Master thesis
Which personality traits influence the ability to be
individually ambidextrous?
Student: G.A. van Ravenhorst - 10278788
Executive Program Business Studies, University of Amsterdam Master Thesis
g.vanravenhorst@upcmail.nl
Supervisor: Dr. Dipl.-Wirt.-Ing. S. Kortmann
Assistant Professor of Strategy and Innovation
2
Abstract
This study tries to find an answer on the question whether personality traits could predict individual ambidexterity. The study is partly explorative because the field of individual ambidexterity is relatively young and little research has been done on the internal factors forming individual ambidexterity. Therefore there still is a lack of a clear framework and definitions. This study tried to bring direction into this quest for a clear framework and clear definitions although future research should focus onto this quest before going deeper into the micro foundations.
Literature shows that individual ambidexterity, as a balance between exploration and
exploitation, could have personality traits as predictor for individual ambidexterity. This study found that both extraversion and openness to new experiences positively relate to individual ambidexterity and therefore prove that internal factors can predict individual ambidexterity. Next to these findings also educational level positively relates to individual ambidexterity which is in line with findings from another study which found a positive relation between intelligence and individual ambidexterity. These results provide input for future research and contributes to our understanding of the concept of individual ambidexterity which is valuable for both practice as theory.
3
Content
Content ... 3 1 Introduction ... 5 2 Literature review ... 7 2.1 Organizational ambidexterity ... 7 2.2 Contextual ambidexterity ... 8 2.3 Exploration vs. exploitation ...10 2.4 Individual ambidexterity ...13 2.5 Personality traits ...17 2.5.1 Openness to experience ...18 2.5.2 Conscientiousness ...19 2.5.3 Extraversion ...20 2.5.4 Agreeableness ...21 2.5.5 Neuroticism ...22 3 Conceptual model ...23 3.1 Moderating effect ...23 3.2 Conceptual model ...25 4 Method ...25 4.1 Sample ...25 4.2 Measurement of variables ...26 4.2.1 Control variables ...27 4.2.2 Personality ...27 4.2.3 Hierarchical level ...27 4.2.4 Individual ambidexterity ...274
4.3 Statistical procedure ...28 5 Results ...34 5.1 Correlation analysis ...34 5.2 Regression analysis ...35 6 Discussion ...35 7 Implications ...40 7.1 Theoretical implications ...40 7.2 Practical implications ...428 Limitations & future research ...43
9 Conclusion ...45
10 References: ...47
11 Appendixes ...52
5
1 Introduction
Firms are constantly searching for the best ways to survive and to win from the competition. Because environments constantly change, these firms search for ways to be successful today but also at the same time think about what brings success tomorrow. This quest for a balance between exploitation and exploration is called organizational ambidexterity (Gibson & Birkinshaw, 2004; Jansen, et al., 2009). This field of study is still open for debate because studies about the need for organizational ambidexterity are done often times, but studies about the way ambidexterity is achieved are still hard to find (Fillipini et. al., 2012; Jansen et. al., 2009; Raisch et. al., 2009; Simsek, 2009). If there is a need for ambidexterity according to researchers, then why is there still so much unclear about how to achieve this? According to the way ambidexterity is described it seems to be a dynamic capability. So extended research could be expected, as dynamic capabilities are studied extensively, but the
integration of ambidexterity and dynamic capabilities is hardly done yet (O'Reilly & Tushman, 2008), so this connection cannot be readily used.
As much as the field of organizational ambidexterity is studied, as little is the field of individual ambidexterity being studied. A reason could be that the focus in this field lies on structural ambidexterity (Good & Michel, 2013) as one of the ways organizational
ambidexterity is formed. The other way organizational ambidexterity is formed is by
contextual ambidexterity. Both ways of organizational ambidexterity differ in the way this is achieved. Structural ambidexterity means an organization forms two separate units, one performing exploration and one performing exploitation, to achieve organizational
ambidexterity. Contextual ambidexterity means an organization tries to create an atmosphere in which every individual becomes ambidextrous and by accumulating those ambidextrous individuals become ambidextrous as an organization (Gibson & Birkinshaw, 2004). In finding an answer on the ‘how’ about becoming ambidextrous this study tries to go deeper into
6
The assumption Gibson & Birkinshaw (2004) make in their theory about contextual ambidexterity is that every individual can become ambidextrous. This can be questioned because of the little research done on individual ambidexterity and therefore little
understanding about individuals being ambidextrous. Studies done in the individual
ambidexterity field are often done on a managerial level (Good & Michel, 2013; Mom et al., 2007; Mom et al., 2009; Yu et al., 2013). The study of Good & Michel (2013) is an exception because this research is done by using an undergraduate population, although it could be questioned if this outcome is generalizable because of the uniform research population. The study of Good & Michel (2013) is also the only study found that tries to link individual
characteristics to the ability to be ambidextrous. They tried to link individual exploitation, individual exploration and cognitive flexibility to individual ambidexterity. This is the first study found really focusing on the individual trying to find differences between individuals that could explain the formation of individual ambidexterity.
Although the study of Good & Michel (2013) focused more on the individual the origin of the individual differences can be studied deeper using individual characteristics or personality traits. This study therefore tries to find differences between individuals in the ability to become individually ambidextrous by using the Five Factor Model of personality traits
consisting of Openness to Experience, Conscientiousness, Extraversion, Agreeableness and Neuroticism (Emotional Stability) (Costa Jr & McCrae, 1992). This model is one of the most influential and most popular models to evaluate a person’s personality (Costa Jr. & McCrae,
1992; Goldberg, 1990; Judge, et. al., 2002; Just, 2011; Vassend & Skrondal, 2011; Widiger & Trull, 1997) and therefore used in this study. Therefore the research question this study tries to answer will be;
‘Which personality traits influence the ability to be individually ambidextrous?’.
In the study of Good & Michel (2013) intelligence was also a part of the study, explaining a part of the ability to be individually ambidextrous. This in combination with the research
7
focused on managerial level individuals (Gupta, et al., 2006; Mom, et al., 2009; Raisch & Birkinshaw, 2008) could implicate that hierarchical level could be a moderating factor. The rationale behind this is the assumption that individuals higher on intelligence also have a higher hierarchical level. Also because of the exploratory nature of this study this moderating effect is tested in this study to find possible implications for further studies to be done in this field to fully understand the underlying foundations of individual ambidexterity.
The contribution this study tries to make is to further open the field of individual ambidexterity by building a better understanding of the construct and create a foundation for further
research. Outcome of this study could also have implications for the contextual ambidexterity field of research by answering the question is every individual is able to become individually ambidextrous. Apart from this theoretical contribution this studies also contributes to the practical field. This information will help managers and organizations to find the proper people when they strive to become contextual ambidextrous as an organization or to assess if the right people are already a part of the organization to become contextual ambidextrous.
2 Literature review
2.1 Organizational ambidexterity
Ambidexterity refers to the ability to perform different tasks simultaneously, like exploration and exploitation (March, 1991), evolution and revolution (Tushman & O'Reilly III, 1996), flexibility and efficiency (Adler, et al., 1999) or alignment and adaptability (Gibson &
Birkinshaw, 2004). In literature these different underlying dimensions are all used to indicate the same phenomenon that is now labeled as ambidexterity (Raisch & Birkinshaw, 2008). Exploration and exploitation will be used mainly in this thesis because March (1991) started with these two dimensions and these terms are used in later studies and therefore gives the most uniform construct.
To further develop the understanding of organizational ambidexterity Gibson & Birkinshaw (2004) proposed two levels of analysis; structural and contextual ambidexterity. The first level
8
is called structural ambidexterity which tries to understand ambidexterity purely from an organizational point of view. This means the organization should structure the responsibility for exploitation or exploration. Ambidextrous organizations where described by Tushman & O'Reilly III (1996) as organizations that could master evolutionary and revolutionary change, thereby giving an accent on the fact that also exploitation is still change, although
evolutionary. This is an important aspect for understanding exploitation as one might think that exploitation has nothing to do with change and gives an idea of a sitting duck.
Contextual ambidexterity should further deepen the understanding of ambidexterity within an individual. Especially ambidexterity at the individual unit of analysis is not yet studied
extensively by researchers (Gupta, et al., 2006; Mom, et al., 2009; Raisch & Birkinshaw, 2008). Gibson & Birkinshaw (2004) further developed the concept of contextual ambidexterity by stating that ambidexterity could be best build by encouraging individuals to make their own judgment on the balance between alignment and adaptation. Gibson & Birkinshaw (2004) also state that when contextual ambidexterity is achieved every individual can deliver value to existing customers and at the same time be on the lookout for possible changes in the task environment and act to it accordingly.
2.2 Contextual ambidexterity
This study focuses on the individual level of ambidexterity. Therefore structural ambidexterity won’t be explained any further in this study, as this is focused on the organizational ability to
structure ambidexterity into organizational departments and not within individuals. Contextual ambidexterity on the other hand is about the context of an organization to build individual ambidexterity and that way become ambidextrous as an organization (De Clerq, et al., 2013; Gibson & Birkinshaw, 2004; Lubatkin et. al., 2006), most of the time managers are the individuals mentioned in this literature. The rationale behind this philosophy is the
accumulation of ambidextrous individuals which form a contextual ambidextrous organization together. Lubatkin et. al. (2006) studied this contextual ambidexterity phenomenon by
9
small and medium sized enterprises are more dependent on contextual ambidexterity because they lack the size to choose for structural ambidexterity. They found that when top management teams where formed by ambidextrous individuals the enterprise as a whole became more ambidextrous. The accumulation of individual ambidexterity formed an ambidextrous top management team, which forms an ambidextrous enterprise (Lubatkin, et al., 2006).
This phenomenon is also described by Adler, et al. (1999) in their case study of model changeovers in the Toyota Production System as employees and managers where able to switch between effiency and flexibility. In the case study the Toyota organization handled model changeovers in their factory differently than competitors resulting in a much quicker changeover process. They relied on individuals and trust was a key aspect, both to suppliers as to their own employees. The organization created an atmosphere where people were able to become ambidextrous. To organize this ambidextrous context an organization should possess enough performance management and social support to stimulate managers to become ambidextrous (Birkinshaw & Gibson, 2004). This field of research is therefore more interesting for this study, although the contextual ambidexterity field of research is still focused on the organizational level.
Gibson & Birkinshaw (2004) conclude that the context of an organization should be characterized by a combination of discipline, stretch, support and trust. A small recap is made for each characteristic. Discipline marks the level of commitment to voluntarily strive to meet the expectations made by commitments that are made. Stretch induces individuals to voluntarily strive for more, rather than less, ambitious objectives. Support induces individuals to assist each other and give each other access to resources available, which means
individuals are open to share knowledge with others. Trust, as the last attribute, gives individuals the courage to rely on commitments of others. Important issues in building trust are fairness and involvement in decision making (Gibson & Birkinshaw, 2004; Goshal &
10
Bartlett, 1994). The last characteristic, trust, is also found in the Toyota case (Adler, et al., 1999).
To become contextual ambidextrous, organizations have to make sure the conditions are proper for individuals to become ambidextrous. Therefore there needs to be enough performance management to push the managers to strive for better results and search for new ways to generate better results. There also needs to be enough social support to give the managers a sense of security because when they search for new ways to generate better results they might mistake (Birkinshaw & Gibson, 2004). When both performance
management and social support are high in an organization the organization gets into a high performance context triggering ambidexterity in managers and that way high performance (Birkinshaw & Gibson, 2004).
As Gibson & Birkinshaw (2004) make the assumption that every individual can be ambidextrous, of which no empirical evidence is given, Mom, et al. (2009) analyses
ambidexterity of managers which is already more individually focused. What is still missing in researching contextual ambidexterity are other, non managerial, employees, although a recent study of Yu, et al. (2013) focuses on front line employees of customer service centers. As studies on contextual ambidexterity have mainly focused on the environmental context needed to foster ambidexterity at the managerial level, this study will focus on the ability of every individual in the organization to be ambidextrous.
2.3 Exploration vs. exploitation
Organizational ambidexterity is about the tradeoff between exploration and exploitation (Gibson & Birkinshaw, 2004). Both exploration and exploitation have been studied in a variety of literatures such as organizational learning, organizational design, knowledge management and adaptation (Lavie, et al., 2010). It has also been studied in different
contexts such as strategic alliances, technology development, product innovation and senior-management teams (Lavie, et al., 2010). The definition of exploration and exploitation itself
11
has also been different over time from a very broad definition to a more narrow one (Lavie, et al., 2010; Nosella, et al., 2012). Lavie, et al. (2010) critique these changing definitions,
because outcomes from different studies cannot be compared because they use different levels and contexts of definitions for exploration and exploitation. This is also the critique and confusion for this study, because when no uniform definition is used for both constructs any definition can be given and that could change the outcome of research dramatically. When for example exploration is explained as doing things more efficient as the day before, or this explanation is used for exploitation, this could significantly change the outcome when the question is asked if an organization or individual is being ambidextrous or not. Therefore it’s
important to know what is understood when using the term exploration and exploitation.
March (1991) gives the following description of exploration and exploitation: “Exploration
includes things captured by terms such as search, variation, risk taking, experimentation, play, flexibility, discovery, innovation. Exploitation includes such things as refinement, choice, production, efficiency, selection, implementation, execution.”(p. 71). Smith & Tushman (2005) say the following: “Successfully building an innovation stream is challenging because
exploring and exploiting are contradictory to one another. Exploitative innovation is associated with efficiency, focus, convergent thinking, and reducing variance; while
exploratory innovation is associated with experimentation, flexibility, divergent thinking, and increasing variance”(p. 523). So it could be concluded that balancing both activities causes
conflict and stress as they are contradictory to each other.
To come to a definition of exploration and exploitation used in this study some additional theoretical background is needed. There seems to be some consensus about the definition for exploration but not for exploitation (Gupta, et al., 2006). A first distinction that has been made between exploration and exploitation is that exploration is more about discovery, experimentation and innovation and exploitation is more about refinement, implementation, efficiency, production and selection (Gupta, et al., 2006; He & Wong, 2004; Lavie, et al., 2010). So there seems to be more risk in exploration then in exploitation. When exploration is
12
more risky, why bother to do so? An answer to this question is given by Tushman & O’Reillly
III (1996) as they describe the tension between evolution and revolution. The pattern of evolution in an industry can be incremental or abrupt, like in the semiconductor industry for example. When sticking to evolution, or exploitation, a company could become best in class in the current technology. But as soon as this technology is overthrown by another company, not necessarily being a competitor, the results of the company sticking to the old technique can quickly erode, eventually ending up in bankruptcy. This example shows the importance of exploitation and therefore the tension between both exploitation and exploration. This is why researchers in the ambidexterity field say there needs to be a balance between the both (Gibson & Birkinshaw, 2004; He & Wong, 2004). The definition of ambidexterity and
exploration and exploitation is still made on an organizational level and there seems to be a lack of definition on an individual level. This study focuses on the individual part of
ambidexterity, so there is a need for a clear definition.
The statement of Gibson & Birkinshaw (2004) that every individual can become
ambidextrous when the right context is created raises a problem if there is no definition for exploration and exploitation on an individual level. One can image that not all people in an organization are busy thinking about new products. People in for example administration or production should be able to become ambidextrous following their statement but most definitions of exploration and exploitation focus on new product development or new technology (Gupta, et al., 2006; He & Wong, 2004). When people in administration and production think of better ways to produce or make their process more efficient this would be exploration following these definitions. This would make it impossible for those people to become ambidextrous. Therefore the definition used in these study for exploration is doing business like the day before and for exploration; thinking of better ways to do business then today. This way getting the process more lean would become an explorative action (widening current knowledge) instead of an exploitative one (deepening current knowledge). This
13
definition is needed to be able to give everybody an opportunity to become individually ambidextrous.
2.4 Individual ambidexterity
Studies done at the individual level are scarce and there is a lack of a conceptually and empirically validated framework of individual ambidexterity (Mom, et al., 2007). A reason why there is a lack of research on individual ambidexterity may be that there is a focus on
structural ambidexterity instead of contextual ambidexterity (Good & Michel, 2013), making it harder to drill deeper into the individual as it is harder, or sheer impossible, for an individual to perform two tasks at the same time.
Finding a definition for individual ambidexterity is still hard, a recent study of Good & Michel (2013) propose the following definition for individual ambidexterity: “the individual-level
cognitive ability to flexibly adapt within a dynamic context by appropriately shifting between exploration and exploitation” (p. 437). This definition gives a good impression of the tension
an individual faces being ambidextrous. One of the difficult things in becoming ambidextrous as an individual is coping with the stress of balancing between exploration and exploitation because of the conflicting nature of both activities (Smith & Tushman, 2005), so an individual needs to cope adaptively with changing circumstances, balancing between short-term
efficiency and long-term innovation and about building existing products and creating new products that cannibalize on the old products (D’Aveni, 1994 in Smith & Tushman, 2005).
A recent study of Yu, et al. (2013) explores the service-sales ambidexterity of front line employees of banking customer service centers. In this study individual level ambidexterity and branch level ambidexterity is studied and it shows the tension individuals face when two structurally separated tasks are combined in an individual. Yu, et al. (2013) studied the forces that facilitate the conversion from a service-only environment to an environment that emphasizes both service and sales. This means individuals needed to make a trade-off between giving service and when to up- or cross-sell. To be better in making this trade-off,
14
individuals need empowerment, team support and transformational leadership (Yu, et al., 2013), these factors are only contextual factors why individuals became ambidextrous but this aren’t individual factors to explain why these individuals became ambidextrous. The study also doesn’t say if every individual was able to make the switch from service only to
service and sales at the same time. In sum, no evidence is given that every individual was able to become ambidextrous and what abilities where necessary to become ambidextrous. It could even be questioned it this behavior can be seen as ambidextrous, or simply
switching between two exploitative tasks as they learned to sell next to giving service. Ambidextrous or not, this study clearly shows the switching process between two distinct tasks and the difficulty managing both at the same time.
Studies of Mom, et al. (2007) and Mom, et al. (2009) have focused on individual
ambidexterity at the managerial level. As Mom, et al. (2007) studied the exploratory and exploitative activities of managers they found evidence for managers to have high levels of exploitation and exploration activities at the same time. Put another way; they became individually ambidextrous. Although not all managers where individually ambidextrous this difference wasn’t studied as well as individual differences between ambidextrous individuals wasn’t studied. An interesting finding from this study though is the fact that at least some
managers became ambidextrous. They performed both exploitative and explorative activities at the same time, the way structural ambidexterity performs both activities at the same time by spatially separating exploitative and explorative activities. From a contextual ambidextrous point of view both exploitation and exploration has to be achieved by a single individual. This individual has to make its own decisions regarding exploration and exploitation. As it seems hard for an individual to perform both tasks at the same time, like Mom, et al. (2007) seems to argue for, it is more likely that an individual uses both exploration and exploitation in a sequential way rather than a simultaneous way (Good & Michel, 2013). This immediately raises an interesting thought; if an individual becomes ambidextrous by sequentially using exploration and exploitation, the pace of this sequential process could show the level of
15
ambidextrous behavior. Hence, when this pace is slow an individual could experience less stress switching from exploration to exploitation and vice versa. When this pace is quick an individual needs to switch quicker and could experience more stress because the time to perform one task is shortened, almost like it happens simultaneous.
So some research tried to step from the organizational point of view of ambidexterity to the individual point of view of ambidexterity. Factors or variables to influence ambidexterity are still externally focused, like team support and transformational leadership (Yu, et al., 2013), an exception is the study done by Good & Michel (2013). They studied variables that accounted for the formation of individual ambidexterity, like intelligence, individual
exploration, individual exploitation and cognitive flexibility. More focused on the individual, he or she should be able to balance and switch between both conflicting activities, cope with the uncertainty that mainly exploration gives and cope with the stress balancing between these abilities gives. According to Good & Michel (2013) individual ambidexterity includes at least the following variables for managing the exploration vs. exploitation dilemma; divergent thinking, focused attention and cognitive flexibility. One of the outcomes of the study done by Good & Michel (2013) is that individual abilities may, at least partially, explain ambidextrous behaviors. Therefore this study will use personality traits, as an extensively used method to asses a persons abilities, to find out if trait factors can predict individual ambidexterity.
To further operationalize the individual ambidexterity construct a measurement should be found to measure ambidexterity, as ambidexterity is a combination of both exploitation and exploration. Ambidexterity consists of a balance between exploitation and exploration, hence doing both activities in an equally balanced way. Although this sounds a logic way to
measure ambidexterity the strength of ambidexterity cannot be measured this way. The strength of ambidexterity means the amount of exploration or exploitation. Somebody scoring high on exploration but low on exploitation can be more ambidextrous, although not in
balance, then somebody scoring low on exploration and at the same time score low on exploitation (see also table 1). The latter one has a low strength but is well balanced. Cao, et
16
al. (2009) call both ways of calculating organizational ambidexterity; ‘the balance dimension of ambidexterity’ (BD) which is measured be extracting exploitation from exploration and ‘the combined dimension of ambidexterity’ (CD) which is measured be multiplying exploration
with exploitation. Cao, et al. (2009) applied these dimensions on the organizational form of ambidexterity. In this study this principle is assumed to be uniform and therefore applicable to individual ambidexterity as well. Although a combination of both dimensions would be best there is no accepted method for measuring both balance and strength found. For this study the combined dimension of ambidexterity seems to be the most applicable measurement. The reason for this is that the balance dimension only tells if somebody is able to balance both exploration and exploitation even if the scores are the lowest possible. To be able the measure if somebody is individually ambidextrous the relative strength is more important, although probably not in perfect balance. In table 1 both ways of measuring ambidexterity are showed and the effect for a few examples. When BD is zero it means exploration and exploitation are perfectly balanced while a negative score means more exploitation then exploration and a positive score exactly the opposite. For CD scoring 1 is weak ambidexterity while scoring 25 means strong ambidexterity.
Example Exploration Exploitation BD CD
A 1 1 0 1 B 1 5 -4 5 C 5 1 4 5 D 2 3 -1 6 E 3 2 1 6 F 5 5 0 25 G 4 5 -1 20 H 5 4 1 20
Table 1: examples of ambidexterity, both balanced (BD) and combined (CD)
In sum, individual ambidexterity is not yet studied extensively, let alone individual abilities or variables which explain individual ambidextrous behavior. Individual ambidexterity is most probably a sequential process of switching between exploitative activities and explorative activities, rather than a simultaneous process. The dimension of measurement in this study will be the combined dimension of ambidexterity. Factors that are important for an individual
17
to become ambidextrous are; divergent thinking, focused attention, cognitive flexibility and coping with stress and uncertainty.
2.5 Personality traits
As mentioned in the paragraph before individual abilities can be used trying to explain individual ambidextrous behavior. A tool often used to asses these abilities are personality traits. Personality trait constructs are studied well over time and therefore are used
extensively because of their rigid foundations (Judge, et al., 2002). Several models have been made (Goldberg, 1999) and one of the most influential and most popular models to evaluate a person’s personality is the Five Factor Model (FFM) or Big Five personality traits
(Costa Jr. & McCrae, 1992; Goldberg, 1990; Judge, et. al., 2002; Just, 2011; Vassend & Skrondal, 2011; Widiger & Trull, 1997). The FFM is a set of five traits: Openness to
Experience, Conscientiousness, Extraversion, Agreeableness and Neuroticism (Emotional Stability). These traits have been validated over a long period of time as the ‘founding’ of the FFM can be tracked back to the early 1960’s (Judge, et al., 2002). The cross-cultural validity
is also studied and findings where that the FFM is valid for different cultures and languages which strengthens the usability of the FFM in this study (McCrae & Costa, Jr., 1997). Although more traits are still debated to be part of the FFM, like Honesty or Hedonism (De Raad & Barelds, 2008), these traits don’t seem to link with the organizational characteristics in this study; discipline, stretch, support and trust and therefore aren’t considered a part of
the FFM.
Although there seems to be relative agreement about these five higher-level factors of personality traits, there seems to be no agreement about the lower-level facets making up these five personality trait factors. These lower-level facets have been developed by several scientists, like the AB5C model proposed by Hofstee, De Raad & Goldberg (1992) which has 45 bipolar dimensions. Other models are the NEO-PI-R from Costa Jr & McCrae (1992) with 30 bipolar facets, the California Psychological Inventory (CPI) with 30 to 35 bipolar facets and there is the 16PF model from Cattel with 16 factors (Goldberg, 1999). Goldberg (1999)
18
tried to develop a common item format which measures all five higher-level factors by a common subset of lower-level facets.
The Revised NEO-Personality Inventory (NEO-PI-R) is used for assessing personality traits in this study (Costa Jr & McCrae, 1992) because of its extensive us of it in literature found. In this test every trait is structured by 6 scales, these scales will be mentioned in the next paragraphs.
2.5.1 Openness to experience
The first trait in the FFM is Openness to experience, which means that a person is open to new experiences and change. People who score low on openness tend to be more
conservative and prefer familiar routines. People who score high on openness tend to have broader interests, have a preference for variety and tend to behave more flexible (Digman, 1990). The underlying 6 facets following the NEO-PI-R (Costa Jr & McCrae, 1992) are: fantasy, aesthetics, feelings, actions, ideas and values.
Especially for exploration, openness to new experiences is needed. People scoring high on openness tend to seek for new ways of working or new products to be sold because of their preference for variety and broader interests (Digman, 1990) and are therefore argued to be better in exploration. Next to exploration, ambidexterity is about balancing exploration and exploitation. Therefore an individual also needs to be flexible in balancing both. Good & Michel (2013) found in their study that cognitive flexibility is one of the factors an individual needs to become ambidextrous. Openness to new experiences positively relates to this factor as people scoring high on openness behave more flexible (Digman, 1990). Another factor important for individual ambidexterity is divergent thinking, as it is one of the
characteristics of exploration (Smith & Tushman, 2005) and individual ambidexterity (Good & Michel, 2013). Divergent thinking is also positively related to openness to experience
19
In sum people scoring high on openness to new experience are argued to score high in exploration, flexibility and divergent thinking. All positively related to individual ambidexterity. Therefore it is hypothesized that openness to experience is positively related to individual ambidexterity because it’s important for an individual to be open to new ways of working
(exploration) next to the already known way (exploitation) and to be cognitive flexible (Good & Michel, 2013).
H1: Openness to experience is positively related to individual ambidexterity.
2.5.2 Conscientiousness
Conscientiousness means that a person is eager to strive for the best and wants to do a task well. People scoring high on conscientiousness tend to be more efficient, organized and systematic in their work. People scoring low on conscientiousness tend to be less goal-oriented and less driven by success. People scoring high in conscientiousness also show high levels of persistence, impulse control and self-regulation (McCrae & John, 1992 in Bartley & Roesch, 2010) and people scoring high in conscientiousness show lower levels of stress because they seem to be better in coping, especially problem-focused coping (Bartley & Roesch, 2010; Besser & Shackelford, 2007) and task-focused coping (Matthews, et al., 2006). The underlying 6 facets following the NEO-PI-R (Costa Jr & McCrae, 1992) are: competence, order, dutifulness, achievement striving, self-discipline and deliberation.
Individual ambidexterity, the balancing between exploration and exploitation, needs a focused attention (Good & Michel, 2013). As stated above, people scoring high on
conscientiousness are more efficient, organized and systematic. They also show high levels of impulse control (McCrae & John, 1992 in Bartley & Roesch, 2010). These characteristics together can be labeled as focused attention and therefore positively relate to individual ambidexterity. Next to a focused attention, balancing between exploration and exploitation is argued to generate stress, because an individual constantly needs to choose between both activities as it is sheer impossible to perform both activities at the same time. Coping with this
20
stress is therefore argued to be an important characteristic to be able to become
ambidextrous. People scoring high in conscientiousness show lower levels of stress because they seem to be better in coping, especially problem-focused coping (Bartley & Roesch, 2010; Besser & Shackelford, 2007) and task-focused coping (Matthews, et al., 2006). Therefore conscientiousness positively contributes to the ability to become individually ambidextrous.
In sum, focused attention and coping with stress are important attributes found in conscientiousness to lead to individual ambidexterity. Therefore higher levels of
conscientiousness are hypothesized to be positively related to individual ambidexterity. The following hypothesis is formulated:
H2: Conscientiousness is positively related to individual ambidexterity.
2.5.3 Extraversion
Extraversion is about the way a person needs social stimulation and the amount a person is outside focused. The more extravert a person is, the more social stimulation it needs. The less extravert a person is, the less social stimulation it needs. A theory about arousal explains why these differences appear. Extraverts have lower arousal levels, therefore they display more sociable, spontaneous and risky behavior in order to get there arousal levels up. For introverts the opposite is applicable as they generally have higher arousal levels and therefore try to avoid extra arousal by being socially reserved, serious and controlled
(Swickert, et al., 2002). Another studied positively linked extraversion to divergent thinking, meaning a person is able to think outside the box and create new ideas (Furnham, et al., 2009). The underlying 6 facets following the NEO-PI-R (Costa Jr & McCrae, 1992) are: warmth, gregariousness, assertiveness, activity, excitement seeking and positive emotions.
As persons scoring high on extraversion tend to have larger social networks (Swickert, et al., 2002), it can be argued that these persons can use their networks to get knowledge and ideas from. Those ideas and knowledge can be used to enhance their own work
21
(exploitation) or come up with new ideas by combining information (exploration) and therefore have a positive relation to individual ambidexterity. Next to having larger social networks extraversion also relates to divergent thinking (Furnham, et al., 2009) and divergent thinking is one of the characteristics of exploration (Smith & Tushman, 2005) and individual ambidexterity (Good & Michel, 2013).
In sum, people scoring high on extraversion tend to have having larger social networks and being good in divergent thinking. Both aspects positively contribute to being ambidextrous. Therefore it is hypothesized that extraversion is positively related to individual ambidexterity.
H3: Extraversion is positively related to individual ambidexterity.
2.5.4 Agreeableness
People scoring high on agreeableness tend to have more empathic abilities, are good-natured, helpful, altruistic and show cooperative behavior towards others (Matzler, et al., 2011). The underlying 6 facets following the NEO-PI-R (Costa Jr & McCrae, 1992) are: trust, straightforwardness, altruism, compliance, modesty and tender-mindedness.
There is a lack of literature found to be able to asses a connection between agreeableness and individual ambidexterity. It can be conclude from the characteristics written above that people scoring high on agreeableness go with the flow, not directly searching for better ways of doing business or come up with new ideas. On the other hand it can be argued that those people are flexible in going the way they are told to go, although flexibility is an important aspect for individual ambidexterity this form of flexibility doesn’t seem to be the kind of
flexibility needed. For sharing knowledge people scoring high on agreeableness are open. Studies on knowledge sharing show that agreeableness is positively linked to affective commitment, which means a person is strongly attached to an organization and is highly motivated to engage in knowledge sharing (Matzler, et al., 2011). This is in line with the contextual ambidexterity factor support and trust, so for contextual ambidexterity
22
would become ambidextrous when scoring high on agreeableness. The logic behind this is that people scoring high in agreeableness are open to knowledge sharing with others, which is good, but it can be doubted if they would come up with their own ideas and dare to go left if everybody goes right and that way explore new territory. Put another way, it is highly doubtable if people scoring high in agreeableness are good enough in exploration.
In sum, people scoring high on agreeableness are expected to go with the flow and therefore lack the ability to be explorative. Therefore it is hypothesized that agreeableness is
negatively related to individual ambidexterity because of the lack of explorative abilities.
H4: Agreeableness is negatively related to individual ambidexterity.
2.5.5 Neuroticism
People scoring high on neuroticism tend to respond poorly on environmental and social stress. These people also tend to have trouble controlling urges, want gratification for their results quickly, have a low self-esteem and have social anxiety (Costa & McCrae, 1987). On the other hand people who score low on neuroticism tend to be more emotionally stable. These people tend to be more calm and respond better on environmental and social stress (Digman, 1990). The underlying 6 facets following the NEO-PI-R (Costa Jr & McCrae, 1992) are: anxiety, angry hostility, depression, self-consciousness, impulsiveness and vulnerability.
For individual ambidexterity it is important to make the tradeoff between exploration and exploitation and this balancing could give tension and stress. Coping with stress can be done by several coping strategies. Emotion-focused coping strategies are believed to be less effective than problem/task-focused coping and neuroticism relates negatively with problem/task-focused coping (Bouchard, 2003; Boyes & French, 2010). As a result individuals scoring high on neuroticism tend to react worse on stress coping situations. People scoring high in neuroticism also have a low self-esteem and social anxiety (Costa & McCrae, 1987). It is assumed that having a low self-esteem and being socially anxious
23
results in less risk taking behavior resulting in low explorative activities (March, 1991) which is needed to become ambidextrous.
In sum, scoring high on neuroticism makes an individual more risk averse, resulting in low explorative activities, and less able in coping with stress, resulting in low ability to cope with the stress generated balancing both exploitation and exploration. Therefore the following hypothesis for the effect of neuroticism on individual ambidexterity is made:
H5: Neuroticism is negatively related to individual ambidexterity.
3 Conceptual model
3.1 Moderating effect
In the conceptual model the direct effect of the five factor model is studied. Individual
ambidexterity is hardly studied yet and studies that have been done in that direction are done on managerial level and not on lower level employees (Gupta, et al., 2006; Mom, et al., 2009; Raisch & Birkinshaw, 2008). An exception is the study done by Good & Michel (2013) who studied characteristics effecting the ability to be individually ambidextrous within an
undergraduate research population. In the study of Good & Michel (2013) intelligence was one of the characteristics, explaining a part of the ability to become individually
ambidextrous. It can also be reasoned that another skill necessary for a manager is the ability to be a generalist instead of a specialist. This assumption follows from the fact that a manager needs to oversee more processes than the average employee and a manager has to take some distance from the processes to oversee the full impact of certain actions. This in combination with the research focused on managerial level individuals (Gupta, et al., 2006; Mom, et al., 2009; Raisch & Birkinshaw, 2008) could implicate that hierarchical level could be a moderating factor. The rationale behind this moderating effect is the assumption that individuals higher on hierarchical level need the ability to balance between activities to succeed in their job and also have a higher average intelligence than the average employee giving them the ability to positively or negatively change their individual traits. For example
24
an individual who has a high intelligence is able to positively change its extroversive behavior. Therefore it can be assumed that managers are able to positively strengthen the effect of their individual traits on the ability to be individually ambidextrous. It is also easier for managers to delegate work to subordinates and therefore makes it easier to balance the exploitative and explorative activities. A manager could for example more easily delegate certain exploitative activities to subordinates in order to focus on explorative activities.
To be able to find out if hierarchical level has an influence between the five factors of the FFM and individual ambidexterity the hierarchical level is added as a moderator in the conceptual model. The hierarchical level is expected to have a positive moderating effect on the relation between the five factors of the FFM and individual ambidexterity. It is expected that a higher level in the hierarchy has a positive moderating effect because of the higher average intelligence. Therefore non-managerial level employees and managerial level employees will be tested on the moderating effect resulting in the following hypotheses:
H6: Hierarchical level moderates the relationship between individual traits and individual ambidexterity, such that the relationship between individual traits and individual
25
3.2 Conceptual model
Fig.1: conceptual model
4 Method
The empirical study is done by a questionnaire. In this chapter the sample for this study, the measurement of variables and the statistical methods are explained. The questionnaire is done in Dutch as all respondents have Dutch as their native language.
4.1 Sample
The sample of this study consisted of Dutch employees working for a company in The Netherlands as a part of a large international company. The Dutch part of the company has 299 employees on the payroll at the moment the questionnaire was sent. The company employs both blue collar workers (160) and white collar workers (139). The population exists of 87,6% man and 12,4% women. These employees all received a questionnaire
electronically via a link in an e-mail which was generated by Qualtrics Survey Software, an online questionnaire tool. To follow the responses more accurately the population was divided into 3 panels. Panel 1: technicians (114), Panel 2: managers / team leaders (41) and Panel 3: others (144). The reason for doing this was that de questionnaire was sent
26
anonymously and it was expected that getting responses from technicians could become a problem and this way responses per group could be monitored. At the end 36 of 114
technicians finished the survey (31,6%) which was assessed as an acceptable response rate for this panel. These panels aren’t traceable in the data because they were only used for
sending the survey. Responses where returned anonymously in 1 response pool and are therefore untraceable to ensure anonymity.
From the 299 employees 182 persons started the questionnaire, 156 respondents finished the questionnaire (85,7%) and 127 respondents completely filled out the questionnaire, but 1 respondent was deleted because of false data giving a response rate of 42,1%. 29
respondents returned the questionnaire incomplete and where therefore excluded by list wise deletion. 85,7% of the respondents work full-time. 83,3% of the respondents were male (n=105). The respondents have different educational backgrounds. 9,5% completed
secondary school as the highest educational level, 59,5% completed education on university of applied sciences level (MBO), 27,8% completed education on university of applied
sciences (HBO) or research university and received their bachelor title and the remaining 3,2% completed an education at research university level and received their masters title. The mean age of the respondents is 42,7 years, with a SD of 10,4 years. The organizational tenure is divided as follows: 20,6% has an organizational tenure of 5 years or shorter; 23,8% 6-10 years; 20,6% 11-15 years; 16,7% 16-20 years; 8,7% 21-25 years; 5,6% 26-30 years; 0,8% 31-35 years and 3,2% 36-40 years.
4.2 Measurement of variables
The questionnaire is in Dutch but not all items used in the questionnaire are from Dutch sources. Therefore items from English studies where translated into Dutch and then
translated back into English by a third person. This method ensures the same test is done in Dutch as it is done in the English version which is validated. The complete questionnaire is added as appendix 1.
27
4.2.1 Control variables
The control variables for this study where asked in the beginning of the questionnaire to give the respondents an easy start to motivate them to complete the entire questionnaire. The control variables for this study are; gender, age, work status (full-time vs. part-time), educational level and organizational tenure.
4.2.2 Personality
To test the effect of personality on the ability to be individual ambidextrous the five factor model is used to test for personality. The personality factors from the five factor model; openness to experience, conscientiousness, extraversion, agreeableness and neuroticism are tested using a 60-item NEO Five-Factor Personality Inventory (FFPI) using 12 items from each factor from the NEO-PI-R of Costa & McCrae (1992). The FFPI is measured on a 5 point Likert-scale rating from 1 (strongly disagree) to 5 (strongly agree). Examples of items included in the questionnaire are: “I often feel inferior to others” for neuroticism, “I laugh easily” for extraversion, “Most people I know like me” for agreeableness, “I strive to be the best in everything I do” for conscientiousness and “I often try new and foreign dishes” for
openness to experience.
4.2.3 Hierarchical level
To test for hierarchical level as a moderator the respondents were asked to answer the question whether they held a managerial level position by answering ‘yes’ or ‘no’. This
question was asked at the beginning of the questionnaire to avoid having completed questionnaires without this question being answered.
4.2.4 Individual ambidexterity
Individual ambidexterity is not yet studied a lot so appropriate scales at the individual level of analysis are not available in big numbers. Mom, et al. (2007) have done research on the exploration and exploitation activities of individual managers. Mom, et al. (2007) came to the same conclusion and developed a scale for measuring both exploration activities and
28
exploitation activities on individual level. To enhance construct validity for both the scales and the items, in-depth interviews where held by Mom, et. al. (2007) with 12 managers resulting in a 11-item scales; five items for measuring exploration (Cronbach’s α = 0,86 in original study of Mom, et. al. (2007)) and six items for measuring exploitation (Cronbach’s α = 0,81 in original study of Mom, et. al. (2007)). Both Cronbach’s α are relatively high so
therefore these items are used for this study and examples of items included in the
questionnaire are: “To what extent did you, last year, engage in work related activities that
can be characterized as follows: Searching for new possibilities with respect to
products/services, processes or markets” for measuring explorative activities and “To what
extent did you, last year, engage in work related activities that can be characterized as follows: Activities of which a lot of experience has been accumulated by yourself” for
measuring exploitative activities. All items are measured on a 5 point Likert-scale rating from 1 (none) to 5 (very often).
4.3 Statistical procedure
Data were collected by an online survey. The survey started at April 4 and was closed at April 18. The software package SPSS (version 22) was used for statistical analyses. After reversed items where recoded the reliability of data has been tested and items were excluded if reliability could be significantly improved (α > .10 or if α < .700). After the
reliability check normality tests were computed for skewness and kurtosis.
Reliability statistics for the items are as follows; independent variables; neuroticism α = .799, extraversion α = .765, openness to new experiences α = .727 after excluding 2 out of 12 items (Q1 and Q8), agreeableness α = .712 and conscientiousness α = .805. For the dependent variables the reliability statistics are as follows; exploration α = .839 and exploitation α = .789.
Skewness and kurtosis tests show the following results (see table 1): skewness is normal for all items except Gender (-1,810), Employment form (2,066) and Function (-1,298). These 3
29
control variables all have only 2 choices which could give an explanation why it is positively or negatively skewed. All other items have values between -0,410 and 0,882 staying well between -1 and 1. For kurtosis 4 items are positively skewed (>1), these are: Gender (1,298), Employement form (2,304), Neuroticism (1,487) and Conscientiousness (2,407). The first 2 items are again control variables with only 2 choices. For neuroticism and conscientiousness both values have a positive kurtosis which could be explained by the fact that people are hired with a specific profile.
30
N Minimum Maximum Skewness Kurtosis
Gender 126 1 2 -1,810 1,298 Age 126 19,00 64,00 -,070 -,677 Employment form 126 1 2 2,066 2,304 Education 126 1 4 ,338 ,279 Function 126 1 2 -1,298 -,322 Organizational tenure 126 1 8 ,882 ,468 Neuroticism 126 1,08 4,17 ,547 1,487 Extraversion 126 2,50 4,50 -,082 -,332 Openness 126 1,90 4,50 ,390 ,031 Agreeableness 126 2,67 4,50 -,264 -,169 Conscientiousness 126 2,08 4,92 -,410 2,407 Exploration 126 1,00 4,40 -,050 ,156 Exploitation 126 1,83 5,00 ,068 -,338 Combined Ambidexterity 126 2,83 18,90 0,294 -0,225 Table 2: check for normality
Correlation analysis (table 3) and regression analysis (table 4 – 8)where done to test the
hypotheses. Regression analysis was done to test the hypothesized moderating effect of ‘function’, as a construct for hierarchical level, on the effect of personality traits on individual
ambidexterity (operationalized as combined ambidexterity). In total 5 regression analysis where run for all five personality traits. In the first step the standardized control variables age, employment form, education and organizational tenure where entered. In the second step the standardized variable function and 1 of 5 traits where entered. In step three the product of the standardized variables function and 1 of 5 traits where entered. The data contains no multi-collinearity issues as this was tested with variance inflation factors (VIF) and tolerance levels. All data contain VIF below 10 and tolerance levels above 0,1.
31
Items Mean SD Skewness Kurtosis 1 2 3 4 5 6 7 8 9 10 11 12
1. Gender 1 1,83 0,37 -1,810 1,298 (-) 2. Age 1 42,73 10,45 -,070 -,677 ,254** (-) 3. Employment form 1 1,14 0,35 2,066 2,304 -,487** ,008 (-) 4. Education 1 2,25 0,67 ,338 ,279 -,155 -,344** -,083 (-) 5. Function 1 1,77 0,42 -1,298 -,322 -,245** -,197* ,115 -,053 (-) 6. Organizational tenure 1 3,05 1,74 ,882 ,468 ,283** ,736** ,002 -,328** -,268** (-) 7. Neuroticism 12 2,15 0,50 ,547 1,487 ,009 ,116 ,118 -,326** ,096 ,192* (.799) 8. Extraversion 12 3,56 0,42 -,082 -,332 -,044 -,106 -,080 ,210* -,128 -,203* -,270** (.765) 9. Openness 10 3,15 0,50 ,390 ,031 -,160 ,117 ,088 ,175 ,032 -,017 -,112 ,104 (.727) 10. Agreeableness 12 3,63 0,38 -,264 -,169 -,213* ,078 ,161 ,152 -,026 ,049 -,352** ,162 -,043 (.712) 11. Conscientiousness 12 3,92 0,41 -,410 2,407 -,134 ,041 ,074 -,023 ,048 -,042 -,402** ,104 ,123 ,221* (.805) 12. Ambidexterity 11 10,01 3,26 ,294 -,225 ,046 -,126 -,115 ,253** -,162 -,082 -,117 ,212* ,176* -,158 ,013 (-) N=126. Variable reliability on diagonal
**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).
32
Individual Ambidexterity
Variable Neurotiscm
Model 1 Model 2 Model 3
Step 1 Gender 0,062 0,034 0,048 Age -0,096 -0,101 -0,115 Employment form -0,064 -0,060 -0,036 Education 0,241* 0,209* 0,205* Organizational tenure 0,051 0,016 0,022 Step 2 Neuroticism -0,019 0,000 Function -0,150 -0,166 Step 3 Neuroticism X Function -0,099 R² 0,079 0,099 0,108 ∆R² 0,002 0,008
N = 126. * regression is significant at the .05 level. ** regression is significant at the 0,01 level. Table 4. Hierarchical regression Neuroticism explaining Individual Ambidexterity
Individual Ambidexterity
Variable Extraversion
Model 1 Model 2 Model 3
Step 1 Gender 0,062 0,042 0,042 Age -0,096 -0,119 -0,119 Employment form -0,064 -0,051 -0,051 Education 0,241* 0,194* 0,194* Organizational tenure 0,051 0,055 0,055 Step 2 Extraversion 0,151 0,151 Function -0,125 -0,125 Step 3 Extraversion X Function -0,001 R² 0,079 0,120 0,120 ∆R² 0,041 0,000
N = 126. * regression is significant at the .05 level. ** regression is significant at the 0,01 level. Table 5. Hierarchical regression Extraversion explaining Individual Ambidexterity
33
Individual Ambidexterity
Variable Openness
Model 1 Model 2 Model 3
Step 1 Gender 0,062 0,062 0,054 Age -0,096 -0,164 -0,170 Employment form -0,064 -0,067 -0,069 Education 0,241* 0,173 0,165 Organizational tenure 0,051 0,039 0,048 Step 2 Openness 0,186* 0,184* Function -0,158 -0,161 Step 3 Openness X Function -0,520 R² 0,079 0,130* 0,133 ∆R² 0,051 0,003
N = 126. * regression is significant at the .05 level. ** regression is significant at the 0,01 level. Table 6. Hierarchical regression Openness explaining Individual Ambidexterity
Individual Ambidexterity
Variable Agreeableness
Model 1 Model 2 Model 3
Step 1 Gender 0,062 -0,005 0,005 Age -0,096 -0,070 -0,084 Employment form -0,064 -0,047 -0,065 Education 0,241* 0,252* 0,239* Organizational tenure 0,051 0,021 0,027 Step 2 Agreeableness -0,190* -0,173 Function -0,158 -0,159 Step 3 Agreeableness X Function 0,130 R² 0,079 0,131* 0,148 ∆R² 0,052 0,016
N = 126. * regression is significant at the .05 level. ** regression is significant at the 0,01 level. Table 7. Hierarchical regression Agreeableness explaining Individual Ambidexterity
34
Individual Ambidexterity
Variable Conscientiousness
Model 1 Model 2 Model 3
Step 1 Gender 0,062 0,039 0,024 Age -0,096 -0,105 -0,064 Employment form -0,064 -0,062 -0,076 Education 0,241* 0,216* 0,237* Organizational tenure 0,051 0,017 -0,004 Step 2 Conscientiousness 0,040 0,083 Function -0,152 -0,165 Step 3 Conscientiousness X Function -0,158 R² 0,079 0,101 0,122 ∆R² 0,022 0,022
N = 126. * regression is significant at the .05 level. ** regression is significant at the 0,01 level. Table 8. Hierarchical regression Conscientiousness explaining Individual Ambidexterity
5 Results
5.1 Correlation analysis
An overview of the descriptive statistics, correlation and reliabilities is shown in table 3. For neuroticism no correlation with individual ambidexterity is found, rejecting H5. A negative correlation is found for education (p<,01, r = -,326), extraversion (p<,01, r = -,270), agreeableness (p<,01, r = -,352), conscientiousness (p<,01, r = -,402). A positive relation with neuroticism is found with organizational tenure (p<,05, r = ,192). For extraversion a positive correlation is found with individual ambidexterity (p<,05, r = ,212), accepting H3, furthermore a positive correlation with education (p<,05, r = ,210) and a negative correlation with organizational tenure (p<,01, r = -,270) is found. Also for openness a positive correlation with individual ambidexterity is found (p<,05, r = ,176), accepting H1, furthermore no
correlation for openness are found. Agreeableness doesn’t correlate with individual
ambidexterity, rejecting H4, but it does correlate negatively with gender (p<,05, r = -,0213) and positively with conscientiousness (p<,05, r = ,221). Conscientiousness does not
35
correlate with individual ambidexterity, rejecting H2. At last individual ambidexterity positively correlates with education (p<,01, r = ,253).
5.2 Regression analysis
Table 4 to 8 show the results of the hierarchical regression analyses. For each personality trait a separate regression analyses is run to see if hierarchical level, or functional level, moderates the effect of each trait on individual ambidexterity. Also traits and control variable which doesn’t correlate with individual ambidexterity are taken into the analyses. Education
was significantly positively related to individual ambidexterity in all models, except for model 2 and 3 in the analyses for openness. Furthermore openness showed a positive relationship with individual ambidexterity for both models. Only model 2 is significant (p<,05), explaining 5,1% of the variance. Agreeableness also positively related to individual ambidexterity in model 2. Model also was significant (p<,05), explaining 5,2% of the variance. Besides these relations no significant relations are found in these regression analyses, rejecting H6.
In sum, positive relations are found for openness and agreeableness, but no moderating effect for functional level was found. Education showed positive relations in almost all models and therefore being a notable control variable, quick analyses for a moderating effect of education gave no effect either.
6 Discussion
The first hypothesis is accepted as a positive correlation is found in this study between openness to new experiences and individual ambidexterity. It is important for an individual to be open to new ways of working (exploration) next to the already known way (exploitation) and to be cognitive flexible (Good & Michel, 2013). A possible explanation for this outcome from the data is the fact that the minimum score on exploitation in this study is 1,83 meaning that every single respondent scored on exploitation. Put in other words, if every individual has exploitative capacities it is only necessary to be able to develop explorative activities to become combined individually ambidextrous. This finding, in combination with the reasoning
36
for this hypothesis that openness to new experiences mainly stimulates exploration (Digman, 1990), gives a plausible explanation why this trait is important to become combined
individually ambidextrous. Surprisingly and confusing enough this study finds no evidence for a correlation between openness and exploration but instead it does find evidence for a positive correlation between openness and exploitation (p<,5). This finding makes it
confusing to understand the micro foundations of the positive relation between openness to experiences and combined individual ambidexterity in relation with findings in earlier studies about the relationship between openness and exploration and exploitation (Digman, 1990). The correlation between openness and balanced individual ambidexterity could perhaps give more insight but when correlation is measured for balanced individual ambidexterity no significant correlation is found for openness to new experiences. The argument that openness stimulates balancing (Digman, 1990; Good & Michel, 2013) therefore cannot be found in this study and does not give more insight in the micro foundations of the positive relationship between openness and combined individual ambidexterity.
Because found micro foundations from earlier studies and results from this study are more or less the opposite it is hard to understand what explains the positive relation between
openness and combined individual ambidexterity. What can be concluded is that the combined version of individual ambidexterity, which is chosen for this study, makes that somebody scoring only a small amount of exploitation becomes ambidextrous as he or she is also capable of being explorative as well, if only for a small amount too. This immediately poses some thoughts about the construct of combined ambidexterity. This discussion will be further deepened at chapter 7 (implications). Next to this construct discussion a possible explanation for the findings in this study could be the isolated population, therefore it would be interesting to repeat this study on other populations to see if these micro foundations are uniform.
So in sum the positive relation found between openness to new experiences and combined individual ambidexterity is in line with what theory proposes but the micro foundations cannot be explained as findings from this study are opposite to what theory proposes.
37
The hypothesized positive effect of conscientiousness on combined individual ambidexterity is not found because the effect, although positive, was not significant. Focused attention (Good & Michel, 2013) and coping with stress are argued to be important attributes found in conscientiousness to lead to individual ambidexterity, especially within balancing. The focus of this trait is therefore more on the balancing side of individual ambidexterity, so to find a possible explanation for this lack of a significant effect a check for correlation with balanced individual ambidexterity was made. The results are not significant either. It was not expected that conscientiousness had no significant effect on individual ambidexterity, especially because this trait was believed to be important in coping with the stress formed by balancing both exploration and exploitation, especially problem-focused coping (Bartley & Roesch, 2010; Besser & Shackelford, 2007) and task-focused coping (Matthews, et al., 2006). A possible argument for this unexpected finding could be that the sample consisted of people scoring high on conscientiousness (mean = 3,92) with a high positive kurtosis (2,407). This gives less variation in the sample and therefore less predictive power, resulting in results which are not significant. The predicted direction of the effect was in line with the
hypothesized direction, so future research with a more diverse sample could produce a significant effect which is expected to support the hypothesized effect.
Extraversion positively relates to individual ambidexterity, so the hypothesis is accepted. This trait has the strongest effect on individual ambidexterity of the traits that influence individual ambidexterity. This effect could be explained by the fact that extravert people tend to have larger social networks (Swickert, et al., 2002), which they use to get new ideas from, and it also follows the reasoning that extraversion also relates to divergent thinking (Furnham, et al., 2009) and divergent thinking is one of the characteristics of exploration (Smith & Tushman, 2005) and individual ambidexterity (Good & Michel, 2013). Here goes the same reasoning as for openness to new experiences, but instead of the findings for the relation between openness to new experiences and exploration, extraversion does positively correlate to exploration (p<0,01) and therefore stimulates the explorative behavior. This
