AoM-Paper_Effectuation R&D-Kuepper-09.

doc

EFFECTUATION IN THE CONTEXT OF R&D PROJECTS: CHARACTERISTICS AND IMPACT ON PROJECT PERFORMANCE

ABSTRACT Innovative products are widely recognized to be an important source of competitive advantage. Both projects with an exploitative and explorative objective have been shown to be mandatory to ensure the long-term success of technology-based organizations. However, many companies have difficulties in finding efficient and successful approaches to different types of R&D projects and particularly to those that involve a high level of innovativeness. The aim of this study is to adopt a new theoretical concept called effectuation that was originally developed in the entrepreneurship research. This concept brings in a new perspective on the issue of how to accomplish successful R&D projects by focusing on the decision making in uncertain situations. First, characteristics of an effectual approach in the context of R&D projects are developed and clearly delineated from conventional prediction-based control strategies hereinafter called causation. Second, a thorough qualitative and quantitative scale development process is applied. Expert interviews and a pilot study (123 R&D projects) are used to develop a multi-factor measurement model of effectuation and causation respectively. These measures are validated in a follow-up study with a larger sample of 400 projects. Third, the new measures are applied to test two central hypotheses: (a) effectuation is positively related with the success in highly innovative contexts, (b) causation approaches are particularly beneficial in projects with a low level of innovativeness. Structural equation modeling is used to test the hypothesized effects. The results suggest that non-predictive control approaches consistent with effectuation are an important predictor of successful highly innovative R&D projects. At the same time causation proves to significantly enhance performance of projects that involve a low level of innovativeness.

Keywords: Effectuation; R&D projects; Project performance

AoM-Paper_Effectuation R&D-Kuepper-09.doc To compete successfully, technology based firms need to be innovative. Considerable resources have to be invested in research and development (R&D) to enable the short term oriented exploitation of the existing product base as well as the long term oriented development of new, highly innovative products. To date it is widely accepted that both exploitation and exploration are required to assure the long term success of organizations (e.g., Benner & Tushman, 2003). However, several authors have stressed that companies are comparably successful in exploiting existing product lines, whereas they face significant obstacles in exploring new businesses and technologies (Dougherty, 1992; Salomo, Weise & Gemnden, 2007). This managerial issue is also reflected in the scientific research. Although project characteristics are known to make a difference, only few studies have systematically analyzed the extent to which specific factors are contingent upon the projects degree of innovativeness. Research is so far dominated by incremental R&D processes. According to the rational plan approach (Brown & Eisenhardt, 1995), studies that analyze incremental R&D projects generally assume stable and predictable environments. Consequently, process management activities with clear project goals, staged task sequencing, pre-defined milestones and formal project reviews are widely considered to be best practices (e.g., Shenhar, Tishler, Dvir, Lipovetsky, & Lechler, 2002). However, it is still not clear if the success factors associated with incremental R&D projects also apply to projects with a high degree of innovativeness. Studies in this field are comparably rare. In addition, the existing studies have not yet converged to generally accepted success factors. For example, the study of Song and Montoya-Weiss (1998) reveals that project planning can be disadvantageous for highly innovative projects. Similarly, Lewis, Welsh, Dehler and Green (2002) observed a negative interaction between a planned style management approach and projects that involve a high degree of technical uncertainty. However, Shenhar et al. (2002)

AoM-Paper_Effectuation R&D-Kuepper-09.doc found that project planning is particularly important for projects that involve a high complexity and uncertainty. Another example is related to the optimum extent of process formality. Griffin (1997) revealed no significant relationship between process formality and development time whereas Salomo et al. (2007) state that process formality is not possible in highly innovative projects. One reason for the variations in research findings may be a lack of theoretical foundation that can be observed in many contributions to the innovation management research. To date no theoretical framework is available to systematically analyze success factors of projects with different degrees of innovativeness and to make sense of seemingly contradictory findings in the current literature. A reason for not having agreed upon one general framework may be found in the fact that one size does not fit all projects (Shenhar, 2001). Project characteristics like the degree of innovativeness are indeed important moderators. Therefore, a universal approach is likely to be inappropriate. With this article we intend to contribute to the theory building in the R&D project management literature in three ways. First, we investigate the effectuation logic as a conceptual basis for R&D projects. We adopt Sarasvathys theory of effectuation that was originally developed in the entrepreneurship research (Sarasvathy, 2001) to the context of R&D projects. We describe the characteristics of such a theory along five key principles. These principles are clearly delineated from conventional planning and prediction focused R&D approaches. We show that effectuation can be an important conceptual basis for analyzing R&D projects under uncertainty. Second, we contribute to the extant literature by precisely developing effectuation and causation measures. This is done on the basis of a generally accepted scale development process. The newly developed measures are of utmost importance for further research in the field of effectuation. They can serve as a basis for future empirical studies that are required to advance the

AoM-Paper_Effectuation R&D-Kuepper-09.doc effectuation research that is so far focused on theory building through experiments and case studies. Third, we study the impact of effectuation on the R&D project performance. Employing a path model, the relationship between effectuation and causation principles respectively and the performance of R&D projects are investigated. We measure the project performance on the basis of process and output criteria. In addition the moderating role of the level of innovativeness is examined in detail. Based on recently suggested measurement scales, we measure the degree of innovativeness as a multi-dimensional concept (Danneels & Kleinschmidt, 2001; Gatignon, Tushman, Smith, & Anderson, 2002; Garcia & Calantone, 2002). We differentiate market and technology innovativeness. In the following section we develop characteristics of effectuation in the context of R&D projects. On this basis we build up our research model and derive research hypotheses by linking effectuation and project success variables. Next, we give a description of the research method and the study context. The development and evaluation of measures is detailed afterwards. We present the results of tests of the measures as well as the whole research model using data from 123 R&D projects in a pre-study and 400 R&D projects in the main study. The article concludes with a discussion of the theoretical contributions and the managerial implications of the studys findings.

THEORETICAL BACKGROUND AND RESEARCH HYPOTHESES The existing R&D literature mostly assumes that a key task of the R&D manager is to discover opportunities and exploit them. With this concept in mind it is easy to see the importance of clear project targets, pre-defined milestones or formal project reviews and other factors that are often cited in the literature. However, the effectuation research offers new thinking that ap4

AoM-Paper_Effectuation R&D-Kuepper-09.doc proaches the challenge from an alternative view. While the existing R&D literature is mostly focused on the exploitation of existing opportunities, effectuation assumes that all opportunities need to be actively created. This means that opportunities are not waiting to be discovered but emerge through the R&D activity of an organization and its partners. Effectuation was first introduced in the entrepreneurial context. It was induced from an experiment of entrepreneurship as a form of expertise (Sarasvathy, 2001). However, it quickly gathered interest in different disciplines including management (Augier & Sarasvathy, 2004), economics (Dew, Sarasvathy & Venkataraman, 2004), psychology (Sarasvathy, 2003), and finance (Wiltbank, Read, Dew & Sarasvathy, 2008). We intend to apply effectuation in the context of R&D project management. The effectuation logic appears to be particularly suitable to form a conceptual basis in the R&D context since the R&D project management can be considered as a specific decision making problem (Dewar & Dutton, 1986). Effectuation in turn shall be a general theory of decision making in uncertain situations (Sarasvathy, 2008: 227) that focuses on the human action as the predominant factor shaping the future (Sarasvathy, 2008: 87). Effectuation prefers control over prediction: to the extent we can control the future we do not need to predict it (Sarasvathy, 2001: 251). These are all aspects that make effectuation promising under uncertainty and in the specific context of R&D projects. Effectuation can be summarized in five key principles that we briefly outline and apply in the context of R&D projects in the following. Principles of Effectuation and Delineation from Causation In order to reduce the uncertainty and to increase the controllability of the future, an effectual R&D approach starts on the basis of given resources and competences. The first principle therefore emphasizes creating a new outcome on the basis of existing means. This is opposed to a causal approach that starts with pre-defined project targets and derives the required means on 5

AoM-Paper_Effectuation R&D-Kuepper-09.doc that basis. In addition to the competences that are already linked to a specific organization or project team, commitments are required to decide about the implementation of an R&D project. On the one hand these commitments are related to the organization itself. The second principle that is often referred to as affordable loss principle (Sarasvathy, 2001) incorporates the potential risk or downside of an R&D project. The decision maker needs to define how much he is willing to lose in a worst case scenario by making in-advance commitments of how many resources he is willing to put at risk. This effectual approach can be clearly delineated from causation that is characterized by a thorough quantification of expected project returns as basis for decision making. In-depth business case modeling is often applied as means to support decision making in those cases. On the other hand commitments of other stakeholders are required to further reduce the risks and increase the controllability of an uncertain future. The third principle of effectuation involves forming partnerships and getting commitments from potential customers, suppliers or external groups of researchers. Such a pre-commitment can for example imply agreements with a customer to take specified quantities of the developed R&D output in the future or to support the R&D project with specified financial resources or competences. Whereas such an approach actively seeks to reduce uncertainty, causation focuses on the identification and the avoidance of uncertainty, e.g., by elaborate market and competitor analyses. The first three principles can be regarded as the decision preparation process. It has to be decided whether or not to implement the R&D project and which option to pursue. Seen from that perspective, the following two principles are related to the implementation of an R&D project. Again the reduction of risks and an increase of the controllability of an uncertain future are in the focus of an effectual approach. The fourth principle deals with unexpected surprises during

AoM-Paper_Effectuation R&D-Kuepper-09.doc the project implementation. Effectuation considers such surprises to be a vital source of opportunities. Contingencies are acknowledged and appropriated by leveraging surprises rather than trying to avoid them, overcome them, or adapt to them (Sarasvathy, 2008: 21). A causal approach that currently dominates the R&D literature strictly follows a linear process that seeks to reach the given project target as efficient as possible and within the given timeframe. The fifth principle of effectuation concerns the whole R&D process. Human agency is considered to be the prime driver of future developments. This has significant implications on the overall organization of the R&D process. E.g., it implies that an organization is more entrepreneurial in terms of initiating new trends rather than relying on analyses about potential developments of the exogenous environment. Generally, a conventional causal approach is characterized by the assumption that future developments and existing trends are exogenously given. Thus, these trends cannot be influenced but can be exploited by thorough forecasts and early assimilation. The five principles together are called effectuation. Effectuation can be regarded as the inverse of causation (Sarasvathy, 2008: 22) in the context of R&D project management. Table 1 summarizes the five principles with its effectuation and causation characteristics respectively. ---------------------------------Insert Table 1 about here ---------------------------------We already hinted at the fact that effectuation is not a static model but a dynamic nonlinear approach. The five principles all represent aspects of such a dynamic model. Given means are considered to be the starting point of an effectual R&D approach. Commitments concerning the maximum affordable loss as well as commitments from stakeholders on the one hand enlarge the decision scope and on the other hand form a framework that leads to a converging process. The sum of the existing means and committed resources are the basis for the decision making 7

AoM-Paper_Effectuation R&D-Kuepper-09.doc and for refining project goals. After having decided about the project option that shall be pursued, the principles 4 and 5 guide an iterative implementation process. New insights and surprises can lead to an integration of new stakeholders that again bring in new ideas together with new commitments. The dynamic process of an effectual decision making is summarized below in Figure 1. ---------------------------------Insert Figure 1 about here ---------------------------------Moderating Effects of Technology and Market Innovativeness After we have delineated effectuation and causation in the precedent section, we still need to clarify the boundary conditions of each concept. Neither effectuation nor causation is hypothesized to be generally preferable. The conceptual work of Sarasvathy (2001) clearly describes an effectual problem space. Central to it is the notion of uncertainty. Under real uncertainty that is also called Knightian uncertainty, probabilities for future consequences cannot be calculated. We define uncertainty to be the difference between the information possessed by an organization and the information required to perform particular tasks (Galbraith, 1977). This means that a prediction-based causal approach might be misleading under high uncertainty whereas a control-based effectual approach might be beneficial. Under these circumstances, future developments cannot be controlled by forecasting and exploiting trends but rather by actively participating in building the future. Differently, in a low-uncertainty context, causation should have advantages. Then future developments are relatively stable and predictable and thus can be best exploited through adaptation. Market analyses and business cases can be used effectively to decide about investing in a specific R&D project.

AoM-Paper_Effectuation R&D-Kuepper-09.doc In the context of R&D projects, the level of uncertainty is closely linked to the degree of innovativeness. Projects that involve a high degree of innovativeness also imply a high level of uncertainty and vice versa. The degree of innovativeness is defined as the difference between the status quo and the aspired R&D outcome. The outcomes of highly innovative R&D projects are, by definition, difficult to predict and therefore uncertain. Different dimensions need to be considered to cover the notion of innovativeness. Although previous research offers a wide range of definitions and perspectives (Danneels & Kleinschmidt, 2001; Garcia & Calantone, 2002; Salomo et al., 2007), it is particularly important to consider both a marketing and technological perspective (Garcia & Calantone, 2002). The technological innovativeness is related to the novelty of the required technical expertise within the organization (Henderson & Clark, 1990; MacCormack & Verganti, 2004; Shenhar et al., 2002; Wheelwright & Clark, 1992a). It is particularly high if the considered organization does not have any experiences with the technical principle (Green et al., 1995). The level of uncertainty increases with the degree of technological innovativeness since the innovativeness reflects the amount of new design work that needs to be undertaken in a project. In the case of high technological innovativeness it is not clear from the beginning whether a technical realization is feasible within a reasonable cost and timeframe. The market-related innovativeness reflects the novelty from the perspective of the marketing department. It prescribes the amount of new market requirements and new customer groups that need to be addressed. The marketing innovativeness increases with a decreasing market familiarity and experience of an organization (Danneels & Kleinschmidt, 2001). Again, the uncertainty increases with an increasing marketing innovativeness. Ex ante, the customer loyalty towards an R&D outcome is unknown and customer requirements as well as their willingness to pay are unclear. Other dimensions beyond the technological and marketing innovativeness like

AoM-Paper_Effectuation R&D-Kuepper-09.doc an organizational or process innovativeness are not considered here since the focus of this study is on product innovations. Process innovations are explicitly excluded from the analyses due to the fact that success determinants can significantly vary between different types of innovations (Wolfe, 1994). On the basis of the discussion of effectuation and causation, the technological and marketing degree of innovativeness can be assumed to have a significant impact on the relevance of each concept. This assumption is not new to the project management research (e.g., Marquis, 1969; Myers & Marquis, 1967; Rothwell et al., 1974; Utterback et al., 1976). The management of and coping with uncertainty is known for a long time to be the central problem of administration (Gifford et al., 1979). However, only recently studies have systematically considered the degree of innovativeness as an important moderator in search of project success factors (e.g., Eisenhardt & Tabrizi, 1995; Lewis et al., 2002; Moorman & Miner, 1998; Shenhar et al., 2002; Song & Montoya-Weiss, 2001; Tatikonda & Montoya-Weiss, 2001). Similar to our dichotomous discussion, Eisenhardt and Tabrizi (1995) developed two different strategies that should be followed depending on the specific context. The first one is called compression strategy and is related to projects that involve a low degree of uncertainty. Under these circumstances projects can be planned and executed in pre-defined steps. Such a process can then be compressed by shortening the time of each step (Eisenhardt & Tabrizi, 1995: 88). The second strategy is related to circumstances with a higher degree of uncertainty. It is called experiential strategy and is characterized by an iterative process to account for foggy and shifting markets and technologies (Eisenhardt & Tabrizi, 1995: 91). The impact of the degree of innovativeness that is involved in a project is increasingly acknowledged in the researcher community. However, the moderating role is complex since it can differ between different analyzed variables; simple di-

10

AoM-Paper_Effectuation R&D-Kuepper-09.doc chotomies like incremental vs. radical are considered inadequate (Song & Montoya-Weiss, 2001). Therefore, many authors claim that it is still not sufficiently understood and remains a question for empirical analysis (Salomo et al., 2007).

Effectuation, Causation and R&D Performance: Model Overview and Research Hypotheses The assumed impact of effectuation and causation on the R&D project performance is detailed in this section. The first step in doing so is to establish what success means. We differentiate two dimensions of R&D project success. Different and partly weak measures in many contributions are one major reason for inconclusive findings in the project management literature (Shenhar et al., 2002). A precise definition of the project performance is therefore particularly important. It enables a comparison of the results with former as well as with the ongoing research. The first success dimension, denoted as process efficiency, refers to the performance of the project implementation. It assesses success in meeting schedule and budget goals as well as the operational and technical performance of the R&D process. This dimension was fully or partly applied in different studies before (e.g., Dvir & Shenhar, 1992; Montoya-Weiss & Calantone, 1994; Shenhar et al., 2002). The second success dimension refers to the projects output and has also been applied in the literature (e.g., Shenhar et al., 2002). It includes two subdimensions, namely one construct that measures the benefits to the organization in terms of newly acquired experiences and competences and a second construct that measures the perceived value and future potentials of the R&D output. The differentiation between process- and output-related success criteria is particularly important in this study as some of the five effectuation and causation principles are more process and others more output related. In the following, two research hypotheses are derived for each 11

AoM-Paper_Effectuation R&D-Kuepper-09.doc principle. One hypothesis is related to projects that involve a low degree of innovativeness and one to projects with a high level of innovativeness. Figure 2 summarizes the research model. ---------------------------------Insert Figure 2 about here ---------------------------------Principle 1: Means versus goal driven initiation of R&D projects. The emphasis of effectuation here is on creating something new with existing means rather than discovering new ways to achieve given goals (Sarasvathy, 2008: 21). According to the original wording of Sarasvathy and a transformation to the context of R&D projects, relevant resources can be existing competencies and project experiences (Who I am), skilled employees and experts in the considered field of R&D (What I know), relations to partners (e.g., R&D networks) (Whom I know), financial means and tangible assets like R&D equipment (What I have). In the case of an effectual approach of a R&D project such resources are used as an initial basis. Project targets are fuzzy at the beginning. This is different when using a causal approach. Causation assumes detailed project targets to derive required resources on that basis. A direct comparison of a means vs. goal orientation is not yet described in the literature. However, on the basis of the effectuation logic it can be assumed that a means driven approach should have a positive impact on the R&D project output particularly in projects with a high degree of innovativeness. Recently, different authors have stated that formal processes are limited or inappropriate (Seidel, 2007: 522) in uncertain project environments, like radical innovation projects. Goals can be unclear and changing, [] the objective may not only be ambiguous to begin with but also may change over time (Seidel, 2007: 522). This means that the often encountered assumption that project targets need to be clearly specified right from the beginning does not hold true in all cases. In projects with a high degree of innovativeness, targets cannot be detailed initially. They rather have to be elaborated during the R&D project implementation. 12

AoM-Paper_Effectuation R&D-Kuepper-09.doc In addition it can be argued that basing an R&D project on the current competencies and resources can have a direct advantage particularly when the future is not clear to the organization. Different authors have found that building up new competencies needs to be grounded on existing expertise (Atuahene-Gima, 2005; Danneels, 2002). That means that a firm must exploit some level of its current competencies to leverage its new competencies to develop radical innovations (Atuahene-Gima, 2005: 79). This altogether leads to hypothesis 1a. Hypothesis 1a. The effectuation characteristic of principle 1 has a positive impact on the R&D output in the context of high market and technological innovativeness. However, whenever the projects degree of innovativeness is comparably low a clear project target should positively impact upon the project output. In this case clear project targets can support the team in guiding towards the aspired R&D output. For instance, interfaces between different sub-systems can then be planed and developed early in the process guaranteeing a superior quality and a consideration of specific customer requirements. Hypothesis 1b. The causation characteristic of principle 1 has a positive impact on the R&D output in the context of low market and technological innovativeness. Due to the specific output orientation of principle 1, neither an advantage of effectuation nor causation is expected concerning the process efficiency. On the one hand a focus on existing means like experiences and competencies should accelerate the R&D process. On the other hand, clear project targets can be beneficial to prioritize activities. This can avoid loosing time with minor important tasks that can slow down the process. Both effects should level out in a way that no clear advantage of an effectual or causal approach on the project efficiency is expected. Principle 2: Consideration of affordable loss versus calculation of expected returns.

13

AoM-Paper_Effectuation R&D-Kuepper-09.doc The effectuation characteristic of this principle prescribes committing in advance to what one is willing to lose rather than investing in calculations about expected returns to the project (Sarasvathy, 2008: 21). This principle is particularly related to the process efficiency. In a project that involves a high level of uncertainty, a consideration of risks should be particularly suitable. Reliable returns cannot be calculated. A limitation of what one is willing to put at risk ensures that the maximum potential loss is in a reasonable relation with the existing risk. Precommitments with regard to the project budget and the project schedule help to avoid permanent overspendings that can often be observed in the managerial practice. Such an effectual approach should therefore have a direct positive impact on the project efficiency under high uncertainty. Via this relation, the effectuation characteristic of this principle is also expected to have an indirect positive effect on the R&D output. A direct positive effect of effectuation on the R&D output however should be difficult to analyze with a project based perspective that we have chosen in this study. Projects that are terminated on the basis of a strict criterion generally have a low project output. However, in the sum of different projects within an organization, this might be beneficial because it avoids spending too much money on the wrong projects. Due to this methodological aspect we limit hypothesis 2a to the positive impact of effectuation on the R&D efficiency. Hypothesis 2a. The effectuation characteristic of principle 2 has a positive impact on the R&D efficiency in the context of high market and technological innovativeness. In contrast, in low uncertainty environments a causal approach should be beneficial. Future opportunities and returns can then be analyzed, estimated and used as a solid basis for the decision of implementing a specific R&D project. The precise knowledge about future developments and trends can help to efficiently manage the project within the given budget and schedule.

14

AoM-Paper_Effectuation R&D-Kuepper-09.doc Hypothesis 2b. The causation characteristic of principle 2 has a positive impact on the R&D efficiency in the context of low market and technological innovativeness. Principle 3: Reduction versus identification of uncertainty. This principle involves negotiating with any and all stakeholders who are willing to make actual commitments to the project, without worrying about opportunity costs, or carrying out elaborate competitive analyses (Sarasvathy, 2008: 21). An effectual approach actively reduces the risk that is inherent in a R&D project with a high degree of innovativeness. This can be achieved through direct commitments of stakeholders like potential customers or R&D partners. Different studies in the R&D and innovation management literature have shown the importance of recognising and working with stakeholders (Elias, Cavana, & Jackson, 2002: 309). It has been shown that external relations and commitments have a positive impact on the performance of a project or the organization (e.g., Coombs, McMeekin, & Pybus, 1998; Eckert, 1996; Miller, 1995; Rogers, 1996; Tipping, Zeffren, & Fusfeld, 1995). According to these studies, external ties reduce the project risk as well as they are a vital source of new competencies that support the innovation activity of an organization. In addition external relations can increase the quality of decision making concerning the implementation of a project and therefore indirectly impact upon the performance (Robins & Atuahene-Gima, 2003). Although these studies have not analyzed the moderating effect of the degree of innovativeness, it can be concluded that stakeholder commitments and external ties are particularly beneficial under high uncertainty. An active reduction of uncertainty is simply not required when the uncertainty is already low. Beyond it, external relations and co-operations are often not aspired in the case of low uncertainty due to the fact that the expected (and probable) benefits would then need to be divided among the partners. Therefore, projects that involve a low degree

15

AoM-Paper_Effectuation R&D-Kuepper-09.doc of innovativeness are expected to rather profit from detailed market and competitor analyses and a corresponding assimilation of their own activities than from stakeholder commitments. This leads to the following two hypotheses. Hypothesis 3a. The effectuation characteristic of principle 3 has a positive impact on the R&D output in the context of high market and technological innovativeness. Hypothesis 3b. The causation characteristic of principle 3 has a positive impact on the R&D output in the context of low market and technological innovativeness. Principle 3 has a rather low impact on the project efficiency. Similarly to principle 1, it is not assumed that an active reduction or an identification of uncertainties has advantages in meeting the defined project budget and schedule. Principle 4: Acknowledge versus overcome the unexpected. Effectuation is about exploiting contingencies and surprises. The effectuator leverages uncertainty by treating unexpected events as an opportunity (Sarasvathy, 2008: 85). Effectuation fundamentally rearranges the relationship between planning, contingencies and uncertainty. Because effectuators often begin with only a very loose notion of their goals, they can make up their plans in an incremental fashion, utilizing uncertainty and contingent information as resources for their goals rather than relying on goals as determining factors of resource acquisition and choice (Sarasvathy, 2008: 85). The ideas expressed in this principle can be best compared to the formal or planned management styles (formality) on the one hand and the more flexible, emergent styles (discretion) on the other hand, that are discussed in the literature. Formality can be related to a causal approach. The term formality in R&D projects denotes the degree to which rules, policies, and procedures govern the R&D project (Tatikonda & Montoya-Weiss, 2001). Discretion refers to the ability to

16

AoM-Paper_Effectuation R&D-Kuepper-09.doc break formal rules and to react to the emergence of unanticipated ideas, which are conceived, shaped, and transformed in the process of implementation (Naveh, 2007). Discretion therefore can at least partly be compared to the fourth principle of effectuation. However, it has to be kept in mind that effectuation rather suggests actively shaping a new opportunity than adapting to a changing environment which is part of the discretion construct. Apart from this aspect that differentiates effectuation and discretion, the discussions in the literature shall be briefly considered here to support the development of research hypotheses. The empirical studies in this field show mixed results. On the one hand researchers prefer an emergent, fluid style of management (Dougherty, 1992; Lewis et al., 2002; Moorman & Miner, 1998). On the other hand a planned style of management in the context of R&D projects is described as a means to accelerate projects (Wheelwright & Clark, 1992b; Zirger & Maidique, 1990), reduce errors (Cooper & Kleinschmidt, 1986; Montoya-Weiss & Calantone, 1994) and increase the financial performance (Ittner & Larcker, 1997; Song & Parry, 1997). Beyond it, there are researchers who try to integrate both perspectives (Eisenhardt & Tabrizi, 1995; Naveh, 2007; Tatikonda & Rosenthal, 2000). Case studies have shown that the balance and the interplay of emergent and planned styles (Lewis et al., 2002: 547) is decisive (Jelinek & Schoonhoven, 1990; Laufer, 1997). Decision making discretion can have a positive impact on the project performance since it may foster spontaneity, encouraging members to improvise and explore market opportunities as they arise (Lewis et al., 2002: 562). At the same time a close managerial supervision and a formalized approach were found to improve the process efficiency (Lewis et al., 2002). Following the argumentation for the first three principles, an impact of the degree of innovativeness could again be expected. However, Naveh (2007) as well as Tatikonda and MontoyaWeiss (2001) did not found a moderating effect of the uncertainty involved in a project. Naveh

17

AoM-Paper_Effectuation R&D-Kuepper-09.doc (2007) assumes that discretions main effect and its interaction with formality may mitigate the effect of uncertainty in R&D projects (Naveh, 2007: 123). This would mean that the impact of uncertainty could be eliminated by managing formality and discretion (Naveh, 2007: 123). This is an important result that is not in line with what is mostly expected in the literature (e.g., Tatikonda & Montoya-Weiss, 2001) and has to be considered in the development of research hypotheses for principle 4. It can be concluded that the degree of innovativeness should not have a significant moderating impact. The literature results suggest at the same time that effectuation as well as causation have a positive effect on the project performance. On the one hand the interplay of both approaches may be important. On the other hand it has to be differentiated between the different performance dimensions. Lewis et al. (2002) hint at the fact that a distinction between the process efficiency and the project output is required. An emergent style of management is positively related to the project output since it builds technical knowledge by exploring novel scientific concepts and experimenting with alternative product designs. Furthermore, such an approach is in line with what is increasingly acknowledged in the literature namely that a focus only on front-end concept generation practices may not be sufficient. Later changes to the concept are not only likely but are also likely to be important (Seidel, 2007: 531). Hypothesis 4a. The effectuation characteristic of principle 4 has a positive impact on the R&D output irrespectively of market and technological innovativeness. Formal processes have clear restrictions concerning the generation of new experiences and competencies (Sethi & Iqbal, 2008). We therefore assume that they do not have a positive effect on the project output. However, a planned style of managerial activities or the causal characteristics of principle 4 should be positively related with the project efficiency. In such an approach managers can make sure that the R&D teams are always focused on the project target. Mile-

18

AoM-Paper_Effectuation R&D-Kuepper-09.doc stones concerning the project schedule and budget visualize the limited resources and lead to an efficient use of those (Wheelwright & Clark, 1992b). Hypothesis 4b. The causation characteristic of principle 4 has a positive impact on the R&D efficiency irrespectively of market and technological innovativeness. Instead of having a moderating effect it can still be assumed that both hypothesized relations are stronger in projects that involve a high degree of innovativeness, i.e., to be successful in terms of the R&D output requires the integration of new experiences particularly in the case of a high degree of innovativeness especially as such projects can face considerable ambiguity and unforeseen challenges (Seidel, 2007: 524). Principle 5: Create versus exploit opportunities. The effectuation characteristic of this principle urges relying on and working with human agency as the prime driver of opportunity (Sarasvathy, 2008: 22). Human action is seen as the decisive factor that shapes the future. In contrast, causation assumes that opportunities are exogenously given. Trends then cannot be influenced by a single organization but can be exploited by most adequately adapting to it. An impact of this principle can be particularly expected with regard to the process efficiency. Effectuation should be beneficial in projects with a high level of uncertainty. In such projects our knowledge about the future is not a deficiency that can be remedied by sufficient study but in fact derives from the originating, creative force of human choice that such uncertainties are even in principle imponderable (Dew & Sarasvathy, 2007: 269f.). Consequently, an approach should have advantages that relies to a less extent on forecasts of an uncertain future but rather relies on means to actively control the future by shaping it. More and more authors conclude that strategies that are based on a prediction of an uncertain environment are not likely to

19

AoM-Paper_Effectuation R&D-Kuepper-09.doc be well informed. Since men can (must) choose how to act, their chosen acts, together with the evolution of the physical world, are continuously creating the emerging future. If this is so (as it must be), then the future cannot be known now (Wiseman, 1989: 268). This means that under real uncertainty or in the case of a high degree of innovativeness trends cannot be reliably identified ex ante. A similar aspect has been analyzed in the comprehensiveness research. Comprehensiveness denotes a concept that captures the extensiveness with which an organizations top executives systematically gather and process information from the external environment in making strategic decisions (Forbes, 2007: 362). Forbes found that comprehensiveness only has a positive effect on the decision quality if enough information is available and if this information is clear in its meaning. Following the results of Forbes, it has to be questioned whether forecasts and trend analyses can have a positive impact due to the fact that under real uncertainty little reliable information is available. Therefore a causal approach that spends time and budget on analyses that have a negligible value in the end is assumed to have a negative impact on the process efficiency which comprises meeting time and budget targets. However, an effectual approach that actively shapes the future and thus tries to control the future should have a direct positive impact on the project efficiency under high uncertainty. Via this relation the effectuation characteristic of this principle is also expected to have an indirect positive effect on the R&D output. A direct advantage of an effectual approach versus a causal approach however might be difficult to analyze with regard to the output. On the one hand effectuation should be preferable due to the fact that an organization that actively creates the future might be able to establish standards and profit from first mover advantages. On the other hand an approach that uses causal characteristics of this principle like prediction and forecasting techniques is not necessarily less successful regarding the project output. This is simply because the project team is not bounded to

20

AoM-Paper_Effectuation R&D-Kuepper-09.doc the forecasts. In other words the causal activities do not have a negative impact on the project output. The only negative impact can be measured with regard to the project efficiency: Additional costs and time efforts that are required for a prediction based approach are disadvantageous. Hypothesis 5a. The effectuation characteristic of principle 5 has a positive impact on the R&D efficiency in the context of high market and technological innovativeness. In projects with a low degree of innovativeness a causal approach should be beneficial. Under these circumstances a detailed process planning and forecasting is possible. The process can be accelerated by analyzing and following technological trends. Hypothesis 5b. The causation characteristic of principle 5 has a positive impact on the R&D efficiency in the context of low market and technological innovativeness.

METHODS Study Context and Samples We used two samples in this study. Both samples were collected in a cross-sectional survey methodology. We applied the first sample for the scale development process that is described below. With the second sample we intended to cross-validate the findings from the first sample. In addition it provided an enlarged data basis for testing the research hypotheses. Sample 1 We drew the first sample from our personal networks and their contacts. To build the sample we asked known senior R&D managers to provide contact information of other experienced German R&D managers from their personal network. All in all contact information of 900 R&D

21

AoM-Paper_Effectuation R&D-Kuepper-09.doc managers were available for the first study. A total of 123 responses led to a response rate of 13.7% using a very moderate approach with only one reminder. Sample 2 The sampling frame for this group was the Bureau van Dijk database of European companies. After we eliminated companies that did not match specified requirements including industry sector (only R&D intensive sectors considered), country and a company size above 50 employees, a total of 5,288 German companies remained. In addition, we checked the contact information via internet and used phone calls to ensure that the companies are carrying out own R&D projects. Service firms as well as overseas companies that only have a European sales representation have not been considered. We furthermore eliminated firms with wrong or incomplete company information which led to a final database of 3,138 organizations. A suitable key informant within each company was either selected from the information given in the Bureau van Dijk database, a telephone call to identify the right contact partner or an internet search using social networks like XING. In addition, we highlighted the importance of having an experienced project manager as key informant in the invitation mail. We formulated clear criteria for participation in order to ensure that the respondent had a broad view of the project and could provide the detailed management and technical information required. We asked each informant to complete a self-administered online questionnaire. From this set, 606 individuals could not be reached (421) or refused to participate (185) resulting in an effective base of 2,532 project managers. A total of 277 responses were obtained, accounting for a response rate of 10.9%. This response rate can be considered satisfactory since only R&D managers with managerial responsibility and experience were asked to participate. Former studies have already shown that survey-

22

AoM-Paper_Effectuation R&D-Kuepper-09.doc ing top managers can reduce the expectable response rate by 50% (Bartholomew & Smith, 2006; Baruch, 1999). Instrument Development and Refinement We used existing scales and items to measure the degree of market and technological innovativeness and the different success dimensions. Scales for the causation and effectuation constructs are not available. Therefore, the first step entailed the development of new measures. We adopted the following three-step approach for this purpose (cmp. Churchill, 1979). The scale development process consisted of a qualitative, a first quantitative and a second quantitative analysis. The objective of the qualitative part was to develop preliminary measures that could be applied in a first quantitative test. To finalize the scale development we used a second quantitative analysis to cross-validate the findings of the preceding tests. Within the first phase, we carried out a rough conceptualization and a qualitative pre-test. First, we developed a basic understanding of effectuation including a delineation from causation. This involved developing a pool of items to start with for each of the five effectuation and causation principles. We contrasted each effectuation item with a causation indicator. We largely used qualitative techniques like literature research, expert discussions and interviews to identify items from different perspectives. Following the recommendations of DeVellis, similarly items were explicitly allowed since removing all redundancy from an item pool or a final scale would be a grave error because redundancy is an integral aspect of internal consistency (DeVellis, 2003: 87). All in all, we generated a full list of 74 items during this process. Next, we carried out a qualitative pre-test to test all items concerning their content validity and comprehensibility. The objective was to improve or eliminate inappropriate items. On the one hand we asked project managers to assess the items for clarity and appropriateness along a 5-point scale ranging from 23

AoM-Paper_Effectuation R&D-Kuepper-09.doc very low to very high. 15 probands were asked to complete a questionnaire that included the items and indicate any ambiguity or other difficulty in responding, as well as offering suggestions for improvement. This approach was iterative meaning that the next respondent had to assess the improved items. On the other hand we asked effectuation researchers to assess the content validity and the delineation of effectuation and causation. The researchers had to assign each item to one of the three categories Clearly effectuation, Clearly causation and Could be both. This test led to a consistent picture. Three effectuation and causation items had to be eliminated due to an insufficient demarcation. In addition the resulting constructs were discussed with academics and scale development experts and finally with yet another set of R&D managers. We asked them to complete a full questionnaire during a phone interview. At this stage, very few concerns were raised and only minor refinements were required. In the second step of the scale development approach we used a self-administered survey to make a first quantitative test of the new measures. We applied sample 1 for that purpose. A set of commonly used analyses was carried out to test the new scales for reliability and convergent validity (Bagozzi & Phillips, 1982), discriminant validity (Bagozzi, Yi & Phillips, 1991), and nomological validity (Peter & Churchill, 1986). First, we assessed the reliability by calculating Cronbachs alpha and item-to-total correlations. Next, we applied an exploratory factor analysis to ensure convergent validity. It was demanded that each hypothesized factor explains at least 50% of the variance of the related items (VAR). In addition we conducted a confirmatory factor analysis (CFA). Again, the objective was to eliminate items in the case that pre-defined cut-off criteria were not met (indicator loadings .4; factor reliability (FR) .6 and average variance extracted (AVE) .5). Hereafter, we assessed discriminant validity demanding that each indicator is clearly associated with the hypothesized factor. Additionally we verified discriminant va-

24

AoM-Paper_Effectuation R&D-Kuepper-09.doc lidity on factor level using AVE demanding that it be greater than the squared interscale correlations of each pair of factors (Fornell & Larcker, 1981). Nomological validity requires that the single factors are confirmed within the context of a larger theory (Bagozzi, 1979: 14). Therefore, we analyzed the overall fit of the model within a full structural equation model using AMOS. In addition we completed an analysis to identify potential misspecifications along the residuals of the covariance matrix and the modification indices. In total, we eliminated eight effectuation and causation items during this process. The third and final step of the scale development process contained the same analyses as completed in step 2. However, the sample size was enlarged by using sample 2. This enabled a cross-validation of the results of the earlier analyses. All results that we show in the following section could be confirmed with this enlarged set of additional empirical data. An additional elimination of items was therefore not necessary. The final scales are provided in the Appendix. Evaluation of Measurement Models Principle 1 that prescribes a means versus goal driven initiation of R&D projects, contains seven items ( = .90; VAR = 64.1%; FR = .91; AVE = .60). As for the other principles, we contrasted effectuation with causation items and used a six-point scale to measure whether the rater rather used an effectual or causal approach. The scale of principle 2 describing a consideration of the affordable loss versus a quantification of expected returns also shows a high reliability and validity ( = .86; VAR = 63.8%; FR = .87; AVE = .58). It uses five items in its final version. Principle 3 that basically denotes reducing versus identifying uncertainty, is measured on the basis of four items ( = .82; VAR = 64.9%; FR = .83; AVE = .56). Good reliability and validity characteristics also shows the fourth principle (acknowledge vs. overcome the unexpected) that includes six items ( = .86; VAR = 59.5%; FR = .86; AVE = .50). The measure for principle 5 25

AoM-Paper_Effectuation R&D-Kuepper-09.doc tests whether the project rather created a new opportunity or if it exploited an existing trend. It consists of four items ( = .80; VAR = 62.7%; FR = .82; AVE = .54). Discriminant validity was confirmed by the fact that, for each pair of factors, the AVE was greater than the squared interscale correlations (see table 2). ---------------------------------Insert Table 2 about here ---------------------------------Strong support for a nomological validity could be achieved through the structural equation modeling that is represented in the results section. Many studies only focus on project efficiency to assess the project performance (e.g., Bonner, Ruekert, & Walker, 2002). We took a broader approach by additionally considering two success dimensions that are related to the project output. We asked the informants to evaluate project performance relative to the expectations. Self-assessment measures were applied. Although such an approach may be prone to bias, it is the most commonly used way to measure the project performance and it has been shown to be sufficiently reliable (Dawes, 1999; Dess & Robinson, 1984; Pearce, Robbins, & Robinson, 1987). The process efficiency scale consisted of three items (e.g., Shenhar et al., 2002) leading to acceptable reliability and validity values ( = .72; VAR = 77.9%; FR = .72; AVE = .56). The first sub-dimension of the output-related performance construct applied three items to measure the experiences and competencies gathered throughout the project ( = .76; VAR = 67.2%; FR = .78; AVE = .58). The second sub-dimension that measures the perceived value and future potential of the R&D output (e.g., Pinto & Mantel, 1990) initially consisted of four items and had to be reduced to three key indicators ( = .85; VAR = 76.9%; FR = .82; AVE = .61). The moderating variables market and technological innovativeness followed prior conceptual and empirical work. All items of the two innovativeness dimensions have been applied and 26

AoM-Paper_Effectuation R&D-Kuepper-09.doc validated in previous research (Danneels & Kleinschmidt, 2001; Hauschildt & Schlaak, 2001; Salomo et al., 2007). We measured the technological innovativeness on the basis of four items and the market innovativeness based on five items respectively (see Appendix). Hypotheses Testing: Structural Model In order to test the hypothesized contingency model, we split the sample into two groups based on the composite score of the technological and market innovativeness scale (Song & Montoya-Weiss, 2001). Using the mean score of these two constructs enabled us to differentiate four categories. We further considered the two extreme cases, namely on the one hand projects that involved a low market and technological innovativeness, and on the other hand projects that involved a high market and technological innovativeness. ---------------------------------Insert Figure 3 about here ---------------------------------This classification scheme led to 136 projects in the high and 136 projects in the low innovativeness sub-group respectively. We used a two-group AMOS model to examine the structural relationships separately and to test for significant differences between the two groups (Bollen, 1989; Song & Montoya-Weiss, 2001). Following the procedure of Keil et al. (2000) significance levels were evaluated based on t-tests (compare Keil et al., 2000 for details and formula). We additionally confirmed the results with commonly used -difference tests. We conducted an overall model assessment for both groups simultaneously (Song & Montoya-Weiss, 2001).

RESULTS Table 3 reports the parameter estimates and significance levels for each path in the twogroup structural equation model. The results suggest that the structural model fits the data very

27

AoM-Paper_Effectuation R&D-Kuepper-09.doc well (/df = 1.59, p < .001, RMSEA = .04 (pCLOSE = 1.0), SRMR = .05, TLI = .95, CFI = .96, GFI = .90). All paths that we hypothesized to be significant differentiators between effectuation and causation show moderated effects (bold cells in Table 3). ---------------------------------Insert Table 3 about here ---------------------------------Specifically, hypothesis 1b is supported (standardized coefficient (SC) .24, p < .01) meaning that a causal approach concerning the characteristics of principle 1 has significant advantages in projects that involve a low degree of innovativeness. The negative path coefficient for the subgroup of highly innovative projects indicates at also supporting hypothesis 1a. However, this path coefficient is not significant. On the other hand, the group comparison shows significant differences in the most successful approaches under low and high innovativeness which in total can be interpreted at least as a weak support of hypothesis 1a. Hypotheses 2a (SC = -.25, p < .09) and 2b (SC = .25, p < .09) can be both supported at significant levels. The effectuation characteristic of principle 2 is preferable in highly innovative projects whereas causation shows significant advantages under low innovativeness. This can also at least partly be concluded when analyzing hypotheses 3a and 3b. The data directly supports 3a (SC = -.14, p < .05), i.e., using the effectual characteristics of principle 3 has a positive impact on the R&D output. Analogical to the discussion of hypothesis 1a, hypothesis 3b can be weakly supported since our group comparison shows that best-practice approaches in the context of low and high innovativeness are significantly different. Hypothesis 4a (Low innovativeness: SC = -.11, p < .09; High innovativeness: SC = -.23, p < .01) is also directly supported whereas the results for 4b are mixed (Low innovativeness: SC = .02, n.s.; High innovativeness: SC = .28, p < .01). As additionally hypothesized, in both cases the relations are stronger in projects that involve a high degree of innovativeness. Hypotheses 5a and 5b are again both supported. The effectuation characteristics of 28

AoM-Paper_Effectuation R&D-Kuepper-09.doc principle 5 have a significant positive effect on the process efficiency in highly innovative projects (SC = -.19, p < .05) and the causal approach shows clear advantages in low innovative projects (SC = .22, p = .15). This result is additionally supported by the group comparison.

DISCUSSION The primary aim of this study was to analyze the effects of effectuation and causation processes on the R&D project performance. We went forward in three steps: First, we defined and delineated characteristics of an effectual approach in the context of R&D projects. Second, we developed measurement instruments to analyze the new effectuation and causation concepts in a large scale empirical study. Finally, we applied the new scales to investigate the performance impact of each concept. Seen from a more abstract level we hypothesized that (a) effectuation is positively related with the success in highly innovative projects and (b) causation is particularly beneficial in projects with a low level of innovativeness. As described above, (a) is supported or partly supported by the research hypotheses 1a, 2a, 3a and 5a. The results also support (b) that was tested through the hypotheses 1b, 2b, 3b and 5b. This studys results therefore provide support for the assumption that the degree of innovativeness is an important moderator that needs to be included when analyzing project success factors. Furthermore, it directly shows that effectuation and causation respectively are important theoretical concepts that can be valuable frameworks for researchers when analyzing decision making related R&D project management topics. The results also have some managerial implications as they bring in a new perspective that is at least partly contrary to what is mostly believed to be best practice.

29

AoM-Paper_Effectuation R&D-Kuepper-09.doc Theoretical Contributions and Future Research Directions The current study enriches and extends researchers understanding in several ways since we achieved the findings by linking two fields of research. We tested the dichotomous concept of effectuation and causation as a potential theoretical framework for the second field of research namely the management of R&D projects. This approach thus provided contributions to both research directions. First, we want to focus on the contributions to the R&D management literature. Early research on R&D management has produced descriptive frameworks and models that consider the R&D process as a sequential, linear system with discrete stages (e.g., Zaltman, Duncan, & Holbek, 1973; Cooper, 1990). More recently, recursive and chaotic frameworks have been developed to acknowledge that R&D progresses in a non-linear way with several feedback loops (e.g., Cheng & Van de Ven, 1996; Constant, 2000). These frameworks resist reductionism and linear analysis. Our concept extends the linear, recursive and chaotic frameworks by viewing R&D processes as effectual or causal approaches depending on the specific circumstances. In addition our framework is not only descriptive, i.e., a statement about the truth or otherwise of a phenomenon in the world but it is also an internally consistent set of ideas that forms a clear basis for action upon the world (Sarasvathy, Dew, Read, & Wiltbank, 2008: 345). With this in mind, the dichotomous concept of effectuation and causation can be regarded as an extension of the existing concepts and also as an independent R&D framework. It takes into account that specific circumstances like different degrees of innovativeness require different R&D approaches and thus provides recommendations for both, incremental and radical innovation projects. It is specifically this integrated view that differentiates our approach from the current R&D literature. Existing frameworks rather focus either on incremental (linear approaches) or on radical innova30

AoM-Paper_Effectuation R&D-Kuepper-09.doc tions (chaotic systems). In addition, the framework presented in this article offers a set of clear recommendations to enhance decision making in certain and uncertain situations. Whereas e.g., the chaotic framework rather describes an R&D approach as a chaotic system that leads to an expanding and diverging process, effectuation suggests how to navigate through complex decision trees. It offers support to make such processes converge. Beyond offering specific directions for action as extension of the existing frameworks, we believe that our approach also directly competes with the others by using different underlying key assumptions. Whereas most of the existing frameworks assume that exogenous factors drive the R&D process, we position the human action and the organization in the center of the process. This acknowledges that control can actively be built up e.g., by using effectual elements like starting the process on the basis of ones means and competences. This is clearly in contrast to the existing frameworks like the recursive one that deal with uncertainty by recommending quick adoptions to any unforeseen developments and surprises. The ability to actively reduce uncertainty as an important characteristic of our approach is not considered. Beside the contributions to the R&D management research, this study helps to develop effectuation towards a general theory of decision making in uncertain situations. We did that by clarifying three critical aspects that can be regarded as important towards building a generally accepted theory (House, 1963). First, we showed that effectuation can be delineated from conventional approaches and that effectuation can be measured. This study is to our knowledge the first large scale empirical analysis that provides newly developed scales of effectuation. Second, it was important to validate effectuation in terms of its relevance. To date only experiments (e.g., Wiltbank et al., 2008), case studies (Sarasvathy & Kotha, 2001) and one meta-analysis (Read, Song, & Smit, 2008) have been used to analyze effectuation. With this study we revealed the

31

AoM-Paper_Effectuation R&D-Kuepper-09.doc relevance of effectuation with a large scale empirical study and found positive impacts of effectuation on performance. Third, we transferred effectuation from its original field of research to another discipline. This helps to broaden the scope of effectuation and thus gives it a more general meaning. With this study we intended to cover a broad field of effectuation and causation characteristics in order to get a first empirical foundation of its key constructs. Of course, this approach was carried out at the expense of a more detailed analysis of each effectuation principle. We suggest additional in-depth analyses on each principle as interesting future research opportunities. Beyond it, the analysis of mediating effects appears to be particularly relevant. In addition, it might be interesting to identify specific antecedents that support effectuation (e.g., organizational antecedents). In order to be able to successfully implement effectual cells in plan dominated large organizations, it can be assumed that parallel organizational structures are required. The design of these parallel structures is yet another promising future research topic. Managerial Implications and Limitations A strict process management including concepts like Total Quality Management (TQM) or Lean Management dominated the managerial practice during the last two decades. However, more and more authors hint at the fact that managers that only seek to achieve productivity gains are not well informed in the long run (e.g., Benner & Tushman, 2003). Proponents of process management concepts have promoted it as being universally beneficial for organizations (Hammer & Stanton, 1999; Harry & Schroeder, 2000). However, researchers have already raised concern about that since the diffusion of process management techniques favors exploitative innovation at the expense of exploratory innovation (Benner & Tushman, 2003: 239). Without doubting that process management activities in the context of R&D projects have clear advan32

AoM-Paper_Effectuation R&D-Kuepper-09.doc tages in exploitative projects, we have shown in this study that the benefits have to be questioned in exploratory projects. In fact we have shown that R&D managers need to differentiate between projects with a low and a high degree of innovativeness. Whereas a strict planned and predictionbased causal approach is suitable under low uncertainty, effectuation offers ways to better deal with high uncertainty situations. Like any investigation, ours has some limitations. One limitation of this study is the use of retrospective data. A valuable approach for future research may be employing a longitudinal methodology to limit the recall time frame and to avoid bias from this source. In addition, multiple respondents might be used to limit common method bias. However, we believe that our findings are not an artifact of common method bias because our analyses with the one-factor test indicated no serious common method problems.

APPENDIX: INFORMATION ON CONSTRUCTS AND ITEMS

Principle 1: Means vs. goals ( = .90) Effectuation
1) Our R&D project was specified on the basis of given means/resources 2) The target of our R&D project was vaguely defined in the beginning 3) Given means/resources have been the starting point for the project 4) The process converged towards a project target on the basis of given means/resources 5) Rather given means than concisely given project targets have been the starting point for our project 6) The project specification was predominantly based on given resources 7) Given means have significantly impacted on the framework of our R&D project

Causation
Our R&D project was specified on the basis of given project targets The target of our R&D project was clearly defined in the beginning Given project targets have been the starting point Required means/resources have been determined on the basis of given project targets A concisely given project target has been the starting point for our project The project specification was predominantly based on given targets Given project targets have significantly impacted on the framework of our R&D project

Principle 2: Affordable loss vs. expected returns ( = .86) Effectuation

1) Considerations about potential losses were decisive for the selection of the R&D option 2) Project budgets were approved on the basis of considerations about acceptable losses 3) The selection of the R&D-option was mostly based on a minimization of risks and costs 4) We mainly considered the potential risk of the project *) We compared different R&D options on the basis of an assessment of risks and costs that we were willing to lose (e.g., due to a lack of possibilities to make concise forecasts of expected returns)

Causation
Considerations about potential returns were decisive for the selection of the R&D option Project budgets were approved based on calculations of expected returns (e.g., ROI) The selection of the R&D-option was mostly based on analyses of future returns We mainly considered the potential odds of the project We compared different R&D options on the basis of return and profit forecasts (e.g., comparison of expected value of return)

33

AoM-Paper_Effectuation R&D-Kuepper-09.doc
*) We did hardly perform systematic analyses of external parameters; we rather decided intuitively 5) Decisions on capital expenditures were primarily based on potential risks of losses We carried out systematic analyses of all external parameters and trends Decisions on capital expenditures were primarily based on potential returns

Principle 3: Reduce vs. identify uncertainty ( = .82) Effectuation

1) We tried to reduce risks of the R&D project through internal or external partnerships and agreements *) We did not perform detailed market and competitor analyses to decide about the realization of the project *) We approached potential stakeholders/partners actively to jointly shape the future 2) We jointly decided with our partners/stakeholders on the basis of our competences 3) Our focus was rather on the reduction of risks by approaching potential partners and customers *) Before starting the project, we received binding promises of our customers, suppliers or other stakeholders 4) In order to reduce risks, we started partnerships and received pre-commitments

Causation
We tried to identify risks of the R&D project through thorough market and competitor analyses We performed detailed market and competitor analyses to decide about the realization of the project We have analyzed the market and external trends to better assess future developments We have taken our decisions on the basis of systematic market analyses Our focus was rather on the early identification of risks through market analyses in order to be able to adopt our approach Pre-commitments from e.g., customers, suppliers or other stakeholders like researcher groups have been rarely used In order to identify risks, we focused on market analyses and forecasts

Principle 4: Acknowledge vs. overcome the unexpected ( = .86) Effectuation Causation

*) We always reacted on surprises if necessary the project target was adjusted 1) We always tried to integrate surprising results and findings during the R&D process even though this was not necessarily in line with the original project target 2) Our R&D process was flexible enough to be adjusted to new findings 3) New R&D findings influenced the project target 4) The project planning was carried out in small steps during the project implementation 5) Despite of potential delays in project execution we were flexible and took advantage of opportunities as they arose 6) We allowed the project to evolve as opportunities emerged even though the opportunities have not been in line with the original project target 7) Potential setbacks or external threats were used as advantageous as possible The project target had the highest priority; We tried to avoid potential surprises (e.g., through upfront market analyses) We only integrated surprising results and findings when the original project target was at risk Our R&D process focused on reaching the project target without any delay New R&D findings did not influence the project target The project planning was basically carried out at the beginning of the project We first of all took care of reaching our initially defined project targets without delays We have always paid attention to reach the initial project target By the use of upfront market analyses we tried to avoid setbacks or external threats

Principle 5: Create vs. exploit opportunities ( = .80) Effectuation

1) Before starting the project, we did not carry out detailed analyses concerning future trends 2) We have rather started a new trend than exploiting exogenous trends *) We have rather build up a new market/R&D-segment with our R&D project than we have exploited exogenous trends 3) We did not forecast the evolvement of our R&D-segment through market analyses; the evolvement was initiated by our project 4) We did not carry out any analyses on future trends because we could control the development and trends in our concerned field of R&D due to the fact that we and our partners have been the active drivers

Causation
Before starting the project, we could identify trends through concise analyses and forecasts The R&D-project was our answer on existing trends We have rather exploited external trends than we have build up a new market/R&D-segment with our R&D project We recognized the evolvement of our R&D-segment through analyses and used this trend We carried out elaborate analyses on future trends because we felt that a better understanding of future trends would give us more control over such developments

Performance: Process efficiency ( = .72)

1) Met project schedule 2) Stayed on budget 3) Met operational and technical performance of the R&D process

Performance: Output Experiences and competencies ( = .76)

1) Learnings and expertise that can be leveraged in other projects 2) Generation of new ideas as starting point of potential future projects

34

AoM-Paper_Effectuation R&D-Kuepper-09.doc
3) Enhancement of competencies and capabilities

Perceived value and future potentials ( = .85)

1) 2) *) 3) Perceived value of the R&D output Opportunities to market R&D output Enhancement of reputation of our organization Quality and performance of the R&D output

Innovativeness Technological innovativeness (formative construct: mean score used to differentiate groups)
1) 2) 3) 4) We did not possess the required technological know-how at the beginning of the R&D project We did not have any practical experiences in the application of the required technological competencies and the technological know-how Our organization could not use existing technological competencies and experiences during the project The R&D expenses for the considered R&D output have been above the average compared to previous projects

Market innovativeness (formative construct: mean score used to differentiate groups)

1) 2) 3) 4) 5) The degree of novelty of the R&D output was very high compared to previous products The R&D output aimed at many new customers to our organization The R&D output catered to new customer needs that we have not served before The new product required to use new sales and distribution channels The sales and marketing expenses for the R&D output have been above the average compared to previous projects

*) Item eliminated during the scale development and refinement process

REFERENCES Atuahene-Gima, K. 2005. Resolving the Capability-Rigidity Paradox in New Product Innovation. Journal of Marketing, 69: 61-83. Augier, M. S., & Sarasvathy, S. D. 2004. Integrating cognition, evolution, and design: extending Simonian perspectives to strategic organization. Strategic Organization, 2(2): 169-204. Bagozzi, R. 1979. The Role of Measurement in Theory Construction and Hypothesis Testing: Toward a Holistic Model, in: Ferrell, O., Brown, S., Lamb, C. (Eds.). Conceptual and Theoretical Developments in Marketing, Chicago: 15-32. Bagozzi, R., Phillips, L. 1982. Representing and Testing Organizational Theories: A Holistic Construal. Administrative Science Quarterly, 27: 459-489. Bagozzi, R., Yi, Y., Phillips, L. 1991. Assessing Construct Validity in Organizational Research. Administrative Science Quarterly, 36: 421-458. Bartholomew, S., Smith, A. D. 2006. Improving Survey Response Rates from Chief Executive Officers in Small Firms: The Importance of Social Networks. Entrepreneurship: Theory and Practice, 30: 83-96. Baruch, Y. 1999. Response Rate in Academic Studies A comparative Analysis. Human Relations, 52: 421-438. Benner, M. J., & Tushman, M. L. 2003. Exploitation, Exploration, and Process Management: The Productivity Dilemma Revisited. Academy of Management Review, 28(2): 238-256. Bollen, K. A. 1989. Structural equations with latent variables. New York: Wiley. Bonner, J. M., Ruekert, R. W., & Walker, O. C. 2002. Upper Management Control of New Prod35

AoM-Paper_Effectuation R&D-Kuepper-09.doc uct Development Projects and Project Performance. Journal of Product Innovation Management, 19: 233-245. Brown, S. L., & Eisenhardt, K. M. 1995. Product Development: Past Research, Present Findings, and Future Directions. Academy of Management Review, 20: 343-378. Cheng, Y. T., & Van De Ven, A. 1996. Learning the Innovation Journey: Order Out of Chaos? Organization Science, 7(6): 593614. Constant, E. 2000. Recursive Practice and the Evolution of Technological Knowledge. In J. Ziman (Ed.), Technological Innovation as an Evolutionary Process: 219-233. Cambridge, UK: Cambridge University Press. Coombs, R., McMeekin, A., & Pybus, R. 1998. Toward the development of benchmarking tools for R&D project management. R&D Management, 28(3): 175-186. Cooper, R.G. 1990. Stage Gate Systems: A New Tool for Managing New Products. Business Horizons, 33(3): 44-53. Cooper, R. G., & Kleinschmidt, E. 1986. An Investigation into the New Product Process: Steps, Deficiencies, and Impact. Journal of Product Innovation Management, 3: 71-85. Danneels, E. 2002. The Dynamics of Product Innovation and Firm Competencies. Strategic Management Journal, 23: 1095-1121. Danneels, E., & Kleinschmidt, E. J. 2001. Product innovativeness from the firms perspective: Its dimensions and their relation with project selection and performance. Journal of Product Innovation Management, 18: 357-373. Dawes, J. 1999. The Relationship between Subjective and Objective Company Performance Measures in Market Orientation Research: Further Empirical Evidence. Marketing Bulletin, 10: 65-75. Dess, G. S., & Robinson, R. B. 1984. Measuring Organizational Performance in the Absence of Objective Measures. Strategic Management Journal, 5:265-273. DeVellis, R. F. 2003. Scale Development Theory and Application, 2nd edition, London: Sage. Dewar, R. D., & Dutton, J. E. 1986. The adoption of radical and incremental innovations: an empirical analysis. Management Science, 32: 1422-1433. Dew, N., & Sarasvathy, S. D. 2007. Innovations, Stakeholders & Entrepreneurship. Journal of Business Ethics, 74(3): 267-283. Dew, N., Sarasvathy, S. D., & Venkataraman, S. 2004. The economic implications of exaptation. Journal of Evolutionary Economics, 14(1): 69-84. Dougherty, D. 1992. Interpretive barriers to successful product innovation in large firms. Organization Science, 3: 179-202. Dvir, D., & Shenhar, A. J. 1992. Measuring the success of technology based strategic business units. Engineering Management Journal, 4(4): 33-38. Eckert, J. L. 1996. Preparing to manage an international project. SRA Journal, 28(1/2): 33-36. 36

AoM-Paper_Effectuation R&D-Kuepper-09.doc Eisenhardt, K. M., & Tabrizi, B. N. 1995. Accelerating adaptive processes product innovation in the global computer industry. Administrative Science Quarterly, 40(1): 84-110. Elias, A. A., Cavana, R. Y., & Jackson, L. S. 2002. Stakeholder analysis for R&D project management. R&D Management, 32(4): 301-310. Forbes, D. P. 2007. Reconsidering the strategic implications of decision comprehensiveness. Academy of Management Review, 32(2): 361-376. Galbraith, J. R. 1977. Organization Design. Reading, MA: Addison-Wesley. Garcia, R., & Calantone, R. 2002. A critical look at technological innovation typology and innovativeness terminology: a literature review. Journal of Product Innovation Management, 19: 110-132. Gatignon, H., Tushman, M. L., Smith, W., & Anderson, P. 2002. A Structural Approach to Assessing Innovation: Construct Development of Innovation Locus, Type and Characteristics. Management Science, 48(9): 1103-1122. Gifford, W. E., Bobbitt, H. R., & Slocum Jr., J. W. 1979. Message Characteristics and Perceptions of Uncertainty by Organizational Decision Makers. Academy of Management Journal, 22(3): 458-481. Green, S. G., Gain, M. B., & Aiman-Smith, L. 1995. Assessing a Multidimensional Measure of Radical Technological Innovation. IEEE Transactions on Engineering Management, 42(3): 203-214. Griffin, A. 1997. The Effect of Project and Process Characteristics on Product Development Cycle Time. Journal of Marketing Research, 34: 24-35. Hammer, M., & Stanton, S. 1999. How process enterprises really work. Harvard Business Review, 77(6): 108-118. Harry, M. J., & Schroeder, R. 2000. Six Sigma: The breakthrough management strategy revolutionizing the world's top corporations. New York: Currency. Hauschildt, J., & Schlaak, T. M. 2001. Zur Messung des Innovationsgrades neuartiger Produkte. Zeitschrift fr Betriebswirtschaft, 71: 161-182. Henderson, R. M., & Clark, K. B. 1990. Architectural innovation: the reconfiguration of existing product technologies and the failure of established firms. Administrative Science Quarterly, 35(1): 9-30. House, R. J. 1963. Research Criteria and Methods for the Development of Management Theory. Academy of Management Proceedings, 7-13. Ittner, C., & Larcker, D. F. 1997. Product Development Cycle Time and Organizational Performance. Journal of Marketing Research, 34: 13-33. Jelinek, M., & Schoonhoven, C. B. 1990. The Innovation Marathon: Lessons from high technology firms. San Francisco, CA: Jossey-Bass. Keil, M., Tan, B. C. Y., Wei, K.-K., Saarinen, T., Tuunainen, V., & Wassenaar, A. 2000. A 37

AoM-Paper_Effectuation R&D-Kuepper-09.doc Cross-Cultural Study on Escalation of Commitment Behavior in Software Projects. MIS Quarterly, 24(2): 299-325. Laufer, A. 1997. Simultaneous management: Managing projects in a dynamic environment. AMACOM, Boston. Lewis, M. W., Welsh, M. A., Dehler, G. E., & Green, S. G. 2002. Product Development Tensions: Exploring Contrasting Styles of Project Management. Academy of Management Journal, 45: 546-564. MacCormack, A., & Verganti, R. 2003. Managing the Sources of Uncertainty: Matching Process and Context in Software Development. Journal of Product Innovation Management, 20: 217-232. Marquis, D. G. 1969. The anatomy of successful innovations. Innovation (November). In M. L. Tushman, W. L. Moore (Eds.), Readings in the Management of Innovation: 79-87. Cambridge, MA: Ballinger. Miller, W. L. 1995. A broader mission of R&D. Research and Technology Management, 38(6): 24-36. Montoya-Weiss, M. M., & Calantone, R. J. 1994. Determinants of New Product Performance: A Review and Meta-Analysis. Journal of Product Innovation Management, 11(5): 397417. Moorman, C., & Miner, A. S. 1998. The Convergence of Planning and Execution: Improvisation in New Product Development. Journal of Marketing, 62: 1-20. Myers, S., & Marquis, D. G. 1967. Technological Innovation: Its Environment and Management. Washington, DC: US Department of Commerce. Naveh, E. 2007. Formality and discretion in successful R&D projects. Journal of Operations Management, 25: 110-125. Pearce, J. A., Robbins, D. K., & Robinson, R. B. 1987. The Impact of Grand Strategy and Planning Formality on Financial Performance. Strategic Management Journal, 8: 125-134. Peter, J., & Churchill, G. 1986. Relationships among Research Design Choices and Psychometric Properties of Rating Scales: A Meta-Analysis. Journal of Marketing Research, 18: 133145. Pinto, J. K., & Mantel, S. J. 1990. The causes of project failure. IEEE Transactions on Engineering Management, 37(4): 269-276. Read, S., Song, M., & Smit, W. 2008. A meta-analytic review of effectuation and venture performance. Journal of Business Venturing, doi:10.1016/j.jbusvent.2008.02.005. Robins, J. A., & Atuahene-Gima, K. 2003. The effects of external relationships on innovative activity within firms. Academy of Management Best Conference Paper, G1-G6. Rogers, D. M. A. 1996. The challenge of fifth generation R&D. Research and Technology Management, 39(4): 33-41. 38

AoM-Paper_Effectuation R&D-Kuepper-09.doc Rothwell, R., Freeman, C., Horlsey, A., Jervis, V. T. P., Robertson, A. B., & Townsend, J. 1974. SAPPHO updated project SAPPHO phase II. Research Policy, 3: 258-291. Salomo, S., Weise, J., & Gemnden, H. G. 2007. NPD Planning Activities and Innovation Performance: The Mediating Role of Process Management and the Moderating Effect of Product Innovativeness. Journal of Product Innovation Management, 24: 285-302. Sarasvathy, S. D. 2001. Causation and Effectuation: Toward a theoretical shift from economical inevitability to entrepreneurial contingency. Academy of Management Review, 26(2): 243-263. Sarasvathy, S. D. 2003. Entrepreneurship as a science of the artificial. Journal of Economic Psychology, 24(2): 203-220. Sarasvathy, S. D. 2008. Effectuation: Elements of Entrepreneurial Expertise. Northampton, MA: Edward Elgar. Sarasvathy, S. D., Dew, N., Read, S., & Wiltbank, R. 2008. Designing Organizations that Design Environments: Lessons from Entrepreneurial Expertise. Organization Studies, 29(3): 331-350. Sarasvathy, S. D. & Kotha, S. 2001. Dealing with Knightian uncertainty in the new economy: The RealNetworks case. In J. Butler (Eds.), Research on Management and Entrepreneurship: 31-62. Greenwich, CT: IAP Inc. Seidel, V. P. 2007. Concept Shifting and the Radical Product Development Process. Journal of Product Innovation Management, 24: 522-533. Sethi, R., & Iqbal, Z. 2008. Stage-Gate Controls, Learning Failure, and Adverse Effect on Novel New Products. Journal of Marketing, 72: 118134. Shenhar, A. J. 2001. One size does not fit all projects: exploring classical contingency domains. Management Science, 47(3): 394-414. Shenhar, A. J., Tishler, A., Dvir, D., Lipovetsky, S., & Lechler, T. 2002. Refining the search for project success factors: a multivariate, typological approach. R&D Management, 32(2): 111-126. Song, X. M., & Montoya-Weiss, M. M. 1998. Critical Development Activities for Really New versus Incremental Products. Journal of Product Innovation Management, 15: 124-135. Song, X. M., & Montoya-Weiss, M. M. 2001. The effect of perceived technological uncertainty on Japanese new product development. Academy of Management Journal, 44(1), 61-80. Song, X. M., & Parry, M. E. 1997. The Determinants of Japanese New Product Successes. Journal of Marketing Research, 34(1): 64-76. Tatikonda, M. V., & Montoya-Weiss, M. M. 2001. Integrating operations and marketing perspectives of product innovation: The influence of organizational process factors and capabilities on development performance. Management Science, 47: 151-172. Tatikonda, M. V., & Rosenthal, S. R. 2000. Successful execution of product development pro39

AoM-Paper_Effectuation R&D-Kuepper-09.doc jects: Balancing firmness and flexibility in the innovative process. Journal of Operations Management, 18: 401-425. Tipping, J. W., Zeffren, E., & Fusfeld, A. R. 1995. Assessing the value of technology. Research and Technology Management, 38(5): 22-39. Utterback, J. M., Allen, T. J., Hollomon, J. H., & Sirbu, M. A. 1976. The Process of Innovation in Five Industries in Europe and Japan. IEEE Transactions on Engineering Management, 23(1): 3-9. Wheelwright, S. C., & Clark, K. B. 1992a. Creating project plans to focus product development. Harvard Business Review, March/April, 70-82. Wheelwright, S. C., & Clark, K. B. 1992b. Revolutionizing Product Development, New York. Wiltbank, R., Read, S., Dew, N., & Sarasvathy, S. D. 2008. Prediction and control under uncertainty: Outcomes in angel investing. Journal of Business Venturing (2008), doi:10.1016/j.jbusvent.2007.11.004. Wiseman, J. 1989. Cost, Choice, and Political Economy. Aldershot: Edward Elgar. Wolfe, R. A. 1994. Organizational Innovation Review, Critique, and Suggested Research Directions. Journal of Management Studies, 31(3): 405-431. Zaltman, G., Duncan, R., & Holbek, J. 1973. Innovations and Organizations. New York: John Wiley & Sons, Inc. Zirger, B. J., & Maidique, M. A. 1990. A model of new product development: An empirical test. Management Science, 36(7): 867-883.

40

AoM-Paper_Effectuation R&D-Kuepper-09.doc Table 1: Delineation of effectuation and causation: Key principles

Principle Principle 1: Means vs. goals Principle 2: Affordable loss vs. expected returns Principle 3: Reduce vs. identify uncertainty Principle 4: Acknowledge vs. overcome the unexpected Principle 5: Create vs. exploit opportunities Effectuation characteristics R&D approach driven by given means R&D approach guided by in-advance commitments to what one is willing to lose Existing uncertainty reduced through partnerships and pre-commitments of stakeholders Contingencies/surprises seen as source of opportunities Human agency seen as prime driver of future developments Causation characteristics R&D approach driven by given project targets R&D approach guided by expected project returns Existing uncertainty identified and avoided through market and competitor analyses Contingencies/surprises avoided or quickly overcome to reach given project targets Development/trends seen as exogenously given that can be exploited by use of forecasts

Table 2: Discriminant validity: Squared interscale correlations are greater than AVE
Principle 1 Principle 1 Principle 2 Principle 3 Principle 4 Principle 5 Proc. efficiency Output AVE .60 .58 .56 .50 .54 .56 .64 .60 .23 .03 .13 .12 .04 .14 Principle 2 .58 .05 .05 .15 .03 .11 Principle 3 Principle 4 Principle 5 .56 .50 .54 Squared interscale correlations .03 .29 .01 .02 Proc. efficiency .56 Output .64

.12 .08 .02

.02 .12

.14

Table 3: Two-Group Maximum Likelihood Path Coefficient Estimatesa, b

Hypothesis a) H1 b) a) H2 b) a) H3 b) a) H4 b) a) H5 b)
a b c

Path From Principle 1 (Effec.) Principle 1 (Caus.) Principle 2 (Effec.) Principle 2 (Caus.) Principle 3 (Effec.) Principle 3 (Caus.) Principle 4 (Effec.) Principle 4 (Caus.) Principle 5 (Effec.) Principle 5 (Caus.) To Output Output Process efficiency Process efficiency Output Output Output Process efficiency Process efficiency Process efficiency

Innovativeness High Low High Low High Low Low/high Low/high High Low

Standardized Coefficient -.03 .24**** -.25** .25** -.14*** .06 -.11**/-.23**** .02/.28**** -.19*** .22*

Significance level of differencec *** *** ** (*) ** ***

Positive signs (+) correspond to Causation and negative signs (-) to Effectuation. Overall fit indices: /df = 1.59; p < .001; RMSEA = .04 (pCLOSE = 1.0); SRMR =.05; TLI =.95; CFI =.96; GFI =.90 The significance of differences between the path coefficients in the low and high innovativeness groups have been tested using the formula following Keil et al. (2000). The results were confirmed with a chi-square difference test (likelihood ratio test). Significant levels: **** p .01; *** p .05; ** p .10; * p .15; (*) p .24

41

AoM-Paper_Effectuation R&D-Kuepper-09.doc Figure 1: Dynamic model of effectuation

Decision/project preparation Decision making Decision/project implementation New insights/ surprises?

Principle 1: Given means

Principle 5: Human action is the predominant factor shaping the future (Sarasvathy, 2008: 87) Principle 2: Affordable loss as criterion for decision making Principle 4: Need to change the initial plan

New means that can impact on goals

Adjustment of initial decision

Principle 3: Interact with others and get stakeholder commitments

New stakeholder/ means required?

Figure 2: Research model

Moderator Innovativeness Effectuation/Causation characteristics Principle 1 Principle 2 Principle 3 Principle 4 Principle 5 high H1a H2a H3a low H1b H2b H3b Output R&D project performance Process efficiency

H4a, H4b H5a H5b

Figure 3: Classification of sub-groups for further analysis

High II N = 64 Technological innovativeness I Domain of effectuation N = 136

Low

III Domain of causation N = 136

IV N = 64

Low High Market innovativeness

42

AoM-Paper_Effectuation R&D-Kuepper-09.doc Additional Appendix for reviewers (not intended for publication) Statistics for newly developed scales (N = 400)
Factor Item Item loadings ,769 ,619 ,867 ,774 ,839 ,863 ,644 ,714 ,759 ,820 ,698 ,810 ,798 ,866 ,647 ,653 ,628 ,753 ,616 ,652 ,783 ,804 ,876 ,560 ,757 ,711 Item-tototal correlation ,706 ,477 ,803 ,734 ,818 ,811 ,646 ,660 ,602 ,731 ,588 ,778 ,622 ,689 ,629 ,625 ,605 ,707 ,552 ,543 ,768 ,757 ,577 ,519 ,675 ,672 Cronbachs Alpha ,903 Variance explained through first factor (exploratory factor analysis) 64,1% Factor reliability ,911 AVE

Principle 1

Principle 2

Principle 3

Principle 4

Principle 5

1 2 3 4 5 6 7 1 2 3 4 5 1 2 3 4 1 2 3 4 5 6 1 2 3 4

,598

,856

63,8%

,873

,580

,820

64,9%

,832

,558

,861

59,5%

,858

,504

,800

62,7%

,821

,540

43