had eigenvalues over Kaiser’s criterion of 1 and in combination explained 65.836% of the variance. In agreement with Kaiser’s criterion, examination of the scree plot revealed a levelling off after the second factor. Thus, two factors were retained and rotated with an Oblimin with Kaiser normalization rotation. Table 5.1 shows the factor loadings after rotation. The items that cluster on the same factors suggest that factor 1 represents employee development by a leader, and factor 2 represents future orientation by a leader. Even though this two-factor solution is contradiction with the original plan and constructs of previous research (see literature review), to use a one-factor solution
(transformational leadership behaviors), the Kaiser’s criterion and scree plot criterion are followed.
Fifth, reliability analysis was executed. The contextual ambidexterity scale has high reliability, with Cronbach’s Alpha = 0.816. The corrected item-total correlations indicate that only the item “The organization in this BU…causes us to use resources on unproductive activities” (AL4) does
not have a good correlation with the total score of the scale (0.086). Also, the item would affect reliability positively with .031 if it were deleted. Therefore this item was deleted, which resulted in a new Cronbach’s Alpha = 0.847. All other variables have high reliability, with Cronbach’s Alphas = 0.883 for TLB: employee development (self-created variable based on transformational leadership behaviors), = 0.891 for TLB: future orientation (self-created variable based on
transformational leadership behaviors), = 0.778 for intra-organizational context, and = 0.762 for performance. For most variables the corrected item-total correlations indicate that all the items have a good correlation with the total score of the scale (all above 0.30). Also, none of the items would substantially affect reliability if they were deleted. Only for the variable
intra-organizational context there were some items that do not have good correlations with the total score of the scale (0.155 and 0.280). However, since deleting the items would affect reliability in a minor way, it was decided to not delete them.
The final step of data preparation was the creating the final scales by taking the means of the means. The means of the items of alignment and adaptability were added to form contextual ambidexterity (“CA”), and of performance management and social context to form intra-organizational context (“IOC”). Also the mean of the items of performance was taken to form performance (“P”).
Table 5.1: Transformational leadership behaviors (factor loadings)
Item
Rotated factor loadings
EmpDev FutOrient
My manager…only criticizes employees with appropriate reasons .930 .151 My manager…sees to the creation of the conditions in which employees can adequately
execute their activities/tasks .825 .016
My manager…delegates challenging responsibilities to employees .769 .008 My manager…involves employees with decisions that are of importance for their work .637 -.250 My manager…talks with employees about what is of importance for them .622 -.242 My manager…gives employees the feeling that they work on an important and common
My manager…has a vision and a direction for the future -.196 -1.014 My manager…is always looking for new possibilities for the organization -.032 -.869
My manager…is able to enthuse others for his/her plans .241 -.623
My manager…shows to be convinced of his/her ideals, conceptions and values .117 -.612 My manager…stimulates employees to develop their talents in the best way possible .313 -.601 My manager…stimulates employees to think about problems in new ways .423 -.510
Eigenvalues 6.701 1.199
% of variance 55.84 9.995
Extraction Method: Principal Component Analysis. a. 2 components extracted.
5.2 Hypotheses testing
Reference is made to table 5.2 for descriptive statistics and correlations. The table shows correlations between the main variables. As can be seen in table 5.2, high correlations exist between performance and contextual ambidexterity (r = 0.562,p ≤0.01), performance and TLB: future orientation (r = -0.440,p ≤0.01), performance and TLB: employee development (r = 0.474, p ≤0.01), and performance and intra-organizational context (r = 0.679,p ≤0.01). With regard to the outcome variable ‘performance’ and the related prediction that contextual ambidexterity positively affects it, these figures show that it is likely that the prediction is true. The figures show correlations between transformational leadership behaviors and performance as well. This was not predicted, but will be further investigated since the knowledge of how to affect
performance is, as mentioned in the introduction (chapter one), valuable. High correlations also exist between contextual ambidexterity and TLB: future orientation (r = -0.438,p ≤0.01), contextual ambidexterity and TLB: employee development (r = 0.340,p ≤0.01), and contextual ambidexterity and intra-organizational context (r = 0.664,p ≤0.01). Based, on these figures it is likely that the predictions regarding (1) a positive effect of intra-organizational context on contextual ambidexterity and (2) that this positive effect is enhanced by transformational
leadership behaviors, are true. Finally, high correlations exist between intra-organizational context and TLB: employee development (r = 0.427,p ≤0.01), and between intra-organizational context and TLB: future orientation (r = -0.501, p ≤0.01). These figures show that these variables are related to each other, which could indicate that the predication that transformational
leadership behavior positively enhances the effect of intra-organizational context on contextual ambidexterity is true.
Table 5.2 also shows tendencies to correlations between some of the main variables and control variables; between performance and operating layer in the business-unit of the respondent (r = 0.219, p ≤0.05) and between contextual ambidexterity and business-unit size (r = 0.199, p ≤0.05). These correlations will be further investigated since it is interesting to know whether there is a difference in perception of performance based on the operating layer within a business-unit you are located. Also it would be interesting to know whether an effect exists between business-unit size and contextual ambidexterity, since it could mean that in order to achieve contextual ambidexterity, one should keep the size at a certain level.
To further test the hypotheses, the modeling tool “PROCESS” by Hayes was used. This tool was used for its integration of many functions of popular statistical tools (Hayes, 2012) which suit the tests needed to be done; mediation, multiple moderation and multiple moderated mediation. Hayes (2013) defines 74 models of which use was made of three for this research.
Model four was used to test the first hypothesis, which regards mediation. The effect of intra-organizational context on contextual ambidexterity
a
1 = 0.755 means that two business-units that differ by one unit on intra-organizational context, are estimated to differ by 0.755 units onrelatively stronger intra-organizational context are estimated stronger in contextual
ambidexterity (for a graphical representation, see graph 5.1). This effect is statistically different from zero, t = 8.616, p = 0.000, with a 95% confidence interval from 0.581 to 0.929. This holds that the effect predicted in hypothesis 1a is found. The effect of b1 = 0.237 would indicate that
two business-units who have the same intra-organizational context but that differ by one unit in their level of contextual ambidexterity, are estimated to differ by 0.237 units in performance. The figure is positive, meaning that those business-units with strong contextual ambidexterity are estimated to have higher performance. However, the effect is not significant, with p = 0.055, with t = 1.943, a 95% confidence interval from -0.006 to 0.479. This holds that the effect predicted in hypothesis 1b is not found. The indirect effect of 0.179 would mean that two business-units which differ by one unit in their reported intra-organizational context, are estimated to differ by 0.179 in their reported performance as a result of the tendency for those who hold the intra-organizational context in high regard which results in a higher balance between alignment and adaptability, which in turn leads to higher performance. However, this effect is not significant, since the 95% BC bootstrap confidence interval is below and above zero (-0.02 to 0.402). This holds that the effect predicted in hypothesis 1 is not found.
Besides the main variables, one of the control variables had a significant effect as well: company size on contextual ambidexterity (coeff. = 0.144, p = 0.037). This means that the larger
companies are, the stronger they will be in their achievement of contextual ambidexterity or the more likely it will be that they are contextually ambidextrous.
Variables M SD 1 2 3 4 5 6 7 8 9 10 11
1. Company size 2.16 .888 -
2. Business-unit size 1.35 .594 .307** -
3. Company date of founding 1970.58 38.180 -.395** -.201 -
4. Work experience in years 14.95 12.074 .062 .054 -.045 -
5. Operating layer in business-unit 1.46 .501 .108 .218* -.071 .172 -
6. Industry 7.29 2.434 .065 .187 .029 -.077 .212* - 7. Intra-organizational context 3.5712 .63891 -.039 .003 .002 -.209* .175 .007 (.778) 8. TLB: employee development .0000000 1.0000000 -.015 .072 .033 -.218* .046 .220* .427** (.883) 9. TLB: future orientation .0000000 1.0000000 -.074 -.115 -.036 .273** -.073 -.058 -.501** -.568** (.891) 10. Contextual ambidexterity 3.5450 .71146 .184 .199* -.157 -.155 .169 .089 .664** .340** -.438** (.847) 11. Performance 3.2569 .76700 -.008 -.064 .003 -.135 .219* .089 .679** .474** -.440** .562** (.762) **. Correlation is significant at the 0.01 level (2-tailed).
default. These were not included in the conceptual model of this research, but the results proved that they should be mentioned. The direct effect of intra-organizational context, c1 = 0.588, is the
estimated difference in performance between two business-units experiencing the same level of contextual ambidexterity but which differ by one unit in their reported intra-organizational context, meaning that the business-unit with stronger intra-organizational context but with the equal contextual ambidexterity, is estimated to be 0.588 units higher in its reported performance (for a graphical representation, see graph 5.2). This direct effect is statistically different from zero, t = 4.381, p = 0.000, with a 95% confidence interval from 0.321 to 0.855.
The total effect of intra-organizational context on performance is c = 0.767, meaning two business-units who differ by one unit in intra-organizational context, are estimated to differ by 0.767 units in their reported performance. The figure is positive, which means the business-unit with a stronger intra-organizational context reports higher performance. This effect is statistically different from zero, t = 7.713, p = 0.000, with a 95% confidence interval from 0.569 to 0.965.
Table 5.3: Mediation results
Consequent
Contextual Ambidexterity (M) Performance (Y)
Antecedent Coeff . SE p Coeff. SE p Intra-organizational context (X) a1 .755 .088 <.001 c1 .588 .134 <.00 1 Contextual Ambidexterity (M) - - - b1 .237 .122 .055 Company size (CV1) .144 .068 .037 - - - constant i1 4.288 3.143 .1761 i2 3.549 .819 R2 = .542 R2 = .482
F(7.84) = 14.199, p<.001 F(8.83) = 9.6336, p<.001
Effect SE p LLCI ULCI
Direct effect c1 .588 .134 <.001 .321 .855
Total effect c1 .767 .1 <.001 .569 .965
Boot SE Boot LLCI Boot
ULCI Indirect effect a1
b1 .179 .108 -0.02 .402
Graph 5.1: Intra-organizational context and contextual ambidexterity Graph 5.2 Intra-organizational context and performance Model two was used to test the second hypothesis, which regards multiple moderation. The regression coefficient for XW is c4 = 0.188, which would mean that the effect of
intra-organizational context on contextual ambidexterity depends on TLB: employee development. However, the effect is not significant since p = 0.119, with t(73) = 1.579. This means that there is no moderation effect from TLB: employee development. The regression coefficient for XZ is c5 =
0.313 and is statistically different from zero, t(73) = 2.905, since p = 0.005. Thus, the effect of intra-organizational context on contextual ambidexterity depends on TLB: future orientation.. This holds that the effect predicted in hypothesis 2 is only partly found. Moreover, 4.75% of contextual ambidexterity is explained by the interaction between intra-organizational context and
TLB: future orientation and in total, this model accounts for 58.9% of variance in contextual ambidexterity.
Table 5.4: Multiple moderation results
Coeff. SE t p
Intercept i1 7.819 3.345 2.338 0.022
Intra-organizational context (X) c1 .752 .106 7.099 <.001
TLB: employee development (W) c2 -.002 .071 -.022 .982
TLB: future orientation (Z) c3 -.023 .078 -.297 .767
IOC*TLB: employee development (XW) c4 .188 .119 1.579 .119
IOC*TLB: future orientation (XZ) c5 .313 .108 2.905 .005
R2 = 0.589, p<.001
F(11.73) = 9.504
Model nine was used to test the third hypothesis, which regards multiple moderated mediation. The results indicate that no evidence is found for a significant multiple moderated mediation effect, so it is not proven that the positive effect of intra-organizational context on performance, through contextual ambidexterity, is enhanced by TLB: employee development and/or TLB: future orientation. The first interaction effect (with TLB: employee development), as described earlier at the multiple moderation results, is not significant with a regression coefficient for XW is a5 = 0.189. The second interaction effect (with TLB: future orientation), as described at the
multiple moderation results, is significant (p = 0.005) with a regression coefficient for XZ is a6 =
0.313. However, the multiple moderated mediation effect is not found since This holds that the effect predicted in hypothesis 3 is not found.
However, this effect is not significant, since the 95% bootstrap confidence interval is below and above zero (-0.018 to 0.169). This holds that the effect predicted in hypothesis 3 is not found.
Table 5.5: Multiple moderated mediation results
Consequent
Contextual Ambidexterity (M) Performance (Y)
Antecedent Coeff. SE p Coeff. SE p
Intra-organizational context (X) a1 .752 .106 <.001 c1 .655 .128 <.001 Contextual Ambidexterity (M) a2 - - - b1 .205 .111 .068 TLB: employee development (W) a3 -.002 .071 .982 - - - TLB: future orientation (Z) a4 -.023 .077 .767 - - - IOC*TLB: employee development (XW) a5 .189 .119 .119 - - -
IOC*TLB: future orientation
(XZ) a6 .313 .108 .005 - - -
constant i1 7.819 3.345 .022 i2 -0,9145 3.377 .7873
R2 = .589 R2 = .558
F(11.73) = 9.504, p<.001 F(4.221) = 11.98, p<.001
Besides tests regarding the hypotheses, some additional tests were performed for significant correlations. Hierarchical multiple regression was performed to investigate the ability of TLB employee development to predict levels of performance, after controlling for company size, business-unit size, work experience, operating layer in the business-unit and industry. In the first step of hierarchical multiple regression, six predictors were entered company size, business-unit size, work experience, operating layer in the business-unit and industry. This model was not significant F (6, 78) = 1.438; p = 2.11. After entry of TLB: employee development at step two the
total variance explained by the model as a whole was 25,8% F (1, 77) = ; p < 0.001. In the final model, one out of seven predictor variables was statistically significant; if TLB: employee development increases for one, the performance of the business-unit will increase for 0.421. So, TLB: employee development has a significant positive effect on performance.
Table 5.6: TLB: employee development and performance
R R2 R2 change B SE β t Step 1 .316 .030 Company size -.080 .102 -.093 -.778 Business-unit size -.089 .151 -.071 -.591 Company date of founding .001 .002 .030 .262 Work experience in years -.010 .006 -.164 -1.480 Operating layer in business-unit .213 .172 .146 1.242 Industry .057 .034 .191 1.650 Step 2 .508 .258*** .158 Company size -.054 .094 -.063 -.575 Business-unit size -.104 .138 -.083 -.754 Company date of founding .001 .002 .048 .457 Work experience in years -.005 .006 -.090 -.873 Operating layer in business-unit .236 .157 .161 1.503 Industry .025 .032 .083 .760 TLB: employee development .305 .075 .421*** 4.054
Note. Statistical significance: *p <.05; **p <.01; ***p <.001
Hierarchical multiple regression was performed to investigate the ability of TLB future orientation to predict levels of performance, after controlling for company size, business-unit size, work experience, operating layer in the business-unit and industry. In the first step of hierarchical multiple regression, six predictors were entered company size, business-unit size,
work experience, operating layer in the business-unit and industry. This model was not
significant F (6, 78) = 1.438; p = 2.11. After entry of TLB: future orientation at step two the total variance explained by the model as a whole was 25,8% F (1, 77) = ; p = 0.001. In the final model, one out of seven predictor variables was statistically significant; if TLB: future orientation
increases for one, the performance of the business-unit will decrease for 0.418. So, TLB: future orientation has a significant negative effect on performance.
Table 5.7: TLB: future orientation and performance
R R2 R2 change B SE β t Step 1 .316 .030 Company size -.080 .102 -.093 -.778 Business-unit size -.089 .151 -.071 -.591 Company date of founding .001 .002 .030 .262 Work experience in years -.010 .006 -.164 -1.480 Operating layer in business-unit .213 .172 .146 1.242 Industry .057 .034 .191 1.650 Step 2 .508 .258*** .158 Company size -.091 .093 -.106 -.971 Business-unit size -.126 .139 -.100 -.906 Company date of founding -6.399 .002 .000 .003 Work experience in years -.003 .006 -.043 -.410 Operating layer in business-unit .154 .158 .105 .974 Industry .052 .032 .176 1.663 TLB: employee development -.308 .076 -.418*** -4.051
Note. Statistical significance: *p <.05; **p <.01; ***p <.001
Furthermore, an independent samples T-test was executed in order to further test the correlation between performance and operating layer within a business-unit. The results indicate that on
average, participants of the middle-management layer within a business-unit have higher
perceptions of performance (M = 3.437, SE = 0.101) than those of the operational layer within a business-unit (M = 3.102, SE = 0.103). This difference, -0.335, BCa 95% CI
[-0.639, -0.075], was significant t(106) = -2.306, p = 0.023.
