新產品研發團隊對新產品品質的影響：台灣與美國之比較

全文

(1)♁ 國立中山大學企業管理學系博士論文. The Effects of New Product Development Teams on New Product Quality：A Taiwanese-American Comparison. 研究生：張倉榮撰指導教授：胡國強 Louis P. White. 中華民國九十四年五月.

(2) Acknowledgments. The dissertation research process is a challenging and growing experience, and a great appreciation for advance learning and research. It is not, however, possible to complete the process without the support, understanding, and guidance from a group of special people in my life. It is with deep and sincere gratitude that I acknowledge all the individuals who helped make this dissertation a reality. First, I would like to thank my advisor, Dr. Hu, who coached me through the steps to completion. His patience, consideration, guidance and support have been the major driving forces for my research. I am also grateful to my co-advisor, Dr. White, who not only provided me with his expertise in research, but also helped coordinate the companies in the U.S. Without their assistance throughout this long journey, this dissertation would have been impossible. My appreciation also goes to Jeremy Verostko, of the Johnson Space Center, NASA, for his assistance in the research. He spared no efforts to help collect data from new product development teams at NASA. His contributions were invaluable to this research. I would also like to express my special thanks to Dr. Allan Chen, who helped conduct the survey at semiconductor companies located in Dallas, Texas. Without his help, I could not have finished my research. I am greatly indebted to him. A very special thank-you is extended to all the companies surveyed, including those in the U.S. and Taiwan, which permitted their employees to help us all learn more about new product development teams’ characteristics and organization contextual factors affecting new product quality. To the people who took the time their busy schedules to complete the surveys, I would also like to express my sincerest gratitude. i.

(3) To my daughter, Sunny, I want to say thank you for always staying interested in the progress of my doctoral studies and dissertation, and encouraging me to go the last mile. There are no words that can express my appreciation and thanks to my wife Li-Hsueh, who encouraged and supported me every step of the way — from preparing doctoral entrance examination to completing my dissertation. She was a true soul mate, and continues to accompany me on my life’s journey.. ii.

(4) Abstract. New product development quality has been found to be a key determinant of the market success and profitability of a new product because of its contribution to superior customer value, the cornerstone of a firm’s competitive advantage in the marketplace. Therefore, enterprises are increasingly utilizing new product development teams for new product development. Previous studies have suggested that the diverse backgrounds of cross-functional team members can increase the amount and variety of information available to design products, thereby improving design process efficiency and product development performance. However, researchers still know very little about how such teams can improve new product quality. Thus, our first research question was; what are the relationships between team characteristics and organization contextual factors and new product quality? In addition, as nationality has been shown to influence individual’s cognitive schema, values, and nonverbal behavior, all of which influence behavior in teams, new product development teams pose a particular type of challenge in different countries. Team-based work designs have been easier to implement in countries with collectivist as opposed to individualist cultures. To address that issue, this study will examine the differences of how team characteristics and contextual influences affect new product quality in collectivist and individualist cultures. Thus, our second research question was; do team characteristics and organization contextual factors affect new product quality differently in collectivist cultures and individualist cultures? Because this was a cross-national research, samples were taken from two countries, one in Taiwan and the other in the U.S.A. We received 184 completed questionnaires from Taiwanese teams, and 176 completed questionnaires from American teams. In order to make sure that the iii.

(5) instrument used in this study could measure what it was intended to measure, we used confirmatory factor analysis (CFA) to examine the validity of the measurement models of both dependent variables and independent variables. We then used hierarchical moderated regression analysis to test the contingency hypotheses. Research results reveal two different outcomes of hypotheses testing. For Taiwanese teams, new product quality is positively affected by the capability of information integration in the team and quality orientation in the firm, but is negatively influenced by speed-to-market pressure and product innovativeness in the firm. Functional and tenure diversity have a moderate no effect on new product quality. The capability of information integration in a team can also reduce the negative effect of speed-to-market pressure on new product quality. In contrast, for American teams, new product quality is positively affected by functional diversity and the capability of information integration in the team and quality orientation in the firm, but is moderately negatively influenced by supplier involvement in the firm. Customer involvement in a firm can increase the positive effect of the capability of information integration on new product quality. In conclusion, the present study contributes to the literature on new product development teams in several ways. First, it extends research on new product development teams across collectivist and individualist cultures and identifies boundary conditions for theories of new product development teams. Second, this study is the first to empirically examine how team characteristics and organization contextual factors affect new product quality across cultures. Third, it provides a thorough and integrative review of the literature from diverse disciplines such as new product development, organizational behavior, and management and innovation as a means of establishing links among team characteristics, organization contextual factors, and new product quality. Finally, the model we have provided will assist managers in identifying the team iv.

(6) characteristics and organization contextual factors needed to assist new product development teams in collectivist and individualist cultures to develop high-quality products. Since these factors can be directly manipulated by managers, they can create the effective conditions, specific to the industry characteristics they are engaged in, for new product development teams to develop high-quality products.. Keywords: new product development teams, cross-functional project teams, new product quality, product development.. v.

(7) TABLE OF CONTENTS CHAPTER 1. INTRODUCTION ....................................................................... 1. 1.1. The Importance of New Product Quality ........................................................................ 1 1.2. The Importance of New Product Development Teams ................................................... 2 1.3. The Effects of Culture on Behavior in Teams.................................................................. 3 1.4. Research Questions............................................................................................................ 4 1.5. Research Objectives........................................................................................................... 5. CHAPTER 2. CONCEPTUAL MODEL AND HYPOTHESES....................... 7. 2.1. Team Characteristics and New Product Quality........................................................... 10 2.1.1. Functional Diversity.................................................................................................. 10 2.1.2. Tenure Diversity .........................................................................................................11 2.1.3. Capability of Information Integration .....................................................................11 2.2. Organization Contextual Factors and New Product Quality ...................................... 12 2.2.1. Speed-to-Market Pressure........................................................................................ 12 2.2.2. Product Innovativeness ............................................................................................ 12 2.2.3. Quality Orientation................................................................................................... 13 2.2.4. Supplier Involvement................................................................................................ 13 2.2.5. Customer Involvement ............................................................................................. 14 2.3. Moderators ....................................................................................................................... 14 2.3.1. Capability of Information Integration as a Moderator on the Speed-to-Market Pressure and Quality Linkage ................................................................................. 15 2.3.2. Capability of Information Integration as a Moderator of the Product Innovativeness and Quality Linkage....................................................................... 15 2.3.3. Quality Orientation as a Moderator of the Capability of Information Integration and Quality Linkage ................................................................................................. 16 2.4. Effects of Culture on How Team Characteristics and Contextual Factors Affect New Product Quality............................................................................................................... 16 2.5. Covariate........................................................................................................................... 20. CHAPTER 3. RESEARCH METHOD............................................................ 22. 3.1. Sample and Data Collection............................................................................................ 22 3.2. Measures and Validation ................................................................................................. 23 3.2.1. New Product Quality ................................................................................................ 23 vi.

(8) 3.2.2. Functional Diversity.................................................................................................. 24 3.2.3. Tenure Diversity ........................................................................................................ 25 3.2.4. Capability of Information Integration .................................................................... 25 3.2.5. Speed-to-Market Pressure........................................................................................ 25 3.2.6. Product Innovativeness ............................................................................................ 26 3.2.7. Quality Orientation................................................................................................... 26 3.2.8. Supplier Involvement................................................................................................ 26 3.2.9. Customer Involvement ............................................................................................. 27 3.2.10. Covariate.................................................................................................................. 27 3.3. Analytical Procedures ...................................................................................................... 27. CHAPTER 4. ANALYSIS AND RESULTS ..................................................... 29. 4.1. Analysis ............................................................................................................................. 29 4.1.1. Scale Validity and Reliability ................................................................................... 29 4.1.2. Multicollinearity Test................................................................................................ 36 4.1.3. Descriptive Statistics ................................................................................................. 36 4.1.4. Hierarchical Moderated Regression........................................................................ 38 4.1.5. Residual and Influence Analysis .............................................................................. 43 4.2. Results ............................................................................................................................... 48 4.2.1. Direct Effects ............................................................................................................. 48 4.2.2. Moderating Effects.................................................................................................... 51 4.2.2.1. The Interactions of Taiwanese Teams .............................................................. 51 4.2.2.2. The Interactions of American Teams ............................................................... 52 4.2.3. Covariate.................................................................................................................... 52. CHAPTER 5. DISCUSSION............................................................................. 53. 5.1. Discussions of Results ...................................................................................................... 53 5.1.1. Discussions of Team Characteristics ....................................................................... 53 5.1.2. Discussions of Organizational Contextual Factors ................................................ 55 5.1.3. Discussions of Moderating Effects........................................................................... 58 5.1.4. Discussions of Cultural Effects ................................................................................ 60 5.2. Managerial Implications ................................................................................................. 65 5.3. Research Limitations ....................................................................................................... 67 5.4. Suggestions for Future Research .................................................................................... 68. CHAPTER 6. CONCLUSIONS........................................................................ 72. REFERENCES..................................................................................................... 74 vii.

(9) APPENDIX........................................................................................................... 85. viii.

(10) LIST OF FIGURES FIGURE 1 A Model of Team Characteristics and Organization Contextual Factors Affecting New Product Quality............................................................................................................................. 8 FIGURE 2 A Comparison of Taiwanese-American New Product Development Team’s Models ............. 9 FIGURE 3 Independent Variable Measurement Model of Taiwanese Teams .......................................... 32 FIGURE 4 Independent Variable Measurement Model of American Teams ........................................... 33 FIGURE 5 Dependent Variable Measurement Model of Taiwanese Teams ............................................. 34 FIGURE 6 Dependent Variable Measurement Model of American Teams .............................................. 35 FIGURE 7 Histogram of Taiwanese Teams.................................................................................................. 43 FIGURE 8 Histogram of American Teams .................................................................................................. 44 FIGURE 9 Residual Plot of Taiwanese Teams ............................................................................................. 45 FIGURE 10 Residual Plot of American Teams.............................................................................................. 46 FIGURE 11 Index Influence Plot of Taiwanese Teams ................................................................................. 47 FIGURE 12 Index Influence Plot of American Teams .................................................................................. 48 FIGURE 13 The Final Regression Model of Taiwanese Teams.................................................................... 63 FIGURE 14 The Final Regression Model of American Teams..................................................................... 64. ix.

(11) LIST OF TABLES TABLE 1 Summary of Hypotheses............................................................................................................. 21 TABLE 2 Fit Indices of Independent Variables of Taiwanese and American Teams............................. 31 TABLE 3 Fit Indices of Dependent Variables of Taiwanese and American Teams................................ 31 TABLE 4 Means, Standard Deviations, and Correlations of Taiwanese Teams.................................... 37 TABLE 5 Means, Standard Deviations, and Correlations of American Teamsa ................................... 38 TABLE 6 Results of Moderated Regression Analyses of Taiwanese Teams............................................ 40 TABLE 7 Results of Moderated Regression Analyses of American Teams ............................................ 41 TABLE 8 Results of Revised Moderated Regression Analyses of American Teams .............................. 42 TABLE 9 Summary of Testing Results of Taiwanese and American Teams’ Models ............................ 62. x.

(12) LIST OF ABBREVIATIONS CAPII. Capability of information integration. CFI. Comparative fit index. CFA. Confirmatory factor analysis. CUSINV. Customer involvement. GFI. Goodness-of-fit index. NNFI. Non-normed fit index. NPQ. New product quality. NPDT. New product development team. QLTYOR. Quality orientation. RMSEA. Root mean square error of approximation. SEM. Structural equation modeling. STRPRE. Speed-to-market pressure. SUPINV. Supplier involvement. xi.

(13) CHAPTER 1. INTRODUCTION. 1.1. The Importance of New Product Quality Product quality has received increasing attention in both practice and research (Menon, Jaworski & Kohli, 1997), as evidenced by the growing number of firms instituting quality programs, and the abundance of coverage in both business-oriented periodicals (Byrne, 1994; Calonius, 1991; Jacob, 1993) and research-based academic literature (Aaker & Jacobson, 1994; Curry, 1985; Fornell, 1992; Phillips, Chang & Buzzell, 1983; Shetty, 1987). The main reason for this increased attention is that product quality delivers superior customer value, which is the cornerstone of a firm’s competitive advantage in the marketplace (Day & Wensley, 1988). As such, new product quality has been found to have a major influence on the market success and profitability of a new product. This concept has been widely recognized by domestic enterprises since Japanese products began capturing a major share of the market in one industry around the world in the 1980s (Bounds, Yorks, Adams, & Ranney, 1994; Clark & Fujimoto, 1991; Garvin, 1988). In order to enhance the quality of their new products, a number of firms in Taiwan have introduced different quality-improvement techniques, such as quality control circles (QCC) and total quality management (TQM) in different phases. The use of these techniques is believed to have helped Japanese companies improve new product quality (Bounds et al., 1994; Clark & Fujimoto, 1991). One of the five TQM interventions prescribed by Juran, Ishikawa, and Deming is the use of cross-functional teams to identify and solve quality problems (Hackman & Wageman, 1995). As a result, giving the product development task to cross-functional teams is rapidly increasing. These teams are composed of individuals from a variety of functional areas, such as manufacturing, engineering, marketing, research and development, and purchasing. 1.

(14) Cross-functional teams are thought to better link upstream and downstream organizational activity (so that, for example, manufacturing considerations can be designed into a product rather than being discovered after prototypes already exist), to push decision making down to those who have the necessary expertise, and to enhance speed and coordination (Ancona, Kochan, Scully, Maanen, & Westney, 1996). 1.2. The Importance of New Product Development Teams Previous studies have suggested that the diverse backgrounds of cross-functional team members can increase the amount and variety of information available to design products, thereby improving design process efficiency and product development performance (Brown & Eisenhardt, 1995; Chang, Hu, & White, 2002; Clark & Fujimoto, 1991; Iansiti, 1993; McDonough III, 2000; Sarin & Mahajan 2001). Few would argue however, that the mere formation of such teams is sufficient to meaningfully enhance new product quality (Bounds et al., 1994; Sethi, 2000a; Sethi, 2000b; Webber 2002; Wheelwright & Clark, 1992). However, in spite of the critical role of quality in influencing the success of a new product and the growing popularity of cross-functional product development teams, there has been little research on how these teams affect new product quality (Menon, Jaworski, & Kohli, 1997). Over the past twenty years, research about quality has been published, but most of it is either descriptive or prescriptive in nature and relies heavily on anecdotes rather than on appropriate research designs (Hackman & Wageman, 1995). There exists some recent firm-level research on the determinants of product quality (Clark & Fujimoto, 1991; Menon, Jaworski, & Kohli, 1997; Morgan & Piercy, 1998; Song, Souder, & Dyer, 1997); however, these studies have mainly concentrated on the effects of macro- or firm-level variables, such as structural and cultural factors, rather than on the influence of micro- or team-level factors on new product 2.

(15) quality. Further, firm-level research has usually focused on aggregate outcomes (e.g., the quality of a firm's products in general, instead of on how the quality of a new product is affected by the team). The knowledge generated by these firm-level studies, though certainly useful, may not be effective in explaining variations in team-level outcomes (Sethi, 2000a). Research on product development teams also provides valuable insights into the effects of teams on outcomes such as market performance of the new product, cycle time, and efficiency in developing innovations (Ancona & Caldwell, 1992; Barczak & Wilemon, 1992; Eisenhardt & Tabrizi, 1995; Thamhain, 1990). Such insights do not directly address how teams influence a different outcome such as new product quality. Later we address the idea that team-related factors that help one type of product outcome may not necessarily facilitate other outcomes and may even harm them. 1.3. The Effects of Culture on Behavior in Teams Although new product development teams play a critical role in influencing the success of a new product, they also pose a particular type of challenge in different countries as nationality has been shown to influence individual’s cognitive schema, values, and nonverbal behavior, all of which influence behavior in teams. In a study of value differences in 40 nations, Hofstede (1980) found that individualism-collectivism accounted for the greatest variance in work goal priorities of the dimensions he studied. Compared to individual cultures, collectivistic cultures place greater emphasis on the needs and goals of the group, social norms and duty, shared beliefs, and cooperation with group members (Triandis, 1989). Collectivists are more likely than individualists to sacrifice personal interests for the attainment of a group goal (Bond & Wang, 1983) and are more likely to enjoy doing what the group expects of them (Bontempo, Lobel, & Triandis, 1990). 3.

(16) As Hill (1994) points out, there are cross-national differences on the individual-collective dimension (Erez & Earley, 1993). In individualist cultures such as the United States and the Netherlands people tend to use personal achievements to define themselves, view relationships as short-term, and value the individual more than the team. In collectivist cultures such as Taiwan and Japan, however, there is high commitment to, and identification with, the team, where group harmony, unity, and loyalty are valued more than individual gain (Hill, 1994). Despite the apparent importance of national culture as it influences behavior in teams, it remains unclear whether team characteristics and contextual factors influence new product quality differently in collectivist cultures than in individualist cultures. Therefore, there is a need for research that examines the effects of cultural differences on the relationships between team characteristics, organization contextual factors and new product quality. 1.4. Research Questions The problems facing new product development teams arise from differences in team characteristics, organization contextual factors and cultures, which cause difficulties in improving new product quality. These challenges are becoming increasingly acute as more firms use new product development teams to improve new product quality. This research examines how team characteristics and contextual influences affect new product quality, and how cultures influence their relationships. Therefore, based on the information previously presented on the positive results garnered from new product development teams and new product quality, two questions motivate this study: 1. What are the relationships between team characteristics and organization contextual factors and new product quality? 2. Do team characteristics and organization contextual factors affect new product quality 4.

(17) differently in collectivist cultures and in individualist cultures? 1.5. Research Objectives To answer these questions, this research builds on previous work which showed the physical composition of new product development teams to have positive effects on team’s performance (Bounds et al., 1994; Clark & Fujimoto, 1991; Garvin, 1988), as well as research indicating that team process is positively related to new product quality (Clark & Fujimoto, 1991; Garvin, 1988; Menon, Jaworski, & Kohli, 1997). It also draws on work by quality management theorists (Bounds et al., 1994) which suggested that the ability of a team to produce a high-quality outcome can also be affected by organization contextual influences on the team. In addition, this research relies on earlier study which indicated that individualist and collectivist cultural values influence work-group composition, processes, and outcomes (Earley & Gibson, 1998; Gibson, 1999), and collectivists perform better working in teams, when performing interdependent tasks, and sharing responsibility. Therefore, there were two objectives in this study. The primary objective was to explore how various aspects of cross-functional teams and organizational context enhance or diminish new product quality. On the basis of previous literature, it can be suggested that the emergence of a quality outcome can be affected by two sets of factors: (a) the characteristics of the team and (b) the organizational context influences on the team. We therefore, examined variables related to these two sets of factors. First, team characteristics were defined by critical team processes and the physical composition of the team (i.e., the homogeneity of organization tenure and the mix of functional specialties). Second, the organizational context influences on the team: (a) the effects of the speed-to-market pressure, (b) the innovativeness of the product, (c) customers’ involvement in the product development process, (d) the suppliers’ involvement in the product 5.

(18) development process, and (e) the quality orientation of the firm were examined. The secondary objective of this study was to examine the differences of how team characteristics and contextual influences affect new product quality in collectivist cultures and individualist cultures. To address this issue, this research was conducted in both the United States (individualism) and Taiwan (collectivism). We begin with an overview of our proposed conceptual model and research hypotheses. Then, we describe our research design and discuss the sample, measures, and data collection procedures. After reporting the study findings, we conclude with a discussion of the implications for practice and future research.. 6.

(19) CHAPTER 2. CONCEPTUAL MODEL AND HYPOTHESES. This chapter presents the conceptual framework for the research. The framework provides an overview of the relationship between team characteristics, organization contextual factors, and new product quality, as shown in Figure 1. Owing to the nature of the research questions examined in this study, the conceptual groundwork for the hypotheses is drawn from four streams of literature: group psychology, quality management, innovation, and cultural psychology. This literature suggests that the emergence of a quality outcome can be influenced by team characteristics and organizational contextual factors, and the conceptual framework can be affected differently by cultural differences, such as individualist and collectivist cultures. In addition, the model also illustrates three moderating relationships: 1. The relationship between speed-to-market pressure and new product quality is moderated by the capability of information integration. 2. The relationship between product innovativeness and new product quality is moderated by the capability of information integration. 3. The relationship between capability of information integration and new product quality is moderated by quality orientation. We first discuss the hypotheses linking team characteristics to new product quality, the organization contextual factors, then the moderator hypotheses, and finally the effects of culture on how team characteristics and contextual influences affect new product quality, as shown in Figure 2.. 7.

(20) FIGURE 1 A Model of Team Characteristics and Organization Contextual Factors Affecting New Product Quality. Team Characteristics Functional Diversity. H1. Tenure Diversity. H2. Capability of Information Integration. H3 ＋. H10. H9. ＋. ＋. ＋. New Product Quality. Organization Contextual Factors. －. Speed-to-market Pressure. H4. Product Innovativeness. H5. Quality Orientation. H6. Supplier Involvement. H7. Customer Involvement. H8. ＋. ＋ H11. 8. －. ＋. ＋. ＋.

(21) FIGURE 2 A Comparison of Taiwanese-American New Product Development Team’s Models. Collectivist Culture. Individualist Culture. Taiwanese New Product Development Team’s Model. Team Characteristics New Product Quality. American New Product Development Team’s Model. ≠. Organization Contextual Factors. Team Characteristics New Product Quality Organization Contextual Factors. 9.

(22) 2.1. Team Characteristics and New Product Quality As mentioned earlier, we considered two important team characteristics in this study – physical composition and team process. As to the physical composition, we focused on functional diversity and tenure diversity, which are likely to be of particular importance for new product development teams. In the quality literature, it is considered important to include several functional areas, which can bring diverse input to the decision-making process (Bounds et al., 1994; Clark & Fujimoto, 1991; Garvin, 1988). Additionally, different team tenure can provide a range of skills and perspectives to obtain higher overall performance (Ancona & Caldwell, 1992). Regarding team process, we focused on the capability of information integration across functional areas. The capability of information integration refers to the degree to which members of the team share, pay attention to, and challenge one another's information and perspectives resulting in new insights about the product (Sethi, 2000a). This can be viewed as a special case of the general concept of integration, which focuses on interaction and collaboration between functional areas at the firm level (Griffin & Hauser, 1996; Gupta, Raj, & Wilemon, 1986; Kahn, 1996; Ruekert & Walker, 1987). 2.1.1. Functional Diversity We defined cross-functional diversity as the number of functional areas represented in the team. The underlying reasoning is that the functional diversity of these teams increases the amount and variety of information available to design products. The increased information helps team members to understand the design process more quickly and fully from a variety of perspectives, and thus it improves design process performance (Brown & Eisenhardt, 1995; Lovelace, Shapiro, & Weingart 2001). As Lutz (1994) noted, cross-functional diversity brings together people from upstream and downstream functional areas so that they can communicate. 10.

(23) and contribute coordinated knowledge to bear on a project. Therefore, cross-functional new product development teams produce better-quality products more quickly and at lower cost. Thus, Hypothesis 1: The more functionally diverse new product development teams are, the higher the new product quality will be. 2.1.2. Tenure Diversity The diversity of team members’ tenure provides the new product development team with a range of experiences, information bases, biases, and contacts. Members who have entered the organization at different times know a different set of people and often have different technical skills and different perspectives on the organization’s history (Brown & Eisenhardt, 1995). We postulate that this range of skills and perspectives improves the probability that a group will generate innovative solutions (Souder, 1987) and have higher overall performance. Thus, Hypothesis 2: The more tenure diverse new product development teams are, the higher the new product quality will be. 2.1.3. Capability of Information Integration A team with a high capability of information integration is more likely to achieve a common understanding and consistency among team members across various decisions. This is because members can share information effectively, carefully attend to one another’s perspectives, and freely challenge these perspectives and their underlying assumptions (Sethi, 2000a). Therefore, developing a common understanding about the product and achieving consistency among decisions made throughout the product development process is considered critical for the development of a good product (Clark & Fujimoto, 1991; Garvin, 1988; Menon, Jaworski, & Kohli, 1997). Thus, Hypothesis 3: The more capability of information integration new product development 11.

(24) teams have, the higher the new product quality will be. 2.2. Organization Contextual Factors and New Product Quality In addition to the team characteristics discussed previously, the ability of a team to produce a high-quality outcome can also be affected by organization contextual influences on the team (Bounds et al., 1994). We therefore, focused on the effect of these influences to facilitate an understanding of how various contextual factors affect new product quality. The five contextual factors we considered in this study are (a) speed-to-market pressure, (b) product innovativeness from the perspective of the firm, (c) quality orientation in the firm, (d) suppliers’ involvement, and (e) customers’ involvement. 2.2.1. Speed-to-Market Pressure Karau and Kelly (1992) suggest that a minimum level of time pressure is useful in spurring individuals to strive for a superior outcome. However, beyond a certain point, individuals are less likely to search for additional information and are expected to restrict the processing of available information, because time pressure creates a need for cognitive closure. It is especially evident when team members strongly perceive time demands with respect to the product to be marketed. Under speed-to-market time pressure, team members may be forced to consider a narrow range of decision alternatives and may not be able to fully consider the various ways to build superiority into the product (Karau & Kelly, 1992). Thus, Hypothesis 4: Beyond a moderate level, speed-to-market pressure has a negative effect on new product quality. 2.2.2. Product Innovativeness If a new product is novel for a firm, it can hinder the emergence of a quality outcome. A very innovative product can require major changes in the existing technology and manufacturing 12.

(25) process and thereby disturb the balance among the product, technology, and manufacturing systems (Clark & Fujimoto, 1991). However, such a balance helps the firm improve quality by bringing the variation in production under control. Because team members who work on a novel product are likely to be overwhelmed with a large diversity of unfamiliar issues, they may not have the necessary frame of mind and the time to work in a focused manner on the process of continuous improvement while balancing the demands of a new technology and the manufacturing process (Imai, 1986). As a result, high product innovativeness can lead to a higher variation in production or lower product quality. Hypothesis 5: The more novel a new product is, the lower the new product quality will be. 2.2.3. Quality Orientation A strong quality orientation in the firm should generally spur the members of a product development team to strive for higher quality outcomes (Bounds et al., 1994; Garvin, 1988). On the contrary, without a commitment to quality on the part of various functional areas in the firm, a team may be less capable of developing a quality product. The resources and facilities of functional areas that lack a commitment to quality are also expected to be relatively low in quality. Thus, Hypothesis 6: The extent of quality orientation in a firm is positively associated with new product quality. 2.2.4. Supplier Involvement Previous research has associated a faster product development process with early (Gupta & Wilemon, 1990) and extensive (Clark & Fujimoto, 1991; Imai, Ikujiro & Takeuchi, 1985) supplier involvement. If extensive supplier involvement in product design can decrease the complexity of the design project, the result would be a faster and more productive 13.

(26) product-development process. Supplier involvement also can produce higher-quality products with fewer defects and improve the resulting product design’s technical capabilities through the incorporation of supplier-oriented innovations (Bonaccorsi & Lipparini, 1994). Thus, Hypothesis 7: The extent of supplier involvement in the product development process is positively related to new product quality. 2.2.5. Customer Involvement Customer involvement has also been shown to improve the effectiveness of the product concept (Cooper & Kleinschmidt, 1987; Sethi, Smith, & Park 2001; Zirger & Maidique, 1990). Hence, it is important to expose members of the product development team to customer needs and the consumption context (Bounds et al., 1994; Clark & Fujimoto, 1991; Hauser & Clausing, 1988). Moreover, it is useful to seek customer feedback about the team's ideas and plans (e.g., the product concept). This feedback serves as a reality check and ensures that the superiority the team has incorporated into the product is considered meaningful by customers (Sethi, 2000a). Thus, Hypothesis 8: The extent of customer involvement in the product development process is positively related to new product quality. 2.3. Moderators In this study, we considered two moderating variables: (a) capability of information integration and (b) quality orientation. Regarding the former, we focused on the moderating effects of speed-to-market time pressure and product innovation. Regarding the latter, we focused on the moderating effect of capability of information integration.. 14.

(27) 2.3.1. Capability of Information Integration as a Moderator on the Speed-to-Market Pressure and Quality Linkage When a product development team’s capability of information integration increases, its members are in a better position to reduce the adverse effect of high speed-to-market time pressure on new product quality (Sethi, 2000a). This is because team members can communicate effectively while finding time for essential activities. This can occur even under conditions of high speed-to-market time pressure, by setting task priorities, doing parallel work on different aspects of the project, and avoiding unnecessary backtracking of the development job between functional areas because of the common understanding they develop about the project (Clark & Fujimoto, 1991; Smith & Reinertsen, 1991; Takeuchi & Nonaka, 1986). The time thus saved by a team with a high capability of information integration can be used to carefully study the input of various functional areas, search for superior ways to satisfy customer needs, and keep variation in the production of the new product to a minimum. Thus, Hypothesis 9: When new product development teams’ capability of information integration increases, the negative effect of speed-to-market pressure on new product quality will be reduced. 2.3.2. Capability of Information Integration as a Moderator on the Product Innovativeness and Quality Linkage A product development team in which members work in a highly integrated manner might be able to minimize the disturbance in the operational balance among the product, technology, and manufacturing processes that an innovative new product creates (Clark & Fujimoto, 1991). Therefore, team members are likely to handle a large diversity of unfamiliar issues that a highly innovative new product can produce. As such, teams with high capabilities of information integration may still be able to help control variation in production, and then effectively reduce the negative effect of high innovativeness on quality. Thus 15.

(28) Hypothesis 10: When new product development teams’ capability of information integration increases, the negative effect of product innovativeness on new product quality will be reduced. 2.3.3. Quality Orientation as a Moderator of the Capability on Information Integration and Quality Linkage As quality orientation in a firm increases, teams with high capability of information integration are expected to produce a higher-quality product. This is because a team with high capability of information integration is likely to have a better and clearer understanding and plan for the development of a quality product. Thus, the team is in a better position to take advantage of other employees' commitment to quality and the superior facilities that a firm with high-quality orientation generally has (Sethi, 2000a). Thus, Hypothesis 11: When quality orientation in a firm increases, the positive effect of information integration on new product quality will be enhanced. 2.4. Effects of Culture on How Team Characteristics and Contextual Factors Affect New Product Quality People of different ethnic backgrounds possess different attitudes, values, and norms that reflect their heritage (Cox, Lobel, & McLeod, 1991). One of the few areas of cultural differences that has been extensively studied is the contrast between individualism and collectivism. Individualism-collectivism is an analytical dimension that captures the relative importance people accord to personal interests and to shared pursuits (Wanger, 1995; Cukur & Carlo, 2004). Because there is no standard definition of culture and no universal set of cultural dimensions (Javidan & House, 2001), there are potentially many ways that cultures can be different. Cultural differences include many dimensions, such as individualism-collectivism, power distance, uncertainty avoidance, assertiveness, and masculinity-femininity. Previous research has primarily focused on the cultural dimension of individualism-collectivism, and its 16.

(29) influence. on. new. product. development. teams’. quality. model.. In. addition,. the. individualism-collectivism cultural syndrome appears to be the most significant cultural difference among cultures (Triandis, 1996). Greenfield (2000) calls it the “deep structure” of cultural differences. While there are a myriad of cultural differences, this one seems to be important both historically and cross-culturally. Therefore, we chose individualism-collectivism as the key dimension of culture that might affect new product development teams’ quality model. Oyserman, Coon, and Kemmelmeier (2002) noted that, the core element of individualism is the assumption that individuals are independent of one another, whereas collectivism is the assumption that groups bind and mutually obligate individuals. As defined by Wanger and Moch (1986), individualism is the condition in which personal interests are accorded greater importance than are the needs of groups. Individualists look after themselves and tend to ignore group interests if they conflict with personal desires. In contrast, collectivism occurs when the demands and interests of groups take precedence over the desires and needs of individuals. Collectivists look out for the well-being of the groups to which they belong, even if such actions sometimes require that personal interests be disregarded. Cross-cultural researchers have suggested that the distinction between individualist and collectivist cultural values influence work-group composition, processes, and outcomes (Earley & Gibson, 1998; Gibson, 1999). This is because collectivist values emphasize social community (Durkheim, 1933), collective goals and harmony (Kluckhohn & Strodtbeck, 1961), and a self-concept rooted in the collective (Hofstede, 1991) parallel attributes of high-performing work groups (Mohrman, Cohen & Mohrman, 1995). For instance, in individualist societies, people are autonomous and independent from their in-groups; they give priority to their personal goals over the goals of their in-groups, they behave primarily on the basis of their attitudes rather than the norms of their in-groups, and exchange theory adequately predicts their social behavior. In 17.

(30) contrast, in collectivist cultures people are interdependent within their in-groups, give priority to the goals of their in-groups, shape their behavior primarily on the basis of in-group norms, and behave in a communal way. People in collectivist cultures are especially concerned with relationships (Triandis, 2001). In conflict situations, collectivists are primarily concerned with maintaining their relationship with others, whereas individualists are primarily concerned with achieving justice (Ohbuchi, Fukushima, & Tedeschi, 1999). Thus, collectivists prefer methods of conflict resolution that do not destroy relationships, whereas individualists are willing to go to court to settle disputes (Leung, 1997). Besides, collectivists emphasize the value of cooperation (Diaz-Guerrero, 1984), and thus prefer cooperative approaches like bargaining and mediation to conflict. resolution. (Leung,. 1988).. Wanger. and. Moch. (1986). also. found. that. individualism-collectivism is related to job type. To illustrate collectivists tended to perform jobs that required teamwork while individualists performed more independent tasks. Therefore, preference for teamwork is also likely to be higher for collectivists than for individualists (Sosik & Jung, 2002). Most previous investigators have examined preference for teamwork as part of a person’s cultural orientation. Earley (1989) tested individual versus group-oriented cultural orientations among U.S. and Chinese people and found that their cultural orientation had a significant effect on their preferred mode of work and group performance. Cox et al. (1991) found that collectivistic group members, assessed by collectivist-individualist orientation scales, displayed more cooperative behavior than did highly individualist group members. By allowing people with high levels of preference for group work perform their tasks in groups, organizations can satisfy people’s social motives and needs (Erez & Somech, 1996). Hill (1994) suggested that in individualist cultures such as the United States and the 18.

(31) Netherlands, people tend to use personal achievements to define themselves, view relationships as short-term, and value the individual more than the team; on the contrary, in collectivist cultures such as Taiwan and Japan, there is high commitment to, and identification with, the team, where group harmony, unity, and loyalty are valued more than individual gain. Oyserman et al. (2002) also found that European Americans showed more individualism—valuing personal interdependence more—and less collectivism—feeling duty to in-groups less—than others, whereas Chinese—from People Republic of China, Taiwan, and Hong Kong—showed less individualism and more collectivism. Although previous research has indicated that Taiwan is a collectivist country, some domestic researchers are still suspicious of its reliability as opposed to general cognition. This phenomenon can be explained by what some researchers refer to as multicultural minds. Hong, Morris, Chiu, and Benet-Martinez (2000) noted that bicultural individuals are typically described as people who have internalized two cultures to the extent that both cultures are alive inside of them. Many bicultural individuals report that the two internalized cultures take turns in guiding their thoughts and feelings (Phinney & Devich-Navarro, 1997). These situations are widely seen in the universities and high-tech companies. Because a lot of faculties and high-tech engineers are educated in the Western countries, mostly in the U.S., they are acculturated with Western social beliefs and values (Bond, 1993). When working in the universities and companies, they are easily activated by English language and American cultural icons, such as American flag, the White House and the photo of President Lincoln, and then tend to be individualistic, and vice versa. This interpretation could explain why some Taiwanese look more individualistic than the previous research has indicated. Previous researchers (Cox et al., 1991) have also indicated that collectivists perform better working in teams, when performing interdependent tasks, and sharing responsibility. 19.

(32) Cross-cultural studies have also shown that northern and western Europeans and North Americans tend to be individualists (Hofstede, 1980; Inkeles, 1983) and that Chinese people (Hsu, 1981), other Asians, Latins, and most east and west Africans tend to be collectivists (Hofstede, 1980). Therefore, collectivists have more cooperative teams and respond more favorably to team goals than individualists do. Such characteristics of high-performing teams may be inconsistent with teams composed of individualists, who perform better when working alone, with individual goals and individual responsibility (Earley & Gibson, 1998). Thus, Hypothesis 12: Team characteristics and organization contextual factors influence new product quality differently in collectivist cultures and individualist cultures. 2.5. Covariate Group research has established that group size is an important variable. It indirectly influences the potential magnitude of the coefficient of variation and may affect group process and communication (Ancona & Caldwell, 1992). We therefore, tested the model with team size as a control variable. The hypotheses of this study are summarized in Table 1.. 20.

(33) TABLE 1 Summary of Hypotheses H1. The more functionally diverse new product development teams are, the higher the new product quality will be.. H2. The more tenure diverse new product development teams are, the higher the new product quality will be.. H3. The more capability of information integration new product development teams have, the higher the new product quality will be. H4. Beyond a moderate level, speed-to-market pressure has a negative effect on new product quality.. H5. The more innovative a new product is, the lower the new product quality will be.. H6. The extent of quality orientation in a firm is positively associated with new product quality.. H7. The extent of supplier involvement in the product development process is positively related to new product quality.. H8. The extent of customer involvement in the product development process is positively related to new product quality.. H9. When new product development teams’ capability of information integration increases, the negative effect of speed-to-market pressure on new product quality will be reduced. H10. When new product development teams’ capability of information integration increases, the negative effect of product innovativeness on new product quality will be reduced.. H11. When quality orientation in a firm increases, the positive effect of information integration on new product quality will be enhanced.. H12. Team characteristics and organization contextual factors influence new product quality differently in collectivist cultures and individualist cultures.. 21.

(34) CHAPTER 3. RESEARCH METHOD. 3.1. Sample and Data Collection We surveyed teams that firms identified as being composed of members developing a new product or a revision of an existing product through the entire development task. Because this was a cross-national research, samples were taken from two countries, one in Taiwan and the other in the U.S.A. In Taiwan, subject teams were drawn from the top 1000 companies (in sales revenue) listed in the Common Wealth Magazine of 2004. Four hundred (400) companies which met our team selection criteria were chosen from among the top 1000, including electronics, electrical, information & communication, cleaner & cosmetics, chemical, transportation, and metal industries. There were two considerations for choosing 400 firms as our samples: (1) Not every firm in top 1000 companies develops a new product by itself. This is because the R&D budget is highly related to the company’s capital and scale. Research has shown that the more capital a company owns, the more money on R&D activities it is willing to invest. Thus, we chose the top 400 in 1000 companies to ensure that these firms have NPDTs to develop new product. (2) In order to meet the criteria of statistical power, generally set to 0.8, we also needed enough samples to survey. Therefore, we calculated the necessary sample size in this study by using the G-power software package, and determined that 113 samples were needed under the conditions that effect size=.15 (medium), α=.05, power=.8, and number of predictive variables=8. We made initial contact with the general manager of each company via e-mail requesting the company’s participation in the study. If they chose to participate in the research, a copy of the questionnaire, a personalized letter, and a return envelope were mailed to the informants. The measures of the questionnaire were translated from English to Chinese by using the conventional method of 22.

(35) back-translation to ensure their relevance to the Chinese context. In the U.S.A., we took our samples from two sources: (1) NASA contractors at the Johnson Space Center in Houston, TX, and (2) semiconductor firms in Dallas, TX. For the sake of comparison between Taiwanese and American teams, we selected 400 teams as our samples. Following consent to participate from these organizations, 400 questionnaires were distributed and collected at the site by our coordinators. We also agreed to provide a report with the findings of the study for those organizations surveyed both in Taiwan and in the U.S.A., including results on how their teams compared to the entire sample, in return for their participation. We received 184 completed questionnaires from 63 firms in Taiwan, of which 16 questionnaires were deemed not usable because of missing data. Thus, the effective response rate was 42 % (168/400). Each company in Taiwan had at least two or more teams participating in the research. Respondents were 10% female and 90% male in Taiwanese teams. Alternatively, we collected 176 completed questionnaires from 25 NASA’s contractors and 50 semiconductor companies in the U.S., of which 15 questionnaires were regarded not usable because of missing data. Each semiconductor firm had only one team taking part in the study, whereas each NASA’s contractor at least had two or more teams. Therefore, the effective response rate was 40.25 % (161/400). Respondents were 15%female and 85% male in American teams. 3.2. Measures and Validation 3.2.1. New Product Quality We used the scale developed by Olson, Walker, and Reukert (1995) to measure new product quality. Ten items asked team managers for their assessment of the degree to which the new product delivers value to the customer and meets the quality control standards laid out for it by the team/organization. These ten items were as follows: (1) Quality of this product compares. 23.

(36) well with similar product offered by your competitors, (2) The product meets the customer’s needs, (3) Complaints have been received regarding the poor performance of this product, (4) The product meets the specifications outlined for it, (5) The product is reliable, (6) This product is of a higher quality than competing products available to customers, (7) The product’s performance shows little deviation from expected standards, (8) Quality of this product compares well with other products developed by our organization, (9) The consumers of this product perceive our product to be better than our competitors’, and (10) This product will deliver benefits to the customers that are not currently available to them. A seven-point Likert scale (1 = “strongly disagree” and 7 = “strongly agree”) was used for each item. The ten items were averaged to produce a measure of new product quality. 3.2.2. Functional Diversity We measured functional diversity with the diversity index recommended by Teachman (1980) and used by Ancona and Caldwell (1992) and Keller (2001): s. H = − ∑ Pi (ln Pi ) i =1. Under this formula, the greater the distribution of new product development team members across different functional areas, the higher the score would be for functional diversity. Membership in a functional area was determined by company organization charts. In this formula, H represents heterogeneity and P is the fraction share of team members assigned to marketing, manufacturing, engineering, and so forth. Thus, the diversity index represents the products of each function’s proportion in the team and the natural log of its proportion. The only exception occurs when an area is not represented (Ancona & Caldwell, 1992; Keller, 2001; Teachman, 1980).. 24.

(37) 3.2.3. Tenure Diversity Allison (1978) and Pfeffer and O’Reilly (1987) suggest that the coefficient of variation (the standard deviation divided by the mean) provides the most direct and scale in-variant measure of dispersion. Therefore, to assess the relative homogeneity of team’s tenure, each team’s standard deviation of tenure was divided by the team mean (Ancona & Caldwell, 1992). 3.2.4. Capability of Information Integration The measure used to evaluate the capability of information integration is based on the conceptualization of integration suggested by Gupta, Raj, and Wilemon (1986) and the measure of market information used suggested by Deshpand and Zaltman (1982). We used four items to measure the degree to which members of a new product development team share, pay attention to, and challenge one another’s information and perspectives to discover new ideas about the product. Members were asked to evaluate the following: (1) Members freely share information and perspectives with one another, (2) When making important project-related decisions, members pay great attention to the information and the perspectives of members from other departments, (3) Members freely challenge the assumptions underlying one another’s ideas and perspectives, and (4) Exposure to the information and perspectives of other departments help members think of new ideas about the product. Items for each construct were scored on a seven-point Likert scale (1 = “strongly disagree” and 7 = “strongly agree”). We averaged these items to produce a measure of capability of information integration. 3.2.5. Speed-to-Market Pressure We used three items developed by Sethi (2000a) to measure speed-to-market pressure that team members experienced during the development of new product. Items for each construct were scored on a seven-point Likert scale (1 = “strongly disagree” and 7 = “strongly agree”). 25.

(38) These items were as follows: (1) On this project, team members often wished they had more time to complete their work, (2) On this project, team members had plenty of time to think carefully about project-related details, and (3) On this project, team members believed they were under a high degree of speed-to market pressure. Items were averaged to produce a measure of speed-to-market pressure on team members. 3.2.6. Product Innovativeness Four items, derived from Booz, Allen and Hamilton (1982), were used to assess how novel the product was for the firm. Respondents were asked to select one of four items: (1) an improvement/modification in one of the firm’s existing product, (2) a new product added to one of the existing product lines, (3) a new product that is a new line for the firm, but similar products are offered by other companies, and (4) a totally new product, and no similar products are offered by other companies, to indicate the perspective of their firm on product innovativeness. A score of four indicates that innovativeness is high and the product is novel for the firm, and vice versa. 3.2.7. Quality Orientation We used three items to measure quality orientation: the extent to which a firm (a) emphasizes product quality (b) creates a commitment to produce a quality product among its employees, and (c) implements total quality management. A seven-point Likert scale (1 = “strongly disagree” and 7 = “strongly agree”) was employed for all items. We averaged these items to produce a measure of quality orientation. 3.2.8. Supplier Involvement Three items asked informants for their assessment of supplier’s involvement in the 26.

(39) development process. Items reflected the extent to which input and feedback from suppliers was relied on in the development of the product idea/concept, the evaluation of product idea/concept, and the development of design. A seven-point Likert scale (1 = “a great extent” and 7 = “not at all”) was employed for all items. Items were averaged to produce a measure of supplier involvement. 3.2.9. Customer Involvement Three items asked informants for their assessment of customer’s involvement in the development process. Items reflected the extent to which input and feedback from customers were relied on in the development of the product idea/concept, the evaluation of product idea/concept, and the development of design. A seven-point Likert scale (1 = “a great extent” and 7 = “not at all”) was employed for all items. Items were averaged to produce a measure of customer involvement. 3.2.10. Covariate Team size was included as a control variable because prior research has found it to be related to cohesiveness and internal communication for teams (Ancona & Caldwell, 1992; Bantel & Jackson, 1989). 3.3. Analytical Procedures To test the relationships that we hypothesized in the model, we conducted several analyses. First, we used the Amos 5 statistical software to conduct confirmatory factor analysis (CFA) in order to examine the validity of the measurement models of both dependent variables and independent variables. Second, we performed reliability analysis (Cronbach’s α) for testing the consistency of measurement. Third, we centered all the data upon variable means to decrease 27.

(40) the likelihood of multicollinearity in equations having multiple interaction terms (Cohen & Cohen, 1983). This procedure also had the effect of making the interaction terms more interpretable. Fourth, we also examined the variance inflation factors (VIFs) associated with each regression coefficient to determine if multicollinearity was present in the model. Fifth, for dependent variables that had multiple predictors, we examined the correlation between the predictor variable and the dependent variable. Sixth, we used hierarchical moderated regression analysis, suggested by Aiken and West (1991) and Jaccard, Wan, and Turrisi (1990), to test the contingency hypotheses. Finally, we conducted residual and influence analyses to test if our data complied with linearity, independence, normal distribution and equal variance assumptions, and the results of regression models were influenced by some observation points. All tests except as noted were performed using SPSS.. 28.

(41) CHAPTER 4. ANALYSIS AND RESULTS. In this section, we present the results of our tests of hypotheses and discuss their substantive interpretation. In turn, we consider the effects of team and contextual factors on new product quality, and the moderator hypotheses. 4.1. Analysis 4.1.1. Scale Validity and Reliability First, we examined the validity of the constructs with confirmatory factor analysis. Given variable attributes, we divided the constructs into two sub-models: independent variables and dependent variable. Each of the five independent variables (i.e., capability of information integration, speed-to-market pressure, quality orientation, supplier involvement, and customer involvement) was modeled as a latent variable measured by multiple indicators, or manifest variables (e.g., capii1, capii2, capii3, and capii4 in the case of capability of information integration). It is not necessary to assess the validity of the measurement model of the other independent variables (functional diversity, tenure diversity, and product innovativeness) since we employed one scale to measure each independent variable. The fit indexes of independent variables are as follows: (1) Taiwanese teams: χ2 = 134.105, p = .000, χ2/df = 1.698, GFI = .907, RMSEA = .065, NNFI = .931, CFI = .955, and (2) American teams: χ2 = 139.062, p = .000, χ2/df = 1.760, GFI = .911, RMSEA = .069, NNFI = .903, CFI = .936. Using the cutoff criteria suggested by Hu & Bentler (1999), an index greater than or equal to 0.90 indicates an adequate fit. Therefore, the fit indexes indicate that the models fit the data well, as shown in Table2, providing evidence of validity. The fit indexes of dependent variables are as follows: (1) Taiwanese teams: χ2 = 56.615, p = .000, χ2/df = 2.265, GFI = .937, RMSEA = .087, NNFI = .937, CFI = .965, and (2) American teams: χ2 = 32.770, p = .137, χ2/df 29.

(42) = 1.311, GFI = .959, RMSEA = .044, NNFI = .971, CFI = .984. As illustrated in Table 3, the fit indices of dependent variables also show that the models fit the data well, providing evidence of validity. All items loaded on their respective constructs, and each loading was large and significant at the .01 level. Second, we tested the reliability of the measurement with reliability analysis. As Table 4 and Table 5 show, Cronbach’s α for multi-item scale ranged from .61 to .91, with all but one (speed-to-market pressure) having alphas over .70, indicating high reliability of the constructs, and thus meeting the requirements of consistency suggested for exploratory research.. 30.

(43) TABLE 2 Fit Indices of Independent Variables of Taiwanese and American Teams Index. Taiwanese Teams. American Teams. χ2 test. 134.105. 139.062. χ2/df. 1.698. 1.760. GFI. 0.907. 0.911. RMSEA. 0.065. 0.069. NNFI. 0.931. 0.903. CFI. 0.955. 0.936. TABLE 3 Fit Indices of Dependent Variables of Taiwanese and American Teams Index. Taiwanese Teams. American Teams. χ2 test. 56.615. 32.770. χ2/df. 2.265. 1.311. GFI. 0.937. 0.959. RMSEA. 0.087. 0.044. NNFI. 0.937. 0.971. CFI. 0.965. 0.984. 31.

(44) FIGURE 3 Independent Variable Measurement Model of Taiwanese Teams capii1 capii2 capii3 capii4. .622 .801. Capability of Information. .690. Integration. .765. .071 .566. stmpre1 stmpre2. .722 .501. .309. Speed-to-market Pressure. .416. .646 .117. stmpre3 qltyor1 qltyor2. .738 .908. Quality Orientation. .411. .659 qltyor3 supinv1 supinv2. .091 .947 .926. .189. .579. .618. supinv3 cusinv1 cusinv2. .251. Supplier Involvement. .993. Customer Involvement. .838 .588. cusinv3. 32.

(45) FIGURE 4 Independent Variable Measurement Model of American Teams. capii1 capii2 capii3 capii4. .685 .810. Capability of Information. .642. Integration. .537. -.135 .795. stmpre1 stmpre2. .642 .323. .109. Speed-to-market Pressure. .496. .861 -.050. stmpre3 qltyor1 qltyor2. .792 .681. Quality Orientation. .274. .585 qltyor3 supinv1 supinv2. -.030 .884 .722. .317. .805. .430. supinv3 cusinv1 cusinv2. .025. Supplier Involvement. .962. Customer Involvement. .984 .358. cusinv3. 33.

(46) FIGURE 5 Dependent Variable Measurement Model of Taiwanese Teams. npq1. npq2. .663 .722. npq3 .624 npq4. npq5. .538 New Product Quality. .652 .737. npq6. npq7. npq8. npq9. .648 .819 .784. .847. npq10. 34.

(47) FIGURE 6 Dependent Variable Measurement Model of American Teams. npq1. npq2. .539 .642. npq3 .266 npq4. npq5. .746 New Product Quality. .590 .761. npq6. npq7. npq8. npq9. .465 .699 .706. .517. npq10. 35.

(48) 4.1.2. Multicollinearity Test To reduce the potential problem of multicollinearity among interaction terms and their constituent terms in the regression model, we centered all the data around variables means, namely, replacing values by deviations from the means (Aiken & West, 1991; Jaccard, Wan, & Turrisi, 1990). Fourth, we examined the variance inflation factors (VIFs) associated with each regression coefficient from two regression models, and found that a range of from 1.04 to 3.32 (<10) for Taiwanese teams and 1.02 to 2.31 (<10) for American teams, suggesting no serious problems with multicollinearity. 4.1.3. Descriptive Statistics Means, standard deviations, and a correlation matrix of all variables are presented in Table 4 for Taiwanese teams and Table 5 for American teams. An examination of correlations among independent variables showed that correlations ranged from 0 to .56 for Taiwanese teams and 0 to .60 for American teams.. 36.

(49) TABLE 4 Means, Standard Deviations, and Correlations of Taiwanese Teams Variable. Mean. s.d.. α. 1. 1. New product quality. 5.55. 0.72. 0.91. 2. Functional diversity. 1.18. 0.45. －. - .07. 3. Tenure diversity. 0.51. 0.23. －. .06. 4. Capability of. 5.80. 0.75. 0.81. 4.73. 0.66. 0.65. 6. Product innovation. 2.50. 0.87. 7. Quality orientation. 6.15. 8. Supplier involvement 9. Customer involvement. .56***. 2. 3. 4. 5. 6. 7. 8. 9. - .02 .00. .04. - .02. .17. .05. .06. －. - .50***. .16*. .07. - .26***. .06. 0.77. 0.81. .40***. - .01. - .06. .45***. .10. - .32*. 4.96. 1.12. 0.87. .16*. - .18**. - .03. .25***. .28***. - .19**. .07. 5.07. 1.12. 0.84. .25***. - .12. .07. .36***. .17*. - .19**. .13*. .55***. 10.45. 5.35. －. .17*. .06. .11. .05. information integration 5. Speed-to-market pressure. 10. Team size a. .50. .36***. n = 168. * p < .05 ** p < .01 *** p < .001. 37. .06. - .01. .06.

(50) TABLE 5 Means, Standard Deviations, and Correlations of American Teamsa Variable. Mean s.d.. α. 1. 2. 3. 4. 1. New product quality. 5.81. 0.63. 0.84. 2. Functional diversity. 0.94. 0.50. －. 3. Tenure diversity. 0.77. 0.27. －. 4. Capability of. 5.78. 0.84. 0.76. .51***. .20*. - .04. 5.47. 0.82. 0.61. .06. .23**. .07. - .13. 6. Product innovation. 2.48. 1.04. －. .07. .10. - .01. .00. 7. Quality orientation. 5.95. 0.77. 0.71. .55***. .16**. - .03. .60***. 5. 6. 7. 8. 9. .33*** - .12. - .03. information integration 5. Speed-to-market pressure. 8. Supplier involvement. 4.14. 1.69. 0.88. 9. Customer involvement. 5.22. 1.17. 0.73. 10. Team size a. 15.79 10.71. －. - .06. .01. - .08. .08. .22**. .00. - .05. .41***. .12. .31***. .00. .01. .15 - .05 .22. ***. - .03 .30***. - .03 .00. .04. - .07. .32***. .36***. - .14. .18*. .44***. .32***. n = 168. * p < .05 ** p < .01 *** p < .001. 4.1.4. Hierarchical Moderated Regression To test our hypotheses, we used two regression models in which we regressed the dependent variable on (a) the direct effect predictor variables in model 1, and (b) the multiplicative interaction terms plus the direct effects in model 2. We then computed the difference in the multiple squared correlation coefficient (R2) of the equation testing the direct effects and the equation with the interaction terms to determine whether the interaction terms existed. As shown in Table 6, model 2, which is with the interaction terms, increased the multiple 38.

(51) squared correlation coefficient (R2) by 2 percent, compared to model 1, indicating the existence of moderating effects in Taiwanese teams. This finding shows that model 2 explained more variance and was a better fit than model 1. We therefore, used model 2 to test our hypotheses for Taiwanese teams. In contrast, we found that in Table 7, model 2 did not increase as much as a percent of its multiple squared correlation coefficient (R2), compared to model 1, suggesting the nonexistence of the same moderating effects in American teams. Thus, we replaced these three interactions by another two interactions, i.e., supplier involvement and capability of information integration and customer involvement and capability of information integration. We then conducted the regression analysis, and found that the multiple squared correlation coefficient (R2) increased by 2 percent in Table 8, compared to model 1, indicating the existence of moderating effects in the revised regression analysis of American teams. This finding also shows that model 2, in Table 8, explained more variance and was a better fit than model 1. Therefore, we used model 2 in Table 8 to test our hypotheses for American teams.. 39.

(52) TABLE 6 Results of Moderated Regression Analyses of Taiwanese Teams New Product Quality Model 1 Independent Variables. Model 2. Β. t. Β. t. 0.08. 1.18. 0.09. 1.49. - 0.03. - 0.48. - 0.07. - 1.01. 0.08. 1.30. 0.07. 1.24. Control variable Team size Direct effects Functional diversity Tenure diversity. 0.39. 5.48. 0.66. 2.15*. Speed-to-market pressure. - 0.05. - 0.76. - 0.17. - 2.19*. Product innovativeness. - 0.38. - 6.18***. - 0.50. - 2.04*. 0.17. 2.58**. 0.59. 2.39**. Capability of. ***. information integration. Quality orientation Supplier involvement. - 0.03. - 0.38. - 0.02. - 0.25. Customer involvement. 0.03. 0.38. 0.01. 0.15. 0.22. 2.05*. 0.13. 0.54. - 0.78. - 1.82†. Moderating effects Capability of information integration × speed-to-market pressure Capability of information integration × product innovativeness Quality orientation × capability of information integration R2. 0.45. 0.47. 2. 0.02***. ∆R. 16.57***. F. 13.48***. † p < .10 * p < .05 ** p < .01 *** p < .001. 40.