[1]
Investigation of the dual character of innovation, i.e. firms’ innovativeness
and innovation impact, by the prism of organizational learning activities.
Master’s Thesis
Student:
Nikoleta Ntinopoulou……….Student’s Number10622527.
University of Amsterdam, Faculty of Economics and Business.
Supervisor:
Mr. Bernardo Silveira Barbosa Correia Lima.
[2] ABSTRACT
This study examines the relationship between organizational learning and innovation outcome. The paper identifies two distinct categories of organizational learning, i.e. exploration and exploitation, and two outcomes of innovation, i.e. firm’s innovativeness and innovation impact. By expressing the organizational learning activities through relative exploration orientation the findings suggest an inverted U-shaped relationship between exploration orientation and each of the innovation outcomes. Furthermore, by introducing as a moderator the environmental dynamism the results propose that the degree of environmental dynamism does not affect the relationship between exploration orientation and firm’s innovativeness and innovation impact. The research examines longitudinal data covering the period between 1976 and 2000. In total 431 U.S. corporations, which were in the 1974 Standard & Poor’s 500 index and belong in the wide industrial category of manufacturing, included for the empirical testing of the hypotheses. The results present also some implications for the exploration and exploitation trade-off.
[3]
TABLE OF CONTENTS
I INTRODUCTION ... 4
II LITERATURE REVIEW ... 6
1. Exploration and Exploitation ... 6
2. Innovation ... 10
3. Literature gap and research question ... 15
III THEORITICAL FRAMEWORK AND HYPOTHESES ... 18
1. Exploration orientation, innovativeness and innovation impact ... 19
2. Environmental dynamism as moderator ... 21
IV METHODOLOGY ... 24 1. Research design ... 25 2. Sample ... 26 3. Dependent Variables ... 27 4. Independent Variable ... 29 5. Moderator ... 31 6. Control Variables ... 31 V RESULTS ... 32 1. Descriptive Statistics ... 33 2. Correlations ... 36 3. Regression Analysis ... 38 VI DISCUSSION ... 47 1. Theoretical Implications ... 49 2. Additional Findings. ... 52 3. Practical Implications. ... 53
4. Limitations and avenues for future research. ... 53
VII CONCLUSION ... 54
[4]
I INTRODUCTION
The phenomenon of exploration and exploitation has been catching the interest of many
scholars who were representatives of various research backgrounds such as organizational learning, innovation and marketing (Auh, & Menguc, 2005; Jansen, Van Den Bosch, & Volberda, 2006; March 1991; Levinthal, & March 1993). The significance, that justifies this intensive focus, stands in the fact that positive organizational achievements, such as more than simple satisfactory performance outcomes, are highly correlated to the degree that corporations are focused on these distinct activities, termed as exploration and exploitation (Jansen et al., 2006; Benner, & Tushman 2002). Interestingly, except for the important results regarding firms performance directly, exploration and exploitation can also affect other organizational outcomes that trigger positive organizational responses on other aspects, such as innovation outcomes (Ahuja, & Lampert, 2001; Kim, Son, & Nerkar, 2012). The fact that innovation is characterized as one of the most important organizational processes which actually leads an organization to the path for value creation, underscores the necessity and the interest to focus and examine the relationships between these two organizational learning activities and the innovative outcome (Deeds, DeCarolis, & Coombs, 2000: Kim et al., 2012).Although in previous researches has been addressed the concept of the impact that exploration and exploitation cause on innovation outcome (Ahuja, & Lampert 2001; Katila, & Ahuja, 2002) few of them have examined the impact on dual character of innovation (Kim et al., 2012; Auh, & Menguc, 2005). Innovation is a multidimensional concept and important role express the dimension of innovativeness and the dimension of innovation impact (Danneels, 2002; Kim et al., 2012). Behind the simple focus of the generation and the offer of something new, different perspectives and deeper understanding can bring on the surface the topics of how much new is the innovation that is generated and how
[5]
big is the influence that is evoked by this innovation on the wider environment of the firm, or in simple terms on the topics of firms’ innovativeness and innovation impact (Danneels, 2002; Kim et al., 2012; Sorenson, & Stuart, 2000). The inherent tensions between exploration and exploitation and the differences that constitute these two activities, which find their roots on the fact that they both compete for important organizational resources (March, 1991), render the examination of the issue of whether exploration or exploitation have more positive results on firm’s innovativeness and innovation impact fundamental.
Additionally, major contribution of this study is the examination of how these organizational phenomena affect the dual character of innovation under different environmental circumstances. Environmental dynamism, or differently the degree that the environmental setting is changing, affect the preference of the firms towards exploration and exploitation (Jansen, et al., 2006; Kim, et al., 2012; Uotila, Maula, Keil, & Zahra, 2009). The main goal here is to identify how exploration and exploitation affect the innovation rates and the innovation impact under turbulent environment. The window in the literature for this study is the importance of exploration and exploitation (March, 1991), the importance of innovation outcome and even more, the deeper digging in the duality of the innovation outcome, by using as a moderator the environmental dynamism; a factor under which these relationships have not been examined yet (Kim et al., 2012). By examining diverse industrial settings, so different environmental contexts, and by using patent data to identify the dependent, i.e. innovation rate and innovation impact, and the independent variables this research will provide a further understanding of the above mentioned concepts by moving a step forward the already established literature.
For reaching empirical evidences regarding the above concepts, industries that are included in the wide category of manufacturing represent the core of the analysis. The number of the
[6]
observations that were active in the data analysis is 3.248 and is consisted of 431U.S. companies which were in the 1974 Standard & Poor’s 500 index. The appropriateness of the examined population is justified by the industrial and the environmental variety which can provide more dynamic and generalized findings. The paper is organized as follows. In the first section, a deep literature review which covers topics regarding exploration, exploitation and innovation go step by step in the gap that current studies have not yet examine. After describing the theoretical framework and the methodology, empirical findings regarding the sample as whole or divided in industries are presented. As last points, follow the discussion of the results, the practical implications and the future research that following researchers can use as starting point.
II LITERATURE REVIEW
In the following paragraphs are presented the core insights of the existing literature regarding
the topic. I first introduce the exploration and exploitation notions and I clarify the significant characteristics for each of these two conflicting organizational concepts. Subsequently, I introduce the innovation concept and the importance of examining and analyzing two elements which consist the dual character that innovation can express; i.e. the innovativeness of a firm and the innovation impact. Finally, all these lead to the literature gap and the research question that this study is based on.
1. Exploration and Exploitation
The notion of exploration and exploitation has been developed through different lens by many scholars in the literature of organizational learning. March’s (1991) framework, which presents exploration and exploitation as two learning processes, consisted the groundwork for the
[7]
following theorists in the literature (He, &Wong, 2004; Jansen et al., 2006; Katila, & Ahuja, 2002). According to March’s approach (Lavie, Stettner, & Tushman, 2010), exploration is defined as a set of terms such as ‘‘search, variation, risk taking, experimentation, play,
flexibility, discovery and innovation’’, whereas exploitation as ‘‘refinement, choice, production, efficiency, selection, implementation and execution’’ (March, 1991, pg. 71).In line with that,
organizations engage in exploration to generate new ideas whereas engage in exploitation to develop and improve things that are already known (Levinthal, & March 1993; March, 1991; Gupta, Smith, & Shalley, 2006). However, many researchers have approach these two terms through different angles. He and Wong (2004, pg. 483) used the terms of exploitative innovation, which is defined as ‘‘technological innovation activities aimed at improving existing
product-market domains’’ and exploratory innovation as ‘‘technological innovation aimed at entering new product-market domains’’. Beldebros, Faem, Leten and Van Looy (2010) in their study refer
to exploitative and explorative technological activities in order to analyze this organizational phenomenon. By exploitative technological activities is expressed the creation of new and different ideas which fall into the already established domains of the firm and by the explorative technological activities is described the generation of new knowledge which are not included in existing domains of the firm. Furthermore, Katila, & Ahuja (2002) attach the notion of exploration and exploitation with two learning activities named search depth and search scope. With search depth is pinpointed the focus of the firm to already established knowledge domains and by search scope the focus of the firm on new gained experimental learning. Although it can be easily noticed that there is ambiguity regarding the exact definition of exploration and exploitation the majority of the studies embrace the idea that both exploration and exploitation are coupled with different types of learning and innovation (Gupta et al., 2006). There is also a
[8]
distinction between the narrow and the broad approach of the notions. Of high importance, which also leads to more accurate results, is the narrow approach, which is focusing on the knowledge based aspect for describing exploration and exploitation (Lavie et al., 2010).
In several studies of organizational learning has been addressed the problem of balancing exploration and exploitation. There are inherent tensions between these two learning processes
which find their roots in the fact that are viewed as ‘‘competing phenomena’’ (Levinthal, & March, 1993; March, 1991; Gupta et al., 2006; He, & Wong, 2004).This means
that exploration and exploitation compete for scarce resources and organizational routines (Andriopoulos, & Lewis, 2009; March 1991; Gupta et al., 2006, He & Wong, 2004) and require different orientations, strategies and capabilities in the organization (Auh, & Mengue, 2005, Benner, & Tushman 2003).
Levinthal, & March (1993) argue that the difficulty of balancing is resulted not only because is complicated to indicate this proper balance, but also because the learning process itself lead to instabilities. Moreover because exploration and exploitation rely on different organizational capabilities, it is easier for the firms to focus and specialize in one of them than to try to be efficient in both simultaneously (Greve, 2007). This unilateral specialization can result in advantages but can also result in severe problems for the firm. An organization that focuses solely on exploitation will have the opportunity to build and reinforce its existing capabilities and thus become more effective and efficient, something that leads to positive short-term effects (Beldebros et al., 2010). On the other side, this exclusive focus can cause obsolescence and rigidity, conditions in which the firm’s long term survival is threatened due to the difficulties that the firm faces to adapt in the industrial and environmental changes (Beldebros et al., 2010, Katila,& Ahuja, 2002; Levinthal, & March, 1993; March, 1991; Uotila, et al., 2009). Excessive
[9]
focus on exploration hold advantages by making the firm keep the pace with the changes in the environment but is also possible to lead to gain losses and thus harm the financial performance of the firm. This can happen due to the endless efforts of the firm to correct possible failures that these changes evoke (Beldebros et al., 2010; Levinthal, & March, 1993; March, 1991; Uotila et al., 2009).More specifically in the article of Levinthal, & March (1993) is mentioned that excessive reliance on exploitation can lead to ‘‘success trap’’ whereas too much reliance on exploration can lead to ‘‘failure trap’’.
The reason why firms have to manage the balance and the precise combination between exploration and exploitation is that in this way it can be created the possibility of synergistic effect between these two competing phenomena, and thus firms can improve their survival and success (Auh, & Mengue, 2005; Andriopoulos, & Lewis, 2009; He, & Wong, 2004; Levinthal, & March, 1993; March, 1991). According to the studies of He, & Wong (2004) and Uotila, Maula, Keil, & Zahra (2009) firms need to find the balance in their explorative and exploitative activities as this can provide them the opportunity to achieve the best possible outcome regarding their performance. The achievement of balance and the oscillation between exploration and exploitation is expressed as the capability that organizations have, through which become capable to manage tensions and competing situations, such as alignment and adaptation or optimization and innovation (Benner, & Tushman, 2003; Gupta et al., 2006; He, &Wong, 2004; Jansen et al., 2006; Venkatraman, Lee, Iyer, 2007).
Furthermore, scholars dichotomize the ways that exploration and exploitation can be executed in organizations. These two activities of organizational learning can either be represented as
‘‘two ends of a continuum’’ or as ‘‘orthogonal activities’’ (Auh, & Mengue, 2005; Benner, &
[10]
of a continuum’’ means that there is a sequential moving from exploration to exploitation and a
balance between these two approaches is necessary for the efficient and effective operation of the organization (Auh, & Mengue, 2005; Benner, & Tushman, 2002). At the other side of the spectrum, scholars from the recent literature refer to exploration and exploitation as independent and interactive activities; i.e. “orthogonal activities” in order to express the ability of at least some organizations to pursue explorative and exploitative behaviors simultaneously (He, & Wong, 2004; Jansen et al., 2006; Katila, & Ahuja 2002; Teece, 2006; Schreyogg, & Sydow, 2010). This approach to exploration and exploitation according to the study of Lavie et al. (2010, pg. 115) evoke underestimation of the inherent trade-offs between them (Lavie et al., 2010; Levinthal, & March, 1993).
To sum up, the research on the literature on the topic of exploration and exploitation reveals the importance of the moving towards exploration and exploitation and the consequences that a one-sided focus brings on the surface for the organizations (Auh, & Mengue, 2005; Andriopoulos, & Lewis, 2009; He, & Wong, 2004; Levinthal, & March, 1993; March, 1991; Uotila et al., 2009). Firms do not set only short-term goals and the reason is that are also concerned about their long term prosperity, thus their survival. Therefore, as exploitation oriented activities play the main role for the short term success whereas exploration oriented activities for the long term success (Uotila et al., 2009) the importance to focus at least on each of them on different stages of organizational life is mandatory.
2. Innovation
The concept of innovation has been tackled by many authors, and it has been expressed extensively through a variety of approaches. The reasons for that are rooted in the continuous
[11]
changes and instabilities of the environment, in the increasing competition, as well as on the fact that the economies nowadays are mainly knowledge based. All these, make innovation a fundamental issue that organizations have to focus on (Keskin, 2006; Sorensen, & Stuart, 2000). Furthermore, innovation is considered as one of the most important organizational outcomes which leads the organization in value creation and growth, has an impact on the viability and the performance of the firms and also creates new paths for change for a whole industry (Kim, et al. 2012; Penrose, 1959; Simpson et al, 2006; Sorensen, & Stuart, 2000). A very broad definition that illustrates this concept could be that innovation is something new; it can be an improvement, an idea, a behavior, a product, a project that changes and brings new prospects in a corporation (Chen, & Taylor, 2009; Jimenez, & Valle, 2011). An important element in the notion of innovation is that it also includes the commercialization of these new ideas (Van de Ven, 1986, cited in Chen and Taylor, 2009; Garcia, & Calantone, 2002).
In the article of Abernathy and Clark (1985) innovation is categorized into four distinct dimensions, using as a criterion its competitive significance. The categories are: architectural, revolutionary, regular, and niche creation. In architectural innovation the “new element” is noticed on the design and the configuration of the current product or the process, while in the revolutionary innovation- which in current literature is also defined as a category of radical or discontinuous innovation (Benner, & Tushman, 2003; Danneels, 2002) - this new idea disrupts existing established domains and thus creates the opportunity for the firm to make its competitors obsolete (Abernathy, & Clark, 1985). In the category of regular innovation, which can be differently expressed as a form of incremental innovation, changes and improvements can be noticed on the already existing product or process. The innovation here is aiming at the improvement of the technical competences of the firm. Finally, the last group, which refers to the
[12]
innovation in niche market creation, sets as a goal the cover of the needs that are offered outside of the present market, thus new market opportunities are created (Abernathy, & Clark, 1985). According to things that already mentioned the categorization of the innovation depends on the departure from existing products, technologies and services and from existing market segments (Abernathy, & Clark, 1985; Atuahene, & Gima, 2005; Benner, & Tushman, 2003; Danneels, 2002; Jansen et al., 2006). So, in a more narrow view innovation can be divided in two different categories between which also exist additional sub categories. Thus, the division is between continuous or incremental and discontinuous or radical innovations (Atuahene, & Gima, 2005; Benner, & Tushman, 2003; Hederson, & Clark, 1990; Danneels, 2002).
Apart from the categorization of innovation in distinct dimensions, the nature of an innovative activity can be dichotomized in different settings. This is determined by the environment that a firm operates and the competition that a firm faces (Utterback, & Abernathy, 1975). In the dominant design theory, a concept that many researchers adopted in order to suggest the right innovative dimension that a firm must engage, proposes two core distinct phases in the technology life cycle; the pre and post dominant design (Christensen, 1997). In the pre- dominant design phase, organizations are in the most creative part, and the product innovation, which usually is more radical, is in the heart of this stage (Christensen, 1997; Utterback, & Suarez, 1993, cited in Klepper, & Simons 2005). In the post- dominant design phase, the product is standardized and thus organizations try to incorporate process innovation, as the focus now turns to the creation of profits and to the offer of more effective products to the customers (Christensen, 1997; Utterback, & Suarez, 1993, cited in Klepper, & Simons, 2005).
From all the points that have been mentioned so far, becomes clear the multi-dimensionality in the notion of innovation and the necessity of organizational efforts to push the firm in
[13]
innovative activities in order to encourage creativity and growth. Besides, the significance of innovation as a whole process, other pieces that compose the whole idea of the term bring substantial insights, which further elaborate and put a step forward this concept. Apart from the introduction of something new, also important role has ‘how much new’ the innovation is (Calantone, Garcia, & Droge, 2003; Sorensen, & Stuart, 2000). The degree and the frequency that an organization innovates or the degree that disrupts current technologies can describe in general terms how innovative is (Garcia, & Calantone, 2002; Calantone, et al., 2003; Sorensen, & Stuart, 2000). “Highly innovative” firms are seen as having high degree of newness, bring something completely different from what is already established, whereas “low innovative firms” are on the other side of the extreme (Garcia, & Calantone, 2002). In accordance with Garcia, & Calantone (2002), firms’ innovativeness lays in the culture of the organization and depends on how open is the firm in new ideas and experimentation. The departure from the existing routines, organizational competences and strategies signal higher innovative activity and accelerate the possibilities for the organization to be away from the risk of obsolescence and inflexibility (Danneels, 2002; Lee, & Tsai, 2005; Sorensen, & Stuart, 2000). Obsolescence and inflexibility can lead to inertia which contributes to the decrease of the organizational efficiency and effectiveness (Lee, & Tsai, 2005; Sorensen, & Stuart, 2000). This remarkable note underscores the substantial role that a high degree of innovativeness accounts for a firm (Lee, & Tsai, 2005; Sorensen, & Stuart, 2000). Emphasis should be also made to the fact that Firms’ Innovativeness has a different meaning from product innovativeness, as the introduction of new products in the market is not a guarantor of generation of a completely new idea (Garcia, & Calantone, 2002). Many innovative products can be a result of successful imitating activities, something which is not included in the term of Firms’ Innovativeness (Danneels, 2002; Garcia, & Calantone, 2002).
[14]
The existence and the introduction by the firm of something completely new, accelerate the probability of firms’ survival especially in hostile dynamic environments (Danneels, 2002).
Apart from the degree of innovativeness that a firm has, another important aspect in the concept of innovation is the impact that an innovation has (Kim et al., 2012). As innovation impact can be defined the level that firm’s knowledge is captivated and recognized by other firms and is absorbed in future developments (Kim et al., 2012; Rosenkopf, & Nerkar, 2001). The impact of an innovation does not only affect the industry in which a firm operates, as sometimes the influential character of an innovation can go beyond the boundaries of a specific industrial setting (Kim et al., 2012; Rosenkopf, & Nerkar, 2001). By this becomes clear that innovation is a phenomenon that influences a whole community, contributes to the evolution of a wider environment and suggests ideas for further developments in various industrial settings (Gittelman, & Kogurt, 2003; Kim et al. 2012). New technological trajectories can ground the basis for future innovations, so movements from new generated innovation to future creative ideas are noticed, and this reveals the role of innovation as driving force for the industrial evolution (Ahuja, & Lampert, 2001; Katila, & Ahuja, 2002; Damanpour, & Gopalakrishnan, 1998; Damanpour, & Wischnevsky, 2006). In this way, innovation can be seen as a “generation process” because on the one hand firms create outcomes with an expanded impact and on the other hand it can be seen as an “adoption process” as other firms assimilate these outcomes (Damanpour, & Gopalakrishnan, 1998).
Admittedly, the degree of innovativeness as well as the influential power of an innovation depend on the organizational learning activities that firms occupy (Kim et al, 2012; Rosenkopf, & Nerkar, 2001; Sorensen, & Stuart, 2000). In March’s (1991) seminal work is sharply expressed that organizational learning, both explorative and exploitative, determines the
[15]
innovation outcome and that the key for success is to strike a balance between these two activities. Exploitative “localized learning” and exploratory “learning by experimentation” contribute to the degree that a firm innovates and also to the impact that these innovations will have in the environment (Kim et al., 2012, Sorensen, & Stuart, 2000). By focusing only on one aspect, such as on path- dependent knowledge and on routines that have been developed during the years, organizations face difficulties to adapt to fast changing environments and so to enhance their innovativeness and the impact of the innovations, two notions that from now on for the needs of this study will formulate the term of dual character that describes innovation (Kim et al., 2012; Sorensen, & Stuart, 2000; Rosenkopf, & Nerkar, 2001). The study of Sorenson, & Stuart (2000) which conducted in semiconductor and biotechnology industries, indicates that high reliance on prior developments contribute to improvements but these improvements are less significant. This is something that according to the empirical evidence of the research accelerates the tendency of the organizations towards inertia (Sorenson, & Stuart, 2000). Hence, the importance of the dual character of innovation, i.e. Firms’ Innovativeness and Innovation Impact, is grounded in the fact that gives to the firm the opportunity for growth, flexibility and continuous adaptation. These elements significantly influence firm’s survival and prosperity (Chan, & Taylor, 2009; Kim et al., 2012).
3. Literature gap and research question
The phenomenon of exploration and exploitation has been examined by many scholars in the literature by employing different measurements, contexts, industries, dependent and independent variables, moderators and mediators, and so on. The variety of the perspectives that it has been expressed, signals the ambiguity of the topic as well as the concerns that can be raised for firms regarding this issue. The majority of the researches address the effect that exploration and
[16]
exploitation have on the performance of the firms (Auh, & Mengue, 2005; Katila, & Ahuja, 2002; He, & Wong, 2004; Uotila, et al., 2009) whereas others present the effect on the innovative outcome, such as the research by Kim et al., (2012) or Katila, & Ahuja, 2002. Although the amount of studies has been increased through the years the results are quite vague. This is because some of the studies highlight the importance of a well- balanced position while others do not take it into consideration. An example is that in the studies of Katila, & Ahuja, (2002); Jansen et al., (2006) and He, & Wong ,(2004) exploration and exploitation have been modeled as orthogonal activities with a positive interaction while in other studies have been modeled as continuum (Benner, & Tushman,2002). According to this, there is a necessity to examine the concept of exploration and exploitation, separately and try to define and reach conclusions about whether and how each of these activities affects the organizational operation of firms. It is important to build a more clear view about these concepts and also to contribute and comment to the phenomenon of balancing (Beldebros, et al., 2010; Gupta et al., Uotila, et al., 2009; Venkatraman, et al., 2007).
The scope that will be used to elaborate on this topic comes from the concept of innovation, which undoubtedly has catch the interest intensively on the literature, as it does not only impact on the survival of the firms but also on wider social and economic changes (Sorensen, & Stuart, 2000). It has been proved that these two organizational behaviors guide the dual outcome that innovation has, as it has been defined on this paper (Katila, & Ahuja, 2002; Kim, et al., 2012). As it has been mentioned above (subsection 2, Innovation) learning that has been gained through experience and learning that has been gained through experimental behavior are mandatory for firms’ innovativeness and for the impact that innovation activities will have (Kim, et al., 2012). Although explorative learning affects more the innovativeness of the firm and the innovation
[17]
impact, is risky for an organization to ignore and to not facilitate exploitative behavior because, as it has already pinpointed, unilateral focus can cause serious damages for the entire corporation (Beldebros et al., 2010, He, & Wong, 2004; Levinthal, & March, 1993; March, 1991; Katila, & Ahuja, 2002; Uotila et al., 2009).
An additional element that will be incorporated on the study is the environment that firms operate. By using the definition environment, there is a clear reference on the wider environment that the firm is surrounded by and actually affects the evolution of the industry and thus influences every component that constitutes it (Jansen et al., 2006). According to Zahra (1996) organizational environments differ significantly in their degree of dynamism. The two extremes include environments that alter in a very low degree and environments that continuously change (Das, & Beard, 1984; Jansen et al. 2006; Zahra, 1996). Frequent technological change might make organizational resources obsolete during a very short period of time (Atuahene, & Gima, 2005; Danneels, 2002; Uotila et al., 2009; Zahra, 1996). Thus, it can be said that the degree of environmental dynamism affects the tendency of the firms towards exploration or exploitation, as in a rapid changing environment firms prefer to invest more on exploration in order to adapt and overcome external changes (Lavie, et al., 2010, Jansen, et al., 2006). According to the various points that this study presented so far, becomes obvious the existence of a causal chain with a starting point the dynamism of the environment, which affects the focus of organizations either to exploration or to exploitation something that is correlated to innovation outcomes.
All these lead us to a clear path for research. It is still a very promising area the examination of the concepts of exploration and exploitation and more specific the impact of them on two innovative outcomes, namely firm’s innovativeness and innovation impact. The conceptual model of the research will be extended by examining how this initially direct effect can be
[18]
affected and on which direction by another characteristic, the environmental dynamism. Hence the role of the environmental dynamism represents the moderator on this paper. Consequently, the research question that is formulated is the following:
“How exploration and exploitation affect firm’s innovativeness and innovation impact and
how this relationship is affected by the degree of the dynamism of the environment?”
III THEORITICAL FRAMEWORK AND HYPOTHESES
This study theorizes that two oppositional elements, exploration and exploitation, compose the phenomenon of organizational learning in a corporation (March, 1991; Levinthal, & March, 1993). Furthermore it has been argued that the appropriate balancing between these two competing activities contribute to a more effective and efficient functioning for organizations (Auh, & Mengue, 2005; Andriopoulos, & Lewis, 2009; He, & Wong, 2004). There is also strong evidence that the explorative and exploitative activity also affects the innovation activity that firms engage in (March, 1991; Kim et al, 2012; Rosenkopf, & Nerkar, 2001; Sorensen, & Stuart, 2000). Furthermore, is presented the dual outcome of innovation activity, named firms innovativeness and innovation impact (Kim et al., 2012; Sorensen, & Stuart, 2000). Firms’ innovativeness affects the organizations survival- especially in environments with high degree of dynamism- where as the innovation impact depicts the importance of an innovation which also can be affected by the dynamism of the environment that a firm operates (Kim et al., 2012; Rosenkopf, & Nerkar, 2001; Sorensen, & Stuart, 2000).Thus the dynamism of the environment will influence the relationship among exploration, exploitation and the innovativeness of a firm and the innovation impact.
[19]
In this section, will be presented the theoretical framework as well as the hypotheses and the model which this academic paper has its basis.
1. Exploration orientation, innovativeness and innovation impact
At this point it is important to be mentioned that in the following part of the study the notions of exploration and exploitation will be expressed through one term which is relative exploration
orientation (Sidhu, Volberda, & Commandeur, 2004; Uotila, et al., 2009). The high degree of the
relative exploration orientation indicates that the corporation is focused more on explorative activity whereas the lower degree shows that the corporation is mainly conducts exploitative activity (Sidhu, et al., 2004; Uotila et al., 2009).
By having high explorative orientation firms can avoid the risk of being trapped in improvements that will not enhance their degree of adaptation and that will become liabilities as their environment changes (Hashman, 2005; Uotila et al., 2009). Firms with high explorative orientation present creative insights which are developed through experimentation and discovery and which could not be developed through already established ideas (Cho, & Pucik, 2005; Jansen et al., 2006; Katila, & Ahuja, 2002; He, & Wong, 2004; March, 1991).Thus, these efforts for exploration will enhance their innovation activity and so their innovativeness (Cho, & Pucik, 2005; Danneels, 2002; Lee, & Tsai, 2005; Sorensen, & Stuart, 2000). This is obviously an advantageous aspect, but over-exploration may cause drawbacks which can diminish firms’ innovativeness (Katila & Ahuja, 2002). Firms might be stuck on a vicious cycle and continuously innovate (Beldebros et al., 2010; Levinthal, & March, 1993; March, 1991; Uotila et al., 2009) without achieving their primary goal which is the implementation and the commercialization of these new ideas in order to introduce new products or services to cover
[20]
customer needs (Haushman, 2005; Jin, Hewitt-Dundas, & Thompson, 2011).Firms departure from that essential goal create a disadvantageous situation that hurts their innovativeness because corporations have to cover excessive costs, without actually having enough returns to cover these costs (Beldebros et al., 2010; Levinthal, & March, 1993; March, 1991; Uotila et al., 2009). Furthermore, corporations may face difficulties to align these new ideas with already established ideas (Katila, & Ahuja, 2002). The process of integration becomes more complex and costly when the difference between existing and new ideas is sharper (Katila, & Ahuja, 2002). By having presented these two contradictory aspects it is clear that an organization must be alarmed and able to recognize that after a point the preference to conduct explorative activity can actually diminish its innovativeness while it is possible to cause various types of damages. According to this the following hypothesis is proposed:
H1: ‘The relative exploration orientation of the corporation exhibits a curvilinear (inverted
U-shaped) relationship to the innovativeness of the firm.’
Furthermore, the time, the effort and the resources that organizations sacrifice for exploration also affect the impact that these innovations will have on their close or distant industrial domains. As already mentioned, corporations with high orientation in exploration sacrifice their effectiveness on the short term (March, 1991) as they do not offer an immediate output, something that incur firms with costs (Beldebros et al., 2010; Haushman, 2005; Jin, et al., 2011; Levinthal, & March, 1993; March, 1991; Uotila et al., 2009). On the contrary, for the long term high exploration orientation may contribute to the production of valuable technologies and methods which can be utilized and influence positively other future innovations (Ahuja, &
[21]
Lampert, 2001; Kim et al., 2012). The empirical results from the study of Kim et al., (2012) also support this view. By the research that conducted for firms in the global pharmaceutical industry is proposed and supported that the higher the scientific intensity, which reflects the tendency of the corporation to focus on exploration, the higher the impact that the innovation will have. Accordingly it is expected that corporations which spend and invest on exploration accelerate the possibility to generate instrumental innovations that will affect various industrial environments in the future. Hence the following hypothesis is formulated:
H2: ‘The relative exploration orientation of the corporation does not exhibit a curvilinear
(inverted U-shaped) relationship with the impact that an innovation may have’.
2. Environmental dynamism as moderator
Unpredictable changes in the external environment that firms operate are defined as environmental dynamism (Des, & Beard, 1984). These turbulences necessitate direct adjustment of the firm in order to meet new customer preferences, variations in technologies or changes in supply resources (Lavie, et al., 2010, Jansen, et al., 2006). The reason behind this is that the already established skills and processes in products and services become obsolete when these changes occur (Jansen et al., 2006; Sorensen, & Stuart, 2000; Yang, & Li, 2011). Hence, this requires orientation of the organization toward exploration in order to keep up with the instabilities and survive (Lant, & Mezias, 1992; cited in Lavie et al., 2010; March, 1991). Furthermore, the degree that firms will invest to exploration depends on the frequency of these environmental swifts (Kim, & Rhee, 2009). In a relatively low or mild dynamic environment, corporations can rely on existing knowledge and processes as the pace of the changes do not require radical modifications (Benner, 2009; Yang, & Li, 2011). On the other side of the blade,
[22]
the higher the degree of turbulence in the environment the more increased the need for the corporation to garner organizational knowledge through exploration (Benner, 2009; Kim, & Rhee, 2009). Due to the difficulty of firms to refine current experiences, as they devalued in such a dynamic environment, exploration-oriented activities are the key to achieve adaptive responses (Kim, & Rhee, 2009; Yang, & Li, 2011). Consequently, in hostile dynamic environments organizations become more innovative to avoid rigidity as they are structured more towards exploration. This is something that obviously affects positively their innovativeness. Accordingly, is proposed that environmental dynamism amplifies the positive relationship between relative exploration orientation and firm’s innovativeness.
H3: ‘The more dynamic is the industrial environment that a firm operates the more positively
will be moderated the relationship between relative exploration orientation and firm’s innovativeness’.
Apart from this resulting positive relationship, dynamic environmental conditions also affect
the impact of firms’ innovative activity (Damanpour, & Gropalakrishnan, 1998). In a turbulent environment the character of organizations is twofold (Damanpour, & Schneider, 2006). On the one hand, firms struggle to survive by focusing more on exploration and so generate innovations, as it is already mentioned (Damanpour, & Gropalakrishnan, 1998). On the other hand organizations try to adjust to these radical changes by adopting innovations that are introduced by other corporations (Damanpour, & Gropalakrishnan, 1998; Damanpour, & Schneider, 2006; Naranjo, 2009). Thus, firms can have a role of innovation generating but also a role of innovation-adopting organizations (Damanpour, & Gropalakrishnan, 1998; Damanpour, & Wischnevsky, 2006). Hence, firms that innovate through exploration can affect a whole industry especially when the environment is characterized by instabilities (Damanpour, & Schneider,
[23]
2006). According to the study of Damanpour, & Gropalakrishnan, (1998) about structure-innovation relationship under different environmental conditions, firms’ adoption of structure-innovation in unstable environments is higher, regardless the predictability or the unpredictability of a change. Also Naranjo (2009) in his paper regarding environmental and organizational factors that affect the approval of innovations in public sector firms, is proved that firms’ adoption of innovations is positively related to environmental uncertainty and market concentration; two elements which constitute environmental dynamism. For the needs of this paper the findings from these studies can be expressed through an alternative aspect. From the perspective of innovation impact, according to these works can be considered that organizations innovative activity has higher impact in other corporations when the environment is dynamic. The alignment to environmental instabilities is mandatory for firms in order to sustain their participation in industries. Hence it is proposed:
H4: ‘The more dynamic the environment that a firm operates the more positively will be
moderated the relationship between relative exploration orientation and the impact than an innovation has.’
Based on the four hypotheses described in the theoretical framework section, the following conceptual model (Figure 1) is presented:
[24]
Figure 1-Conceptual Model
IV METHODOLOGY
In the following chapter are explained in details the research approach and analysis of this study. First, it is explained why the use of secondary data was the identical choice as data source. Then the section continues with a description regarding the sample characteristics. Finally, an in depth operationalization of the variables is outlined, which also includes the description of the measurement instruments. Relative Exploration Orientation Exploration Exploitation Environmental Dynamism Dual Outcome Firms’ Innovativeness Innovation Impact Control Variables Unit Size R&D intensity Tobin’s Q ROA
[25]
1. Research design
The research technique that was used on this paper is based on secondary data. The already
existing dataset was constructed by a combination of data that were created for the purposes of a specific prior work and of data that were already available on COMPUSTAT. The matching between these two databases generated the dataset that is also used on this academic paper. The nature of the data is quantitative as all of the variables are interpreted by numbers (Sauneders, & Lewis, 2012). Furthermore, the data constitute a form of time series secondary data, as they cover an extended time period in which same companies represent different observations through the years. This allows having an extended view regarding the behavior of the companies in different periods of time, something that is very important for this specific analysis as the presence of longitudinal data contributes to more accurate results (Kim et al., 2012; Sauneders, & Lewis, 2012). Regarding the suitability of the data on this specific analysis, is important to be mentioned that all the data were collected for a prior study that had as main topic area the explorative and exploitative organizational learning activities as well as the innovative actions that firms engage in. Thus, it is obvious that the research interest and the original purpose of these two papers are very close, something that accelerates the fact that this database is adequate for this piece of work (Sauneders, & Lewis, 2012). Additionally, the measurements that were used by the prior study for the calculations of variables that represent the core of this work are in match with the calculations that this paper intended to use. This is another reason that also reinforces the relevance of the data for this new study (Sauneders, & Lewis, 2012). According to the argumentation that is made, the use of this database is an optimal and reliable choice for this type of research because of the large number of observations and the fact that these data have been calculated and used to test similar relationships in the same area of interest.
[26]
2. Sample
The sample of this empirical research covers the period between 1976 and 2000. The number
of the observations that were active in the data analysis is 3.248 and is consisted of 431 U.S. companies which were in the 1974 Standard & Poor’s 500 index. Three industrial settings that differ on the degree of environmental dynamism are included in the population. These distinct industrial domains are separated by a two-digits sic number (28, 35 and 36). On this way each company of the sample is classified according to a specific industrial identity. The industries that are encompassed belong in the wide category of Manufacturing and are respectively divided to (28) Chemical & Allied Products, (35) Industrial Machinery & Equipment and (36) Electronic & Other Electric Equipment. Approximately the 39% of the observations belong to the industry with sic code 28, the 30% to the second industrial category, with sic number 35 and the last 31 % of the companies are part of the 36 sic categorization. The existence of various industrial environments in the sample is significant as it puts a step forward the analysis of how the relationship among the independent and the dependent variables is moderated by the environmental dynamism. Apart from industrial variety, diversity can be also noticed in the size of the companies which is expressed by the number of employees. With a mean of 33.000 employees, the range of the number of employees has minimum and maximum value 0 and 341.000 employees respectively. The size (58.522 observations) and the time frame (from 1974 until 2008) of the original database allowed cutting down the population in a smaller but still representative sample. The selected case that was used in the whole analysis had the following limitations: match=1, patent_ count>=15, year>= 1976 and year<=2000. The limitation, match=1, was active in order to ensure that will be used complete data with narrow range of missing values. This indicates a match in the information that was included in the database of
[27]
COMPUSTAT and the initial constructed database. The restrictive choice of the years contributed to the avoidance of truncation problems, which are caused from missing observations in patent measurements due to the fact that patents from recent years might have not yet been granted (Hall, Jaffe, & Trajtenberg, 2001). The restrictions of the number of patents to more than 15, was used in order to avoid problematic results as the number of patents was used in the construction of important measurements of the study. A characteristic example is patent values which equal to zero. Obviously these values could undermine the calculation of variables with main role in the analysis.
3. Dependent Variables
Both dependent variables incorporate measurements which have on their core patent data. The employ of patent data have been widely accepted and have been adapted by many scholars, as they are considered one of the most valid measurements for innovation (Buderi, 1999; Katila, & Ahuja, 2002; Kim et al., 2012). The ease of collecting data of this nature, as well as their availability makes them one of the most appropriate measurements (Hall, et al., 2001). Additionally, the fact that patent data cover a huge time range put them in the first choice for studies that depend on longitudinal data (Hall, et al., 2001; Katila, & Ahuja, 2002; Kim et al., 2012). On the contrary, it is widely acknowledged that many limitations arise regarding the use patent data (Hall, et al., 2001). An example is the existence of truncation issues or the fact that many inventors prefer other mechanisms to protect their innovations, such as secrecy, instead of patents (Hall et al., 2001). For this study this type of issues is minimized as the database is delimited in order to avoid truncation problems and the majority of the firms which belong on the already mentioned industrial environments are turning to patenting in order to protect their innovations (Hall, et al., 2001; Kim et al., 2012).
[28]
Innovativeness
Firm’s Innovativeness is one of the dependent variables of this study and describes how innovative is a corporation (Danneels, 2002; Lee, & Tsai, 2005; Sorensen, & Stuart, 2000). For the proper mathematical interpretation of this notion this study incorporates the term of
originality that also Hall et al., (2001) introduce in their work. Originality is presented through
patent data, where in cases that previous patents are part of a narrow technological set the originality score is low whereas in cases that previous patents are part of a broader technological set then originality scores are higher (Hall, et al, 2001). This in the current study is expressed through low and high innovative behavior, respectively. The mathematical approach that used is the following:
Originality i =1 – Σni j s2 ij
The S ij parameter express the percentage of citations received by patent i that is a component of patent class j, out of ni patent classes (Hall, et al, 2001). In other words, the variable of Originality measures how distant is the introduction of an innovation from the current standards (Hall, et al., 2001; Trajtenberg, Henderson, & Haffe 1997). This is in accordance with the characterization of a firm as innovative, as also this term indicates the departure from existing practices and the creation of new and alternative methods (Danneels, 2002). This match in the terminology between these two notions makes even more appropriate the usage of originality as a mean to measure innovativeness.
Innovation Impact
Innovation impact, has also a citation based measurement which is computed in the same way as Originality, i.e.
[29]
Generality i =1 – Σni j s2 ij, (Hall, et al., 2001) Citations per patent indicate the impact that an innovation has, thus in other words according to the study of Hall, et al. 2001, how generalized can be the innovation (Kim, et al 2012). More analytically, if a patent is cited by patents of firms that belong in a wide range of industrial settings the generality score will be higher, or the impact of the innovation is high (Hall, et al., 2001; Kim et al., 2012).Conversely, if a patent is cited by subsequent patents from firms that do not express intense variation in their industrial background then the impact of the innovation is considered low (Hall, et al., 2001; Kim et al., 2012). Hence, Generality provides the extent that technical advances are adopted by different technological fields, rather than being absorbed in just few of them. As such, the notion represents the wider acceptance of the usefulness that a new method provides (Hall et al., 2001; Trajtenberg, et al., 1997). This mirroring the term innovation impact, which includes the ratio of the citation that the patents of a firm are cited by other organizations (Kim et al., 2012). Also in this case, the fit in between these two concepts makes Generality the appropriate variable to calculate the impact that innovations have.
4. Independent Variable
Relative Exploration Orientation
The relative amount of exploration versus exploitation is the main independent variable of this research and it was calculated by a mathematical formula after the construction and the measurement of two separated independent variables; i.e. exploration and exploitation. The measurements for exploration and exploitation are following the logic of the two independent variables that Katila, & Ahuja, (2002) present in their study, termed search depth and search scope. For computing them patent data were incorporated, which according to many studies
[30]
represent a valid indicator for expressing the search activity that firms adopt (Hall et al., 2001; Katila, & Ahuja, 2002; Rosenkopf, & Nerkar, 2001). As highlighted above there are many advantages and disadvantages regarding the use of patent data. For this study patents are an appropriate mechanism also for the measurement of the independent variable, as the focus is specifically on three diverse industrial domains and differences in the patenting activity will mainly pinpoint the variety and the heterogeneity among these industries.
The variable of exploitation, or search depth, describes the focus of firms in traditional ways and paths that have been learned from the past (March, 1991; Katila, & Ahuja, 2002; Uotila, et al., 2009). This can be interpreted through patenting activity as “the number of times a firm
repeatedly used the citations in the patents it applied for” (Katila, & Ahuja, 2002, pg. 1187).
The formula that was used in the study of Katila, & Ahuja (2002) and was also used in this study is the following:
∑
The calculations of the variable exploitation are based on the average number of times that each citation in a base year t-1 was repeatedly used during a time period of the past five (5) years (Katila, & Ahuja, 2002, pg. 1187).
The variable of exploration, which interpretation is correlated to the search scope notion, expresses the tendency of the firms to search for and adapt to new methods and skills (March, 1991; Katila, & Ahuja, 2002; Uotila, et al., 2009). The function that was used for the calculations of this variable is (Katila, & Ahuja, 2002):
[31]
According to that mathematical expression exploration “was the proportion of previously unused
citations (new citationsit-1) in a firm’s focal year’s list of citations that could not be found in the
previous five (5) year’s list of patents and citations by that firm” (Katila, & Ahuja, 2002).
After, computing these two variables the creation of relative exploration orientation, which is the main independent variable of this study, was generated by the division of exploration by the sum of exploration and exploitation for each company annually (Uotila, et al., 2009). Thus the formula that adopted is:
REO =
5. Moderator
The moderator of this study is the dynamism of the environment. The research will focus separately on how the environmental dynamism affects the relationship between relative exploration orientation and firm’s innovativeness and the relationship between relative exploration orientation and innovation impact. Industrial R&D intensity was the element that was incorporated to operationalize the notion of environmental dynamism (Uotila et al., 2009). The measurement is in match with the measurement that Uotila et al. (2009) purposed on their study. Industrial R&D intensity was calculated “as the logarithm of the industry’s total R&D
expense divided by the total amount of industry sales” (Uotila et al., 2009, pg. 225-226). This
variable is expressed on the database as environmental dynamism.
6. Control Variables
In the empirical analysis, in order to control for confounding effects, are also included the
[32]
Firms’ behavior towards exploration varies according to their size (Jansen, et al. 2006). To control for firms size the variable of Company’s size used and was measured as a logarithm of the company’s number of employees (Uotila et al., 2009). Furthermore, the focus on exploration also depends on Company’s R&D intensity (Uotila, et al., 2009). This control variable was calculated by dividing the R&D expenses by sales and was treated to be zero in cases that the value was missing (Uotila et al., 2009). As well, the performance of the firms can also navigate their actions in organizational learning activities (Jansen, et al., 2006; Katila, & Ahuja, 2002). Returns on assets (ROA), a control variable that express firm’s financial performance used for this purpose, and it was calculated by the total yearly net income of a firms divided by its total assets (Katila, & Ahuja, 2002). An alternative, market based performance measure that also used to control the study, is the Tobin’s Q variable. Tobin’s Q was constructed by the ratio of the market value of assets divided by the replacement value of the assets (Uotila et al., 2009).
V RESULTS
The following section describes the results of this research. First an overview of the variables is included, by applying descriptive statistics in all the main variables of the study. Then a correlation analysis is carried out and all the correlations among all the variables are reported. Hierarchical multiple regressions were performed and are presented further in the section as a mechanism to test the hypotheses that were formulated in the theoretical framework of this paper.
[33]
1. Descriptive Statistics
All of the main variables (the dependents, the independent and the moderator) that are included on this analysis are continuous. Since all of them are interpreted through continuous numbers, descriptive statistics was used to obtain summary statistics, such as mean, standard deviation, minimum and maximum values. The analysis initially describes the characteristics of each variable for the whole sample (by taking into consideration the following limitation: match=1 & year>=1976 & year<=2000 & patent_ count>=15) and then is divided by having as main criterion the industrial identity of the firms (in the sample this is expressed by the two-digit sic number).
For the dependent variable, Firm’s Innovativeness, the values are ranging between .31 and 20.04, with a mean 1.86 and a standard deviation 1.60, when in the sample all of the three industrial domains are selected. In case that is incorporated the restriction in the sample for a specific industry, such as sic=28, then the minimum and the maximum values are .45 and 17.55 respectively, with a mean 1.89 and a standard deviation 1.54. In general, for all the industries the mean does not express sharp differences. The results are presented on Table 1:
[34]
Table 1. Descriptive Statistics_ Firm’s Innovativeness.
For the second dependent variable of the study, i.e. Innovation Impact, the observations, in the entire population of the sample, have minimum value .000 whereas the maximum value reaches .753, with a mean .40 and a standard deviation .11. Also, in this case the differences in the statistical data among distinct industrial zones are not intense as the minimum and the maximum values, as well as the mean and the standard deviation demonstrate very low variances. On the table 2, are included the descriptive statistics for Innovation Impact.
Table 2. Descriptive Statistics_ Innovation Impact.
Industry N Min Max Mean St. Dev
All 3248 0.00 0.75 0.40 0.11
28 1254 0.00 0.75 0.39 0.12
35 973 0.08 0.72 0.42 0.11
36 1021 0.00 0.73 0.41 0.11
The Relative Exploration Orientation, which is the only independent variable of this study,
has low degree of variety among distinct industrial domains. The range of the values is between .02 and 1.00 for all the three industries, and the mean value and the standard deviation are .63 and .24 respectively. The low scores in standard deviation show that the mean is a valid
Industry N Min Max Mean St. Dev
All 3248 0.31 20.04 1.86 1.60
28 1254 0.45 17.55 1.89 1.54
35 973 0.43 15.73 1.95 1.74
[35]
representation of the population (Field, 2005). On table 3, a summary of the descriptive statistics is presented.
Table 3. Descriptive Statistics_ Relative Exploration orientation.
Industry N Min Max Mean St. Dev
All 3248 0.02 1.00 0.63 0.24
28 1254 0.02 1.00 0.54 0.22
35 973 0.06 1.00 0.67 0.23
36 1021 0.04 1.00 0.71 0.22
The last main variable of this study is the moderator that actually affects the relationship
between the independent and each of the dependent variables. By viewing the descriptive statistics only for this variable the range of the values is between .00 and .11, while separately on each of the industrial domains the lowest value is the same, thus is equal to zero. Difference in the value of the mean is only noticed for the industry with sic number 28, whereas the values of standard deviation for all the different scenarios are close to zero.
Table 4. Descriptive Statistics_ Environmental Dynamism.
Industry N Min Max Mean St. Dev
All 3248 0.00 0.11 0.02 0.01
28 1254 0.00 0.06 0.01 0.01
35 973 0.00 0.11 0.02 0.02
36 1021 0.00 0.07 0.02 0.01
Apart from this general descriptive statistic analysis regarding the characteristics of the main variables in the sample, it is important to be mentioned, before the further elaboration of the analysis, how the missing values of the data were handling. On the table 5, it is easily noticed
[36]
that the percentages of the missing values for the dependents, the independent and the moderator are zero. For the control variables the percentages of missing values are below 5%. This indicates that the most appropriate way to overcome the difficulty of missing values is by the use of list-wise exclusion (Field, 2005; Pallant, 2005). In every case that the list-wise selection is activated the observations of the sample are N= 3.046.The appropriateness of this method is based on the fact that the database is large and so the possibility of this mechanism to misinterpret the data is minimized, while only the cases with complete set of data are used (Field, 2005; Pallant, 2005). Additionally the fact that the data are missing randomly also reinforces the suitability of this decision (Field, 2005; Pallant, 2005).
Table 5. Missing Values.
Variable N Missing Values Count Percent Explr.Orientation 3248 0 0.0% Firms' Innovativeness 3248 0 0.0% Innovation Impact 3248 0 0.0% Env. Dynamism 3248 0 0.0% R&D intensity 3236 12 0.4% ROA 3240 8 0.2% Tobin's Q 3084 164 5.0% Company's Size 3207 41 1.3% 2. Correlations
Table 6_ Correlations, indicates the results of the bivariate correlation analysis which express the strength and the direction of the relationships among the variables of the study (Field, 2005; Pallant, 2005). Pearson product-moment correlation coefficient is the key factor for this analysis (Pallant, 2005). The asterisk on the results signals the most significant correlations.
[37]
The correlation result for the independent variable and the Impact of an Innovation indicates a sufficient positive relationship between them which is also statistically significant (r=.110, n=3406, p< .005). For the other dependent variable of the study, i.e. Firm’s Innovativeness, the results exhibit an even better correlation, as they support the existence of a stronger relationship (r=.320) which is also significant (n=3406, p< .005). Regarding the strength and the direction of the relationship between the Relative Exploration Orientation and the Environmental Dynamism, again a sufficiently strong positive relationship is noticed which as well has statistical significance (r=.099, n=3406, p<.005). Also for Innovation Impact and Environmental Dynamism, the result are on the same direction as the correlation indicator is statistically significant (n=3406, p<.005) and scores r=.098. On the contrary, for Firm’s Innovativeness the relationship with the moderator does not exhibit any significance (r= .005, n=3406, p>.005). With regard to the control variables of the study, Tobin’s Q variable reveals a strong negative relationship with both of the dependent variables, which is also statistically significant. More specifically for Innovation Impact the results are r=-.168, p<.005 and for Firm’s Innovativeness r= -.167, p<.005, n=3406. All the results regarding the correlations among all the variables of this academic paper are presented analytically in the table below.
[38] Table 6. Correlations Variable ( 1 ) ( 2 ) ( 3 ) ( 4 ) ( 5 ) ( 6 ) ( 7 ) (1) Exploration Orientation (2) Innovation Impact .110** (3) Firm's Innovativeness .320** .076** (4) Env. Dynamism .099** .098** .005 (5) R&D Intensity -.011 -.051** -.038* -.012 (6) ROA .022 .066** .079** -.067** -.260** (7) Tobin's Q -.119** -.168** -.167** -.050** .029 .067** (8)Company’s Size .127** -.018 .166** -.078** -.208** .274** -.254** **. Correlation is significant at the 0.01 level (2 -tailed).
*. Correlation is significant at the 0.05 level (2 - tailed). c. Listwise N=3046
3. Regression Analysis
To clarify how much of the variance in the dependent variables each of the independent
variables (the predictor and the control variables) is able to predict, hierarchical multiple regressions were incorporated in order to generate the appropriate models. Each of the dependent variables was examined separately and the results are presented on distinct tables. On this way the first two hypotheses of the study are tested further on the section, as well as the third and the fourth hypotheses that were presented on the theoretical framework. All the findings from all the analysis are included.
Before starting to elaborate on the analysis is important to be mentioned that the size of the sample, which is considered to be large, helps to overcome some biases on the results that can be created by smaller samples. An example is the issue of generalizability that on this case is
