تاثیر عوامل ساختاری و متنی بر شکل گیری اعتماد در تیم های توسعه محصول

This study examines antecedents of trust formation in new product development (NPD) teams and the effects of trust on NPD team performance. A theoretical framework relating structural and contextual factors to interpersonal trust and project outcomes was built, including task complexity as a moderating variable. Hypotheses from this model were tested with data on 93 product development projects carried out in Turkey. The findings showed that structural factors such as moderate level of demographic diversity, proximity of team members, team longevity, and contextual factors (procedural and interactional justices) were positively related to the development of interpersonal trust in NPD teams. The findings also revealed that interpersonal trust had an impact on team learning and new product success, but not on speed-to-market. Further, the findings showed that the impact of interpersonal trust on team learning and new product success was higher when there was higher task complexity. Theoretical and managerial implications of the study findings are discussed.

An extensive body of academic business literature has shown the importance of cross-functional teams in developing new products that stand out in highly competitive markets (Brown and Eisenhardt, 1995, Gupta et al., 1986, Keller, 2001, Lovelace et al., 2001, Nijssen et al., 1995 and Sherman et al., 2000). Since cross-functional teams often involve people with different views, perspectives, backgrounds and knowledge, the majority of research studies in this stream have focused on the collaborative functioning of teams. In most of these studies, antecedents of collaborative functioning of teams and the impact of collaborations on team effectiveness and efficiency have been extensively investigated (Hoegl and Gemuenden, 2001, Hoegl and Proserpio, 2004, Larson and Gobeli, 1988 and Sethi and Nicholson, 2001). Sethi and Nicholson (2001), for instance, examined the impact of structural and contextual factors on the collaborative behavior of teams. Despite this large body of research on collaborative functioning, much less attention has been paid to the role of trust on the effective functioning of new product development (NPD) teams. Research on organizational effectiveness indicates that trust plays a significant role in the collaborative behavior of organizations. In a classic study, Thompson (1967) notes that under conditions of uncertainty and complexity requiring mutual adjustment, sustained effective coordinated action is only possible where there is mutual confidence or trust. An extensive literature has emerged on the importance of trust between partners in the distribution channel (e.g., Anderson and Narus, 1990, Garbarino and Johnson, 1999, Mohr and Spekman, 1994, Moorman et al., 1993 and Morgan and Morgan, 1994). In the NPD team setting, trust potentially can have several effects on the team's performance (it could increase likelihood of product success and/or accelerate speed-to-market, for example); these multiple possible consequences of increased trust within the NPD team have not yet been adequately researched. Similarly, although the importance of trust has been acknowledged in the product innovation literature (Akgun et al., 2005, Koskinen et al., 2003 and Madhavan and Grover, 1998), the matter of how it develops within the NPD team has received little systematic theoretical attention. If indeed increased team trust positively affects project outcomes, it is important to NPD team leaders to understand the antecedent factors that can boost levels of trust among their team members. The research objective of the present study is to develop and test a theoretical model that integrates the literature on antecedents of trust and outcomes of increased trust. The development of trust in NPD teams is modeled as an outcome of two types of antecedents (see Fig. 1). The first type of antecedents consists of structural factors: diversity in NPD teams, physical proximity of team members, and team longevity. These factors have emerged as important structural factors in group research, and senior management of organizations has some degree of control over them (Sethi & Nicholson, 2001). Functional and demographic diversity in teams has been considered one of the main factors having a significant impact on team trust and effectiveness (Knouse and Dansby, 1999, Spector and Jones, 2004 and Tolbert et al., 1995). Similarly, physical proximity is known to help team members have a better chance to communicate and coordinate effectively (Khan and McDonough, 1997 and Pinto et al., 1993). Team longevity is also considered as an important factor that helps team members develop a shared history and bond among themselves (Akgun and Lynn, 2002 and Katz, 1982). Beside these structural factors, as indicated in previous research (Brewer and Miller, 1984; Mackie and Goethals, 1987 and Kramer, 1999), contextual factors, such as procedural and interpersonal justice, may also contribute to the development of trust among team members. We argue that honesty, fairness and respectfulness of managers to team members as well as fair procedures would create a climate where interpersonal trust in teams can be fostered. Increased interpersonal trust is anticipated to have a positive effect on multiple project outcomes: it should increase team learning, accelerate speed-to-market, and increase likelihood of product success. Finally, we anticipate that task complexity will have a moderating impact on the effect of trust on team performance. In complex tasks, more cooperation and coordination among team members is required, therefore higher levels of trust may be necessary to achieve greater team success (Akgun et al., 2005). In the next section, we build our theoretical model of the antecedents and consequences of interpersonal trust on the NPD team, including task complexity as a moderator, and develop testable hypotheses. In subsequent sections, we test our hypotheses using a sample of teams from Turkish firms engaged in product innovation. We conclude by deriving theoretical conclusions and implications for managers.

5.1. Theoretical contributions The theoretical framework of Fig. 1 proposes that both structural and contextual factors act as antecedents to interpersonal trust within the NPD team, and that greater levels of interpersonal trust have positive consequences in terms of team learning, speed-to-market, and product success. The framework also proposes that the relationship between interpersonal trust and outcomes is mediated by task complexity. The findings of this study support most of the hypotheses drawn from this theoretical framework about formation of interpersonal trust, its antecedents and consequences, and the outcomes of interpersonal trust. Our results provide evidence of the centrality of interpersonal trust within collaborative functioning of new product development teams. We also show that the process of formation of interpersonal trust is complex. We acknowledge that most managers would recognize the importance of having NPD team members trust each other, as this should lead to better teamwork and improved outcomes. But achieving desirable levels of trust among team members can be an elusive goal. Our theoretical model, largely supported by our empirical results, shows that there are many significant precursors to interpersonal trust: some demographic, some related to team characteristics (and controllable to some extent by management) such as longevity or proximity, and some having to do with perceived levels of justness and fairness in interactions and procedures. To substantially improve trust among NPD team members, senior management can ignore none of these antecedents. First, the results indicate that the beliefs of team members about the trustworthiness of peers can be measured along two dimensions, the extent of cognitive-based trust and the extent of affect-based trust. Specifically, the findings reveal that each form of interpersonal trust functions in a unique manner in despite of causal connectedness between cognitive- and affect-based trust (McAllister, 1995). Second, the findings demonstrate that interpersonal trust can be developed in a NPD project team when team members remain on the team till at least for a certain period of time, team members are geographically located in a closer proximity, there are fair decision-making procedures during the product development process, and the quality of interpersonal treatment team members receive from their supervisors is high. Moreover, the results show that a moderate amount of demographic diversity appears to be optimal for the development of interpersonal trust. These findings are consistent with previous study. In the following section, integration of the findings with related prior research will be discussed. Regarding demographic diversity, as found by Knouse and Dansby (1999), the findings show that interpersonal trust is fostered by the moderate level of demographic diversity. This finding generally supports the psychological minority phenomena (Davis, 1980) in the literature. A moderate level of diversity should create greater synergy as dissimilar members interact and develop trust. However, as the level of diversity increases, such interactions and trust disappear. Contrary to our expectation, interpersonal trust is not affected by functional diversity. A possible reason for this finding would be the structural feature of cross-functional NPD project teams. Unlike other types of teams in organizations, NPD teams often involve people with different functional backgrounds and knowledge since inputs of several functional units within an organization are required during the new product development process in order to develop a successful product. As expected, team member proximity is positively associated with interpersonal trust. This finding supports the validity of findings of previous research examining the relationship between collaborative functioning of teams and the vicinity of team members (Hoegl & Proserpio, 2004). The close proximity allows team members to have face-to-face encounters that are considered irreplaceable for both building trust and repairing shattered trust (Nohria and Eccles, 1992 and O'Hara-Devereaux and Johansen, 1994). The findings also support the positive impacts of procedural and interactional justices on interpersonal trust. In the organizational justice literature, such positive relationships are explained with several different models as discussed earlier. According to these models, for instance the rational model (Lind and Tyler, 1988 and Tyler and Lind, 1992), if team members perceive the legitimacy of team's leadership and seek to comply with team rules and decisions, they could remain more attached to the team and show more trustworthiness. Similarly, honesty, fairness and respectfulness of managers to team members would create a climate where interpersonal trust in teams can be fostered. Third, this study also tested the impact of interpersonal trust on team performance, such as team learning, speed-to-market, and new product's success in the market place. The bivariate correlations demonstrated that interpersonal trust has a positive association with team learning and new product success but not with speed-to-market. As team members develop trust, they develop the new product with fewer technical problems, find and solve the product problems areas with which customers are dissatisfied, and develop product better. On the other hand, it should be noted that the impact of interpersonal trust on team performance is contingent upon the project's complexity; when there the tasks rely on established bodies of knowledge, and there is a clearly defined way to do the major types of project's work, the impact of interpersonal trust on team learning and new product success will be less. However, when the repetitiveness of the elements of project's tasks is low (team members' work is non-routine), the impact of interpersonal trust on team learning and new product success will be higher. Non-routine work requires high interpersonal trust among team members more than routine work, because less repetition of the elements of project's tasks requires team members to communicate and support each other on new tasks, but such kind of interaction and support cannot be achieved without adequate level of interpersonal trust in a team. According to a classic study of Hofstede (1980), Turkish society is characterized by close, stable, long-term social relations. Hence, it is expected that trust should be stronger within Turkish society compared with other cultures, such as that of the United States, where individualism and self-interest are distinctive cultural values. However, even though Japanese society is characterized much the same as Turkish society on these cultural dimensions, Yamagishi and Yamagishi (1994) reviewed survey evidence that Japanese people often report lower levels of trust. To resolve this anomaly, they proposed a distinction between generalized trust and assurance. Mutual assurance is based on stability of interpersonal and interorganizational relationships within the society. It is believed that because of this high degree of perceived stability, social uncertainty in transactions is less common in the Japanese society than U.S. or other Western societies. In U.S. society, in cases where social uncertainty in transactions tends to be greater, trust is one of the main concerns of people in order to reduce uncertainty. We argue that comparable sense of stability or assurance is not readily available in Turkish organizations, although there is stability and closeness in social relations. In other words, Turkish managers may not perceive as high a degree of stability as their Japanese counterparts do. Hence, as in U.S. society trust is a major main concern of Turkish managers at work. The findings of this study also support our argument. Our findings imply that team longevity and team proximity are significant factors for the development of interpersonal trust, since Turkish managers do not perceive a high degree of stability and closeness in their organizations. Hence, they want more stability and physical proximity so that they can have some interactions in order to develop trust. In addition, perceived levels of justice in Turkish organizations may also be an important factor leading to greater levels of interpersonal trust. Turkish managers may feel, for example, that justice is not as readily available in Turkish organizations yet as it is already in Japan. There is little research to support this claim at this point, however, and this observation must be considered only speculative. Nevertheless, the role of perceived justice, and how this differs across countries and cultures, would be an interesting avenue for future research. 5.2. Managerial implications Some practical implications for management can be drawn from the study. Our study finds that a moderate level of demographic diversity (e.g., age, sex, ethnicity) contributes positively to the development of interpersonal trust. Demographic dissimilarities among team members tend to increase the formation of interpersonal trust and relations, but beyond a certain level the impact can become detrimental to the development of interpersonal trust. This finding presents an interesting practical implication to the manager, as what constitutes “moderate” diversity is not evident. It suggests that, in order to improve interpersonal trust among team members, a manager should consider not only the mix of functional diversity on the team, but also the demographic diversity. Observation of the current demographic makeup of the team might suggest adjusting demographic diversity (in either direction), perhaps in ways previously not considered. In an attempt to add to the functional diversity of the team, for example, the manager might realize the team lacks some demographic diversity (e.g., all men; all United States nationals; all older employees). Further, the adjustments may need to be implemented with care to human relations issues; for example, adding two females to an eight-man team in the interest of increasing demographic diversity does not guarantee that the males will not exclude the females. Another critical factor will be the proximity of team members. The high mean value for the proximity of team members (5.12 out of 7.00) implies that even low levels of geographic dispersion can have detrimental effects on the development of interpersonal trust. Consequently, the manager must be aware that geographic dispersion may come at a cost to the development of interpersonal trust (Hoegl & Proserpio, 2004). Thanks to today's computer-mediated communications methods, managers can easily overcome difficulties of geographic dispersal through the use of virtual teams. A new product development team may include members located on multiple continents. Firms can execute round-the-clock produce development by passing information among partners based in North America, Japan or India, and Europe. For example, Xerox coordinates product development among employees in Shanghai, Hong Kong, and its Rochester, N.Y. headquarters, and Boeing similarly develops new airplanes through joint efforts of team members located in dispersed areas (Ozer, 2003 and Sethi et al., 2003). Some evidence shows that virtual teams speed up new product development or lead to cost efficiencies (Sethi et al., 2003), and even make better decisions than face-to-face teams in some contexts (Schmidt, Montoya-Weiss, & Massey, 2001). Furthermore, the mangers should keep the level of turnovers in a team at the minimal level in order to accelerate the development of interpersonal trust. If the turnover is too high, the effectiveness of communication and interactions among team members will never be at the desired level, with the result that trust among team members cannot be established. We should also note, however, that there is some empirical evidence that, after a certain point, longevity can begin to have a negative impact on team performance. This is the structural holes argument: if a team has been together too long, ideas can become stagnant and the team has fewer new product successes (Burt, 2004). Further research might examine the potential long-term adverse effects of team longevity.1 Moreover, the managers concerned with having successfully developed products and helping teams to be more involved in learning activities during the product development process should foster interpersonal trust among team members. This can be managed if a favorable justice climate exists in NPD teams. In order to create such an environment, team members should first be encouraged to become more involved in key decisions. Second, they should be provided with opportunities to appeal decisions. Third, managers should make their decisions consistently and neutrally, and deal with team members honestly and politely. Additionally, we note that other additional measures of performance might have been included. The product literature finds that the most successful firms in product development consider strategic as well as financial external success criteria (Cooper, Edgett & Kleinschmidt, 1998). Strategic criteria may be more qualitative in nature and will vary by firm, but might include the product's ability to strengthen relationships with key customers, or its complementarity with other products in the firm's portfolio. We did not gather data on any of these strategic or qualitative measures of success, but recognize their importance in assessing a firm's new product performance and recommend that they be included in future research endeavors. Finally, we note that there were several hypotheses that were surprisingly not supported. Team trust was not significantly related to speed-to-market; it is possible that other factors, such as avoiding functional silos and facilitating face-to-face or virtual communication among team members, may have a greater role on speed-to-market than team trust. The speed-to-market literature suggests many important antecedents to accelerated product development (see Sherman et al., 2000, e.g.), and the effect of team trust may be overshadowed. We also find that trust is an important precursor to team learning regardless of the level of complexity. This finding is unexpected, but it suggests that in cases of low complexity, the importance of establishing high trust levels among team members cannot be ignored. Finally, we do not find a relationship (linear or non-linear) between functional diversity and trust. This bears discussion as it seems to run counter to the new product literature. While it is not possible to disentangle all the possible causes of this finding, it is perhaps due to (1) cultural differences due to the Turkish business environment, or (2) lack of sufficient variation in functional diversity in the sample, or both. Further research could shed further light on this important issue. 5.3. Limitations and recommendations for further research Retrospective reporting and cross-sectional research design are potentially methodological limitations of this study. Because of a time lag (over one year) between projects' completion and data collection, there might have been a recall issue in survey questions. However, as suggested by Miller, Cardinal, and Click (1997) the use of retrospective data is acceptable if reported measures are reliable and valid. The measures used in our research demonstrated the criteria of reliability and validity. The use of a single informant, while common in NPD research, may also lead to biases. A respondent may make his or her company “look good” by upwardly biasing all of the responses, for example. We believe we have controlled for the possibility of such biases to a great extent through the use of key informants (Phillips, 1981), though we cannot rule out the possibility that such biases did not occur. All respondents were knowledgeable sources of information about the team's trust levels and its performance. The key informants can be counted on to provide viable and useful responses pertaining to their team. Several recent studies have shown that, as long as the key informant is of a senior enough level within the firm and possesses a high level of knowledge and involvements regarding the project, he or she will provide reliable and valid data on strategy and performance that is not too different from secondary, objective data (Kumar et al., 1993, Menon et al., 1996 and Zahra and Covin, 1993). We noted earlier that we ran a Harman test to determine whether common method bias was a potential limitation. The Harman test results suggested that common method variance (CMV) was not likely to be a concern. Though CMV cannot be completely ruled out using this test, we can bolster the argument that it is not a major concern here, because we are predicting a moderating effect. It is unlikely that survey respondents would be able to guess the nature of the moderation or interaction effect, and therefore, would be unlikely to provide responses that can be seen as contributing to CMV.2 In addition, using a cross-sectional design in this study was also a potential limitation. This research design may not provide true results about flow of knowledge. A longitudinal research design would allow us to track the flow of knowledge over time and would help to overcome this limitation. We suggest this as a possible direction for future research. We acknowledge that the sample size is relatively small and limited in scope to one country. The study was conducted in Turkey where collectivism is one of the prevailing cultural values (Hofstede, 1980). Hence, social belonging to groups and teams is more highly appreciated in Turkey than most western cultures. Previous literature on organizational trust (e.g., Lind and Earley 1992), however, indicates that people's trust perceptions vary from culture to culture. Thus, generalizations of the findings of this study should be kept limited to cultures that share cultural values with Turkey, particularly in terms of collectivism (Hofstede, 1980). Researchers should examine other facets of diversity in addition to functional and demographic backgrounds. Diversity in terms of ability levels, experience, attitudes, values, and personality should be investigated in future research studies (McGrath, Berdahl, & Arrow, 1995). In addition, the impact of structural and contextual factors on the development of interpersonal trust was tested in this study. Further research should also examine other ways of fostering the development of interpersonal trust. Certain management practices or reward systems may enhance the development of interpersonal trust in a NPD team. The impacts of other forms of justice on interpersonal trust should be investigated as well in future research. In particular, aggregate perceptions of interpersonal or informational justice could have a variety of beneficial effects in the development of interpersonal trust (Colquitt, Conlon, Wesson, Porter, & Ng, 2001). Finally, in order to capture the whole aspects of NPD projects, other dependent variables, such as product quality, process proficiency and project cost, could be tested. Using different moderating (e.g., organizational culture and types of work groups) and/or mediating factors (e.g., environmental turbulence and improvisation) could be investigated in future study.