A Technology Acceptance Model: Mediate and Moderate Effect

Transcription

1 Asia Pacific Management Review 14(4) (2009) A Technology Acceptance Model: Mediate and Moderate Effect Wen-Chin Wei Department of Public Finance and Taxation, National Kaohsiung University of Applied Sciences, Taiwan Accepted 24 December 2008 Abstract Based on a Technology Acceptance Model, this study investigates and increases the understanding of how perceived technological attributes (perceived technological implicitness and compatibility) and internal beliefs (perceived usefulness and perceived ease of use) will influence the behavioral intention of accepting licensing in new technology. This study adopts the Structural Equation Modeling approach. Based on the 244 correspondents from Taiwan s four high technology industries, 63% of the variance of licensees behavioral intention can be accounted for by perceived usefulness, perceived technological implicitness, and compatibility in one theoretical mediate model. However, the evidence from Taiwan s 95 biotechnology companies indicates that 64% of the variance of behavioral intention can be explained by an alternative theoretical mediate and moderate model that consists of perceived technological compatibility, perceived usefulness and perceived ease of use. Keywords: technological attributes, inward belief, behavior intention, structural equation modeling 1. Introduction 1 Recent decades have seen an increase in competition in the technology market, especially in technology-intensive industries such as bioengineering, computers, and electronics (Di Benedetto et al., 2003). Inward licensing-in technologies that upgrade a firm s innovative capability have become the prevailing phenomenon, accounting for 20-33% of strategic alliances across industries (Young, 2005). Through licensing-in technologies, firms can enter new markets, serve new customers, and compete with licensors (Mitchell and Singh, 1992). Prior evidence (Howard, 2001; Young, 2005) has shown the important advantages of licensing-in advanced foreign technologies in upgrading a firm s innovative capability. The Technology Acceptance Model (TAM) is widely used in technology acceptance behavior. However, the critical external variables that are associated with internal beliefs, and that will influence a firm s behavioral intention (BI) of accepting licensing-in technologies is a field that has received little attention. This study, based on TAM, samples Taiwan s high technology industries through Structural Equation Modeling (SEM) under both mediate and mixed (mediate and moderate) models to investigate the relationships among perceived technological attributes, internal beliefs, and the BI of licensing-in advanced foreign technologies. It also explores whether the result will vary across industries and concludes with some implications for licensors and licensees (Malhotra et al., 2005). This paper is organized in the following sequence. Section 1 is the introduction. Section 2 presents the theoretical model and hypotheses. Section 3 describes the research method. Section 4 presents the empirical results of the mediate model. Section 5 presents the mediate and moderate model. * Corresponding author. wwc@cc.kuas.edu.tw 461

2 Section 6 discusses the theoretical contributions and management implications. Section 7 offers a conclusion and directions for future research. 2. Theoretical model and hypotheses Davis et al. (1989) developed the technology acceptance model (TAM) to specify the relationships between two key beliefs, related attitude and behavior: perceived usefulness (U) and perceived ease of use (EOU), and users attitudes (A), behavioral intention (BI) and actual computer adoption behavior. Davis et al. (1989) used TAM to establish a more parsimonious model (Hybrid Intention Model, HIM), in terms of U, EOU, to explain and predict the intention to use one system. In HIM, A (an individual s feeling about the target behaviour) was verified not to mediate between the belief and intention; thus, A was excluded. Di Benedetto et al. (2003) insist that after being suitably modified, TAM may be a behavioral model to explain and predict the adoption behavior of technology by firms that are located in developing countries. Since this study is based on TAM and HIM, these models will be illustrated in Sections 2.1 and 2.2 respectively. 2.1 Technology acceptance model (TAM) TAM was modified to explain and predict user acceptance of information systems (Davis et al., 1989). Di Benedetto et al. (2003) insist that as a behavioral model, TAM can describe and explain the adoption of foreign-developed product technology by firms in developing countries. TAM postulates that beliefs U and EOU are the main determinants of computer acceptance behaviors. U is the prospective user s subjective probability of using an application system that will improve his or her job performance. EOU describes the degree to which an innovation is perceived as difficult to use (Moore and Benbasat, 1991; Venkatesh et al., 2003). U and EOU are postulated as statistically distinct constructs (Hauser and Shugan, 1980; Larcker and Lessig, 1980). The U-A link is as follows: when one person holds a higher subjective probability that perceives that the adoption of one technology will improve his or her job performance, then this subjective belief associated with the evaluation response will lead to the form of A, as shown in Figure 1. usefulness (U) External variables Attitude toward using (A) Behavioral intention to use (BI) Actual system use ease of use (EOU) Figure 1. Technology acceptance model. TAM postulates that BI is determined jointly by the person s attitude about using the system (A) and its perceived usefulness (U). In addition, TAM s U and EOU are postulated as one priori, and are referred to as the general determinants of user acceptance. TAM postulates U and EOU as two distinct constructs, thus enabling the comparison of the relative importance of each belief in determining A becoming possible (Davis et al., 1989). 462

3 2.2 Hybrid intention model Davis et al. (1989) uses factor analysis to anatomize the structure of the beliefs that underpin user acceptance. They integrate the beliefs of TAM into the Hybrid Intention Model (HIM), which yields a better perspective on the determinants of BI. They show that U had a direct influence on BI in both periods (time 1, one hour introduction of the Writeone software, and time 2, 14 weeks later), and EOU had a direct effect on BI at time 1 and an indirect effect via U at time 2. This verifies that A does not intervene between beliefs and intentions. Davis et al. s HIM is a more parsimonious, causal structure that is powerful in predicting and explaining user behavior based on three theoretical constructs: BI, U, and EOU. Therefore, HIM is accepted as the theoretical basis of this study. BI is used as a surrogate for usage for two reasons: first, TAM postulates BI rather than the actual use as the main determinant of usage behavior; and second, usage (measured by the self-report scale) has been criticized for oversubjectivity (Thompson et al., 1994). Hence, this research based on HIM contemporaneously considers important external perceived technological attributes to investigate the relationships of internal beliefs (U and EOU) and the BI to accept one new licensing-in technology (including knowhow) (Horng, 2006). The technology refers to the related knowledge, during the process of converting scientific knowledge and theory into market products, associated with research and development inputs (Zhao, 1992). These links (U-BI, EOU-U, EOU-BI), and perceived technological attributes (U-EOU) will be explained in Sections 2.3 to 2.6. One theoretical model is then developed, given in Figure U-BI usefulness (U) is the prospective user s subjective probability that using a specific application system will improve his or her job performance in an organizational setting (Davis et al., 1989). The A-BI link in TAM signifies that, all things being equal, people form intentions to perform behaviors that will benefit them. The U-BI link indicates that, within organizational settings, people form intentions toward behaviors they believe will improve their job performance. This is because enhanced performance is instrumental to attaining rewards, such as pay increases and promotions (Vroom, 1964). The cognitive rule of such means-end behaviors does not need to reevaluate how better performance will serve one s purpose. Hence, the U-BI link in TAM postulates that people form intentions toward using computer systems based on a cognitive evaluation of how they will improve their performance. The direct effect of U on BI has been certified via theoretical perspective and empirical evidence (Bagozzi, 1982). In TAM, U has a significant effect in different periods and presents an increased influence on BI over A (Davis et al., 1989). Szajna (1996) insists that unless one user first perceives one system as useful, its EOU has no influence on the formation of BI. Thus, for performance purposes, organizational members usually emphasize the influence of U over A on BI. For parsimonious straightforward purpose, the link of U-BI is adopted instead of a U-A-BI link in this study. From a technology licensing-in perspective, the individual in an organization forms certain behaviors for his or her job performance, such as earning promotions and raises (Vroom, 1964). Therefore, when one prospective licensee considers licensing-in one new technology that will bring individual positive job performance, that U will motivate the individual s belief and increase the licensee s BI to accept that technology. Thus, H1 hypothesis is confirmed. H1: When one firm uses licensing-in one technology, U will have a positive influence on licensee s BI. 2.4 EOU-U 463

4 EOU is the degree to which the use of an innovation is perceived as difficult (Moore and Benbasat, 1991). At the beginning, when EOU influences the individual because of selfefficacy, one will appraise how successful he or she will become after using the new system. Self-efficacy refers to the user s feeling (Bandura, 1982) and personal control (Lepper, 1985); it is the individual s ability to use the system in question. Bandura (1982; 1986) insisted that the self-efficacy effect will dominate individual reasoning, feelings, and motive, further influencing individual behavioral intention. Over time, this concern will diminish as the individual learns to operate the new system. Then, EOU turns into one more instrumental concern about how adopting the new system will generate improvement in performance. In TAM, when one organization improves EOU, it will lead to performance benefits, since the effort saved by the increased EOU via redeployment will enable one individual to produce more work with the same effort (Davis et al., 1989). Thus, increased EOU will improve performance, and thus EOU will have a positive impact on U. When one organization considers licensing-in one system or technology to be easier to use, an individual might believe that he or she possesses a higher self-efficacy (Bandura, 1982) and has greater personal control over that system or technology (Lepper, 1985). If the organization makes an effort to increase EOU, then the U of the licensing-in technology can be expected to increase via the redeployment of inputs saved via the increased EOU (Wei, 2008). Thus, H 2 is theorized as follows. H2: When one firm adopts licensing-in one technology, increased EOU will make a positive influence on licensee s U. 2.5 EOU-BI At the start, EOU influences the individual via the effect of self-efficacy: how successful the person will be after using the new system. Self-efficacy is the user s feeling about (Bandura, 1982) and personal control over (Lepper, 1985) his or her ability to operate the system in question. Later, EOU s effect is indirectly via U, due to learning, and the A-BI link becomes nonsignificant. When one foreign-developed product technology is perceived to be easy to adopt and transfer, then related costs and time of implementation will be perceived to be less, thus stimulating the customer s intention to adopt the technology. In addition, licensees are concerned about their ability to learn and use a new technology. EOU tends to be more firm-specific, and thus has a direct effect on a licensee s BI of adopting a new technology (Di Benedetto et al., 2003). Davis et al. (1989) has argued that when dealing with more complex and difficult systems, EOU will have more significant impacts on intentions. Thus, the following EOU-BI relation is theorized to exist. H3: When one firm considers licensing-in one technology, EOU causes a positive influence for licensee s BI to accept that technology. 2.6 t technology attributes, U and EOU Davis et al. (1989) insisted that although EOU will have a direct impact on U, U will also be affected by external variables over and above EOU. For example, if two systems have the same EOU, but one provides better accuracy, higher quality or market value, then that system is perceived as more useful. Hence, the design characteristics of one system can have both a direct effect on U and indirect effects via EOU. From system characteristics, such as being more explicit, if the system provides a touchable screen, detailed manual and figure documentation, then it makes a positive contribution to EOU and reduces the related adaptive costs. Thus, related resources can be saved to provide more products and benefits. Alternatively, perceived technological implicitness involves cognitive and technical elements (Nonaka, 1994). For instance, diagnosing why customers are dissatisfied with the tone of their voice and facial expressions involves the cognitive elements of implicit knowledge. In 464

5 addition, the ability to ride a bike requires the procedural skills of pedaling and maintaining balance, which are the technical components of implicit knowledge (Linderman et al., 2004). On the one hand, Linderman et al. (2004) argued that cognitive elements involve mental models, such as schemata, paradigms, and beliefs. On the other hand, the technical component of implicit technology involves skills and crafts for specific contexts. Thus, when one new technology involves more cognitive and technical elements, it will have negative impacts on perceived ease of use. When one firm licenses-in more implicit technologies, from a resourcebased school s perspective, that technology contains more expertise, uniqueness, ambiguous causality, and complexity that are hard to imitate, the new technology will provide higher market value (Barney, 1991; Grant, 1996). Thus, perceived technological implicitness causes organizational staffs to expect new technology that will bring a positive U (Wei, 2008). The causal relationship between perceived technology implicitness-eou and perceived implicitness-u is theorized as follows. H4: When one firm licenses-in one technology, perceived technological implicitness causes a negative contribution to EOU. H5: When one firm licenses-in one technology, perceived technological implicitness bring about a positive influence for U. Rogers (1962) defined compatibility as the extent to which the innovation fits a customer s experiences or activities. Premkumar et al. (1994) insisted that perceived technological compatibility is the cognitive consistency degree between innovative technologies and existing value, past experiences, or demand. If the gap between new and existing technologies narrows when firms use licensing-in technologies, then the function of the new technology is more easily expanded (Rogers and Shoemaker, 1971). Alternatively, if the degree of compatibility increases, the effort the licensee needs to put into new technology decreases. For instance, physicians are likely to consider a technology easy to use when it does not require them to make major changes in their practices. Significant incompatibility necessitates major process and service changes that require considerable learning and unlearning on the part of physicians who are likely to perceive the technology to be difficult to use (Chau and Hu, 2001). Hu et al. (2003) studied the hardware and software aspects of compatibility and insisted that there was a positive effect of compatibility on perceived ease of use. The proposed effects of compatibility on perceived ease of use (EOU) have been identified in prior studies (Al-Gahtani and King, 1999; Chau and Hu, 2001; Hu et al., 2003; Wei, 2008). Thus, it is theorized that when the compatibility of licensing-in technology becomes higher, EOU degree of the new technology increases (Robertson, 1971; Rogers and Shoemaker, 1971). H6: When one firm licenses-in one new technology, perceived compatibility causes a positive contribution on EOU. When the foreign-developed technology is perceived as more compatible with a process, then the technology transfer cost will be less (Di Benedetto et al., 2003). Moreover, when compatibility increases, the new technology will be more rapidly diffused, thus facilitating the adoption of an innovative product by the users, and generating competitive benefits for the firm (Rogers, 1962). The improved technological compatibility will increase the use of the technology and the achievement of greater technology (e.g. increases in product quality) and economic benefits (e.g. increased competiveness). Rogers and Shoemaker (1971), from a function perspective, insist that when the compatibility increases, the function of the new technology is easier to expand. Di Benedetto et al. (2003) argue that compatibility tends to be more firm-specific and more likely to stimulate overseas managers BI to adopt a new technology. Thus, the new technology raises the expectations for the licensee, and increases his or her BI to accept that technology. Thus, the study theorizes the following hypothesis. 465

6 H7: When firms use licensing-in for one technology, perceived compatibility causes a positive contribution to the BI of the acceptance of a new technology. This study is based on the Hybrid Intention Model that is associated with empirical evidence, logical reasoning, and considering the parsimonious purpose and integrating related hypotheses, thus the study theorizes the theoretical mediate model (TBIM) as shown in Figure 2. implicitness (ξ 1 ) H5 (+) usefulness (U, η 2 ) H1 (+) H4 (-) H2 (+) BI (η 1 ) compatibility (ξ 2 ) H7 (+) H6 (+) ease of use (EOU,η 3 ) H3 (+) Figure 2. technological attributes and inward beliefs that explore licensee s behavior intention- initial theoretical mediate model (TBIM). 3. Research method 3.1 Data analysis method In the mediate model TBIM, U is a latent dependent variable in H2 and H5; and EOU is a latent dependent variable in H4 and H6, respectively; however, in H1 and H2, both U and EOU become independent variables. In such a recursive and nested model, Structural Equation Modeling is appropriate for systematically investigating the relationships between latent and observable variables. Anderson and Gerbing s (1988) two-step SEM procedure is adopted via the CALIS procedure of SAS software. In this study, Maximum Likelihood (ML), Weight Least Square (WLS), and General Least Square (GLS) are conducted simultaneously to examine the consistency under different estimation procedures. In addition, this study uses AMOS 6.0 software to conduct one alternative mediate and moderate model (mixed model) analysis illustrated in Section The measurement scale of observable variables A seven-point Likert scale is used to answer questions (from strongly likely = 7 to strongly unlikely = 1 ); summated scales consist of statements that express either a favorable or an unfavorable attitude toward the object in question. 3.3 Measures of the BI of the acceptance of one licensing-in new technology The measurement of BI includes potential licensees of technologies. This study adopts Atuahene-Gima s (1993) following three items: (a) a company might need to license-in new technologies in two years; (b) a company might use the licensing-in technologies to enter a new market; (c) a company might use the licensing-in technologies to expand its market. 3.4 Measures of U 466

7 The measure scales of U are adopted from Davis et al. (1989): (a) using the new technology would improve individual performance in related jobs; (b) using the new technology would enable the individual to complete his or her job more easily and thus enhance his or her productivity. 3.5 Measures of EOU EOU can be measured by the effort to which the adoption of licensing-in technology is perceived to be required. This effort includes the difficulty of technology transfer, time for transfer, the problem of the new technology s application into production facilities, and time required to learn to use the licensing-in technology. Hence, this study adopts the scale of Berg, Duncan, and Friedman (1982) to measure EOU: (a) the difficulty one may encounter when licensing-in technologies from foreign countries. (b) individuals will need to spend time learning the licensing-in technology. 3.6 Measures of perceived implicitness The two constructs, touch characteristic and non-documentation, have been used to measure perceived technology implicitness and obtain their reliability and validation from prior studies. This study adopts the items suggested by Li et al. (2002) to measure perceived technology implicitness: (a) both licensor and licensee must transfer new technologies, through face-to-face communication or learning-by-doing model; (b) technologies can be transferred only via documentation or words. However, Question 2 is a reverse item; therefore, it required the form 8 minus the value of old version of item 2 to recode it (Hatcher, 2004). After this step, the average of both is considered the value of technology implicitness. 3.7 Measures of perceived compatibility The following three constructs have been used to measure perceived compatibility: equipment compatibility, material compatibility and operation habits compatibility. This study adopts Berg et al. s (1982) three question items to measure perceived compatibility: (a) the compatibility of licensing-in technologies with existing equipment; (b) the compatibility of licensing-in technologies with a company s materials; (c) the compatibility of licensing-in technologies with operational habits. 4. Empirical results of mediate model (TBIM) Di Benedetto et al. (2003) insisted that developing countries that are facing competition can produce globally competitive products via rapid technology transfer from developed countries. Taiwan s electronic, mechanical, computer, and biotechnology industries are selected as the research subject. Based on the directories of these industries, 1,100 questionnaires were mailed to technology licensing departments, with follow-up telephone reminders to encourage return of the surveys. After excluding 42 unavailable questionnaires, 244 usable questionnaires (one from each company) were collected; this yielded an effective response rate of 22%. Of these 244 respondents, 32% were CEOs, 19% were R&D managers, 8% were marketing managers, 5% were manufacturing managers, and 36% were administrative staff. Biotechnology companies comprise the largest percentage (57%), with electrical and electronic companies second (20%), and mechanical companies third (19%). 4.1 The overall fitness of the theoretical model The fitness indexes under ML estimate show that GFI, AGFI, CFI, NFI all exceed 0.90, and that RMSEA is However, the fitness of GLS, except for GFI exceeds Other fitness indexes are below the acceptable level of 0.90 and the RMSEA is.08, above the acceptable level of Under the Weighted Least Square (WLS), the GFI, AGFI, CFI and 467

8 NFI are all below 0.90 and its RMSEA is These fitness indices indicate the result of Maximum Likelihood (ML) present the best overall fitness, so the ML estimation is adopted to analyze the theoretical mediate model. Table 1 shows the overall fitness indexes of ML, GLS and WLS respectively. Table 1. The fitness indexes of TBIM from ML, GLS and WLS estimations. Fitness GFI AGFI CFI NFI RMSEA estimation ML GLS WLS Reliability and validity Related internal fitness indices were calculated, with ML estimation, for the measurement model. The indicator reliability is defined as the square of the correlation between a latent construct; that indicator represents the percentage of variation in the indicator explained by the factor it is supposed to measure (Long, 1983). The highest and lowest indicators reliability for the BI are 0.86 (personal adeptness, y 7 ) and 0.60 (lead-in ability, y 6; non-documentation, x 2 ), respectively. However, the highest and lowest composite reliabilities are 0.93 (perceived technology compatibility, ξ2) and 0.80 (perceived technological implicitness, ξ 1 ) respectively, all of which exceed the acceptable level of Composite reliability is analogous to coefficient alpha, and reflects the internal consistency of the indicators measuring a given factor. Both indicator reliability and composite reliability reach the minimally acceptable level 0.60 suggested by Fornell and Larcker (1981).The highest and lowest variance extracted estimates are 0.86 (U, η 2 ) and 0.66 (perceived technological implicitness, ξ 1 ), respectively, which exceed the 0.50 acceptable level. In addition, all standardized factor loadings of the indicators are significant at p < 0.001; hence, the convergent validity is supported. From the analysis of the exogenous variables of the measure model in which all factors are allowed to co-vary, the highest correlation between BI (η 1 ) and perceived usefulness (U, η 2 ) is 0.74, and its confidence interval of two standard errors is between 0.66 and This interval does not include the value of 1.0, meaning it is very unlikely that actual population correlation between BI and U is 1.0. This finding supports the discriminant validity of the measures (Anderson and Gerbing, 1988). 4.3 The fitness and parsimony indices of TBIM According to the stepwise multivariate Wald test, the path from η 3 to η 2 (Pη 2 η 3, H 2 ), the path from ξ 2 to η 3 (Pη 3 ξ 2, H 6 ), and the path from η 3 to η 1 (Pη 1 η 3, H 3 ) could be eliminated. After the chi-square difference test finding, the three paths can be eliminated without hurting the fitness of the model. When these three paths are dropped in sequence, the modified model M r1, M r2, and M r3 is obtained. Next, the Lagrange multiplier test indicates that Pη 1 ξ 1 (H new ) can be added and improves the fitness; this new model is written as M r4 (Figure 3). The chi-square difference test shows that M r4 (with a new path Pη 1 ξ 1 ) provides a fitness that is significantly superior to that of M r3. Although the parsimony ratio (PR) of 0.73 for M r4 is lower than the PR of 0.74 for M r3, the slight decrease in parsimony is offset by an increase in fit. This improvement is reflected in the Non-normed-fit index (NNFI) of 1.00 and in the relative normed-fit index (RNFI) of 0.95 for M r4 (in the structural model). Therefore, this study accepts M r4 as the final model for 468

9 TBIM. 4.4 The path coefficients of TBIM Table 2 shows evidence of the final model for TBIM (M r4 ) under three estimation methods (ML, WLS, and GLS). Almost all path coefficients are significant at p < The empirical evidence indicates that H 1, H 4, H 5, H 7, H new are significantly supported. This new path (Pξ 1 η 1 ) is also defended from the resource-based theory that technologies with higher implicit knowledge will contain more knowhow, provide higher market value, and be harder to imitate, hence making a positive contribution to BI. The final model of TBIM (M r4 ) is depicted in Table 2 and Figure 3. Further analysis and implications will be presented in Section 6. An alternative mediate and moderate mix model is developed and investigated in Section 5. implicitness (ξ 1 ) H5 (+) H new (+) usefulness (U, η 2 ) H1 (+) H4 (-) H2 (+) BI (η 1 ) compatibility (ξ 2 ) H7 (+) ease of use (EOU,η 3 ) Figure 3. technological attributes and inward beliefs that explore licensee s behavior intention- final theoretical model (TBIM, M r4 ). Note: --- represent the insignificant hypotheses, and indicates the significant hypotheses. Table 2. Standardized path coefficients of TBIM. Dependent variable/ Independent variable M BI:ML M BI:WLS M BI:GLS Hypotheses tested η 1 (BI) η 2 (U) 0.50 *** 0.38 *** 0.50 *** H 1 (supported) ξ 1 (new) 0.29 *** 0.23 ** 0.29 *** H new (supported) ξ *** 0.20 ** 0.20 ** H 7 (supported) η H 3 (omitted) η 2 (U) ξ *** *** H 5 (supported) η H 2 (omitted) η 3 (EOU) ξ *** ** *** H 4 (supported) ξ H 6 (omitted) Notes: N = 244, * p <0.05; ** p < 0.01, *** p < η 1 : BI, η 2 : U; η 3 : EOU, ξ 1 : perceived technological implicitness, ξ 2 : perceived technological compatibility; M BI:ML, M BI:WLS, M BI:GLS represents the evidences of TBIM via ML, WLS and GLS estimation respectively. 469

10 5. Mediate and moderate model Thompson et al. (1994) insist that when the perceived technological implicitness of one system increases, it will be perceived as more difficult to understand and use. Thus, perceived implicitness forms one dimension of the complexity of one system, and becomes one of the measure items of effort expectancy (Venkatesh et al., 2003). Effort expectancy is the ease that is associated with the use of the system. When technological implicitness increases, the technology will become more difficult to understand and use; thus, it can be incorporated into perceived ease of use. Therefore, technological implicitness is integrated into perceived ease of use and other relationships (H 1, H 2, H 3, H 6, H 7 ) are maintained the same as the mediate model (TBIM). Moreover, when the compatibility of new technology with existing value, past experience, or demand increase, then the efforts to narrow the gap can be saved via reallocation to decrease cost. The reduced cost represents the increase in user job-related performance. Hu, Clark and Ma (2003) argued that the hardware and software compatibility may have a positive effect on teachers perceived usefulness of PowerPoint. Chau and Hu (2001) have found that physicians are likely to recognize the usefulness of a technology when they perceive it as compatible with their workplace (Chau and Hu, 2001; 2002). Chau and Hu (2002) insisted that the compatibility of telemedicine technology perceived by a physician positively affects the perceived usefulness of that technology. Hence, the following hypothesis is established. H8. When one firm licenses-in one new technology, higher technological compatibility will make a positive contribution on U. Venkatesh et al. (2003) found that experience has moderate effects on behavioral intention. In addition, Lai (2003) insisted that larger firms, with greater financial capability and a higher industrial position, have adopted a professional management system. Thus, an exploratory perspective is taken to establish H 9-1 and H 9-2, as shown in Figure 4. H 9-1 : When one firm licenses-in one new technology, the influence of perceived compatibility on licensee s behavior intention will be moderated by firm experience. H 9-2 : When one firm licenses-in one new technology, the influence of perceived compatibility on licensee s behavior intention will be moderated by firm size. H 8 usefulness (U, η 2 ) H 1 compatibility H 6 H 7 H2 Behavior intention (BI, η 1 ) H 9-2 H3 H 9-1 Firm size ease of use Experience Figure 4. Theoretical model: Mediate and moderate effects. To investigate whether this alternative model can provide an ideal explanation and predict capability in a specific industry, the biotechnology industry is chosen by considering the increasing potential and the sample sufficiency. 470

11 5.1 Sample characteristics The effective sample size of biotechnology company is 139. The descriptive statistics reveal that R&D supervisors occupy the highest percentage (33.7%), followed by the CEO (31.6%), and other departmental supervisors (25.3%). Regarding the employee number, 86.3% of the sample companies have fewer than 200, 6.3% have , and 7.4% have more than 700. As for ownership, 78.9% of the sample companies are domestically owned, 5.3% are foreign-owned, and 15.8% are owned partly by foreign or domestic individuals and a legal entity. 5.2 Reliability, validity and model fitness The standardized coefficients of all measure paths indicate that even the smallest one still exceeds 0.8; however, tests of the construct reliability also find that except for the smallest one (perceived ease of use) close to 0.7, all others far exceed 0.7; therefore, the construct reliability is supported. Because all factor loadings of the indicators are statistically significant at p = 0.001, according to Anderson and Gerbing (1988), the convergent validity of all indicators is supported. Next, the two highest correlated constructs, behavior intention and perceived usefulness, are selected to compute their variance; extracted estimates are 0.66 and 0.74, respectively. Because the variance extracted estimate of these both constructs is greater than the squared correlation 0.34, discriminant validity is demonstrated (Netemeyer et al., 1990). As for the fitness of the model, since the chi-square / df (degree of freedom) = / 30 = 1.95 less than 3, CFI = 0.96, NFI = 0.92, the overall fitness of the model is acceptable. 5.3 Result analyses and moderated effects The empirical evidence shows that perceived technological compatibility and perceived usefulness have a significant positive impact on behavioral intention to licensing-in new technology. In addition, perceived ease of use has significant positive influence on perceived usefulness and the behavioral intention of licensing-in new technology. However, the indirect effect of perceived technology compatibility on perceived usefulness has to be transmitted via perceived ease of use. In addition, the R 2 indicates that 0.64 of the variance of behavioral intention, slightly higher than that of TBIM, can be explained by perceived technological compatibility, perceived usefulness, and perceived ease of use. The pure moderate effects of the two discrete moderators (experience and company size) are examined. In the first run of SEM, the measure path, measure intercept, structural path, structural intercept, structural mean, structural variance, and structural covariance matrix are constrained to be equal across groups, called the structural covariance model. This run generates an overall Chi-Square value (χ 2 H). In the second run of SEM, the constraints are removed, resulting in a Chi-Square value (χ 2 N) with additional degree of freedom. At the moderate effect of experience, the difference value of Chi-Square value (χ 2 H - χ 2 N) is 54.75, with correspondent degree of freedom, 14, which exceeds the critical value, at p = Therefore, it is inferred that there is significant moderate effect of experience (H 9-1 ). Next, regarding the moderate effect of firm size, the Chi-Square value (χ 2 H- χ 2 N) is with degree of freedom, 25, which is greater than the critical value, Therefore, this indicates a significant moderating effect of firm size (H 9-2 ) (Hsu and Liu, 2008). In addition, H 1, H 2, H 3, H 7, and H 8 are significantly supported and H 6 is not supported, as shown in Table

12 Dependent variables ease of use (EOU) Direct effect Indirect effect Overall effect usefulness (U) Direct effect Indirect effect Overall effect Behavior intention (BI) Direct effect Indirect effect Overall effect Table 3. The effects of mediate and moderate model. Independent variables Technological ease of use usefulness compatibility (EOU) (U) Effect C.R. Effect C.R. Effect C.R. H 6 (unsupported) H 8 (supported) *** *** H 7 (supported) *** *** *** H 2 (supported) ** ** H 3 (supported) *** *** *** H 1 (supported) *** *** Notes:* p < 0.05; ** p < 0.01; *** p< Critical ratio (C.R.) represents the ratio of the unstandardized path coefficients to the standard error ssociated with that estimate. This is tested against a z distribution. 6. Theoretical contributions and management implications 6.1 Theoretical contributions This study contributes to the theory of technology management in the following ways. First, examining the theoretical gap of TAM of external technological attribute variables sheds new light on their roles in the behavioral intention of licensing-in new technology. Second, providing new insights of how perceived usefulness bridges perceived technological attributes and the behavioral intention of licensing-in new technology. Third, this study demonstrates that the role of perceived technological compatibility should not be ignored when licensing-in new technology. Fourth, this study demonstrates that individuals instrumental concern for new technology should not be ignored; such means-end perception has a strong association with the behavioral intention of licensing-in technology. Fifth, this study illuminates the mediate-moderated role of experience and firm size, for licensing-in technology behavioral intention. 6.2 Management implications R 2 shows that 63% of the variance of the BI of the acceptance of licensing-in technologies can be accounted for by U, perceived technological implicitness, and compatibility in TBIM. The R 2 also shows that 64% of the variance of the BI of the acceptance of licensing-in technology can be explained by perceived compatibility, perceived usefulness, and perceived ease of use in the mediate and moderate model. This study shows that U provides direct significant influence on BI, as shown in Tables 2 and 3. This point matches the assertion of Venkatesh et al. (2003) that performance expectancy appears to be a determinant of intention in most situations, and is consistent with Bagozzi s (1982) and Vallacher and Wegner s (1985) arguments that individual intention is purpose-oriented and that individual promotion and wage increments determine licensing-in technology behavioral intention. In addition, it echoes Chau s (1996) finding that perceived near-term usefulness, improving job performance, or enhancing job satisfaction had significant influence on the behavioral intention to use technology. This finding implies that when licensing-in one new technology firm s staff will 472

13 evaluate the job performance and wage increase. Therefore, when licensing out, licensors should show prospective licensees how the new technology will improve their job performance and thus make a positive contribution to perceived usefulness. The EOU-BI link is significant in the mediate and moderate model (Table 3). The evidence of both models is consistent with the finding of prior studies (Davis et al., 1989) that EOU initially has a direct effect on BI. Matching Lu and Gustafson s finding (1994) that EOU was a significant determinant of the BI to use computer technology. Lastly, this finding also matches Di Benedetto s (2003) finding that indicates the importance of EOU on managers BI to adopt one new technology. However, the insignificant outcome of TBIM has two possible explanations: first, when the technology becomes more user-friendly, learning to use it becomes much easier (Chau, 1996), and second, the effect of EOU on BI might be moderated by gender and age (Venkatesh, 2003). In practice, licensing-in technology firms need to pay close attention to the design characteristics in addition to the self-efficacy and instrumentality of those technologies. This study s evidence concurs with the findings of prior studies (Di Benedetto et al., 2003; Rogers and Shoemaker, 1971) findings that perceived compatibility has direct significant positive influence on BI (Tables 2 and 3). This implies that when a firm s production equipment, materials, and operation habits have a greater compatibility with one new technology, then its function will be easier to expand. Di Benedetto et al. (2003) also argued that technology compatibility tends to be more firm-specific and is more likely to have direct effect on overseas managers intention to adopt new technology. Therefore, the licensee should collect related information, ensuring that existing production, materials, and operation habits are compatible with new technologies and ensuring the fit between the employees work style and the use of the system in the organization (Venkatesh, 2003). Alternatively, the licensor should make efforts toward the technology compatibility of target firms when implementing marketing plans. In addition, the licensor should build in required features or performance levels, benefits and emphasize these characteristics to support their promotion efforts (Di Benedetto et al., 2003; Talukder et al., 2008). This study finds that technological implicitness has a significant positive impact on perceived usefulness (U), then via U generates important positive effects on the BI of accepting one new technology (Table 2). These results are consistent with those of prior studies (Barney, 1991; Grant, 1996; Wei, 2008) that firms perceive that one new technology with implicit knowledge will generate significant contributions on personal business performance. The empirical evidence indicates that when the licensing-in technologies are associated with higher implicitness, the new technology contains more knowhow, uniqueness, and resistance to imitation. Then, the new technology is more likely to enable the firm to provide specialized products or services with unique competitive advantage (Morgan-Thomas and Bridgewater, 2004), thus making a positive contribution to the BI of accepting the new technology. The evidence suggests that perceived implicitness brings negative significant impact on EOU of new technologies (Table 2). These results are in line with those of previous scholars (Diericks and Cool, 1989; Reed and DeFillippi, 1990; Wei, 2008). When technological implicitness or causal ambiguity of one new technology forms its own technological dynamics is hard to imitate, then the implicitness of this new technology increases and firms will encounter more difficulty and spend more adaptive cost and time to realize the new technology and to expand its function (Cort et al., 2007). Thus, licensees should develop more concrete cooperation plans and activities to reduce the potential negative impacts from new technologies. Hence, the licensee can consider such processes as learning-by-doing models, and figure documentation and related cooperation to increase the EOU of new technologies. 473

14 7. Conclusion and future directions The evidence shows that the BI of accepting foreign technologies of Taiwan s four high technological industries is influenced directly by perceived technological implicitness or via U. technological compatibility presents strong positive contributions on BI in TBIM and mediate and moderate models. Lastly, the evidence shows that technological implicitness results in strong negative influence on EOU derived from new technologies. On the one hand, this study shows that EOU results in no significant impact on U in TBPM, and might be due to the organization s lack of effective and efficient underlying process or deployment to conserve the efforts saved to increase job performance (Linderman et al., 2004). On the other hand, the evidence of TBIM indicates that EOU has no significant influence on BI. One possible reason may be that the education level of the participants in this study is higher than that in other samples. Another potential reason may be that the correspondents in this sample are more experienced than those in other studies. Therefore, they are not as concerned about the capability to learn and use new technology (Venkatesh et al., 2003). These two issues are potential avenues for future research. Acknowledgements This author sincerely appreciates for the precious comments from the three anonymous reviewers and the financial aid from the National Science Council(NSC H (Received, 26 September 2007; received in 1 st revised form 19 June 2008; received in 2 nd form 6 October 2008; received in 3 rd form 16 December 2008; accepted 25 December 2008) References AL-Gahtani, S.S., King, M. (1999) Attitudes, satisfaction and usage: Factors contributing to each in the acceptance of information technology. Behavior & Information Technology, 18(4), Anderson, J.C., Gerbing, D.W. (1988) Structural equation modeling in practice: A review and recommended two-step approach. Psychological Bulletin, 103 (3), Atuahene-Gima K. (1993) Determinants of inward technology licensing intensions: An empirical analysis of Australian engineering firms. Journal of Product Innovation Management, 10(3), Bagozzi, R.P. (1982) A field investigation of causal relations among cognitions, affect, intentions, and behavior. Journal of Marketing Research, 19(4), Bandura, A. (1982) Self-efficacy mechanism in human agency. American Psychologist, 37(2), Bandura, A. (1986) Social Foundations of Thought and Action: A Social Cognitive Theory. Prentice-Hall, Englewood Cliffs, NJ. Barney, J. (1991) Firm resources and sustained competition advantage. Journal of Management, 17(1), Berg, S.V., Duncan, J., Friedman, P. (1982) Joint Venture Strategies and Corporate Innovation. Oelgeschiager, Gunn & Hain, Cambridge, MA. Chau, P.Y.K. (1996) An empirical assessment of a modified technology acceptance model. Journal of Management Information Systems, 13(2), Chau, P.Y.K., Hu, P.J.H. (2001) Information technology acceptance by individual professionals: A model comparison approach. Decision Sciences, 32(4),

15 Chau, P.Y.K., Hu, P.J.H. (2002) Examining a model of information technology acceptance by individual professionals: An exploratory study. Journal of Management Information System, 18(4), Cort, K.T., Griffith, D.A., White, D.S. (2007) An attribution theory approach for understanding the internationalization of professional service firms. International Marketing Review, 24(1), Davis, F.D., Bagozzi, R.P., Warshaw, P.R. (1989) User acceptance of computer technology: A comparison of two theoretical models. Management Science, 35(8), Di Benedetto, C.A., Calantone, R.J., Zhang, C. (2003) International technology transfer model on an exploratory study in the People s Republic of China. International Marketing Review, 20(4), Dierickx, I., Cool, K. (1989) Asset stock accumulation and sustainability of competitive advantage. Management Science, 35(12), Fornell, C., Larcker, D.F. (1981) Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18(1), Grant, R. (1996) Toward a knowledge-based theory of the firm. Strategic Management Journal, 17, Winter Special Issue, Hatcher, L. (2004) A Step-by-Step Approach to Using the SAS System for Factor Analysis and Structural Equation Modeling. SAS Institute Inc, Cary, NC. Hauser, J.R., Shugan, S.M. (1980) Intensity measures of consumer preference. Operation Research, 28(2), Horng, C. (2006) An empirical study of knowledge transfer and knowledge reciprocation in Taiwanese manufacturing industry. Asia Pacific Management Review, 11(1), Howard, P. (2001) The role of acquisition of foreign technology in Taiwanese growth. Industrial and Corporate Change, 10(3), Hsu, C.W., Liu, H.Y. (2008) Corporate diversification and firm performance: The moderating role of contractual manufacturing model. Asia Pacific Management Review 13(1), Hu, P., Clark, T., Ma, W. (2003) Examining technology acceptance by school teachers: A longitudinal study. Information and Management, 41(2), Lai, S. T.A. (2003) Learning strategy of small and medium-sized Chinese family firms: A comparative study of two suppliers in Singapore. Management Learning, 34(2), Larcker, D.F., Lessig, V.P. (1980) usefulness of information: A psychometric examination. Decision Science, 11(1), Lepper, M.R. (1985) Microcomputers in education: Motivational and social issues. American Psychologist, 40(1), Li, W.R., Cao, W.Z., Zhan, X.L., Chen, Y.Y. (2002) The influence of alliance characteristics on knowledge, technology transfer performance. Industry Financial Quarterly, (115), Linderman, K., Schroeder, R.G.., Zaheer, S., Liedtke, C., Choo, A.S. (2004) Integrating quality management practices with knowledge creation process. Journal of Operations Management, 22(6), Long, J.S. (1983) Confirmative Factor Analysis: A Preface to LISREL. Sage University Paper Series on Quantitative Application in the Social Science. Sage, Beverly Hills, CA. Lu, H.P., Gustafson, D.H. (1994) An empirical study of perceived usefulness and perceived ease of use on computerized support system use over time. International Journal of Information Management, 14 (5), Malhotra, N.K., Wu, L., Whitelock, J. (2005) An overview of the first 21 years of research in the International Marketing Review, International Marketing Review, 22(4),

16 Mitchell, W., Singh, K. (1992) Incumbents use of pre-entry alliances before expansion into new technical subfields of an industry. Journal of Economic Behavior & Organization, 18(3), Moore, G.C., Benbasat, I. (1991) Development of an instrument to measure the perceptions of adopting an information technology innovation. Information Systems Research, 2(3), Morgan-Thomas, A., Bridgewater, S. (2004) Internet and exporting: Determinants of success in virtual export channels. International Marketing Review, 21(4/5), Netemeyer, R.G., Johnston, M.W., Burton, S. (1990) Analysis of role conflict and role ambiguity in a structural equation framework. Journal of Applied Psychology, 75(2), Nonaka, I. (1994) A dynamic theory of organizational knowledge creation. Organization Science, 5(1), Premkumar, G., Ramamuthy, K., Nilakanta, S. (1994, Fall) Implementation of electronic data interchange: An innovation diffusion perspective. Journal of Management Information Systems, 11(2), Reed, R., DeFillippi, R.J. (1990) Causal ambiguity, barriers to imitation and sustainable competitive advantage. Academy of Management Review, 15(1), Robertson, T.S. (1971) Innovative Behaviors and Communication. Holt, New York. Rogers, E.M. (1962) Diffusion of Innovations. Free Press, New York. Rogers, E.M., Shoemaker, F.F. (1971) Communication of Innovation: A Cross-Cultural Approach. Free Press, New York. Szajna, B. (1996) Empirical evaluation of the revised technology acceptance model. Management Science, 42(1), Talukder, M., Harris, H., Mapunda, G. (2008) Adoption of innovations by individuals within organizations: An Australian study. Asia Pacific Management Review, 13(2), Thompson, R.L., Higgins, C.A., Howell, J.M. (1994) Influence of experience on personal computer utilization: Testing a conceptual model. Journal of Management Information Systems, 11(1), Vallacher, R.R., Wegner, D.M. (1985) A Theory of Action Identification. Erlbaum. Hillsdale, N.J. Venkatesh, V., Morris, M.G., Davis, G.B., Davis, F.D. (2003) User acceptance of information technology: Toward a unified view. MIS Quarterly, 27(3), Vroom, R.H. (1964) Work and Motivation. Wiley, New York. Wei, W.C. (2008) Inward beliefs in high technological performance: An empirical study. Journal of the Chinese Institute of Industrial Engineers, 25(2), Young, J.K. (2005) The impact of firms and industry characteristics on technology licensing. Advanced Management Journal, 70(1), Zhao, L.A.R. (1992) Toward meta research on technology transfer. IEEE Transaction on Engineering Management, 39(1),