Study on the Effect of Organizational Members Network Interaction on Knowledge Resources in the Tourism Industry of Taiwan

Transcription

1 Study on the Effect of Organizational Members Network Interaction on Knowledge Resources in the Tourism Industry of Taiwan Wen-Tsao Pan 1, Fei-Xiong Ma 2 *, Wen-Cheng Wang 3, Yungho Leu 4 1. Department of Information Management, National Taiwan University of Science and Technology 2. School of Business, Guangdong University of Foreign Studies 3. Department of Business Administration, Hwa Hsia University of Technology Abstract: The members of enterprises have different experiences and backgrounds, and their skills, knowledge and work experiences are also different. For enterprises, the enhancement of members exchange of knowledge resources is the key of success. Knowledge resources are based on interpersonal interaction. This study treated 30 members of the Marketing and Customer Service Department of eztravel in Taiwan as the subjects, and conducted network centrality analysis, network faction analysis, and MRQAP. According to the analytical result, a positive communication environment and encourage member interaction can strengthen members exchange of knowledge resources and result in optimal customer service. Keywords: Social network analysis; Network centrality; Network faction; Knowledge resource; MRQAP 1. Introduction Previous studies have found that knowledge innovation is based on organizational members interaction, and that by this exchange, innovative knowledge can be developed. Therefore, it is important to explore knowledge management from the perspective of corporate organizational members interaction. Social network analysis is an analytical method based on corporate organizational members network structure. Its main purpose is to probe into corporate organizational members interaction. Hence, by social network analysis, this study attempted to find how the network interaction of members in a customer service department of the tourism industry in Taiwan would influence knowledge resources. In addition, data collection was based on the network members evaluations of each other and the nominal responses of all network members in the enterprise, which was then used as the individual members personal data. Therefore, in comparison to the personal data gathered using traditional self-evaluations based on subjective cognition, the result of the measurement was more objective and difficult. The remainder of this paper is organized as follows. Section 1 introduces the research motives and purposes; Section 2 presents the literature review; Section 3 describes the research samples and research methods; Section 4 discusses the empirical results; Section 5 offers conclusions. 2. LITERATURE REVIEW 2.1Knowledge resources Hung et al.(2015) analyzed 83 empirical studies selected from the top 30 journals in information management, five well-known international conferences (namely, AMCIS, ECIS, PACIS, ICIS, HICSS), and two journals in which knowledge sharing has been discussed extensively, during The results indicate the virtual community members, eight factors affect the frequency of knowledge sharing knowledge self-efficacy, structural social capital, identification, relational social capital, social interaction, altruism, reputation, and commitment. Five factors affect quality of knowledge sharing: trust, shared language, identification, social interaction, and altruism. Three factors affect knowledge-sharing attitude: altruism, reciprocity, and trust. Four factors affect knowledge-sharing intention: reputation, knowledge-sharing attitude, knowledge self-efficacy, and the subjective norm of knowledge sharing. For employees, two key factors affect the frequency of knowledge sharing: commitment and trust. Three factors affect knowledge-sharing attitude: reciprocity, information technology, and extrinsic rewards. Two factors affect knowledge-sharing intention: subjective norm of knowledge sharing and knowledgesharing attitude. 2.2 Social network theory Wang(2013) attempts to follow this idea to study the power distribution and its impact factors of Taiwan s National Health Insurance (NHI) sustainability between 1998 and In regard to research methods, in-depth interviews have been used to collect data. In conclusion, first, for those who are searching for localised power map to analyse policy processes in Taiwan, this study is a primary but useful beginning for applying social network analysis. Second, with regard to the most powerful actors of Taiwan s NHI domain, there is no doubt that the alliance between the public sectors and the providers show critical influence in the policy making process. Wang et al.(2012) applies social network analysis (SNA) for identifying policy brokers and what their characteristics are by developing a conceptual framework about brokerage roles in policy processes. In theoretical development, this research is expected to demonstrate how technique of SNA is helpful for not only identifying policy brokers but also showing how measure different type of brokerage positions in policy processes. In practice, the analytical results increase the current understanding, in part, how these political elites play their brokerage roles and enhance their influence over policy-making and implementation in Taichung urban development domain from 1986 to Network centrality Yu et al.(2013) conducted content analysis and network analysis on a sample of 547 master s theses from eight departments of the College of Journalism and Communications of Shih Hsin University to examine the relationships between the advisors and committee members as well as the connections of research topics. The results showed that the topic lifestyle have attracted cross-department research interests in the college. The academic network of the college is rather loose, and serving university administration duties may have broadened a faculty member s centrality in the network. The Department of Communications Management and the Graduate Institute of Communications served as the bridges for the inter-departmental communication in the network. One can understand the interrelations among professors and departments through study on network analysis of thesis as to identify the characteristics of each department, as well as to reveal invisible relations of academic network and scholarly communication. 2.4 Research samples, structure, and method Journal of Residuals Science & Technology, Vol. 13, No. 7,

2 2.5 Sample data and variables In 2013, the tourism industry in Taiwan contributed 5.3% of the country s GDP, and it was increasing year by year, thus making it critical for Taiwan s economy. eztravel has a leading position in the tourism industry in Taiwan, with its service scope covering online ticket reservation, payment, inquiry, and access to a global system. With its advanced technology and excellent service and efficiency, eztravel can be a representative case for research. This study treated members from the Marketing and Customer Service Department of eztravel as the subjects, and explored the reasons for the company s excellent service and efficiency. In June 2014, questionnaires were distributed. After eliminating invalid samples, there were 30 valid samples. Based on existing theories and literature reviews, the definitions and measurement of items were created, as shown in Table 1. There were three dimensions, and each dimension had three items. In addition, consultation networks, trust networks and cognition networks in MRQAP were defined as independent variable (X), while knowledge resource (know-how) was defined as dependent variable (Y). Based on the independent variable and dependent variable, MRQAP analysis was conducted. The descriptive statistics of the samples collected in this study are shown in Table 2. Where N of Obj: the number of group members with possible relationships. Min, Max and Avg: the minimum of connection, maximum of connection and average of connection of group members resource sharing in the social network. Sum: the number of actual relationships of the group members. Std and Var: the standard deviation and variance coefficient of the group members network matrix. Based on the result of the statistical analysis, this study found that the first interviewee is the contact window of the market, whether for business or for this survey. Hence, he is familiar with all other members of the department. Therefore, the number of connection of group members was nine, which was higher than the other members. Research structure This study generalized a framework of the relationships, as shown in Figure 1, in which professional skills of consultation network refer to the skills and knowledge related to travel goods marketing and sales, and knowledge resources refers to the colleagues intention to share goods marketing or sales knowledge. <<Figure 1>> Research method The research methods of this study included network centrality analysis in social network analysis, network faction analysis, coreperiphery analysis, and MRQAP. Borgatti (1998) classified social network analysis into two categories. The first category is self-centered networks based on individuals. This category emphasizes the connections of one member (including individuals or organizations) to others. The key to the analysis of this category is investigating the connections and position of an individual in the network to identify strong ties, weak ties and centrality, etc. The second category treats the total network as the key for analysis. It emphasizes the structural distribution of all members (groups, organizations or communities) in network. The analytical content includes network density, concentration, and factions, etc. Nahapiet & Ghoshal (1998) treated consultation networks, trust networks and cognition networks as types for network analysis. Cognition networks refer to the social networks constructed by group members with matching goals and similar values, and network analysis is based on the pair relationships of all actors in the whole network. Network centrality analysis This study refer to Yu et al. (2013) treated the Marketing and Customer Service Department of eztravel as the target. First, it probed into the network centrality of individual networks in eztravel, and attempted to identify the organizational members who represented the decision-making core of the organizational network, as shown in Figure 1. In the figure, when the point in the diagram is larger, it means the organizational member was closer to the decision-making core. <<Figure 2>> 5. Network faction and core-periphery analysis Based on social network analysis, Weng (2011) explored the evolution of a technological social network in the process of the development of a technology in an organization. Two key findings of Weng (2011) included (1) the core positions with highly dense structures in a technological social network are the current focuses of the acquired technologies of the organization; (2) the peripheral positions with sparse structures in a technological social network may contribute in acquiring new technologies for the organization. This study then analyzed the factions (informal small groups in the organization) of the whole network and tried to find the organizational members in these factions. The core-periphery analysis identified the members at the core of the group, as well as those who were less important. Figure 2 is the output result of the network faction analysis. There were a total of four factions. The diagram also shows four categories. By studying these factions, it was possible to explore the organizational members cohesion and comprehend the operational or managerial performance. <<Figure 3>> 5. MRQAP Using the MRQAP, Kurvers et al. (2013) calculated partial matrix regression coefficients for a response matrix on several explanatory matrices. Then they used a large number of random node label permutations within matrices to generate a sampling distribution and assigned P values. From UCINET they obtained P values together with mean estimates and SEs for all variables. This study adopted MRQAP (multiple regression quadratic assignment procedure) to analyze the correlation between several network matrixes and one network matrix. One matrix was treated as the dependent variable and the other matrixes were treated as the independent variable to analyze the explained power and significance level of the regression model (Chen, 2004). MRQAP is suitable for analyzing the correlation of network matrix. Hence, this study applied MRQAP to the correlation among consultation networks, trust network cognition networks and knowledge resources. However, organizational members mutual consultation is also a type of interaction. When two people consult each other, their exchange of knowledge resources will increase. On the contrary, when mutual consultation is low, they will not have the interaction of knowledge resources. Therefore, this study developed the following hypothesis: H1: When group members degree of mutual consultation is high, the exchange of knowledge resources is higher. Thomas (1998) suggested that trust is usually associated with the expectation or trust that others will take actions by predicable Journal of Residuals Science & Technology, Vol. 13, No. 7,

3 measures and will not be completely concerned about personal benefits. Nonaka, Toyama & Konno (2000) suggested that the condition of organizational members knowledge sharing represents trust and commitment. Nahapiet & Ghoshal (1998) indicated that trust allows individuals to acquire and exchange intellectual capital and allows them to more effectively expect the value of the exchange. Particularly in fuzzy and uncertain situations, knowledge exchange must rely on trust in social relationships. Therefore, the difference of organizational members mutual trust will influence individuals acquisition of social resources in a network. In other words, when two people tend to trust each other, there will be an exchange of knowledge resources; on the contrary, when they do not trust each other, they will avoid the exchange of knowledge. Hence, this study developed the following second hypothesis: H2: When group members tend to trust each other, they are more likely to exchange knowledge resources. From the perspective of the similarity-attraction paradigm, this study suggested that demographic variables, attitudes, values, cognition, common goals, and benefits allow them to recognize the potential value of resource exchange. Therefore, organizational members with a common vision would be more likely to become partners to share and exchange resources. Thus, this study developed the third hypothesis, as follows: H3: When group members have more similar values, they will be more likely to exchange knowledge resources. 6. EMPIRICAL ANALYSIS This study combined all the items for consultation networks, trust networks, cognition networks and knowledge resources to conduct network centrality analysis, network faction analysis, and core-periphery analysis on the subjects. This study then separated the items, and treated consultation networks, trust networks and cognition networks as independent variables (X), while knowledge resources was treated as dependent variable (Y). MRQAP was conducted using the independent variables and dependent variable in order to recognize the effects of the characteristics of members of eztravel s Marketing and Customer Service Department, as well as pair interactions in the network based on the exchange of knowledge resources. 7. Network centrality analysis of the Marketing and Customer Service Department <<Figure 4>> This study analyzed the network centrality of group members in the enterprise in order to find organizational members who represented the decision-making core of the corporate network. The analytical result is shown in Figure 4, in which a larger point represents the organizational member being closer to the decision-making core. The analytical result showed that the point for Member 1 was the largest, because Member 1 was the leader of the department. The result matched the expectation of this study. In addition, the points for Member 15 and Member 17 were the second largest. After consulting the leader of the department, this study realized that the two members were secondary supervisors in the department and were responsible for contacting other members of secondary departments. According to network centrality, this study was able to find the decision-making core members of the corporate network. Therefore, network centrality analysis was important for analyzing the group members communication in the business department. This study measured the Krackhardt GTD of the group members. It explored the correlation, degree, efficiency and current limit of eztravel s Marketing and Customer Service Department. The result revealed that the value of correlation was , the degree was 0, the efficiency was , and LUB was 1, all of which indicated that the department members were close to each other and had frequent interaction. Therefore, the exchange of professional skills and knowledge was rapid. There was no difference of degree. Hence, the members exchanged knowledge and helped each other at the same degree, and they rarely contacted each other using third parties, thus making the communication high efficient. When there was conflict among the coworkers, the leader would help deal with the problem. Hence, the department was highly efficient and aggressive. 8. Network faction analysis of the Marketing and Customer Service Department <<Figure 5>> This study next conducted network faction analysis on the members of the Marketing and Customer Service Department in order to find organizational members who were in the same faction. The analytical result is shown in Figure 5. According to the figure, there were six factions in the enterprise, and each faction had three members. The first faction included Member 1, Member 4 and Member 5. The second faction included Member 1, member 15 and Member 24. The third faction included Member 7, Member 8 and Member 9. The fourth faction included Member 12, Member 15 and Member 24. The fifth faction included Member 16, Member 17 and Member 18. The sixth faction included Member 21, Member 22 and Member 23. In these factions, the members had close interaction and knowledge exchange. Noticeably, Member 1 was in the first faction and the second faction of the secondary level. This individual tended to interact with the first and second factions and enhance the knowledge exchange of both factions. The second and fourth factions shared two members, including Member 15 and Member Core-periphery analysis of the Marketing and Customer Service Department <<Figure 6>> Journal of Residuals Science & Technology, Vol. 13, No. 7,

4 Figure 1. Research framework Figure 2. Diagram of network centrality Figure 3. Diagram of network faction analysis Figure 4. Diagram of the analytical result of network centrality Journal of Residuals Science & Technology, Vol. 13, No. 7,

5 Figure 5. Tree of the network faction analysis result Consultation (X1) Figure 6. Result of network core-periphery analysis Table 1: Operational definitions and measurement of variables Construct Operational definition Items References 1. Who are the colleagues you consult when you encounter obstacles at work? Network based on 2. In daily business, who network mutual consultations of are the colleagues you often professional competence. discuss issues with? 3. In your team, who are the colleagues with the professional skills needed to deal with the job? Trust network (X2) Network based on members mutual trust in the organization. 4. In your project team, who are the colleagues you trust the most? Burt (1984); Krackhardt (1993); compiled by this study Journal of Residuals Science & Technology, Vol. 13, No. 7,

6 (Y) Cognition network (X3) Knowledge resources Network based on members shared goals and similarity of values in the organization. The experience, value and words shared by other members in the network. 5. In your team, who are the colleagues that do not take advantage of you and who are concerned about your benefit? 6. In your team, who are the colleagues that are honest to you? 7. In your team, who are the colleagues that share the same ambition? 8. In your team, who are the colleagues that share the same goals? 9. In your team, who are the colleagues that share similar values? 10. Who are the colleagues you share information with when encountering new technology? 11. When you obtain new concepts or theories, who are the colleagues you share this information with? 12. Who are the colleagues you share your work experience and skills with? Burt(1984); Krackhardt (1993); compiled by this study Nonaka & Takeuchi (1995); compiled by this study Table 2: Descriptive statistics of the members in the Marketing and Customer Service Department No Obs Min Max Sum Avg Std Var No Obs Min Max Sum Avg Std Var Table 3: MRQAP regression analysis of consultation networks, trust networks and cognition networks on knowledge resources Coeff Sign Consultation network * * * Trust network * * * Cognition network * * * R 2 of total model Adjusted R *p<0.05; * *p<0.01; * * *p<0.001 Journal of Residuals Science & Technology, Vol. 13, No. 7,

7 This study conducted network core-periphery analysis on the group members of the department to find the core employees. The analytical result is shown in Figure 6, in which the upper left is the block of the core. This study found that there were four core members, including Member 1, Member 12, Member 15 and Member 24. The rest of the members were in a marginal block. The analytical result matched the expectations. Since Member 1 was the leader of the Marketing and Customer Service Department, this individual was also a core member. Based on Figure 4, the network positions of Member 12, Member 15 and Member 24 were close to Member 1. Thus, they were the core employees of the network. The block density of the core periphery was associated with the overall fit of the block. In the overall fit, the density of the upper left block was 1, and the density of the lower right block was 0. The analytical result showed that the density of the upper left block was and the density of the lower right block was The result demonstrated that the block was close to the total fit. MRQAP analysis of the Marketing and Customer Service Department This study treated consultation networks, trust networks and cognition networks as independent variables (X), and treated knowledge resources as dependent variable (Y) for MRQAP analysis. As shown in Table 3, the adjusted determination coefficient (Adj R-Sqr) of the total regression model was This demonstrated that the explained power of the independent variables (consultation network, trust network and cognition network) on the dependent variable (knowledge resources) was 62%. The P value of the total model was The figure was significantly lower than the 0.05 significance level; therefore, the regression model was statistically significant. In Model 1, the regression coefficient of consultation networks on knowledge resources was and the P Value was , which was significantly lower than the 0.05 significance level. The result showed that the group members interaction in consultation networks had a positive relationship with the exchange of knowledge resources. In other words, H1 was supported. In Model 2, the regression coefficient of trust network on knowledge resource was and the P value is , which was significantly lower than the 0.05 significance level. Therefore, group members interaction in trust networks had a positive correlation with the exchange of knowledge resources. In other words, H2 was supported. In Model 3, the regression coefficient of cognition network on knowledge resources was and the P value is , which was significantly lower than the 0.05 significance level. The result showed that the similarity of group members values had a positive correlation with the exchange of knowledge resources. Therefore, H3 was supported. <<Table 3>> In sum, the centrality analysis of a network facilitates the recognition of top-level decision-makers; the faction analysis of a network allows us to recognize the members of different factions; the core-periphery analysis of a network helps us to distinguish core members from peripheral members. With the centrality analysis, we can identify the relationship among different members of an organization, which are useful for the decision making of management personnel. GENERAL CONCLUSIONS AND CONCLUSION The main contribution of this study was the exploration of Taiwan s eztravel to find the reason why its efficiency and customer service were superior to other enterprises. Through network centrality analysis, network faction analysis and core-periphery analysis, this study found that in the company s Marketing and Customer Service Department, Member 1, Member 15 and Member 17 were at the core of the network. After measuring the Krackhardt GTD, this study found that the knowledge transmission of professional skills was rapid and there was no difference of levels. This study recognized six factions in the enterprise, of which each faction had three members. Four of these members were core members, including Member 1, Member 12, Member 15 and Member 24. The result of the MRQAP regression analysis demonstrated that in the Marketing and Customer Service Department, members pair interaction in consultation networks, trust networks and cognition networks would influence the exchange of knowledge resources (know-how). In other words, when members interaction in organizational networks were more frequent, their exchange of knowledge resources would be more significant. Thus, this study suggested that if organizations create a positive communication environment, encourage members interaction, and develop members cohesion and common values, they will strengthen members knowledge resource exchange and result in optimal administration efficiency and customer service. Acknowledgements Funded by the social science foundation of Guangdong Province (296-GK162018) References [1] G.X. Wang, Analysis on power pathway measurement in public policy: the case study of national health insurance s crucial policy events, Public Administration & Policy, No.57, (2013) [2] G.X. Wang, R.M. Hsung, Using network analysis for researching brokerage roles in policy processes: the case of Taichung City s development domain before and after the lifting of martial law, Conference of Taiwan Association for Schools of Public Administraion & Affairs (2012). [3] J.T. Chen, The emergences and consequences of intra-organizational social networks--social capital perspective. Doctoral dissertation, Graduate Institute of Human Resource Management, National Sun Yat-sen University (2004) [4] S.Y. Hung, Y.W. Huang, H.M. Lai, A meta-analysis of critical factors for knowledge sharing, Journal of Information Management, Vol.22, No.4, (2015) [5] I. Nonaka, R. Toyama, N. Konno, SECI,Ba,and Leadership:a unified model of dynamic knowledge creation. Long Range Planning, Vol.33, 5-34 (2000) [6] D. Miller, Shamsie J, The resource-based view of the firm in two environments: The hollywood firm studios from 1936 to 196. Academy of Management Journal, Vol.39, Vol.3, (1996) [7] R. Hall, The strategic analysis of intangible resources. Strategic Management Journal, Vol.13, (1992) [8] J.C. Spender, Making knowledge the basis of a dynamic theory of the firm, Strategic Management Journal, 17:5-9 (1996) [9] Y.N. Yu, Yuan MSA network analysis of the teachers and graduate students research topics in the field of mass communication, Journal of InfoLib & Archives, Vol.11, No.1, (2013) [10] C.W. Thomas, Maintaining and restoring public trust in government agencies and their employees. Administration and Society, 30: (1998) Journal of Residuals Science & Technology, Vol. 13, No. 7,

8 [11] R.S. Burt, Structural holes: the social structure of sompetition. Harvard University Press: Cambridge, MA (1992) [12] D. Krackhardt, MRQAP: Analytic versus permutatioin solutions. Working paper, Carnegie Mellon University (1993) [13] I. Nonaka, H. Takeuchi, The knowledge-creating company NY: Oxford University Press, Inc (1995) [14] C.S. Weng, The exploration and exploitation of technologies in between of the core and the periphery-insurance business method patents, Journal of Management, Vol.28, No.3, (2011) [15] H.J.M. Kurvers Ralf, M.A.P. Adamczyk Vena, H.S. Kraus Robert, Hoffman Joseph I, Wieren Sipke E van, Jeugd Henk P van der, Amos William, Prins Herbert HT, Jonker Rudy M, Contrasting context dependence of familiarity and kinship in animal social networks, Animal Behaviour Vol.86, (2013) Journal of Residuals Science & Technology, Vol. 13, No. 7,