Introduction

Defined as the utilization and use of vocational knowledge in a team’s value-adding process (Gharakhani & Mousakhani, 2012; Shujahat et al., 2019), knowledge application has been widely recognized by industry practitioners and scholars as valuable intangible capital for professional work teams. Great team performance can be attributed to the effectiveness of vocational knowledge that is applied to improve and enhance a team’s outcome (Alavi & Tiwana, 2002). Knowledge application (i.e., vocational knowledge application) strengthens a team’s vocational knowledge in order to improve its activities, decision-making, efficiency, and consequently collective performance. Since problems in vocational situations and activities are often ill-structured, it is critical to know how and when to easily retrieve diverse knowledge as needed by the team and to solve the problems in order to create teamwork value (Billett, 1996; Lin & Lu, 2022). A team that excels at knowledge application is relatively competitive and likely able to transform its intellectual assets into greater team performance. Put differently, the greater the knowledge application is practiced in a team, the greater is the likelihood that there will be quality business processes by the team that actually facilitate higher team performance (Lin et al., 2016).

Despite the key role of knowledge application for teams, it remains an understudied research topic in the vocational learning and team dynamics literature (Alavi & Tiwana, 2002). For instance, the literature has argued that relatively insufficient attention has been paid to knowledge application and its potential predictors (Milicevic et al., 2015), suggesting a critical gap that this study aims to fill. Drawing upon the social network theory and social cognitive theory, this study explores what drives knowledge application to influence a team’s performance outcome and if there exists any critical moderator in the development of knowledge application.

This study differs from previous literature in two important ways. First, it complements the literature by deriving knowledge application grounded in the social network theory to examine critical predictors that have been rarely discussed in knowledge application research. The social network theory is appropriate for justifying knowledge relevant issues as it scrutinizes relationship patterns between a team and its social surroundings in an industry network (e.g., resource availability and social relationship) (Li et al., 2015). Second, this study complements previous research by proposing an important moderator in a knowledge specific discussion from a socio-cognitive perspective. Specifically, while previous research has applied the social cognitive theory to justify self-efficacy as a moderator in information seeking effectiveness (Kelly & Kumar, 2009) or creativity (Jaiswal & Dhar, 2015) at the level of individuals, this paper extends the research to present the moderating role of efficacy in knowledge application from a team-level aspect.

Research Model and Hypotheses

This study develops a research model and relevant hypotheses regarding knowledge application and team performance. In the model, team performance relates to three predictors of knowledge application (i.e., team learning orientation, cluster resources availability, and cluster social relationship) via the mediation of knowledge application. At the same time, the influences of the predictors on knowledge application are moderated respectively by collective learning efficacy. Collective learning efficacy is defined herein as team workers’ collective beliefs in their ability to underpin the proactivity of learning new knowledge (Knapp, 2010), thus catalyzing the formation process of knowledge application.

These three predictors are simultaneously examined, because they represent complementary facilitators (in terms of learning, resources, and social relationship) that jointly enable a team to make good use of its diverse knowledge. These topics have not been jointly examined in previous studies. Team learning orientation, also known as team-level learning goal orientation, is defined as the way a team develops its profession by strengthening its own skills and competencies to perform teamwork (Yi & Hwang, 2003). In terms of industry clusters such as industrial zones or science parks, cluster resources availability is defined as a team’s ease of obtaining external resources such as human resources (e.g., recruitment of high-quality employees) and technical resources (e.g., innovation and technology information) that support its activities, whereas cluster social relationship is defined as social connections and communications that contribute to information sharing and cooperation between a team and its business exchange partners nearby (Li et al., 2015). Overall, while the social cognitive theory (Bandura, 1986) emphasizes the importance of goal-setting learning as a continuous process (i.e., learning goal orientation) that motivates and regulates collective behavior, the social network theory concentrates on resources and social relationship in an industry cluster (i.e., cluster resources availability and cluster social relationship).

Knowledge application plays a mediating role in this study, because it represents a team’s next logical step (Nesheim et al., 2011) after internal and external resources (e.g., team learning orientation and cluster resources) are appropriately taken. Through knowledge application, teams can turn such resources into knowledge integration methods to solve problems in vocational activities for the purpose of improving team performance. The vocational practices, methods, and techniques in question should be applied by teams in such a way that changes their collective outcomes (e.g., Nesheim et al., 2011). The importance of knowledge application is the extent to which the team acquires potentially useful knowledge about, for example, cluster resources availability and applies the knowledge in its own activities (Ode & Ayavoo, 2020). Hence, the exploitation of a team’s efforts requires use and application of actual experiences and knowledge.

Previous studies have suggested that team learning orientation can encourage adaptive behavior that drives team performance (Bunderson & Sutcliffe, 2003; Chen & Mathieu, 2008; Johnson et al., 2011). Knapp (2010) extends the work of Bandura (1997) to conceptualize team learning and discusses the conditions necessary to develop team learning orientation. Team learning orientation motivates team members to keep learning new things and polishing skills (Bell et al., 2012) in a collective fashion and inspires the team to take on challenging work that ultimately provides chances to apply their knowledge (Yi & Hwang, 2003). Previous research has indicated that a team with strong learning orientation is often persistent and task-focused (Bunderson & Sutcliffe, 2003) and shows increased interest in applying what they learn in workplaces (i.e., knowledge application) to eventually improve performance outcomes (e.g., Printrich, 2000). In other words, team performance is likely boosted when team learning orientation, which increases newly learned skills and ability, is turned into knowledge utilization methods that create more value in vocational activities (i.e., knowledge application) (e.g., Ode & Ayavoo, 2020). To sum up, the first hypothesis is derived below.

  • H1: Team learning orientation positively relates to team performance via the mediation of knowledge application.

Industry clusters have substantial influence on work teams. The literature has discovered the importance of industry cluster effects when assessing team project management (Algeo & Connell, 2017), human capital inputs of work teams (Shu & Simmons, 2018), and team performance (Biggiero, 2016). This is understandable, because team workers are likely to generate innovative ideas by making good use of industry clusters through networking for achieving better performance (Som, 2007). This study discusses how work teams are affected by such external resources as industry clusters to facilitate knowledge application and consequently team performance.

Two influential factors of knowledge application relevant to industry clusters are cluster resources availability and cluster social relationship (Lai et al., 2014). According to the social network theory, cluster resources that form a pool of industry-related resources such as innovation, up-to-date information, and skilled personnel can help work teams apply knowledge more easily in a changing and competitive environment (Lorenzen & Mudambi, 2012). Therefore, teams that are able to easily obtain cluster resources (i.e., high cluster resources availability for the teams) can acquire supportive resources (e.g., know-how inputs and alliance channels) and/or approach diverse talents to increase the synergy of their knowledge system (Zhu et al., 2015). The synergy of the knowledge system represents the way team workers share their expertise and understandings about specialization, industry, and environment and interpret and give meanings to their collective professional experiences.

Cluster resources availability allows a team to easily acquire quality workers (Lorenzen & Mudambi, 2012) or needed information (or insights) (Huang & Wang, 2018) that assist the team at improving knowledge application and ultimately team performance (Cooke et al., 2007). In other words, team performance is likely enhanced when cluster resources availability, which facilitates human capital and talent management, increases a team’s pace of employing the knowledge into new products or processes (i.e., knowledge application) (e.g., Cegarra-Navarro & Martelo-Landroguez, 2020). To sum up, cluster resources availability and its influence are hypothesized below.

  • H2: Cluster resources availability positively relates to team performance via the mediation of knowledge application.

The social network theory suggests that a team’s extensive social relationship within the network of an industry cluster can substantially enhance market information sharing between the team and its upstream and/or downstream exchange partners (e.g., Kim & Shim, 2018), thus refining its knowledge application. In fact, the literature has indicated that interaction and exchange based on a social relationship within an industrial cluster center on knowledge (Lai et al., 2014). The use of cluster social relationships by work teams to seek cooperation and exchange up-to-date information with their external work groups (Connell et al., 2014) can improve the effectiveness of knowledge application and consequently team performance. Accordingly, team performance is likely achieved when a cluster social relationship that facilitates social capital in helping a team communicate with others and better understand shared knowledge (Hu & Randel, 2014) enables the team to use market knowledge in its value-adding process (i.e., knowledge application). Hence, a team’s cluster social relationship and its influence are hypothesized below.

  • H3: Cluster social relationship positively relates to team performance via the mediation of knowledge application.

From the perspective of industry clusters, collective learning efficacy refers to team workers’ shared beliefs in their learning capabilities to leverage the motivation (e.g., team learning orientation), external industry platforms (e.g., resources and social relationship), and courses of action needed (e.g., knowledge application) that help achieve their performance goals (Almousa & Hejazi, 2022). While team learning orientation represents an approach or a strategy of learning by undertaking tough and challenging tasks to build up a team’s capability, collective learning efficacy reflects a team’s shared confidence (i.e., collective conviction or morale) in its persistent learning. Specifically, teams with strong learning orientation may still hesitate to apply their learned knowledge if they possess weak collective learning efficacy (i.e., a lack of confidence) that discourages them from actively seeking applicable ways to derive value (Zaccaro et al., 1995). On the contrary, teams with strong collective learning efficacy are more inclined to stick to their shared conviction by persistently making good use of team learning orientation for different ways of knowledge application. This suggests a positive moderating effect of collective learning efficacy on the relationship between team learning orientation and knowledge application. Accordingly, the interaction of collective learning efficacy and team learning orientation is hypothesized below.

  • H4: Collective learning efficacy positively moderates the relationship between team learning orientation and knowledge application, such that the relationship is stronger when collective learning efficacy is higher.

The literature has indicated that a team’s collective learning efficacy may alleviate workplace predicaments by having the confidence necessary to utilize the advantage of industry cluster resources when dealing with, for example, novel technology or technical difficulties (Salanova et al., 2003). In other words, collective learning efficacy helps generate an intervening effect by providing a team with the useful means of empowerment and strong mindset with a focus on what it can do (Jex & Bliese, 1999; Weathers et al., 2016) to coordinate available cluster resources for ameliorating knowledge application. As a result, stronger collective learning efficacy is more likely to enhance the positive effect of cluster resources availability on knowledge application. This is theoretically justifiable, because a lack of efficacy is often a barrier (Farnsworth et al., 2022) that inhibits a team’s learning readiness to exploit cluster resources, consequently weakening its use of learned knowledge for application. In other words, such a team is often not confident in obtaining and using cluster sources for achieving knowledge application even if the sources may be already available for it. Overall, the hypothesized interaction of collective learning efficacy and cluster resources is derived below.

  • H5: Collective learning efficacy positively moderates the relationship between cluster resources availability and knowledge application, such that the relationship is stronger when collective learning efficacy is higher.

Teams with high collective learning efficacy are likely to persist in the face of obstacles, have high performance expectations, and become confident in making good use of a cluster social relationship to increase the utility of their knowledge and expertise (Tasa et al., 2011). More specifically, a team with stronger collective learning efficacy possesses a shared belief in its conjoint capabilities (Danna, 2021; Zhao & Peng, 2022) to positively leverage a social relationship required to produce given levels of teamwork process (e.g., knowledge application). This phenomenon suggests a positive interaction of collective learning efficacy and cluster social relationship, which jointly facilitates a team’s knowledge application. In other words, teams with higher collective learning efficacy expect their efforts to produce positive outcomes (Fransen et al., 2015), expend greater effort in leveraging a cluster social relationship for executing teamwork, and persist longer in the face of adversity related to knowledge application than do those with lower collective learning efficacy (e.g., Little & Madigan, 1997). On the contrary, teams with weak collective learning efficacy tend to lack confidence in learning how to build up cluster social ties and may instead exhibit social withdrawal, consequently debilitating the positive effect of a cluster social relationship on knowledge application. Hence, the hypothesized interaction of collective learning efficacy and cluster social relationship is derived as follows.

  • H6: Collective learning efficacy positively moderates the relationship between cluster social relationship and knowledge application, such that the relationship is stronger when collective learning efficacy is higher.

Methods

Subjects and Procedures

The research hypotheses of this study were statistically assessed using a survey of work teams in a large high-tech industrial zone in north Taiwan. We invited part-time EMBA students from eight leading high-tech firms related to the field of computer and communications to help our data collection. The high-tech zone was chosen, because the technology industry represents the most important part of Taiwan’s economy. This study randomly selected 100 sample teams for investigation from a total pool of available 262 work teams listed by the high-tech firms. Simple random sampling is appropriately adopted herein, because it has little chance of sampling bias, and all the work teams of this study from the same industry share homogeneous characteristics. Supported by the human resources (HR) departments of our sample firms, the research investigators randomly distributed the questionnaires to team workers and leaders. Those who received the questionnaires decided on their own whether or not to fill out the questionnaires. In each team, four members and their team leader were surveyed (i.e., a total of five participants from each team).

Of the questionnaires distributed to 100 teams, 330 usable questionnaires from a total of 68 teams were collected, including 68 questionnaires from leaders and 262 questionnaires from members. Overall, the team-level response rate is 68%. The total participants included 215 males (65.15%), 146 people at the age of 40 or older (44.24%), 223 people with a bachelor degree or above (67.58%), 187 married people (56.67%), and 234 people with job experience of 10 years or above (70.91%).

To alleviate the unpredictable effect of common method variance (CMV) by obtaining data from two different sources (i.e., dyadic data from team members and their leaders), this study arranged team members to measure the remaining three variables (i.e., team learning orientation, knowledge application, and team performance) and team leaders to measure three factors (i.e., collective learning efficacy, cluster social relationship, and cluster resources). Previous research has suggested that managers (e.g., team leaders) are responsible for establishing industry cluster linkages with relevant knowledge resources and finding ways to develop and to benefit from external cluster social relationship (Giuliani, 2013). The precautionary method of surveying different sample subjects to measure different factors for reducing CMV is better than any other post hoc statistical remedies for merely detecting CMV (Baruch & Lin, 2012). Previous research has argued that the best way to minimize a potential bias caused by CMV is to arrange different survey subjects to measure their familiar objects or issues (Chang et al., 2010). Table 1 shows the correlation matrix of empirical data.

Table 1 Correlation matrix
Full size table

Measures

The variables in this study were measured using Likert scales modified from the literature by a focus group of three professors and one graduate student who were aptly versed in knowledge management and team dynamics research. Before the actual survey, this study conducted two pilot tests with exploratory factor analysis to evaluate the quality of scale items. Based on the factor analysis, this study refined the words and statements in the survey questionnaire to improve content validity before the actual survey. Appendix A presents all the finalized scale items.

The measures from the literature were initially chosen, because they had been pre-validated to effectively measure psychometric properties of working professionals. Whereas some measures in the literature were directly used in this study, others were slightly reworded by this study to better fit the teaming context in the high-tech industry cluster. For example, the item “our team looks for opportunities to develop new skills and knowledge” for measuring team learning orientation by Bunderson and Sutcliffe (2003) was slightly reworded as “our team can look for opportunities to develop new knowledge together” in this study. The item “the company can easily obtain individuals with talent and with high educational levels” for measuring “industry cluster resources” by Lai et al. (2014) was rephrased as “being located in the science park enables our team to obtain employees with talent more easily” to better measure cluster resource availability in this study. The item “I am confident in meeting the quality demands of the job” for measuring team efficacy by Lin et al. (2012) was reworded to “our team is confident in learning to meet the quality demands of the job” in this study. Note that team performance was directly measured by workers in this study, because team-related objective figures such as the commercial earning of team projects, their scale, and their success rates are not available due to business confidentiality.

Data Analysis

This study performed data analyses in four steps. First, the intraclass correlation of the survey data was statistically justified to support appropriate data aggregation at the team level. Second, the team-level data were analyzed using confirmatory factor analysis (CFA) to examine the reliability and validity of leaders’ data and members’ data respectively. Third, team-level hierarchical moderated regression analysis was performed to test the hypotheses. Fourth, bootstrapping analysis was conducted to double-confirm the mediation effects of knowledge application. As our sample size (i.e., 68 teams) is far less than the minimum size of 200 required for the analysis of SEM (Structural Equation Modeling) (Fabrigar et al., 2010; Lin & Huang, 2022), regression analysis is more appropriate for this research. Specifically, since the maximum likelihood estimation in SEM is used to estimate parameters, a small sample size will not suffice to test interaction terms (i.e., moderating effects) (Lin et al., 2021). The detailed analytic results based on the above four steps were presented sequentially in the following.

The data collected by this research were aggregated to the level of teams for analysis based on the justified intraclass correlations. This approach is recommended by the literature (e.g., Baruch & Lin, 2012; Rousseau, 1985), which states team-level analysis should be adopted given the collective group focus of the research (e.g., Dirks, 2000). The analytical results of this study showed (1) ICC1 values were larger than the recommended level of 0.12 (James, 1982), (2) ICC2 values were larger than the recommended level of 0.60 (Baruch & Lin, 2012), and (3) rwg values were larger than the recommended level of 0.70 (James, Demaree, & Wolf, 1984).

The data were analyzed using confirmatory factor analysis (CFA) before further empirical tests were conducted (see Tables 2 and 3). Table 2 exhibits goodness-of-fit indices based on team members’ data. The figures of NFI, NNFI, and CFI were larger than 0.9. The figure of RMSEA was smaller than 0.08. Table 3 shows goodness-of-fit indices based on team leaders’ data. Although the figure of NFI was slightly smaller than 0.09, the figures of NNFI and CFI were both larger than 0.9. The figure of RMSEA was smaller than 0.08. Collectively, these indices in Tables 2 and 3 were all acceptable to support that the data fitted our research model well.

Table 2 Standardized loadings and reliabilities of members’ data (N1 = 262)
Full size table
Table 3 Standardized loadings and reliabilities of leaders’ data (N2 = 68)
Full size table

In Tables 2 and 3, all the factor loadings were significant at p<0.001, and the average variance extracted (AVE) of every factor was larger than 0.50. Moreover, the reliability of every factor was larger than 0.70. These results support convergent validity of the empirical data. With regard to discriminant validity, chi-square difference tests were conducted to verify the data from team members (see Table 4) and the data from team leaders (see Table 5). The chi-square difference statistics in Tables 4 and 5 show that each pair of research factors met the overall significance level at 0.01 or lower, thus confirming discriminant validity.

Table 4 Chi-square difference tests on the data from team members (N1 = 262)
Full size table
Table 5 Chi-square difference tests on the data from team leaders (N2 = 68)
Full size table

Results

This study adopted hierarchical moderated regression analysis to test the hypotheses. To increase the precision of the estimates in the statistical analysis, this study included five control variables relevant to teaming contexts (e.g., social desirability). In Table 6, Model 1 included the control variables and three antecedents to explain knowledge application (i.e., the mediator), showing that only team learning orientation and cluster social relationship positively related to knowledge application. Model 2 tested the direct association between knowledge application and team performance, revealing that knowledge application positively related to team performance. Model 3 tested the relational strength between the three antecedents and the outcome (i.e., team performance) when the mediator (i.e., knowledge application) was also included in the regression model. If full mediation exists instead of partial mediation, then the relation between the antecedents and the outcome should be insignificant given the mediator in the same statistical model. In Model 3, the test result revealed that the effects of the three antecedents on team performance were all insignificant when knowledge application remained significant in the same model. This result suggests that knowledge application mediated the indirect relationships between team learning orientation and team performance and between cluster social relationship and team performance (i.e., H1 and H3 are supported, but H2 is not supported).

Table 6 Team-level hierarchical regression analysis
Full size table

To test its hypothesized moderation, Model 4 (see Table 6) included three interaction terms of collective learning efficacy, revealing that collective learning efficacy positively moderated the relationships between team learning orientation and knowledge application and between cluster resources availability and knowledge application (thus, H4 and H5 are supported). However, collective learning efficacy did not moderate the relationship between cluster social relationship and knowledge application (thus H6 is not supported).

This study additionally conducted post hoc bootstrapping analysis to double-confirm the mediation of knowledge application. The test result in Table 7 demonstrated that the relationships between team learning orientation and team performance and between cluster social relationship and team performance were mediated by knowledge application. This result was consistent with what the result obtained in the preceding regression analysis. Finally, Table 8 summarizes the results of hypotheses in this study.

Table 7 The results of the mediation using bootstrapping
Full size table
Table 8 Empirical results of hypotheses
Full size table

Discussion

Implications for Research

This study applied the social network theory and social cognitive theory to develop a team-level model manifested with three predictors that simultaneously motivate knowledge application and team performance. The findings of this study complement previous research that examines the mediating role of knowledge application from the theoretical perspective of trait activation (Lin & Lu, 2022). In other words, this study helps expand the understanding of knowledge application driven by not only inside the team (e.g., team learning orientation), but also from external network resources it possesses (e.g., industry clusters). After all, teams knowing how to exploit industry clusters likely turn out to be high performing ones that effectively deal with any drastic environmental changes.

While prior research based on the experiential learning theory (Almousa & Hejazi, 2022) has proposed the antecedent role of collective learning efficacy for absorptive capacity that eventually leads to knowledge exploitation and exploration, this study analyzes the moderating nature of collective learning efficacy in the formation process of knowledge application. Furthermore, the moderation of collective learning efficacy verified by this study provides important supplementary evidence in previous research (Wu, 2016) that explores the moderation of collective knowledge efficacy in the development of knowledge application and team creativity. Collectively, this study demonstrates that collective learning efficacy is a key lubricant that makes knowledge application run more smoothly in order to boost team performance.

Although there are similarities in the function of learning efficacy beliefs between individual level of analysis and team level of analysis, it has been found that there exist unique interaction effects of self- and collective learning efficacy (e.g., Jung & Sosik, 2002). This study is one of the few to examine the moderating influence of collective learning efficacy at the level of teams. To sum up, this study provides important supplementation for relevant literature based on the social cognitive theory, which has found, for example, that collective learning efficacy moderates the relationship between information exchange and innovation performance (Liu et al., 2015), and that a moderating effect of efficacy on goal achievement exists in the association between the leadership competency of innovation and knowledge transfer (Yoon & Han, 2018). Since these prior studies have not taken into account the system of industry clusters, the results of this study combined with previous findings can help achieve a holistic view of complex industry clusters and generate a learning strategy for the improvement of team performance. Team dynamics cannot be fully comprehended without team workers paying close attention to industry clusters as antecedent forces that impact knowledge application.

Implications for Practice

The empirical result of this study that shows knowledge application as a full mediator suggests team leaders should use knowledge application as a regular check point for predicting team performance. Specifically, it is recommended that team leaders periodically observe the strength of knowledge application in their team so as to take appropriate remedies necessary for guiding the team in an effective manner. Furthermore, team leaders should be provided with up-to-date trainings related to inspirational leadership and supply chain relationship management in order for them to amplify the benefits of team learning and industry clusters and ultimately increase knowledge application and team performance. For example, inspirational leadership represents a team leader’s ability to be a positive influence on team members and inspire them to be aware of cluster resources availability so as to implement the mechanism of knowledge application that helps upscale their teamwork quality. As another example, supply chain relationship management, which represents the process of managing a team’s strong cluster social network, is important as many projects require a team to perform comprehensive analyses of supply chain structure, information and physical flows, and how they communicate and collaborate with cluster social partners to learn new knowledge for specific applications (e.g., Kopczak & Fransoo, 2000).

The empirical result regarding the positive effect of team learning orientation on knowledge application provides additional evidence that supports previous research suggesting that, for example, team learning orientation mediates the relationship between HRM system and knowledge sharing (Yang et al., 2016). Since learning orientation enables teams to pursue cognitively demanding tasks and goals and actively adopt sophisticated learning strategies, stronger learning orientation would be more prominent and predominant among teams with high collective learning efficacy, but ineffectual among teams with low collective learning efficacy. For the purpose of improving knowledge application, a team leader may guide team members to retrospect their prior success, inspire the team’s shared belief in accomplishing teamwork (i.e., increased collective learning efficacy), and also inspire team learning orientation by assigning a variety of challenging and meaningful tasks with varying options. By doing so, team members will be more inspired at developing new teamwork skills and new ideas to find out what works the best (i.e., enhanced team learning orientation).

The moderating effect of collective learning efficacy on the positive relationship between cluster resources availability and knowledge application is an important finding that complements previous research in which people’s beliefs in their own coping efficacy determine how much strain is experienced when demanding conditions for resources (e.g., cluster resources) are attainable (Salanova et al., 2003). A key implication of this study is that industry resources availability may have weaker effects on knowledge application when teams are working with low levels of collective learning efficacy. Therefore, a team leader who offers team members feedback on prior teamwork and reinforces their resolution in achieving goals can facilitate a virtuous confidence-building cycle, thus positively leveraging knowledge application to boost team performance.

Limitations of the Study

It is worth noting three limitations in this study. The first limitation concerns its generalizability associated with industry clusters. Due to the highly delimited nature of our subject sample in the high-tech industry of Taiwan, inferences drawn from our data may not be fully generalizable to teams from traditional industries (e.g., food industry) or those from other countries with completely different national cultures. Second, general empirical findings based on the field survey data collected in this research might constrain causal inferences (e.g., Chou et al., 2016). Third, although the method of arranging different research groups (i.e., either team members or team leaders) to measure different factors in this study is advantageous for reducing CMV, it may have a drawback whereby key factors are only measured from the perspective of the chosen group. Future research may invite both groups to measure all the factors that can be eventually compared across groups in further analyses. Future scholars can consider observing work teams and their activities longitudinally, collect the primary data of team performance, and measure dyadic social relationships in industry clusters so as to further advance our understanding about the synthesis of team learning orientation and industry clusters.

Appendix A Measurement Items

Team performance (Source: Janssen & Van Yperen, 2004)

  1. 1.

    Our team always completes the duties specified in its job description.

  2. 2.

    Our team meets all the performance requirements of the job.

  3. 3.

    Our team fulfills all responsibilities required by its job.

  4. 4.

    Our team never neglects the job that it is obligated to perform.

Team learning orientation (Source: Bunderson & Sutcliffe, 2003)

  1. 1.

    Our team can develop new teamwork skills together.

  2. 2.

    Our team can look for opportunities to develop new knowledge together.

  3. 3.

    Our team is willing to take risks together in order to try new ideas.

  4. 4.

    Our team is willing to take risks together in order to find out what works.

Knowledge application (Source: Choi et al., 2010)

  1. 1.

    Our team members apply vocational knowledge learned from experience.

  2. 2.

    Our team members use vocational knowledge to prevent problems from becoming worse.

  3. 3.

    Our team members apply vocational knowledge to solve new problems.

  4. 4.

    Our team members use vocational knowledge to make decisions.

Cluster resources availability (Source: Lai et al., 2014)

  1. 1.

    Being located in the science park enables our team to obtain employees with talent more easily.

  2. 2.

    Being located in the science park enables our team to obtain experienced employees more easily.

  3. 3.

    Being located in the science park enables our team to recruit employees with required core technique more easily.

  4. 4.

    Being located in the science park enables our team to obtain industry resources we need.

Cluster social relationship (Source: Lai et al., 2014)

  1. 1.

    Our team has close cooperation with upstream and downstream firms in our science park.

  2. 2.

    Our team has quality social connection with upstream and downstream firms in our science park.

  3. 3.

    Our team has information exchange and sharing with upstream and downstream firms in our science park.

  4. 4.

    Our team can easily enhance interpersonal relationship with upstream and downstream firms in our science park.

  5. 5.

    Our team can easily develop strategic alliances with upstream and downstream firms in our science park.

Collective learning efficacy (Source: Lin et al., 2012)

  1. 1.

    Our team is confident in learning to meet the quality demands of the job.

  2. 2.

    Our team is confident in learning to follow all of the safety rules on the job.

  3. 3.

    Our team is confident in learning to maintain job performance.

  4. 4.

    Our team is confident in learning to keep up with the operational pace of my firm.