چگونگی رانش انگیزه، فرصت و توانایی با به اشتراک گذاری دانش : مدل عامل محدود
|کد مقاله||سال انتشار||تعداد صفحات مقاله انگلیسی||ترجمه فارسی|
|4845||2008||20 صفحه PDF||سفارش دهید|
نسخه انگلیسی مقاله همین الان قابل دانلود است.
هزینه ترجمه مقاله بر اساس تعداد کلمات مقاله انگلیسی محاسبه می شود.
این مقاله تقریباً شامل 14409 کلمه می باشد.
هزینه ترجمه مقاله توسط مترجمان با تجربه، طبق جدول زیر محاسبه می شود:
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Journal of Operations Management, Volume 26, Issue 3, May 2008, Pages 426–445
We introduce and empirically test a theoretical metamodel that explains knowledge-sharing behavior among employees. Building on the well-established motivation–opportunity–ability (MOA) framework, we posit that knowledge sharing among employees is a function of their MOA to do so. Existing literature suggests that the interaction among motivation, opportunity, and ability drives knowledge-sharing behavior. In contrast, we specify a new model in which the “bottleneck” or constraining factor among the MOA variables determines the degree of knowledge sharing that occurs. This constraining-factor model (CFM) fits the data better than the traditional multiplicative model and reveals a new, qualitatively different portrait of knowledge sharing that resolves some of the puzzles in the previous literature. The CFM provides macro-level insights with respect to how operations managers can improve employee knowledge sharing by focusing on the bottleneck MOA variable. As a result, the CFM can help set strategic directions of related policies. The model emphasizes that, counter to conventional wisdom, the MOA variables should not be addressed independently, but rather in a dynamic and coordinated way.
The notion of resource constraints has been extensively investigated in the operations management (OM) literature, in part because the identification of constraints enables managers to plan more effective interventions. Bottleneck analysis, for example, identifies constraining resources in a process, so that the capacity of the process can be increased by adding capacity at the bottleneck (Chase et al., 2004). Critical path analysis identifies the set of activities taking the longest time in a project, so that the project can be shortened by crashing activities on the critical path. Although OM-based knowledge of physical resource and time constraints is extensive, much less is understood about how behavioral constraints act in OM contexts. Yet, such constraints can severely limit the effectiveness of managerial interventions (Boudreau et al., 2003). For example, process improvement and just-in-time (JIT) programs can be fruitless without the full participation and motivation of employees (Hackman and Wageman, 1995 and Shah and Ward, 2003). Similarly, behavioral responses can offset the advantages of worker flexibility programs (Schultz et al., 2003 and Siemsen et al., 2007a). In this study, we present a metamodel that can be used to identify behavioral constraints, and we empirically test this meta-model within the operational context of inter-employee knowledge sharing. Our research provides a way of conceptually and empirically identifying and addressing such behavioral constraints. The foundation for our research is the well-known motivation–opportunity–ability (MOA) framework, which has been applied in various management disciplines. Broadly speaking, motivation captures the individual's willingness to act; opportunity represents the environmental or contextual mechanisms that enable action. Ability represents the individual's skills or knowledge base related to the action (Rothschild, 1999). Providing a novel perspective, our research posits that it is the constraining factor among these three MOA variables that ultimately determines behavior. Thus, changes in motivation only affect behavior and outcomes if motivation is the constraining factor; they have little or no impact if either opportunity or ability is constraining. We develop a new modeling approach, which we call the “constraining factor model” (CFM), that embodies this bottleneck perspective. We then empirically test this model's ability to explain knowledge-sharing behavior and evaluate how the CFM performs compared to alternative, existing specifications of the MOA framework in the literature. Our study focuses on the specific context of one-way employee knowledge sharing in a dyadic work relationship. There are four reasons for choosing this context. First, OM researchers have emphasized the importance of better understanding the dissemination of operational know-how and learning (Hayes et al., 1988, Leonard-Barton, 1992, Roth et al., 1994, Roth, 1996, Mukherjee et al., 1998 and Ferdows, 2006). Previous studies have highlighted that employees on the shop floor do not always share their knowledge with their peers (Aeppel, 2002), which makes this context particularly interesting to OM. Second, practitioners have employed many different approaches to the management of knowledge in their organizations (Hansen et al., 1999), but approaches that neglect behavioral constraints are not always successful (Dixon, 2000). Third, existing research has questioned the role of motivation in knowledge sharing (Szulanski, 1996). The CFM enables us to clearly state under what conditions motivation plays less of a role in promoting employee knowledge sharing, thereby clarifying conflicting perspectives in the literature. Finally, the perceptions of individual employees about their intentions to share work-related knowledge with a coworker can be considered a primary building block in this area. Practically speaking, it is usually only known to the employee whether or not she chooses not to share. Even in cases where an employee attempts to share, it is not always clear whether the coworker involved always picks up the knowledge being shared. Arguably, understanding the barriers to an individual's propensity to share knowledge is an important, but understudied area in OM. To illustrate the managerial implications of our research, consider the following three real life examples of knowledge-sharing initiatives. A large public utility was facing a brain drain, as the old guard of engineers was close to retiring. There was a generational gap between these experienced engineers and the junior employees that were hired to replace them. Operations, quality and human resource managers attempted to increase knowledge sharing between the experienced workers and the new recruits, but were stymied. Due to the urgency of the situation, a corporate initiative was put into place to consider what knowledge management initiatives should be prioritized. Should they focus on training, on changing their incentive system, or on providing the time and infrastructure for knowledge sharing to occur? Clearly, the corporate management team needed a guiding framework and empirical data to further their strategic planning process and set directions. Our research provided the firm with both the framework and concrete suggestions that were employed in their strategy. Consider a second example. A management consultancy suffered from high employee turnover. The partners felt that their organization constantly generated and lost important knowledge. To address this situation, they implemented a large-scale corporate intranet, providing instant intranet access for consultants to document and share the lessons they learned from projects. Further, the company employed communication and information experts to help consultants document their knowledge. However, when the system went online, consultants only contributed knowledge of little importance. Even though the system provided an easy opportunity to share knowledge, and experts were readily available to support consultants who lacked the ability to codify their knowledge, the consultants simply had little motivation to share their important knowledge with a broader community. Thus, the managerial intervention did not address the real behavioral bottleneck, resulting in overall failure. The third example has a slightly different context, and pertains to a credit union that desired to improve its overall customer satisfaction levels. Customer satisfaction ratings had reached a plateau at 85%. The former CEO spent precious resources trying to motivate tellers, loan officers, and customer service representatives with various types of monetary and other incentives. Motivational signs and slogans were visible in the break room. The CEO and her management team provided time and space for customer contact employees to meet and have a dialogue. There was very little turnover, since working conditions were good. Yet none of these efforts moved the bar, despite top management commitment. A new CEO was brought in and given the same goal by the Board of Directors to improve customer satisfaction. Soon after his arrival, he made staff development and training his top priority. He brought in external consultants, who actually provided formal training in service quality, teamwork, and process improvement over a 2-year period. Not surprisingly, customer satisfaction rose to 96%—the best in the industry. In this case, whereas ability was the bottleneck, it was not detected because the credit union's employees were technically very competent and experienced—and the former CEO did not realize that service quality training was as important as technical competence. This scenario is consistent with the extensive quality management literature that highlights the importance of training employees in the use of process improvement tools and customer requirements. The MOA framework is well established as a theoretical basis for the explanation of work performance (Blumberg and Pringle, 1982 and Boudreau et al., 2003). It has been successfully employed to explain a wide array of behaviors such as consumer choice (MacInnis et al., 1991), firm-level decision making (Wu et al., 2004), and social capital activation (Adler and Kwon, 2002 and Binney et al., 2006). More recently, it has been used as a conceptual organizing framework for knowledge-management practices (Argote et al., 2003). Despite the popularity of the MOA framework, there are two important puzzles that endure in the related literature. Our research addresses these puzzles. First, the MOA framework highlights the importance of motivation as a driver of behavior (action), or more specifically in the context of this research, an employee's propensity to share knowledge. Clearly, many operations’ infrastructural policies take the importance of motivation for granted. Yet, some empirical research on the dissemination of best practices has questioned the importance of motivation in explaining the successful sharing of knowledge (Szulanski, 1996 and Szulanski, 2000). The puzzle, therefore, is this: if motivation is a construct of theoretical importance in the context of inter-employee knowledge sharing, why does the empirical evidence raise questions about its role as a major driver of knowledge sharing? Second, work performance theory indicates that motivation, opportunity, and ability should play complementary1 roles in influencing behavior (Cummings and Schwab, 1973). In this view, without ability or opportunity, motivation alone should not lead to knowledge-sharing behavior. Yet again, there is little empirical evidence supporting the existence of such complementarity (Terborg, 1977). This inconsistency leads us to the second puzzle: if, as work performance theories predict, there is complementarity among motivation, opportunity, and ability in driving behavior, why have existing empirical tests of the MOA framework often failed to reveal this complementarity? We attempt to resolve these two puzzles using the CFM. Importantly, our research is part of a larger research effort that investigates the theory and practice of interemployee knowledge sharing from analytical and empirical perspectives. We view the CFM as a meta-model of employee knowledge-sharing behavior that provides a strategic perspective. We therefore focus only on the meta-model of employee knowledge sharing propensity, in order to address this first-order question: how do motivation, opportunity, and ability influence employee knowledge-sharing behavior? The answer to this question indicates where operations managers should prioritize their investments and scarce resources to achieve desired knowledge sharing behaviors. Thus, we do not explicitly study antecedents of motivation, opportunity, or ability in this research. Other related work delves into the complexities of the antecedents of motivation, especially by exploring the impact of individual and group incentives associated with different job designs (Siemsen et al., in press), the impact of social identity and competence similarity (Siemsen et al., 2007a), and the effects of psychological safety and codifiability of knowledge (Siemsen et al., 2007b). We identify three key findings in this study. First, our results show that the CFM provides a superior explanation of knowledge-sharing behavior compared to the traditional multiplicative model. This finding clarifies the appropriate functional form for the MOA framework. It helps resolve the first puzzle by suggesting that the lack of empirical evidence for complementarity among the MOA variables may be due to the nature of the models of complementarity employed in prior research. Second, our analysis suggests that extreme complementarity exists among motivation, opportunity, and ability to share, such that the degree to which knowledge is shared strongly depends on which of these three variables is the constraining factor. This result has important managerial implications. For example, managerial interventions that aim to improve ability, such as training workers to better communicate and document their “tricks of the trade,” will only be effective in enhancing employee knowledge sharing if ability is the constraining factor in that setting. Third, we find that motivation plays a pivotal role in explaining successful knowledge sharing between individuals. This helps resolve the second puzzle we referred to earlier, as it shows that motivation is indeed an important factor explaining successful knowledge sharing—but only when it is the constraining factor (i.e., the minimum among the three MOA variables). In Section 2, we develop the theory behind the CFM model. The research methods, data, and measurement issues are discussed in Section 3; the empirical tests and results are described in Section 4. In Section 5, we conclude with a discussion of the contributions of this paper, its limitations, and directions for future research.
نتیجه گیری انگلیسی
In this paper, we have presented a theoretical model and an empirical test of the way that motivation, opportunity, and ability together drive knowledge-sharing behavior. Drawing from theory, observations in practice, and conceptual arguments, we specified four competing models: the linear, multiplicative, constraining-factor, and combined models. Of these, the constraining factor model (CFM) captures a new perspective introduced in this paper—a bottleneck theory of knowledge sharing. We developed and administered a survey to collect data on employee KS behaviors in four companies and established the reliability and validity of the newly developed measures. The results from the empirical estimation support the superiority of our proposed constraining-factor model over the linear and multiplicative models. Our analysis and findings have implications for both research and practice. 5.1. Research implications We argue from an examination of the literature and discussions with practitioners that there were still conceptual difficulties with the MOA framework that needed to be understood if it is to be useful for operations management. For example, as solely an organizing framework, MOA makes no assumptions about how to prioritize investments in motivation, opportunity or ability. Interventions regarding motivation may be very different than those providing opportunities to share, such as improving workgroup leadership versus making time available to workers. Moreover, the functional form of how MOA drive behavior is left unspecified. This severely weakens the explanatory elements of the framework. This weakness is bothersome, as it has already led some research to call the framework a truism (Bell and Kozlowski, 2002), something that ought to be true but lacks proper empirical evidence. Our research helps to clarify these issues. As a result, the theoretical aspects of the MOA framework are strengthened. Secondly, our results highlight the pivotal role that motivation plays in explaining successful knowledge sharing. While casual reasoning and straightforward logic suggest that nothing would happen in the absence of motivation, some past studies have found only limited support for the role of motivation in explaining successful knowledge sharing. The constraining-factor model and the empirical findings in this paper resurrect the strategic importance of motivation, while simultaneously providing a deeper understanding of the conditions under which motivation fails to play a role in explaining successful knowledge sharing. Specifically, extreme complementarity is shown to be empirically valid in that if either opportunity or ability is the constraining factor, changes in motivation have no impact on behavior. This empirical result reveals a possible explanation for the low importance attributed to motivation in other studies (Szulanski, 1996). The divergence of our result from past research highlights the importance of the research context and research approach in studies of this kind. For example, there are three major differences between previous work on knowledge sharing by Szulanski (1996) and the present investigation. First, respondents in Szulanski's study were instructed to rule out practices that could be performed by a single individual. In contrast, our study focuses specifically on individual employees. Second, Szulanski studied the transfer of well-known best practices in the company; we investigate “tribal knowledge” that may be unique to the individual—knowledge that company may not even be aware of. Third, Szulanski examined knowledge transfers that were known to have taken place but encountered difficulties. We do not assume that a transfer has taken place, but instead argue that a lack of motivation may cause individuals not to share in the first place. Some of these differences might explain why Szulanski's study attributed a much lower importance to motivation. Especially because Szulanski studied transfers that were known to have taken place, it could be possible that his study resulted in observations for which motivation was not the constraining factor; this would explain the low statistical effect that changes in motivation had on the outcome of knowledge transfer. These differences suggest that Szulanski's findings and those in this paper are not necessarily inconsistent—rather, in studies of this kind, researchers must pay careful attention to the influence of contextual variables such as the operational setting, the type of knowledge shared, and the details of the survey methodology on the ultimate findings. A further contribution of our research is the establishment of a new functional form that describes how motivation, opportunity, and ability interact to drive KS behavior. We confirm the finding of previous studies that a multiplicative model does not exhibit a significantly improved fit over a linear model. However, we demonstrate that the new theoretical perspective we offer – the constraining-factor model – significantly improves model fit over both the linear and multiplicative models presented in the prior literature. This finding strengthens the MOA framework as a metamodel. Finally, in the methodological context, the CFM proposed in this paper expands the tool set of researchers by providing an alternative approach to the conceptualization and modeling of interactions that emphasizes extreme complementarity among the variables. We have shown how the CFM can be empirically tested and compared to standard multiplicative models. Specifically, it would be interesting to apply the CFM in other OM research settings where complementarity among variables could theoretically be expected to exist, but empirical findings have had difficulty establishing it. A particular example of such a setting in operations management would be the study of JIT practices. Researchers might investigate to what degree practices in the JIT bundle have to be applied together, and whether a lack of implementation of any of the typical JIT practices would reduce the overall effectiveness of the program (Shah and Ward, 2003). In the field of organizational behavior, one could apply the CFM to Vroom's expectancy model. In general, the CFM offers a viable alternative to existing ways of modeling interactions. 5.2. Managerial implications Many organizations are currently engaged in knowledge management and knowledge sharing initiatives. Our model offers a strategic view of the knowledge-sharing landscape from the employees’ perspective. It also provides broad-scale guidance about what interventions in terms of programs, practices and tools will advance knowledge sharing for the particular business entity. First, our CFM results suggest that before expending significant resources on designing knowledge-sharing initiatives, managers must carefully study the workplace to identify whether motivation, opportunity, or ability, or some combination of these variables, represents the bottleneck in the knowledge-sharing process. This will be an ongoing process, as these bottlenecks might shift over time, or go unnoticed. Discovering and widening the bottlenecks is important for two reasons. In the presence of a bottleneck, resources allocated to enhancing the levels of the other variables are likely to be unproductive. Managerial interventions aimed at MOA factors that are not constraining in the organization are less likely to be effective. Also, the nature of investments to be made will vary widely depending on which variable or combination of variables constitutes the bottleneck that needs to be addressed. For example, training people how to communicate their knowledge may improve their ability to share knowledge. Providing proper incentives to share (Siemsen et al., in press) or improving psychological safety in teams (Edmondson, 1999) can improve employee motivation to share knowledge. However, neither of these interventions will result in tangible benefits if they do not address a constraining factor. Second, our analysis provides insights into the kinds of metrics and measurement systems that can be used in the implementation of knowledge-sharing initiatives and in assessing their relative benefits versus costs. Managers can periodically measure levels of employee motivation, opportunity, and ability to share knowledge within a well-defined context and empirically relate those measurements to the actual knowledge sharing that occurs. Then, utilizing insights from this analysis, managers can identify which variable constitutes a bottleneck, and direct their scarce resources and efforts at improving this variable. The multi-item measurement scales developed in this paper serve as a practical first step towards the creation of an assessment tool that would support such an undertaking. Take, for example, company 2: For 40% of respondents, motivation was the constraining factor; for 32% of respondents, ability was the constraining factor, and for 28% of respondents, the constraining factor was opportunity. The managerial advice for this company would be to redirect their knowledge-management efforts on the provision of proper incentives (Siemsen et al., in press) and an environment that is psychologically safe (Siemsen et al., 2007b) in order to enhance their employees’ motivation to share. In company 3, in contrast, 53% of respondents showed opportunity to be the constraining factor, while 30% of respondents in that sample reported ability and only 17% reported motivation as the constraining factor. A recommendation for this company would be to prioritize their efforts on interventions that increase employees’ opportunities for knowledge sharing and increase employees’ ability to share. For example, they can reduce employee utilization and specifically integrate knowledge sharing into the regular workflow (Roth et al., 1994); and they can train employees to better communicate and document their tricks of the trade. In company 4, 24% of respondents indicated that motivation was the constraining factor; 50% reported the constraining factor of opportunity; and 26% showed ability to be the constraining factor. Again, employees in this company seem to be severely time-constrained at work, and a prerequisite for any other knowledge management activity would be either to free up time and space for workers to share their knowledge or to explicitly make the sharing of “lessons learned” a part of regular work processes. 5.3. Limitations and future research Like all studies of this kind, our research has a number of limitations. First, given our research questions and study setting, it was difficult to survey multiple respondents per incident without risking substantial nonresponse bias (see Section 3). Other data collection contexts or experimental conditions might be able to overcome this limitation and attempt to replicate our results in a multi-respondent setting. In addition, our measure of opportunity (time availability) is only a proxy for opportunity. There may be other opportunity-related considerations – such as space or resource availability – that would apply in our research setting. Future implementations of the CFM should keep this in mind and gather data on such alternative situational constraints if possible. For example, in contexts such as knowledge sharing among coworkers on globally dispersed new product development teams, other means for informal dialogue may be relevant proxies for opportunity. Another limitation of our research is that we identify the constraining factor by comparing scale averages of standardized variables. This approach neglects both the measurement error inherent in these scales and the fact that standardizing the variables makes them only imperfectly comparable. We paid careful attention to the creation of reliable and valid scales, thus minimizing the impact of measurement error on parameter estimates. In addition, we also tested the robustness of our results to “close calls” to test whether the scaling mattered. Nevertheless, these limitations could potentially be overcome by a latent variable model that posits the constraining factor as the minimum among latent variables. Future methodological work is warranted that will develop a model of this type and establish procedures for identification and estimation using maximum likelihood techniques. We have tested the constraining factor model within the context of inter-employee knowledge sharing. However, we believe that the CFM is applicable in other contexts as well. Since we only study one context, we cannot rule out the possibility that our test of the CFM is influenced by particularities of this particular setting. Further research applying the CFM in different contexts, such as assessing the propensity of employees to participate in formal process improvement programs and reporting errors, would test the generality of the model. Getting employees to share knowledge is an enduring problem, one that has increased in relevance over time. Our aim in this study has been to provide a fresh perspective on knowledge sharing from both the research and managerial viewpoints. The proposed constraining-factor model developed here has potential as a tool for planning managerial interventions to improve knowledge sharing and may have applications in many other areas of organizational and operational research. We hope our thoughts in this paper will stimulate further research on, and use of, the CFM approach.