پابرجایی شرکت : کنش کارآفرینانه و تأثیر آن بر عملکرد پروژه R & D
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|9483||2012||18 صفحه PDF||سفارش دهید||13310 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Journal of Business Venturing, Volume 27, Issue 2, March 2012, Pages 167–184
Innovative products are widely recognized as an important source of competitive advantage. However, many companies have difficulties finding efficient and successful approaches to different types of R&D projects, particularly those that involve a high level of innovativeness. Therefore, the present study moves effectuation theory from the entrepreneurial context to R&D research. First, the characteristics of an effectual approach in the context of R&D projects are developed and differentiated from those of conventional prediction-based strategies (causation). Second, using a thorough qualitative and quantitative scale-development process to capture particularities of effectual and causal dimensions in the R&D context, expert interviews and a pilot study (123 R&D projects), the study develops a multi-factor measurement model of effectuation and causation. 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 to success in highly innovative contexts, (b) causation approaches are beneficial in projects with low levels of innovativeness. Overall, this study moves the effectuation logic from the entrepreneurial to the corporate R&D context, captures its particularities, and investigates its performance outcomes.
The dichotomous concepts of effectuation and causation provide a framework for building processes in contexts with varying degrees of uncertainty. Effectuation refers to processes that start with “a set of means as given and focus on selecting between possible effects that can be created with that set of means” while causation builds on prediction and processes that “take a particular effect as given and focus on selecting between means to create that effect” (Sarasvathy, 2001; p. 245). With innovativeness providing significant uncertainty in R&D projects (Tatikonda and Montoya-Weiss, 2001), the setting offers a unique environment to examine the iterative effectual problem-solving process and contrast it with what can be referred to as the causal rational plan approach (Brown and Eisenhardt, 1995). In extant R&D literature, there is some agreement about success factors of corporate R&D projects, but it is limited to projects with low levels of innovativeness (e.g. Lewi and Smith, 2007 and Khurana and Rosenthal, 1998). In particular, planning activities have shown to be beneficial for these R&D projects (Brown and Eisenhardt, 1995 and Hayes et al., 1988). For more innovative R&D projects, however, the literature produces contradictory findings. Fredrickson and Mitchell (1984) study the role of rationality in organizations and identify that comprehensive planning efforts are negatively related to firm performance when innovativeness is high. On the other hand, Shenhar et al. (2002) find in their exploratory, data-driven approach that planning efforts are particularly important for projects that involve higher levels of innovativeness. In addition to the contradictory findings on planning as a success factor in more innovative R&D projects, there is a dearth of literature on broader R&D project success factors beyond planning, such as how processes and problem-solving strategies shape project outcomes (Song and Montoya-Weiss, 2001 and Eisenhardt and Tabrizi, 1995). Through the lens of effectuation and causation we gain meaningful contributions to the burgeoning literature on R&D projects. At the same time, we are able to add empirically grounded theoretical insights to the discourse on effectuation which has – after its initial delineation by Sarasvathy (2001) – largely been of conceptual and qualitative nature (e.g., Sarasvathy and Kotha, 2001). In 2009, first steps towards quantitative assessment were made (e.g., Read et al., 2009b and Wiltbank et al., 2009) and in 2010, Chandler, DeTienne, McKelvie and Mumford developed a measurement scale in the entrepreneurial context. Our research adds to the literature in three important ways: First, we apply the theoretical lens of effectuation to the empirical setting of the corporate R&D context in order to gain new theoretical insights into the relationship of effectuation and processes of innovation. Second, we develop measurement models to show the applicability of effectuation theory in the corporate context and link components of the effectuation theory to performance measures in order to validate empirically its effect on performance. Third, since effectuation theory has not yet been applied to the corporate context, we adhere closely to Sarasvathy's (2001) original conceptualization of effectuation by integrating the four key dimensions: means, affordable loss, partnerships and acknowledging the unexpected (Read et al., 2009a and Read et al., 2009b) to extend the application of effectuation into a meaningful new context of larger firms. We proceed as follows: First, we present the theoretical foundations and lay out why effectuation theory is applicable to the R&D context. Then, we derive a research model that links effectuation and causation practices with R&D performance. Next, we describe the research method and the study context, followed by the development and evaluation of measures. We develop a research instrument that operationalizes Sarasvathy's (2001) articulation of effectuation and causation. We then apply the scale to test a set of hypotheses about the relationship of R&D processes and project performance. The scale development is necessary in order to accommodate the fact that effectuation processes may look different in R&D than in entrepreneurial contexts. Next, we present the results of tests of the measures and of the integrated 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 its contributions to the literatures on R&D processes, innovation and effectuation.
نتیجه گیری انگلیسی
The objective of this study is to analyze the effects of effectuation and causation practices in the corporate R&D context. In this section we discuss the contributions of our study, first in terms of its contribution to R&D literature and then in terms of its contribution to effectuation literature. Subsequently, we derive managerial implications, present the study's limitations, and suggest avenues for further research. 7.1. Contributions to R&D literature Our study takes into account that specific circumstances, such as different degrees of innovativeness, require different R&D approaches, but it is the integrated view of effectual and causal processes that differentiates our approach from the current R&D literature. The concepts of effectuation and causation offer guidelines for decision-making in R&D projects with varying degrees of innovativeness. Further, whereas most of the existing frameworks assume that exogenous factors drive the R&D process, we position human action and the organization at the center of the process, thereby acknowledging that control can be actively built using effectual elements, such as starting the process based on individual means and competences. This positioning is clearly in contrast to the existing frameworks, which deal with uncertainty by recommending quick adaptations to unforeseen developments and surprises. To illustrate our view and to shed more light on the role of effectuation in R&D literature, we discuss the implications of effectuation for five major concepts in this literature stream: complex organizational dynamics in innovation, cross-functional integration, ambidexterity, the fuzzy front-end, and experiential strategy. These literature streams include elements of effectuation, so they have the potential to be advanced through the effectuation lens. The study contributes to complexity literature, which has generated insights on the emergence of innovations (Dooley and Van de Ven, 1999 and Fleming and Sorenson, 2001). This literature stream states that the emergence of innovations occurs in nonlinear and temporally complex manners (Lichtenstein et al., 2007). Dooley and Van de Ven (1999) discuss four patterns of organizational dynamics in innovation emergence: white noise, pink noise, periodicity, and chaos. Periodicity and chaos occur in the later stages of the innovation process when control or cooperation mechanisms are in place so that only few variables interact. In contrast, white and pink noise occur in early-stage innovation processes and refer to a large number of participants in the innovation search process who are either loosely connected through feedback mechanisms (white noise) or somewhat connected through local feedback loops (pink noise). The system moves from white noise to pink noise when participants start making use of their personal networks, thereby employing effectual ideas along the means-based and stakeholder principles. Since Dooley and Van de Ven (1999) complain that “organizational scientists have not considered more complex and non-linear dynamic models of organizational change” (p. 369), we see the potential for effectuation to add explanatory value to the discussion of the reasons that non-linear and interdependent processes, such as those in early or uncertain innovation stages, occur. A major tenet of R&D research is that cross-functional integration is beneficial, especially in cases where high levels of uncertainty occur, since different functional perspectives provide different approaches to problem-solving (e.g., Olson et al., 2001). Cross-functional integration claims that, when crucial resources are missing and uncertainty is high, the integration into the process of other functions that may be able to provide the missing resources is a major driver of organizational performance. Our findings add that these stakeholders can be either within the organization or outside it. Since the integration of outside experts is not an inherent part of the cross-functional integration concept because this concept focuses on internal functions or departments of the organization, the use of effectuation in the R&D context adds some new elements to the cross-functional integration construct. Our study also contributes to the literature on ambidexterity (Duncan, 1976). Ambidexterity suggests the necessity of combining planned approaches (causation) and adaptive approaches, but the discussion of ambidexterity has not stretched toward the inclusion of effectuation approaches. Gibson and Birkinshaw (2004) develop the concept toward contextual ambidexterity, referring to potential changes in the environment of an organization as a source of uncertainty, while our study suggests broadening this view to internal processes that also create sources of uncertainty. R&D projects with high innovativeness are usually chosen to strengthen the innovation base of the organization, but they are characterized by uncertainty, creating the necessity for organizations to develop structures that combine effectual processes with causal and adaptive processes. Effectual cells (Wiltbank et al., 2009) are just a first conceptualization of an organizational element that can support effectuation processes in the R&D context. Transferring the concept of effectuation comprehensively to the R&D context facilitates reconsideration of the early process stages of R&D: the fuzzy front-end (Reid and de Brentani, 2004). The fuzzy front-end is defined as “the period between when an opportunity is first considered and when an idea is judged ready for development” (Kim and Wilemon, 2002; p. 270), a description that matches that of the problem space of highly innovative R&D projects in this study. Although the discussion of the fuzzy front-end is still dominated by causal reasoning and an adaptation of causal instruments to the restrictions of the fuzzy situation, this literature stream touches upon elements that resemble effectuation. Kim and Wilemon (2002) mention the importance of notions of means (a knowledgeable team, internal cooperation and support), affordable loss (consider low-cost possibilities until the best product emerges), partnerships (customer involvement, cooperation with suppliers and intermediaries, and horizontal cooperation with competitors) and acknowledging contingency (consider many possibilities) as important in successful fuzzy front-end management. Our results suggest that there is also value in looking at effectuation as a means to manage the fuzzy front-end of the innovation process. Finally, in a comparable approach, Eisenhardt and Tabrizi (1995) pick up on the notion of experiential strategy in the face of fuzziness, which they label “fogginess.” They develop two strategies to follow, depending on the specific context. The first, called “compression strategy,” is related to projects that involve a low degree of innovativeness, which can be planned and executed in pre-defined steps. The second strategy, called “experiential strategy,” is related to circumstances with a higher degree of innovativeness and is characterized by an iterative process to account for “foggy and shifting markets and technologies” (Eisenhardt and Tabrizi, 1995; p. 91). The authors find that the experiential strategy accelerates product development in the foggy context because of its reliance on individual means and the idea of partnership (a powerful project leader and a multifunctional team). They also propose multiple design iterations and extensive testing, both of which hint at the notion of acknowledging the unexpected. However, in contrast to affordable loss as the regulatory element in effectuation, the causal element of frequent project milestones complements the experiential strategy. In sum, we conclude that effectuation not only has a place in R&D literature but that it opens the door to an exploration of a variety of existing concepts through the lens of effectuation. 7.2. Contributions to effectuation literature This study contributes to effectuation literature in several ways: First, we transfer effectuation from its original field of entrepreneurship research to another discipline, adhering to the general understanding of effectuation theory but incorporating the particular context of R&D projects into a scale-development process. The study broadens the context of effectuation theory to the corporate R&D context, thereby increasing its generalizability to what might become a theory of decision-making and action under uncertainty. Our descriptive statistics show that the effectual dimensions are already applied in the R&D context (Table 4). Second, while Chandler et al. (2011) develop an experimentation measure that captures various facets of the effectual process, we differentiate between dependent and independent variables. In our case, the independent variables contain the description of the effectual vs. causal process, while the dependent variables are the outcomes. Acknowledging and examining performance differentials increase the relevance of effectuation theory and expands it from a theory of mere description of entrepreneurial behavior to a theory that identifies performance-enhancing measures. Three out of four dimensions (affordable loss, partnerships, and acknowledging the unexpected) are related to either R&D output or efficiency. Third, while analogies to the resource-based view and the absorptive capacity construct of “means-driven” suggest that “means-driven” can be positively related to R&D output, our findings do not find a relationship between “means-driven” and performance, possibly because it is less important that R&D projects focus on means (as opposed to goals) than on what is being done with existing means. This explanation is in line with Chandler et al. (2011), who name “experimentation” as the first dimension of effectuation in their scale-building process, defining experimentation as finding the best purpose for some given means. Further, Read et al. (2009b) find that only those means that relate to the “domain-specific expertise” (p. 582) of a venture are relevant to performance outcomes. Thus, based on Read et al. (2009b), we can explain our non-significant relationship between “means-driven” and R&D output as indicating, first, that not just means per se, as measured in our study, but only very specific means become performance-drivers and, second, that it is not as important that an R&D project focuses on means as it is that the R&D team experiments with the given set of means. Fourth, our research departs from Chandler et al. (2011) by adhering more closely to the original understanding of the effectuation concept. In particular, we include the notion of co-creation in the partnering dimension by inquiring into joint decision-making. 7.3. Managerial implications A strict process management approach that includes concepts like Total Quality Management and Lean Management has dominated the managerial practice during the last two decades. However, researchers are increasingly hinting that managers who seek only to achieve productivity gains are not well informed in the long run (e.g., Benner and Tushman, 2003). Proponents of process management promote a productivity focus as being universally beneficial for organizations, but researchers raise concerns that the “diffusion of process management techniques favors exploitative innovation at the expense of exploratory innovation” (Benner and Tushman, 2003; p. 239). Without doubting that process management activities in the context of R&D projects provide clear advantages for exploitative projects, we show that exploitative benefits have to be questioned in exploratory projects because R&D managers need to differentiate between projects with low and high degrees of innovativeness. Whereas a strictly planned and prediction-based causal approach is suitable at low levels of innovativeness, effectuation offers better ways to deal with projects with high levels of innovativeness. In addition, our discussion on barriers to innovation suggests rethinking internal processes in favor of forming effectual cells that allow for a more means-oriented innovation approach. Leaving room for individual projects that are means-based – that is, entrepreneurial – may be the key to driving corporate innovation capabilities. 7.4. Limitations and avenues for further research This study covers a broad field of effectuation and causation characteristics in order to get to a first empirical foundation of its key constructs for the corporate R&D context. However, this approach was carried out at the expense of a more detailed analysis of each dimension, so we suggest additional in-depth analyses on each dimension as future research opportunities. Specifically, we see potential in intensifying the discussion of the means dimension with regard to its potential for overcoming barriers to innovation and of the affordable loss dimension with regard to its potential to have a diversification effect on the R&D project portfolio. More detailed research is also needed for the other two dimensions. We do not find relationships for competitive market analysis and R&D output in cases of low innovativeness, and for acknowledging the unexpected and R&D efficiency in cases of high innovativeness. The former suggests that a preference for competitive market analysis is not sufficient and either does not play the key role in the low innovativeness context at all or needs to be complemented by other actions, e.g. building partnerships. The latter points to the fact that openness and flexibility in fact do not only have positive efficiency effects in terms of saving time for gathering information and major adaptations in the process. They could also show negative efficiency effects in terms of establishing a flexible structure in the first place and moving as a group through a flexible process, which certainly demands a lot of communication along the way. Research on the notion of performance is also needed. Since both the causal and the effectual process must fulfill predominant and well-established company rules, we were confident in using causal success measures in our study, and effectual dimensions proved to be at least partly positively related to these measures. Still, we see potential in broadening the discussion of performance toward more effectuation-related measures (Sarasvathy, 2008), including measuring the success of individual project leaders or teams over more than one project. Further, future studies could incorporate objective performance measures; as we argued earlier, comparing R&D projects in terms of their individual financial measures (e.g., profitability) can be difficult given divergent project objectives (Blindenbach-Driessen et al., 2010), but objectives on the firm level tend to be similar because all companies, need to achieve a certain level of profitability. Therefore, future studies could envision measuring effectuation and causation (e.g., as an average of all R&D projects pursued) on a company level and incorporate objective performance measures on this level. The analysis of mediating effects may also be relevant. The existing literature examines mediating effects between causation-oriented planning approaches and performance measures and shows that strategic orientations can be a major mediator between causal dimensions and performance measures (e.g., Pulendran et al., 2003). However, one may also argue that the relevant orientations that form mediating effects can differ between causal and effectual dimensions and their effects on performance. While, for example, a “competitive market analysis” could strengthen the market orientation of a firm or a project, R&D stakeholders from a university could generate a stronger innovation orientation. In addition, identification of specific antecedents that support effectuation (e.g., organizational antecedents) would be useful in efforts to implement effectual cells in plan-dominated large organizations since parallel organizational structures are likely to be required. While our study builds upon cross-sectional data, one could argue that it includes non-recursive relationships; for example, while a means-driven approach can lead to a project staying within its budget, pursuing the objective of staying within the budget may lead to the application of a means-driven approach. Following Rindfleisch et al. (2008), future studies could generate longitudinal datasets to elaborate on these reciprocal relationships in greater detail. Further, we use a group comparison method in structural equation modeling, while interaction effects also allow for the depiction of moderating effects. However, while the literature offers stochastic methods to depict interaction effects in structural equation modeling, such as the Bayesian approach (Arminger and Muthén, 1997), the Exact Maximum Likelihood Estimation (Lee and Zhu, 2002) and the Quasi Maximum Likelihood Estimation (Klein and Muthén, 2007), no proven software packages are available to estimate these procedures (Lee et al., 2004). When such software packages are accessible, future studies should examine the robustness of our findings using these alternative methods. Finally, while we used uni-dimensional measures to accommodate feedback from pre-tests, future studies should measure effectuation and causation separately with multi-item constructs to facilitate examination of the consequences of a simultaneous pursuit of effectual and causal dimensions. Further, in order to gain a clear understanding of a respondent's preference for either effectuation or causation, we used even 6-point Likert scales. The literature indicates that valid results can be expected when five or more answer categories are used and that various parameters (e.g., parameter estimations of confirmatory factor analysis; Chang, 1994, DiStefano, 2002 and Weng, 2004) are largely independent of the number of response categories. While future studies should verify whether our findings hold, for example, for odd scales (e.g., 5 point Likert scales), these extant studies suggest that our findings are not likely to be substantially biased by the number of response categories we used.