Technological Forecasting & Social Change

Transcription

1 Technological Forecasting & Social Change 76 (2009) Contents lists available at ScienceDirect Technological Forecasting & Social Change Intellectual capital and new product development performance: The mediating role of organizational learning capability Ya-Hui Hsu a,, Wenchang Fang b,1 a Department of Business Administration, Ming Chuan University, 11F, No.318, Fuhe Rd., Yonghe City, Taipei 234, Taiwan b Department of Business Administration, National Taipei University, 69, Sec 2, Jian-Kuo N. Rd, Taipei 104, Taiwan article info abstract Article history: Received 29 June 2007 Received in revised form 15 November 2007 Accepted 26 March 2008 Keywords: Intellectual capital Organizational learning capability New product development performance Partial Least Square Previous studies rarely examined the relationship between intellectual capital and organizational learning capability. Moreover, most studies neglect the mediating effect of organizational learning capability in the relationship between intellectual capital and new product development performance. This study uses interviews and the survey method to discuss the relationships governing intellectual capital, organizational learning capability, and new product development performance. Results are based on empirical data from Taiwan's IC design industry, and are generated by the Partial Least Squares (PLS) method. Results show that human capital and relational capital actually improve new product development performance through organizational learning capability. Although structural capital positively affects organizational learning capability, managers should pay attention to possibly negative effects of structural capital on new product development performance. Relational capital is the greatest factor among these three types of intellectual capital in Taiwanese IC design companies, structural capital is second, and human capital is last. Comparing three types of intellectual capital of Taiwan's large enterprises with those of Taiwan's small and medium enterprises (SMEs) reveals that the relational capital of Taiwan's SMEs is marginally less than that of large enterprises Elsevier Inc. All rights reserved. 1. Introduction Intellectual capital is becoming a crucial factor for a firm's long-term profit and performance in the knowledge-based economy as more and more firms identify their core competence as invisible assets rather than visible assets [1]. Nonaka and Takeuchi [2] point out that future society is a knowledge-based society in which knowledge storage and application are the basis of economic growth and accumulated capital. Industries in such a society do not rely on traditional production factors for their competitive advantage, but on knowledge management and integration. This trend stresses the importance of organizational learning capability and how to create, manage, and evaluate intellectual capital. Despite the fact that intellectual capital and organizational learning capability are so important to firms' development, few studies focus on how these two factors relate to each other and affect new product development performance. Griffin's research posits that about 32.4% of company sales is generated by new market products [3]. New product development is necessary for firm survival and competitive advantage [4], especially salient in the high-tech industry. Many studies focus on new product development performance [5 8], yet analysis of factors affecting new product development performance remains incomplete. For example, although Chen et al. [9] discuss the relationship between intellectual capital and new product Corresponding author. Tel.: ; fax: addresses: Yahui_hsu@msn.com (Y.-H. Hsu), fang@mail.ntpu.edu.tw (W. Fang). 1 Tel.: ; fax: /$ see front matter 2008 Elsevier Inc. All rights reserved. doi: /j.techfore

2 Y.-H. Hsu, W. Fang / Technological Forecasting & Social Change 76 (2009) development performance, the study does not address how intellectual capital and organizational learning capability simultaneously affect new product development performance. Studies often neglect organizational learning capability as a firm's mediating role. Therefore, this study fills the research gap by exploring the influence of three types of intellectual capital i.e. human capital, structural capital, and relational capital and organizational learning capability upon new product development performance. This study examines the relationship between intellectual capital including human capital, structural capital, and relational capital and organizational learning capability. Then, the mediating role of organizational learning capability in the intellectual capital new product development relationship is discussed. Third, this study uses the Partial Least Squares (PLS) method to analyze the joint effect of intellectual capital and organizational learning capability on improving new product development performance. Hopefully, these research results can help governments or managers and contribute to relevant studies and future research. 2. Literature review 2.1. Definition of intellectual capital Intellectual capital, a term first introduced by economist John Kenneth Galbraith in 1969, refers to the difference between an organization's market value and book value. Many researchers have come to regard intellectual capital as a firm's primary means of creating competitive advantage. The abstract and dynamic nature of intellectual capital makes it difficult for scholars to define [10]. Moreover, Guthrie [11] even notes that some consider intellectual capital and intellectual assets or intangible assets as synonyms. Previous studies indicate that intellectual capital is the product of dynamic business operation processes, and is closely related to knowledge management or organizational learning [12 15]. Some researchers also contend that accumulating intellectual capital is beneficial to creating competitive advantage or business values [16 19,12,14,15,20]. Following the above-mentioned literature, this study thus defines intellectual capital as the total capabilities, knowledge, culture, strategy, process, intellectual property, and relational networks of a company that create value or competitive advantages and help a company achieve its goals Classification of intellectual capital Previous researchers have their own intellectual capital classification due to research subjects and background, and there seems to be no consistent formulation that comprises all intellectual capital evaluation methods [19]. However, with increasing discussions on intellectual capital, most studies follow the framework proposed by Roos et al. [14], Bontis [21], Johnson [22] and Bozbura [23], adopting human capital, structural capital, and relational capital as the three basic dimensions of intellectual capital, also adopted in this study. Furthermore, this research defines structural capital as including process capital and innovation capital, following the studies of Johnson [22] and Van Buren [24]. Human capital comprises all business capital embedded in employees and not owned by the organization. This capital may be taken away by employees, and includes employees and managers' competence, experience, knowledge, skills, attitude, commitment, and wisdom. Process capital is defined as workflow, operation processes, specific methods, business development plans, information technology systems, and cooperative culture, etc. Innovation capital is defined as intellectual property within an organization, including patents, copyrights, trademarks, and knowhow, etc [24,25]. This term also includes R&D expenses and employees an organization invests in new product development performance. Relational capital includes all value of stakeholders, customers, and supplier relations [22] Organizational learning capability Although previous researchers provide several definitions of organizational learning capability, they often emphasize only part of the concept. This study divides organizational learning capability into absorptive capability and transformative capability for compatibility, following the works of Cohen and Levinthal [26] and Garud and Nayyar [27]. Concerning absorptive capability, Cohen and Levinthal [26] emphasize the external element of capability, noting that the ability to evaluate and utilize outside knowledge is largely a function of prior related knowledge. Prior knowledge confers an ability to recognize the value of new information, assimilate it, and apply to it to commercial ends. Regarding transformative capability, Garud and Nayyar [27] emphasize the internal element, indicating that transformative capability is the ability to choose technologies, maintain them over time, and reactivate and synthesize them with ongoing technology development efforts. Based on previous studies, this study defines organizational learning capability as an organization's ability to absorb and transform new knowledge and apply it to new product development with competitive advantage and high production speed New product development performance Experts and scholars propose different definitions for a new product based either on the product [28], the producer [29], the consumers [30], or product life cycles [31]. This study focuses on the firm, adopting the definition of a new product for a firm, no matter if it is new to the market or not, proposed by Souder [29]. New product development performance is a multidimensional construct [32]. Researchers use various performance evaluation measures based on different research focuses. New product development performance in this study includes market performance, financial performance, customer performance, and product performance.

3 666 Y.-H. Hsu, W. Fang / Technological Forecasting & Social Change 76 (2009) Key success factors in new product development Many studies discuss key success factors of new product development. Several key success factors include: (1) the firm must own a high-quality new product process and design new products with customer orientation; (2) the firm must be able to define new product strategy, including goals, objectives, and areas of strategic focus; (3) senior management must make the necessary resource commitment to new products and product development; (4) the firm must have high-quality new product teams, including a dedicated team leader, strong and frequent communication and interaction, quick and efficient decision making, etc; (5) the firm must possess an innovative climate and culture [33]. Benedetto [34] also indicates that key success factors of new product development comprise: (1) cross-functional teams to make key marketing and manufacturing decisions; (2) high-quality selling efforts, advertising, and technical support; (3) involving logistics early in planning; (4) good launch management and good support program management; (5) forecasting right timing for the launch; (6) information gathering regarding market testing, customer feedback, advertising testing, etc. Key success factors of new product development based on Cooper and Kleinschmidt's [33] and Benedetto's [34] discussion include human capital, structural capital, and relational capital and might relate to organizational learning capability. Intellectual capital might influence new product development performance through organizational learning capability, including the capabilities in launch, marketing, forecasting, information gathering, etc. discussed above. In another words, previous literature implies organizational learning capability as a possible mediating role between intellectual capital and new product development performance. 3. Framework and hypotheses Analysis in this study is at the corporate level, and this research was conducted in Taiwanese IC design companies. Fig. 1 shows the conceptual framework, indicating three dimensions of intellectual capital affecting each other and influencing organizational learning capability. The framework also examines intellectual capital effects of these three dimensions on new product development performance. Moreover, this framework examines whether or not intellectual capital affects new product development performance via the mediating role of organizational learning capability. Previous studies only emphasize the relationship between intellectual capital and new product development performance; however, the whole process should include organizational learning capabilities. Without organizational learning capabilities of a firm, intellectual capital per se cannot achieve performance. The following explains specific hypotheses concerning these relationships, together with their underlying rationale Intellectual capital Managerial activities related to intellectual capital should complement each other. Edvinsson and Malone [19] point out that human, structural, and customer capital reciprocally circulate and affect each other. Their collaboration in generating knowledge value creates synergy. Stewart [12] also notes that human capital, structural capital, and relational capital complement each other, and intellectual capital will be most effective only when these dimensions support each other. Based on the discussion above, this study offers the following hypothesis. Fig. 1. Conceptual framework.

4 Y.-H. Hsu, W. Fang / Technological Forecasting & Social Change 76 (2009) Hypothesis 1. Human capital, structural capital, and relational capital reciprocally affect each other. The research findings of Bontis et al. [35] show that human capital significantly affects customer capital in all industries. Human capital significantly affects structural capital in non-service industries, while customer capital affects structural capital in both service and non-service industries. Employee abilities also affect a firm's process efficiency and innovation processes. Higherquality employees will be more pleasant and provide more information to customers and business partners. These employees will attract good customers and business partners as a result. In other words, human capital positively affects relational capital. A firm maintaining a good relationship with its customers and business partners enables the employee to discuss business processes or innovations with customers and business partners. In other words, relational capital positively affects structural capital. Based on the discussion above, this study offers the following hypotheses. Hypothesis 1a. Human capital positively affects structural capital. Hypothesis 1b. Human capital positively affects relational capital. Hypothesis 1c. Relational capital positively affects structural capital Intellectual capital and organizational learning capability Intellectual capital is the valuable knowledge of an organization [36]. Wiig [37] notes that intellectual capital comprises all assets created by intellectual activities, including knowledge acquisition, innovation, and creation. Intellectual capital also affects and advances knowledge management and improves organizational learning capability. Based on these discussions, this study offers the following hypothesis. Hypothesis 2. Intellectual capital positively affects organizational learning capability. Knowledge innovation is a key element in the creation of product value and economic growth in a knowledge-based economy [38,39]. Knowledge innovation depends on organizational learning capability improvement. Baker [40] contends that high-quality employees are the most important factor for a firm in this economy. Managers train their employees to be knowledgeable workers, improve their overall quality, and increase organizational learning capability. Pfeffer [41] notes that the manner in which a firm retains and trains its best human resources is the most important competitive strategy within the knowledge economy context. Kought and Zander [42] contend that organizational learning depends on the exchange and integration of extant information, knowledge, and ideas exerted by organizational members. The better the employee quality, the more knowledge generated to improve organizational learning capability. Based on the discussion above, this study offers the following hypothesis. Hypothesis 2a. Human capital positively affects organizational learning capability. This study proposes that the system and process of resolving problems and creating value in a firm belongs to the organization's structural capital. These systems and processes include a firm's overall process, organizational structure design, and the capacity to utilize information technology, intellectual assets management, and information system structures. Persaud et al. [43] consider information technology as a means of acquiring external knowledge. Corporate culture, one element of structural capital, may also be helpful in developing an organization's learning capability [44,45]. For example, a firm may provide an open learning environment to encourage informal learning in its members. Employees may be more willing to share knowledge with others in such a culture as a result. A firm's efforts to establish an intellectual assets management system will also help it better utilize patents and intellectual property rights, pursuing value maximization and improving organizational learning capability. Focusing structural capital on establishing a databank will improve a firm's organizational learning capability and reduce the costs of decision making and misjudgment due to insufficient information. Based on the discussion above, this study offers the following hypothesis. Hypothesis 2b. Structural capital positively affects organizational learning capability. Persons with more relationships and networks may exhibit better information acquisition and resource allocation [46].Hansen[47] and Tsai [48] note that employees with better communication skills and more connections with outside people have more opportunities to access different resources. This connectedness may improve a firm's technology development capability. Maintaining and creating stable relational capital requires that a firm build up good interactive relationships with its stakeholders, customers, and suppliers. Better relationships provide greater possibilities that business partners will share their professional technology with each other. Firms gain more new knowledge from customers to improve their own organizational learning capability in the meantime. Firms can also acquire various innovative knowledge resources and seek business opportunities using different cooperative network types, including alliances among the same or different industries and conducting studies with research institutes, to reap new benefits and advantages [49]. Based on the discussion above, this study offers the following hypothesis.

5 668 Y.-H. Hsu, W. Fang / Technological Forecasting & Social Change 76 (2009) Hypothesis 2c. Relational capital positively affects organizational learning capability Intellectual capital and new product development performance Intellectual capital improves new product development performance because it transforms knowledge into value [18]. Harrison and Sullivan [50] note that a firm's value stemming from intellectual capital may include: (1) profit, the income from the product or service; (2) strategic positioning, such as market share, leadership (innovation), standard-setting and name recognition; (3) acquiring the innovations of others; (4) improving customer loyalty; (5) reducing costs; (6) and improving productivity. The values mentioned above also serve as indicators of new product development performance. Chen et al. [9] consider that when a firm has more intellectual capital, inclusive of human capital, structural capital and relational capital, it has more innovative competence to further raise its new product development performance. Based on the discussion above, this study offers the following hypothesis. Hypothesis 3. Intellectual capital positively affects new product development performance. Hurwitz et al. [51] show that a firm's profit comes primarily from human capital. A company hiring capable employees and experienced managers will have higher new product development performance because its manpower can bring skills and capabilities into full play [52,9]. Outstanding managers with knowledge and vision help firms quickly respond to the market in highly turbulent conditions. Thanks to managers, firms may lead the market, and increase their market share. Moreover, Dakhli and De Clercq [53] indicate a positive relationship between human capital and innovation. Human capital accumulation determines teamwork performance [54], and the better the employee quality, the higher the innovation performance [55]. Based on the discussion above, this study offers the following hypothesis. Hypothesis 3a. Human capital positively affects new product development performance. Structural capital includes two dimensions: process capital and innovation capital [22]. Edvinsson and Malone [19] indicate that process capital supports employees and improves organization productivity, while simultaneously maintaining business orders for goods and product quality. A learning organization improves knowledge management efficiency and builds up a sharing and communicating culture, positively affecting new product performance [56]. Bontis et al. [35] empirical study on intellectual capital and firm performance in Malaysia shows that structural capital positively affects firm performance in the service industry but not in the non-service industry. Pena [57] also argues that structural capital affects new business survival and growth. Specifically, research and development expenses a major element of innovation capability enhance a firm's future competitiveness. Aboody and Lev's [58] study on eighty-nine chemistry companies from 1980 to 1999 shows that one dollar spent on research and development produces two dollars profit after ten years. Based on the discussion above, this study offers the following hypothesis. Hypothesis 3b. Structural capital positively affects new product development performance. Relational capital refers to a firm's relationship with its stakeholders, customers, and suppliers. Hill and Jones [59] note that organization stakeholders often provide the organization with important resources. For example, customers provide income, suppliers provide raw materials, and distributors provide sales channels. Moreover, firms' strategic partners are beneficial for new product development [60 62,9]. Firms should therefore understand and satisfy stakeholder needs to improve their new product development performance. Many theories and empirical studies argue that customer collaboration is very valuable to achieving innovation and economic success. This cooperation benefit may come in the form of new product ideas that improve product development efficiency and even reduce new product launch time [63 66]. Previous studies also confirm that customer participation in the early stages of new product development is a key element to new product success [67]. Schulz [68] further indicates that common values and trust in a network not only resolve conflicts, but also improve communication and coordination between network members. These characteristics are vital to new product development performance. Based on the discussion above, this study offers the following hypothesis. Hypothesis 3c. Relational capital positively affects new product development performance Organizational learning capability and new product development performance Nonaka and Takeuchi [2] argue that a firm's success lies in its knowledge, creative capabilities, and techniques. Hunt and Morgan [69] point out that organizational learning capability is a very important and complicated resource which can create competitive advantages. Organizational learning involves developing and using new knowledge to improve organizational performance [70]. Organizational learning positively affects innovation. A firm possessing the ability to attain new knowledge and integrate extant knowledge with different methods will perform well in terms of product innovation and manufacturing processes. In other words, the better the organizational learning capability, the better the innovation capability [71]. Arora [72] contends that growth and learning in the new product development process enhances knowledge innovation and further improves new product performance. Organizational learning capability in the high-tech industry represents organizational capability. Low organizational capability in a firm may result in its rival obtaining an order for goods or its business partners

6 Y.-H. Hsu, W. Fang / Technological Forecasting & Social Change 76 (2009) canceling an order due to numerous alternative buyers. New product development lags in this scenario due to the firm's insufficient capability. Whether or not organizational learning capability adds more value to cooperative firms is the criterion for partners to judge their future cooperation in the competitive high-tech industry. A firm with stronger capability creates more value added, and thus more cooperative firms emerge, leading to its better performance. In summary, learning capability positively affects new product performance. An organization acquiring useful knowledge or techniques will achieve greater things and quickly reach its innovative activity goals. Previous studies stress that new product development is the result of organizational learning. This study offers the following hypothesis based on the above discussion. Hypothesis 4. Organizational learning capability positively affects new product development performance. 4. Method 4.1. Sample and data collection This study concentrates on new product development projects in the Taiwanese integrated circuit (IC) design industry. Selecting this single industry for the research subject helps eliminate interruptions from other variables (e.g., the industrial environment). Several factors led to choosing the IC design industry as the sample. First, the IC design industry is an intellectually capital-intensive industry, and either human capital, structural capital, or relational capital plays a very important role. Second, IC design companies face a turbulent environment and an extremely short life cycle and must continuously develop new products and are well known for good product innovation. Third, the IC design industry exhibits knowledge-intensive and high value added characteristics, and many researchers have used this industry as an important research sample in the knowledge management field [73,74]. Fourth, IC design companies not only emphasize external competitive capabilities, but also stress internal integration, knowledge application, and the process of knowledge internalization in the new product development process. Whether or not intellectual capital and organizational learning capability affect new product development performance remains an important issue in this industry. Finally, the output value of the Taiwan IC design industry accounts for 28.2% of the global total, and ranks second in the world [75] making it a valuable industry to research. This study uses a two-stage research design, including interviews and a research survey. The purpose of the interviews is to obtain compatible measures of intellectual capital and new product development performance in the IC design industry. The research survey tests the hypotheses in this study. The first stage interviews seven project managers in charge of new product development in IC design companies to generate measures of intellectual capital and new product development performance. Many authors have used this interview method to investigate innovative management research in large and small companies [76,77]. The second stage utilizes a survey method. The sample pool consists of 247 IC design companies listed in the 2005 Semiconductor Industry Yearbook published by the Department of Industrial Technology of the Ministry of Economic Affairs in Taiwan. A firm's most recently completed project must have introduced a product on the market for a certain period to be eligible. This limitation increases reporting accuracy on the resulting performance data. and telephone pre-surveys filtered the total down to 229 firms eligible to participate in this research, based on their accessibility. Product managers, project managers, and R&D engineers are the key informants in this study. Following Olson et al. [78], Song and Parry [79], and Bonner and Walker [80], this study relies on these managers to answer questionnaire items because they typically participate in new product development. Each informant received a personalized letter explaining the purpose of the study and a questionnaire by mail or . Two weeks later, non-respondents received a reminder and a second questionnaire. These efforts elicited 123 useable responses, creating a final useable response rate of 53.71%. An extrapolation procedure suggested by Armstrong and Overton [81] assessed the non-response bias. No significant differences exist between early (first quartile) and late (fourth quartile) respondents on the constructs used in this study Measures For consistency, all responses were measured using a Likert-type scale, with 1= strongly disagree, 4= neutral, and 7= strongly agree. The questionnaire comprised four parts. The first part contained measurements of intellectual capital (including human capital, structural capital, and relational capital); the second part measured organizational learning capability, and the third part measured new product development performance. The final part gathered company descriptive data (including total number of employees, R&D intensity, main product, etc.; Appendix A contains the details). Construct measurements are further described as follows Intellectual capital Interviews with the IC design companies helped generate items since no scales for this construct in the IC design industry are available. This study first collected measurements from previous literature, which provided a total of 125 items. Second, to avoid overlap between intellectual capital and new product development performance, this study only incorporated input factors (expenses and inputs) and excluded output factors, belonging to the new product development performance category. As a result, thirty-eight items formed the basis for interviews. Finally, this study generated fifteen items according to their interviewee-rated importance. Human capital measurement comprises the following four items: (1) employee empowerment is high in my company; (2) employees in my company have excellent professional skills; (3) the company provides well-designed training programs; (4) the employees of my company have unique and new ideas [24,21]. Structural capital measurement comprises the following seven items; the first three items belong to process capital, and the rest are part of innovation capital: (1) whether or not my

7 670 Y.-H. Hsu, W. Fang / Technological Forecasting & Social Change 76 (2009) company emphasizes IT investment; (2) whether or not my company is willing to invest in business development; (3) whether or not my company has an easily-accessible information system; (4) whether or not my company invests a high proportion of its money in R&D; (5) whether or not my company invests a high proportion of its money in patent maintenance; (6) whether or not my company emphasizes new market development investment; (7) whether or not my company has a high proportion of R&D employees [19,21,24,82]. Relational capital measurement comprises the following four items: (1) whether or not my company maintains long-term relationships with customers; (2) whether or not my company has many excellent suppliers; (3) whether or not the market that my company is in has the potential to grow; (4) whether or not my company has strong strategic alliances [22,21,23] Organizational learning capability Organizational learning capability has two dimensions: absorptive and transformative. Absorptive capability is a firm's absorptive and searching capability. This measurement was adapted from Cohen and Levinthal [26]: (1) whether or not my company has the capability to search for external information and knowledge; (2) whether or not my company can identify the usefulness of external information and knowledge; (3) whether or not my company can correctly predict the future development of core knowledge. Transformative capability is how a firm applies its knowledge, and this measurement was adapted from Kought and Zander [42,27]: (1) whether or not my company has the capability to integrate internally extant knowledge; (2) whether or not my company has the capability to apply knowledge to a specific problem or task; (3) whether or not my company has the capability to effectively categorize knowledge and wait to use it in the future; (4) whether or not my company can effectively and flexibly utilize extant or newly acquired knowledge to cope with a turbulent environment New product development performance Griffin and Page[83] note that no single measurement can evaluate new product success; therefore a multidimensional measurement is needed. This study collected sixty-five items from previous studies, and eliminated similar or same items, reducing the total to thirty-three items. These thirty-three items formed the basis of the interviews. Finally, this study generated fifteen items based on their interviewee-rated importance and classified them into four dimensions. Market performance items include (1) whether or not this new product meets revenue goals [84]; (2) whether or not this new product meets market share goals [84];(3)whetherornotthis project team achieves high market forecast accuracy [85]; (4) whether or not this new product contributes significantly to market leadership [86]. Financial performance items are (1) whether or not the overall profitability of this new product is high [83]; (2) whether or not the overall profitability of this new product is higher than that of my firm's other new products [79];(3)whetherornot this new product generates a high investment return [84]. Customer performance items include (1) whether or not customers are satisfied with the new product [84]; (2) whether or not this new product generates high customer acceptance [87];(3) whether or not many customers buy this new product [83]. Finally, product performance items comprise (1) whether or not the overall quality of our new products is higher than that of our competitors [88]; (2) whether or not this new product contributes significantly to technology leadership [86]; (3) whether or not this new product provides competitive advantages [83]; (4) whether or not this new product is highly innovative [83]; (5) whether or not this new product launched on time [84]. 5. Analyses 5.1. PLS method This study uses the Partial Least Squares (PLS) method to test its hypotheses. Based on the Ordinary Least Squares (OLS) algorithm, this technique analyzes empirical data with insufficient supporting theories and little available information [89,90]. Compared to the Linear Structural Relations (LISREL) model, PLS is less rigid in variable normality and randomness [91], and more flexible in sample size when estimating path coefficients [92]. When a construct incorporates many similar measurement variables in regression models, multicollinearity is always a problem. Conversely, PLS deals with measurement errors, so multicollinearity is not a problem. This study regards human capital as the formative indicator and structural capital, relational capital, organizational learning capability, and new product development performance as reflective indicators Reliability, validity, and descriptive statistics Table 1 lists descriptive statistics including means, standard deviation, correlations and Cronbach's α value. Table 1 clearly shows that new product development performance has significantly positive correlations with human capital, structural capital, relational capital, and organizational learning capability. The Cronbach's α for human capital, structural capital, relational capital, organizational learning capability, and new product development performance is 0.77, 0.87, 0.86, 0.94, and 0.97, respectively. All values exceed the 0.7 criteria, indicating that the measurement has good reliability [93]. This study generated the measures from important academic journals, and interviewed experts and managers in the IC design industry. These interview results formed the basis for item modification, showing content validity for the measures in this study. Table 1 also shows that the correlation of paired constructs is significantly less than 1, implying the discriminate validity suggested by Bagozzi et al. [94] and Segars and Grover [95]. Table 2 shows the measurement analysis results, including weights, factor loadings, composite reliabilities, and average variances extracted. The composite reliabilities were calculated using procedures suggested by Fornell and Larcker [96], and the composite reliabilities for the five constructs range from 0.87 to All values exceed 0.7, the acceptable level suggested by

8 Y.-H. Hsu, W. Fang / Technological Forecasting & Social Change 76 (2009) Table 1 Mean, SD, correlations and Cronbach's α of the constructs. Constructs Mean SD HC SC RC OLC Cronbach's α Human capital Structural capital (0.08) Relational capital (0.08) (0.08) Organizational learning capability (0.07) (0.07) (0.09) New product development performance (0.07) (0.05) (0.08) (0.09) Note: All correlations are significant at pb0.01. Numbers in parentheses represent standard errors. All differences from unity were statistically significant at pb0.05. Table 2 Weight, factor loading, composite reliability and average variance extracted for constructs. Constructs Weight Factor loading Composite reliability Average variances extracted Human capital Structural capital Relational capital Organizational learning capability New product development performance Bagozzi and Yi [97]. Average variance values extracted from each construct (ranging from 0.57 to 0.73) also exceed the threshold level (0.5) suggested by Bagozzi et al. [94]. All item loadings ranging from 0.64 to 0.90 are significant at the five-percent significance level, indicating convergent validity [94]. The PLS method also creates weights and loadings in the measurement model, where weights explain the formative indicators and loadings explain the reflective indicators [98]. We used this measurement model to assess individual item reliability. All items exceed the acceptable level of [99], except for one structural capital item. However, many researchers argue that this rule should not be so rigid [100,98], and the item was kept. The weight of the first human capital item in this study was lower than the accepted threshold, and was excluded. Since the data was collected from the same informants and the constructs use subjective measures, a possibility of common method bias (CMB) exists. This study used one of the more commonly used methods to check for CMB, as reviewed by Podsakoff et al. [101]. The model was estimated again with a single-method first-order factor added to the construct indicators. Results reveal that CMB is not a significant issue in the study Results This study tested the mediated effects of organizational learning capability between three types of intellectual capital i.e. human capital, structural capital, and relational capital and new product development performance. Table 3 summarizes the structural model produced by PLS analysis. This table charts the explained variance by model (R 2 ), the standardized path Table 3 Effects and variance explained for all endogenous variables. Effects on endogenous variable Path coefficients (β) Variance explained (R 2 ) t value (bootstrap) Non-mediated model Mediated model Non-mediated model Mediated model Effect on relational capital H1a: Human capital 0.75 (17.62) 0.75 (18.98) Effect on structural capital H1b: Human capital 0.41 (5.32) 0.41 (5.52) H1c: Relational capital 0.47 (5.69) 0.47 (5.69) Effect on organizational learning capability 0.59 H2a: Human capital 0.25 (2.62) H2b: Structural capital 0.14 (1.61) H2c: Relational capital 0.43 (4.53) Effect on new product development performance H3a: Human capital 0.47 (3.93) 0.35 (3.76) H3b: Structural capital 0.09 ( 0.96) 0.16 ( 1.53) H3c: Relational capital 0.30 (2.43) 0.11 (1.25) H4: Organizational learning capability 0.46 (4.63) pb0.01; pb0.05; pb0.1 (two-tailed).

9 672 Y.-H. Hsu, W. Fang / Technological Forecasting & Social Change 76 (2009) Table 4 Paired t test between human capital, structural capital, and relational capital. Constructs Mean A B B C C A Mean of human capital (A) ( 5.182) ( 1.677) (6.229) Mean of structural capital (B) Mean of relational capital (C) Note: Number in the sample n=123. The number in the brackets is the t value. pb0.01; pb0.05; pb0.1. Table 5 Difference analysis between large enterprises and SMEs. Constructs Mean of large enterprises (D) Mean of SMEs (E) D E Human capital (0.5367) Structural capital (0.6126) Relational capital (1.3727) Note: Number in the sample n=123, including 35 large enterprises and 88 SMEs. The number in the bracket is the t value. pb0.01; pb0.05; pb0.1. coefficients (β), and t values observed with the level of significance achieved using the bootstrap approach. The table also lists results with and without the mediating effect to discuss the mediating effect of organizational learning capability on the relationship between intellectual capital and new product development performance. Results support hypotheses H1a H1c for the intellectual capital variables in this model. These results show how human capital links to relational capital (β=0.75,pb0.01) and structural capital (β=0.41,pb0.01), and how relational capital links to structural capital (β=0.47, pb0.01). With respect to organizational learning capability, the results support H2a and H2c, establishing the influence of human capital (β=0.25,pb0.01) and relational capital (β=0.43,pb0.01). Results on the other hand, marginally support H2b (β=0.14, pb0.1), establishing a link between structural capital and organizational learning capability, although the standardized regression coefficient shows a low significance. Human capital (β=0.35,pb0.01) and organizational learning capability (β=0.46, pb0.01) affect new product development performance most significantly, supporting H3a and H4. Structural capital shows a marginally negative relationship with new product development performance, which does not support H3b (β= 0.16, pb0.1). Relational capital effect is not significant, indicating no support for H3c. The study subsequently applied a paired t test to analyze possible differences between the three types of intellectual capital human capital, structural capital, and relational capital in Taiwan IC design companies. Table 4 shows obvious differences between human capital and structural capital, and human capital and relational capital; however, marginal difference exists between structural capital and relational capital. Results show relational capital as the most significant among these three types of intellectual capital in Taiwanese IC design companies, structural capital is next, and human capital is the least significant. This finding indicates two things: first, IC design companies in Taiwan stress interactive relationships with their stakeholders, customers, and suppliers; second, Table 3 shows that human capital has a significantly positive relationship with new product development performance. Nevertheless, human capital is the least significant among the three types of intellectual capital. Therefore, it is imperative for Taiwanese IC design companies to create and develop their human capital to raise new product development performance. According to Standards for Identifying Small and Medium-sized Enterprises by the Ministry of Economic Affairs of Taiwan, this study defined a Medium & Small Enterprise (SME) as one where the number of regular employees in the manufacturing industry does not exceed 200 persons (whereas the number of regular employees of a large enterprise exceeds 200 persons). A total of 123 samples were used in the study, including eighty-eight SMEs and thirty-five large enterprises. This study compares the three types of intellectual capital of Taiwan's large enterprises with those of Taiwan's small and medium enterprises (SMEs). Table 5 shows that relational capital in Taiwan's SMEs is marginally significantly less than that of large enterprises in Taiwan's IC design companies. Relational capital positively influences new product development performance via organizational learning capability, as mentioned in Table 3. Therefore, it is imperative for SMEs in Taiwan to develop their relational capital to increase new product development performance. 6. Discussion 6.1. Intellectual capital The empirical results indicate that human capital significantly affects relational capital. The model explains 56% of the data variance, showing that a firm's human capital investment promotes relational capital growth [35]. Both human capital and relational capital significantly affect structural capital. Both variables account for 69% of structural capital variance, indicating that a firm's investment in human capital and relational capital significantly affects its accumulation of structural capital, including process capital and innovation capital [22]. Relational capital, however, has a more obvious effect on structural capital than human capital.

10 Y.-H. Hsu, W. Fang / Technological Forecasting & Social Change 76 (2009) Intellectual capital and organizational learning capability Human capital, structural capital, and relational capital significantly affect organizational learning capability. These three variables explain 59% of variance, showing that the model provides a powerful explanation. The finding that human capital significantly affects organizational learning capability is consistent with Baker's study [40]. Cohen and Levinthal [26] also note that numerous and good quality employees help an organization generate internal knowledge and absorb external technological knowledge. This, in turn, leads to improved absorptive capability. As a result a firm investing more in human capital will gain better quality employees, helping to enhance its organizational learning capability. Structural capital positively affects organizational learning capability, although the result shows a low significance. Empirical data shows that investment in process and innovation capital improves organizational learning capability. Persaud et al. [43] point out that information technology investment encourages an organization to acquire external knowledge. Davenport and Prusak [102] also note that information technology investment helps an organization expand its scope of knowledge and increase knowledge transfer speed. Information technology in the organizational learning process, therefore speeds up and stimulates learning. A firm with a patent or research and development focus also accumulates organizational knowledge and learning capability. Relational capital positively affects organizational learning capability. Persaud et al. [43] note that knowledge creation and diffusion correlates with knowledge media. Customer and partner relationships improve when a firm invests in more relational capital, leading to better interchange and knowledge sharing [47,48]. Relational capital thus positively affects organizational learning capability. Smith [103] argues that an organization could spillover, accumulate, and diffuse its knowledge via information interaction and exchange with other organizations. Relational capital, however, affects organizational learning capability more than human capital and structural capital (β=0.43nβ=0.25nβ=0.14). An organization should build good relationships with its stakeholders, customers, and suppliers to improve organizational learning capability. New product benefits come more from relational capital than from investing in human or structural capital Intellectual capital, organizational learning capability, and new product development performance Human capital and organizational learning capability significantly affect new product development performance. This model explains 50% of new product development performance, indicating good explaining power. Human capital positively affects new product development performance (β=0.35, pb0.01), so managers should invest more in human capital to achieve better performance. This result is consistent with Strober [54] and Becker [104]. Structural capital negatively affects new product development performance (β= 0.16, pb0.1) at a low significance level. This negative relationship may be due to firms spending too much on information system or R&D expenditures, in turn undermining their financial performance and ignoring their customers, leading to reduced new product development performance. A firm investing in process or innovation capital should accurately evaluate profit and loss to avoid investing too much without a good investment return. Organizational learning capability positively affects new product development performance (β=0.46, pb0.01). Hargadon and Sutton [105] emphasize that a firm may use inter-organizational knowledge searching, sharing, transferal, and integration to improve its new product development performance. A firm developing new products cannot ignore organizational learning capability accumulation. That is, a firm with better capability to acquire and integrate knowledge will perform better in the product or development process, thus achieving better new product development performance Mediating effect of organizational learning capability Comparing the model without mediating effect to the model with mediating effect, the direct effect of human capital on new product development performance decreases when incorporating organizational learning capability mediating effect (β=0.47 decreases to β=0.35). Moreover, human capital effect on organizational learning capability and organizational learning capability effect on new product development performance are significant. This result shows that human capital affects new product development performance through organizational learning capability. Comparing with- and without-mediating effect models, the direct effect of relational capital on new product development performance becomes insignificant when incorporating the mediating effect of organizational learning capability. Moreover, relational capital effect on organizational learning capability and organizational learning capability effect on new product development performance are significant. This result shows that relational capital affects new product development performance through organizational learning capability with obvious effect. Finally, comparing with- and without-mediating effect models, the power of explaining variance increases from 42% to 50%. This result implies that the model in this study has good predictive power. Furthermore, the endogenous variables achieved R 2 values never less than Conclusions Intellectual capital has recently attracted much attention from scholars, enterprises, and governments in practical applications and further discussion. However, no relevant studies explore the mediating role of organizational learning capability in the relationship between intellectual capital and new product development performance relationship. This study therefore focuses on and fills this research gap. The study explores the influence of three types of intellectual capital, i.e. human capital, structural capital, and relational capital, on new product development performance, and regards organizational learning capability as a