Li, L., and Buhalis, D., 2008, Influential Factors of Internet Users Booking Online in China's Domestic Tourism, China Tourism Research, Vol.4(2), pp..172 - 188.

Journal of China Tourism Research, 4:172–188, 2008 Copyright # 2008 The Haworth Press ISSN: 1938-8160 print / 1937-8179 online DOI: 10.1080/19388160802313761 Influential Factors of Internet Users Booking Online in China’s Domestic Tourism 使用互联网在线预订的影响因素—以中国国内旅游 为例 LI LI DIMITRIOS BUHALIS Based on empirical data, this article reports factors that influence Chinese Internet users’ purchasing behaviors in the tourism industry. The type of travel website most visited, self-efficacy, domain-specific innovativeness, and perception of the Internet are found to be the significant predictors of Chinese eCustomers. The study affirms the importance of reasoned actions and planned behavior theories and the theory of innovation diffusion in predicting customers’ purchasing behaviors. Managerial implications for Chinese tourism companies are discussed. KEYWORDS. eCommerce, internet, prediction, eCustomer, China 本文以实证数据，报告影响中国旅游业中互联网使用者购买行为的因素。研究 发现，最常浏览的旅游网站类型、自我效能、特定网域的创新以及对互联网的 感知是预测中国电子消费者的重要自变量。本文证实理性行动理论、计划行为 理论和创新扩散理论对预测消费者的购买行为非常重要。最后本文讨论了中国 旅游企业的管理问题。 关键词: 电子商务, 互联网, 预测, 电子消费者, 中国 Introduction The Internet-based technologies have triggered enormous revolutionary changes in the business world, and eCommerce has become increasingly important. As one of the key sectors of eCommerce, eTourism is developing rapidly (Buhalis, 2003; Marcussen, 2004; Sheldon, 1997). Understanding Internet travellers and their adoption of eShopping, thus draws the attention of marketers and researchers. China, an emerging cyber-market, possesses huge potentials of development in eTourism due to its large population and economical burgeoning. However, research Li Li is a Research Student of Faculty of Management and Law at the University of Surrey, Guildford, Surrey GU2 7XH, UK (E-mail: li.li@surrey.ac.uk). Dimitrios Buhalis is Established Chair in Tourism and Deputy Director of the International Centre for Tourism and Hospitality Research at the School of Services Management at Bournemouth University, Talbot Campus, Fern Barrow, Poole, Dorset, BH12 5BB, UK (E-mail: dbuhalis@bournemouth.ac.uk). 172 Journal of China Tourism Research 173 on Chinese eCustomers and their adoption of eTourism products seems to be very limited (Ma, Buhalis, & Song, 2003). This article outlines the current development of eCommerce in China’s tourism market, and explores theories and predictive studies that explain eShopping adoption. It examines Chinese internet users adopting the internet for purchasing travel products by looking into seven independent variable sets, which are socio-demographics, travel-related behaviors, internet usage patterns, perception of the internet, customer domain-specific innovativeness (DSI) and selfefficacy (Anckar & Walden, 2000; Citrin, Sprott, Silverman, & Stem, 2000; Chau, Cole, Massey, Montoya-Weiss, & O’Keefe, 2002; Christou & Kassianidis, 2002; Morrison, Jing, O’Leary, & Cai, 2001; Vijayasarathy, 2004). Guanxi is also examined as it affects Chinese consumers’ buying behavior (Efendioglu & Yip, 2004; Luk, Fullgrabe, & Li, 1999; Merrilees & Miller, 1999). This article aims to identify what determines Chinese eCustomers booking travel online. Findings of an online survey are reported and discussed. eCommerce in China’s Tourism China has been established as one of the fastest growing tourism markets due to its sizable market, continuous economic growth, and increase of educational level (Chow, 1988; Chu, 1994; Li, 2000; National Bureau of Statistics of the People’s Republic of China, 2004; Xiao, 1997). While China’s tourism greatly contributes to the country’s modernization process (Choy, 1984; Oudiette, 1990; Richter, 1983; Sofield & Li, 1998), it is expected to have a significant impact on the international tourism. The World Tourism Organization estimates that by the year 2020, China will be the number one tourist destination and number four international tourists generating country United Nations (World Tourism Organization, 2000). Chinese have now become the main force of internet users in Asia. In contrast to 620,000 Internet users in 1997, there are over 137 million in mainland China in 2006 (China Internet Network Information Centre [CNNIC], 1998, 2006). This online population represents only a small proportion of the total inhabitants in China, demonstrating the potential for growth. The large online customer base provides an opportunity for eCommerce. According to CNNIC (2004), over 40% of Chinese internet users have purchased goods and services through online websites in 2003; around 10% of these cyber-buyers have purchased tourism products. Offline confirmation for online bookings is reported as the common practice in China’s eCommerce (Zhang, 2002). The net has become an important information source with its rich travel information and diverse tourist products in this country. The advent of independent travel and self-driving holidays in this country has further boosted the demand for use of the net for information search and/or travel bookings (Zhang, 2002). Theories of Innovation Adoption and Eshopping Intention-based theories, such as the theory of reasoned actions (TRA) (Fishbein & Ajzen, 1975) and the theory of planned behavior (TPB) (Ajzen & Madden, 1986), are valuable tools in explaining the adoption of eShopping because they are empirically validated. According to TRA, an individual’s behavior is determined by his/her behavioral intention, which is governed by individual attitudes towards performing the behavior and subjective norms (Fishbein & Ajzen) (see Figure 1). Building upon 174 Li Li and Dimitrios Buhalis Figure 1. Theory of reasoned actions. Source: Adopted from Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior. London: Addison-Wesley. this theory, planned behavior theory introduces perceived behavioral control as a determinant of behavior. Both TRA and TPB emphasize the belief-attitudeintention-behavior relationship. They are widely used to predict or explain cognitive and affective behavior in social psychology (Shih, 2004). Many predictive studies are based on these theories, providing an insight into the adoption of eShopping for the European and American markets (Citrin et al., 2000; Davis, 1986; Szajna, 1996; Vijayasarathy, 2004). The intension-based theories postulate that one’s attitude is based on the set of beliefs about the most important consequences of engaging in the behavior. The consequences are typically represented as product attributes or characteristics (Fishbein & Ajzen, 1975). Rogers’ (1995) innovation diffusion theory is found to be useful in explaining what attributes of an innovation, such as eShopping, influence the rate of its adoption. The theory suggests that five perceived characteristics of an innovation and other factors result in a favorable or unfavorable attitude toward the innovation, which leads to a decision of adopting or refusing it (Rogers, 1995). Among these characteristics, perceived relative advantage, trialability, and observability have a positive relationship with the rate of adoption whilst perceived compatibility and complexity have a negative association with the rate. That is, a user’s attitude toward the internet as an information and reservation tool for travel bookings is affected by its characteristics as perceived by the user. It is reported that customers’ perception of the internet has a direct relationship with their intention to book online (Christou & Kassianidis, 2002). This notion is consistent with many other intention-based studies (Gefen & Straub, 1997; Shih, 2004; Straub, 1994). Furthermore, perceived relative advantage, compatibility, complexity, and observability/communicability are found to influence eCustomers’ attitude toward eShopping prior to booking online (Christou & Kassianidis, 2002). According to TPB, perceived behavioral control is governed by control beliefs, which relate to the perceptions of the availability of skills, resources and opportunities, and perceived facilitation (Ajzen & Madden, 1986; Mathieson, 1991). The probability of an individual engaging in a specific behavior is a function of the person’s belief about his/her capability to perform the behavior (Anckar & Walden, 2000). The notion of perceived behavior control can also be expressed by using the concept of self-efficacy. In the research field of eCustomers’ buying behaviors, Vijayasarathy (2004, p. 5) defines self-efficacy as ‘‘a consumer’s Journal of China Tourism Research 175 self-assessment of his/her capabilities to shop on-line’’ and reports that it has a positive relationship with the intention to shop travel services online. In addition to prestated theories, relevant research has identified other influential factors of customers’ buying behavior in eCommerce. For instance, Citrin et al. (2000, p. 298) cite that it is ‘‘important to consider consumer’s domainspecific innovativeness when trying to understand and predict a consumer’s propensity to adopt the internet for shopping in relationship to their prior Internet usage.’’ Open-processing innovativeness does not influence the use of the internet for commerce (Citrin et al., 2000; Park & Jun, 2003). They report that a potential eCustomer is likely to be an experienced internet user, and be innovative within the domain of the internet and web. A number of cultural and economic factors determine the level of eCommerce in different regions (Chau et al., 2002; Christou & Kassianidis, 2002; Park & Jun, 2003). Unlike most of predictive models that tend to disregard cultural difference in eShopping adoption, Park and Jun (2003) observe that Korean internet users show higher perceived risks on privacy, security, and product than American users, but still purchase goods online frequently. To explain this, they employ the cushion effect suggested by Hsee and Weber (1999). It is claimed that in collectivist cultures, like China and Korea, family and other members will offer help when anyone in the group suffers losses after selecting a risky option. By contrast, people in individualist culture, such as in northern Europe and the U.S., are expected to bear the consequences of their own decisions. Guanxi is an aspect of Chinese cultural values, shaping Chinese consumers’ buying behaviors. It is described as ‘‘a network of relationships embedded with mutual obligations through a self-conscious manipulation of ‘‘face,’’ ‘‘Renqing’’ (favor) and related symbols’’ (Wong & Tam, 2000, p. 58). ‘‘Renqing’’ is a set of social norms, meaning that when accepting a favor, the recipient owes Renqing to the person and is obliged to pay back the debt of gratitude to him/her; ‘‘Ganqing’’ (friendship) implies expectations and obligations of getting/granting favorable responses from/to one’s friends (Luk et al., 1999). Merrilees and Miller (1999) and Luk et al. (1999) disclose that Renqing, Ganqing and face can explain the acceptance of direct selling by the Chinese. Moreover, Chinese eRetailors use customers’ ‘‘moral obligation to return a favor,’’ which reflects the cultural characteristics of Guanxi, to facilitate their online sales (Efendioglu & Yip, 2004). The usage pattern of an internet user is also vital to predicting eShopping (Chau et al., 2002). The more frequently customers use the net, the more likely they are to become eShoppers (Morrison et al., 2001; Sexton, Johnson, & Hignite, 2002). It is observed that bookers spend more time online than those booking offline (Weber & Roehl, 1999). American travellers search online travel services more often than principals’ websites to compare travel prices (Morrison et al., 2001). There are also travel-related factors that affect internet travellers’ buying behavior. It is reported that American internet users who travel longer distances tend to be more affluent and better educated, and are apt to spend more for travel services, compared with non-Internet users who travel shorter distances to the same destination regions (Furr, Bonn, & Hausman, 2002). The findings are roughly consistent to those recorded by Bonn, Furr and Susskind (1999). Frequency of travel and the number of trips yearly are revealed to be important indicators of lookers’ likelihood of being bookers (Morrison et al., 2001). In addition to this, membership of Frequent Flyer programs can partially predict lookers’ probability of booking online. 176 Li Li and Dimitrios Buhalis Research Design The primary research adopts the conceptual model presented in Morrison et al.’s study (2001). As the model is tourism-focused, it provides a clear blueprint of the development of internet travellers. Internet travel lookers (hereafter referred as to ‘‘lookers’’) are those who retrieve information about travel products via the net. When lookers purchase tourism products online, they become internet travel bookers (hereafter referred as to ‘‘bookers’’). Hypotheses A research model is proposed as illustrated in Figure 2. It takes into account perception of the internet, DSI, self-efficacy, Guanxi, socio-demographics, internet usage patterns, and travel-related variables. Seven hypotheses were developed as stated below: H1: There is a relationship between DSI and lookers’ likelihood of booking travel online. H2: There is a relationship between perception of the internet and lookers’ likelihood of booking travel online. H3: There is a relationship between self-efficacy and lookers’ likelihood of booking travel online. H4: There is a relationship between Guanxi and lookers’ likelihood of booking travel online. H5: There is a relationship between socio-demographics and lookers’ likelihood of booking travel online. H6: There is a relationship between internet usage patterns and lookers’ likelihood of booking travel online. H7: There is a relationship between travel-related variables and lookers’ likelihood of booking travel online. Questionnaire Development Table 1 details all the variables, of which DSI, perception of the internet, selfefficacy, and Guanxi were measured on a seven-point Likert scale to capture more variances. In this research, DSI is defined as customers’ innovation within the Figure 2. Projected research model. Journal of China Tourism Research Table 1. Variable Inventory. 177 Socio-demographics Gender, Age, Marital Status, Area of Residence, Highest Educational Attainment, Household Monthly Income, Occupation, Household size (HSH). Internet Usage Pattern Personal payment for Internet service, Weekly Internet usage, Length of time using the Internet (LUS), Commercial travel website visited most often (WEB), Frequency of using the Internet for travel information a year (FRE). Travel-Related Variables Membership of frequent flyer program (FFP), Number of domestic trips taken last year, Number of international outbound trips taken last year. Customer Domain-Specific Innovativeness DSI1: I am among the first in my circle of friends to visit a new travel-related website. DSI2: If I heard that a new travel website was available on the Internet, I would be interested enough to shop from it. DSI3: Compared to my friends, I seek out relatively more information over the Web. DSI4: I am the first in my circle of friends to know of any new travel websites. DSI5: I will visit a new tourism company’s website even if I have not heard of it before. Guanxi RQG: Sometimes, I have to purchase tour package from the person/party whom I have owned Renqing to even though the same tour or a similar one can be bought from a travel website. GQG: When I intent to purchase a product presented on a travel website, I would ask a friend of mine who may be able to get a discounted rate to book it for me. Self Efficacy EFF1: I am proficient in using the Internet for shopping travel product/services. EFF2: I feel confident that I can use the Internet for shopping travel product/ services. Perceptions of the Internet Perceived Relative Advantage PRA1: I don’t feel safe to use credit card online. PRA2: Generally travel websites offer tourism products at cheap prices. PRA3: I would buy discounted travel products online. PRA4: I would book travel online more often if incentives, such as frequent flyer miles or points, are provided. Perceived Complexity PCX1: I feel I am not clear with online reservation procedure. PCX2: I feel it is not easy to book travel online. Perceived Compatibility CMP1: Using the Internet to shop travel product/services is compatible with the way I like to shop. CMP2: Using the Internet to shop travel product/services fits with my lifestyle. CMP3: I am used to travel agents or toll-free numbers. CMP4: I intent to book travel online more often. Communicability CMM1: I have heard about people booking travel online many times. CMM2: Many friends have booked travel online. 178 Li Li and Dimitrios Buhalis domain of travel websites. Building upon Citrin’s et al. (2000) and Park and Jun’s (2003) works, five statements were developed to measure this concept (see Table 1). In this research perceived relative advantage, compatibility, complexity, and observability/communicability are defined as follows: 1. 2. 3. 4. Perceived relative advantage – the degree to which consumers perceive eShopping to ‘‘be superior to in-store traditional travel agent shopping’’; Perceived compatibility – the extent to which a consumer believes that shopping tourist products matches with his/her lifestyle, needs, and shopping preference; Perceived complexity – the degree to which consumers feel that eShopping tourist products is difficult to understand and use in practice; Perceived communicability – the degree to which the benefits of eShopping travel products are observable or describable to others. According to Rogers’ (1995) innovation diffusion theory, perception of relative advantage, and compatibility of electronic travel shopping with lifestyle should positively associate with customer’s intention to book travel online, whereas increasing complexity should reduce the intention. That is, the easier the booking online is for individuals, the more likely they are to perform the behavior. Adapting from Christou and Kassianidis’ work (2002), 12 statements were employed to tap participants’ perception of the internet as listed in Table 1. Anckar and Walden (2000) identify that an internet user’s self-efficacy affect his/ her attitude toward self-booking of travel service online. In this research, selfefficacy is regarded as an internet traveller’s self-assessment of the capabilities to perform eShopping. Adapting Vijayasarathy’s (2004) measurement two statements are developed to measure the concept in this study (see Table 1). Recognizing that there are limited empirical studies on the effects of Guanxi in a business-to-customer context, this research makes an effort in this domain by examining the relationships of Renqing and Ganqing with Chinese internet users’ intention to shop for tourist products through the websites. The statement used to measure Renqing is ‘‘Sometimes, I have to purchase tour package from the person/ party whom I have owned Renqing to, even though the same tour or a similar one can be bought from a travel website.’’ Ganqing is measured by asking respondents’ view on this statement: ‘‘When I intend to purchase a product presented on a travel website, I would ask a friend of mine who may be able to get a discounted rate to book it for me.’’ In China, urbanization is densely located on the east coast and in the southeast region, where three major urban belts are based. These are the Beijing and Tianjin Belt, the Yangtze River Delta covering Shanghai, and the Pearl River Delta surrounding the Guangdong Province. Economy and tourism are better developed in these areas; hence potential Chinese tourists are often generated here (Scandinavian Tourism Board, 2002; Wang & Ap, 2003; Xiao, 2003). Within the country, Beijing, Shanghai, Tianjin, and Guangdong province represent relatively higher internet penetration rates, at 28%, 26.6%, 14.4%, and 12%, respectively, in 2003 (CNNIC, 2004). The majority of internet users in China are in their 20s and 30s, and have secondary or above educational qualifications (CNNIC, 2004). Thus, it is expected that the majority of Chinese internet travel bookers are young, well-educated, and live in the three urbanization regions. A wired lifestyle can be measured by months/years of internet experience, what is being bought, and why (Chau et al., 2002). Internet usage has three dimensions, Journal of China Tourism Research 179 namely, frequency of internet usage, amount of daily internet usage, and diversity of internet usage (Igbaria, Schiffman, & Wieckowshi, 1994). Travel-related variables include membership of a frequent flyer program, number of domestic trips, and number of international trips over the 12 months. The first section of the questionnaire examined respondents’ socio-demographic characteristics. The second part was to capture respondents’ ‘‘wired lifestyle.’’ This was followed by the statements for examining DSI, self-efficacy, Guanxi, and perception of the internet. The final section aimed to measure respondents’ travel-related behaviors. All the questions were created in English and then translated into Chinese. A pilot study was conducted to ensure the accuracy of translation. A group of 12 Chinese students was invited to give feedback on the wording and translation of the questions. Since these individuals represented a different market segment and did not belong to the research sample, they were not asked to answer the questions. Two minor amendments on Chinese translation were then made accordingly. Sampling and Data Collection The population of the study was the mainland Chinese nationals who use the internet for travel-related activities, including lookers and bookers. Accessing individual members of the public in China is difficult for a variety of reasons. In order to reach this particular segment effectively, several trusted travel companies with eCommerce capabilities (e.g., Ctrip.com, eLong.com, and China Southern Airline) and two internet portals (Sina.com and 163.com) in China were contacted. eLong.com, a leading online travel company in China, agreed to e-mail the online questionnaire to its registered customers who were randomly selected from its database. The company made effective contacts to Chinese travel lookers across the country for this research. Because, when the recipients opened the e-mail they were taken to the online survey page directly without having to click a hyperlink to the questionnaire URL. Initially, it was not clear what response rate of an online survey in China would be. Hence, a sample of 103,000 registered customers was selected. The online survey was carried out in June 2004 and achieved 872 replies. The questionnaire in Chinese was e-mailed to the sampled individuals. When opening the e-mail, the recipients were prompted to a filter question—‘‘Have you ever used the internet for searching travel information?’’ If a subject gave a negative answer to this question, s/he would be excluded from the sample when analyzing data. Respondents were then asked ‘‘Have you ever booked travel service online?’’ This question allowed obtaining the values of dependent variable (DV) by distinguishing bookers and lookers-only among the lookers. Data Analysis Method To ensure the consistency of the instrument for measuring DSI, self-efficacy, and perception of the internet, reliability tests were performed. A factor analysis on the measurement of perception of the internet was carried out to produce a modified instrument that would measure the concept effectively with fewer items. This analysis identified underlying dimensions of the concept, thus the validity of this measurement was established. The use of the Varimax rotation method in the factor analysis produced unrelated factors, so that a collinearity problem in the logistic 180 Li Li and Dimitrios Buhalis regression analysis was overcome (Field, 2000). A logistic regression analysis was used to predict a binary DV from independent variables (IDV). This technique overcame the problem of violating the assumption of linearity between variables by expressing the ‘‘multiple linear regression equation in logarithmic terms’’ (Field 2000, p. 164). It was not realistic to assume the IDV were multivariate normal as most of the variables in this study were dichotomous, or with restricted ranges, meaning discriminate analysis would not be suitable for the purpose of this research. Methodological Limitations A model generated from logistic regression analysis always fits better to a particular sample than to the population (Norusis, 1993). Also, given the large online ˇ population in China, the sampling method and low response rate (about 1%) reduces the ability to generalize findings. The limited generalizability may also apply to other markets, as Chinese market is governed by unique socio-cultural and economic characteristics. Findings Out of the 872 questionnaires received, 644 were usable while another 10 were excluded because respondents claimed themselves nonlookers. Among the 634 lookers, 563 respondents already booked travel online (e.g., bookers) and 71 were lookers-only. Over 55% of the respondents were single and more than 65% had university qualifications. Over 60% of the subjects were aged between 21 and 30. It appears that the sample is bookers-biased, and mainly reflects the market niche formed by the white-collar Chinese graduates of the young generation. Reliability of Measurements As summarized in Table 2, while the Alpha for the DSI scale is just over the required minimum value of 0.7 (SPSS FAQ, 2004), the other two scales have demonstrated very Table 2. Reliability of Scales: Coefficient Alpha. Concept Self-Efficacy Perception of Internet Cronbach’s Alpha 0.9155 0.8176 EFF1: 0.8455 PRA1: 0.2065 PRA2: 0.4190 PRA3: 0.5990 DSI1: 0.6170 DSI2: 0.5619 Corrected Item-Total Correlation EFF2: 0.8455 PRA4: 0.5316 PCX1: 0.3689 PCX2: 0.3890 DSI3: 0.5315 DSI4: 0.7064 DomainSpecific Innovativeness 0.7780 CMP1: 0.6530 CMP2: 0.6229 CMP3: 0.3384 DSI5: 0.3733 CMP4: 0.5813 CMM1: 0.5075 CMM2: 0.4282 Journal of China Tourism Research 181 good reliabilities. Furthermore, statistics in the Corrected Item-Total Correlation column indicate the contribution of each item to the corresponding measurement. Item PRA1 (‘‘I don’t feel safe to use credit card online.’’) has the lowest value and should, in theory, be eliminated from the scale. However, concern over using a credit card online is one of the core issues in eCommerce that affects users’ attitude towards eShopping. For this reason this item is retained within the scale. Identified Dimensions of ‘‘Perception of the Internet’’ The calculated Kaiser-Meyer-Olkin statistic of greater than 0.74 and highly significant p value in Bartlett’s test (p , 0.001) confirms that factor analysis is appropriate for these data. Disregarding variables with an eigenvalue . 1, a screen plot reveals four components/factors named compatibility, financial advantage, complexity, and communicability. Setting the cutting value at 0.5, 10 items remained in the modified scale as outlined in Table 3. The four-factor solution explains a total of 67.51% of the variance with each factor contributing over 15%. The components show a number of strong loadings, and all the variables loaded substantially on only one factor. The results of this analysis support the use of the compatibility, financial advantage, complexity, and communicability items as separate scales. Also, the interpretation of the four factors is consistent with the perceived attributes of an innovation stood in IDT (Rogers, 1995). Logistic Regression Results – Selected Predictive Model Based on a computed variance inflation factor and tolerance values for all the IDVs, there is not a serious collinearity problem in the research data. Originally, 35 variables were entered into the analysis. During each of the total 22 steps, the variable that had the least significant contribution to the model was eliminated. Table 3. The Modified Perception of the Internet Scale: Factor Analysis. Factor Loading Factor 1 CMP2 Compatibility CMP1 Factor 2 Complexity PCX1 PCX2 Factor 3 CMM2 Communicability CMM1 Factor 4 Financial PRA1 Advantage PRA3 PRA2 PRA4 Eigenvalue 2.473 % of variance 20.61 explained 0.899 0.872 0.898 0.898 0.916 0.898 0.660 0.634 0.630 0.623 1.905 15.87 1.898 15.82 1.883 15.69 67.99a Communalities 0.847 0.809 0.808 0.815 0.878 0.871 0.519 0.621 0.474 0.612 Note: Extraction method - principal component analysis; rotation method - Varimax with Kaiser normalization; items loadings less than 0.5 were omitted; rotation converged in 5 iterations. acumulative % of variance explained. 182 Li Li and Dimitrios Buhalis Table 4 presents all the variables that are retained in the final model. It is evident that all the variable sets have some predictive powers in explaining the probability of lookers booking travel online, with the exemption of Guanxi. Hence, H1, H2, H3, H5, H6, and H7 are accepted. Guanxi does not seem to predict Chinese internet users adopting eShopping for tourist products. Perhaps this is caused by the low level of personal interaction between customers and eRetailors in online transactions. The increasing information transparency on the internet may also attribute to the reduced dependence on personal networks. Users’ self-efficacy, DSI, perception of the internet, and the type of travel website most visited are the significant predictors with p values of less than 0.05. When setting the confidence interval at 90%, all the variables in the equation are significant in explaining the probability, with the exemption of ‘‘frequency of using the Internet for travel information a year’’ (FRE). Placing computed bs into the logistic regression equation, the predictive model can now be presented as: Á À ð1Þ PðYÞ~1 1ze{Z where Z 5 2 0.90478 2 0.7648HSH + 0.60851FFP 2 0.99056LUS + 0.004453FRE 2 0.9789WEB + 1.241151EFF1 2 1.19874EFF2 + 0.521376DSI2 2 0.28904DSI3 2 0.64678DSI5 + 0.351525PRA1 + 0.43865PRA3 2 0.58407PCX1 + 1.172533CMP2 P(Y) is the probability of lookers becoming bookers. e is the base of natural logarithms. Self-Efficacy. Lookers’ self-efficacy is positively associated with the likelihood of booking travel online (the magnitude of change in the odds associated with ‘‘I am proficient in using the Internet for shopping travel products’’ (EFF1) is greater than Table 4. Selected Predictive Model. Variables in the Equation WEB EFF1 EFF2 DSI2 DSI5 PRA1 PRA3 PCX1 CMP2 HSH FFP LUS FRE DSI3 Constant b 20.9789 1.241151 21.19874 0.521376 20.64678 0.351525 0.43865 20.58407 1.172533 20.7648 0.60851 20.99056 0.004453 20.28904 20.90478 Sig. 0.008 0.000 0.000 0.003 0.000 0.012 0.017 0.000 0.000 0.050 0.080 0.060 0.117 0.081 0.539 Exp (b) 0.38 3.46 0.30 1.68 0.52 1.42 1.55 0.56 3.23 0.47 1.84 0.37 1.00 0.75 0.539 Notes: backward likelihood ratio method. Journal of China Tourism Research 183 that related to ‘‘I feel confident that I can use the Internet for shopping travel products’’ (EFF2)), holding all the independent variables constant. When EFF1 increases by 1 unit on the 7-point Likert scale, the odds of the person booking travel online are 3.5 times higher than before. As Exp(b) for EFF2 is 0.3, one unit increase in EFF2 leads to a decrease in the odds. Therefore, the higher a looker’s selfassessment of his/her capabilities to purchase travel online, the more likely the user will book online. Domain-Specific Innovativeness. Holding other variables constant, one unit increase in ‘‘If I heard that a new travel website was available on the internet, I would be interested enough to shop from it’’ (DSI2) results in increased odds by 1.68 times. This means that the more interested an internet user is in shopping from a new travel website, the more likely the person is to become a booker. ‘‘I will visit a new tourism company’s website even if I have not heard of it before’’ (DSI5) has a negative effect on the likelihood of eShopping, which is understandable as competition on the net is just few clicks away. A website’s customer can easily be redirected to another site through a hyperlink (e.g., customer leakage), which consequently lowers the possibility of online transaction fulfil/ment on the original site. In spite of the Exp(b) values for ‘‘Compared with my friends, I seek out relatively more information over the web’’ (DSI3) and DSI5 being less than 1, DSI2 has a relatively stronger influence on the odds of an event occurring (Sig. 5 0.003, Exp(b) 5 1.19). Controlling for other variables, domain-specific innovativeness appears to have an overall positive influence on the odds, which confirms Citrin’s et al. (2000) proposition that a potential eCustomer is likely to be innovative within the domain of the internet and web. Perception of the Internet. ‘‘Using the Internet to shop travel products fits with my lifestyle’’ (CMP2) represents a relatively strong and positive predictive power in the selected model (Sig. , 0.001, Exp(b) 5 3.23). When a user’s perception of compatibility of shopping online with his/her lifestyle raises by 1 point, the odds of him/her eShopping travel increase by 3.2 times, holding other variables constant. This positive association is consistent with the notion of IDT—the more compatible with the values and norms of a social system the innovation is, the more rapid it will be adopted. Controlling for other variables, when ‘‘I feel I am not clear with online reservation procedure’’ (PCX1) increases by one unit, a 0.58 decrease in the log-odds of booking travel online is expected to occur. This indicates that a less complicated online booking procedure will increase internet users’ likelihood of eShopping, echoing that a simple innovation will be adopted faster than a complicated one (Rogers, 1995). Increases in PRA3 (‘‘I would buy discounted travel products online’’) lead to increased odds of booking travel online, which suggests that Chinese lookers are price-conscious and may be attracted by discounted products. Similarly, increases in PRA1 (‘‘I don’t feel safe to use credit card online’’) result in increased odds. This implies that the participants purchase tourism products despite their concern over use of credit cards online, which reflects the presence of cushion effect. Internet Usage Patterns. Given Exp(b) value for ‘‘commercial travel website visited most often’’ (WEB) is under 1, the odds of booking travel online decrease when the user visits the websites of an online travel agency and others most often. People who visit tourism product suppliers’ websites most often are more likely to become bookers. 184 Goodness-of-Fit Li Li and Dimitrios Buhalis The classification table is created to justify the model’s fitness (see Table 5). With 98 out of 100 times, the model correctly produces the probability of a user booking online. When predicting those who do not book travel service online, the model represents a less desirable correct prediction rate. Overall, it performs well in predicting the probability with a correct prediction rate of 91.5%. Generalizability When the confidence interval is set at 95%, the model represents a good generalizability theoretically. The values of lower and upper Exp(b)s for respective significant predictors are both over 1, or both under 1. Taking the Exp(b) values for ‘‘I am proficient in using the Internet for shopping travel product’’ (EFF1) as an example, as the confidence interval for this value ranges from 2.06 to 5.8, it can be confidently said that the value of Exp(b) in the population lies somewhere between these two figures. Thus, statistically, EFF1’s positive relationship with the odds of lookers booking travel online in this sample is true of the population of Chinese lookers. However, there is a 5% chance that the odds of the event occurring may fall outside the value range. Discussion and Managerial Implications Although the research findings are, in general, consistent with those of other similar studies, caution is are required when interpreting and generalizing them given that the selected sample in this study has generated data that may be biased. Nonetheless, the study does confirm the visions of IDT (Rogers, 1995) and TRA (Fishbein & Ajzen, 1975). Referring to Equation (1), the function of users’ perception of the internet in the model has reaffirmed the significance of this concept in explaining innovation adoption. The intention of a potential Chinese eCustomer purchasing tourist product online is partially determined by his/her attitude towards eShopping, which is based on the set of beliefs about the most important consequences of the behavior. ‘‘I do not feel safe using a credit card online’’ (PRA1), ‘‘I would buy discount travel products online’’ (PRA3), ‘‘I feel I am not clear with online reservation procedure’’ (PCX1), and ‘‘Using the internet to shop for travel products fits with my lifestyle’’ (CMP2) reflect the beliefs of sampled Chinese lookers about shopping travel online. Relating to the factor analysis results, PRA1 and PRA3, Table 5. Classification Table. Predicted Have you ever booked travel service online? Observed Have you ever booked travel service online? Overall percentage Note: The cut value is .500. % Correct 31.58 98.56 91.54 no yes no 18 7 yes 39 480 Journal of China Tourism Research 185 PCX1, and CMP2 represent a financial advantage, perceived complexity, and compatibility of eShopping travel products, respectively. It informs us that the attributes of the internet/web as a reservation tool contribute to the prediction of how rapidly the adoption of innovative use of the media is. The second most important predictor in the model is the internet user’s self-efficacy. The concept acts for a positive predictive power in the equation, meaning that it has a positive association with the likelihood of booking online. This echoes the postulation of TPB (Ajzen & Madden, 1986), and is consistent with the findings reported in Vijaysarathy’s (2004) and Anckar and Walden’s (2000) studies. The research reveals that the respondents’ innovativeness within the domain of travel websites partially explains the probability of booking travel online. The functionality of domainspecific innovativeness in the equation confirms Citrin’s et al. (2000) notion. The higher a user’s DSI is, the more likely will the person adopt eShopping, controlling for other variables. In addition to the theoretical inputs, this research exercise also informs a number of practical managerial implications in spite of embedded limitations stated previously. The findings suggest that Chinese eCustomers’ self-efficacy, innovativeness within the domain of travel websites, and their attitudes toward purchasing travel through the Internet are critical determinants of lookers becoming bookers. Hence, to effectively increase online sales revenues, China’s tourism firms may wish to seek various initiatives to help their customers build up their capabilities and confidence in using online booking engines. As what many eCommerce frontiers in western economies have done, Chinese entrepreneurs can introduce visual demonstrations on their websites to educate users how to make an online booking and, thus, improve lookers’ proficiency in using the internet for eShopping. The demos will surely need to be catered to the Chinese who are inclined to distinctive colors, such as red, and animations. As Chinese consumers are generally price-conscious, discounted tourist products in the cyber-market should form a driver of Chinese internet users booking online. However, this may lead to a price war and jeopardize companies’ profitability, and may also cause distribution channel conflict on prices. To overcome these problems, Chinese companies will need to differentiate their offerings by providing integrated and value-added services, for instance, through strategic partnerships facilitated and supported by advanced internet-based technologies. The most challenging situation is probably that Chinese tourism companies need to react proactively and reactively to the fast-changing eBusiness environment, and be innovative and creative in managing business tactically and strategically. However, at the moment, many Chinese entrepreneurs do not appear to appreciate the good practices that eBusinesses in Europe and the U.S. have been exercising. Unless skilled personnel with future thinking joins the management team of China’s tourism eBusinesses, China’s eTourism will not be able to expand and take full advantage of the opportunities. Conclusion Taking into account seven independent variable sets, this article identified and quantified the determinants of Chinese eCustomers purchasing tourist products online. It has confirmed the theories and previous studies of this type, and provided in-depth information about sampled lookers. It has also disclosed the function of 186 Li Li and Dimitrios Buhalis cushion effect in China’s eCommerce. The type of travel website most visited, selfefficacy, DSI, and perception of the internet were found to be the significant predictors. The findings initiate encouraging solutions, which can be achieved by utilizing internet-based applications. While recognizing the problems in China’s tourism industry, the potential of business development in the cyber-market are tremendous for the Chinese tourism firms. The greatest strength of the work is reflected by the quantification of effects of individual identified predictors. Theoretical frameworks that are developed in the West can be applied to the Chinese market to some extent. However, to fully understand a market with unique characteristics like China, researchers need to be adaptive and take into account socio-cultural, regional, and political differences. Future research should develop a comprehensive measurement for Guanxi online, so that its contribution to eShopping in China can be fully understood. References Ajzen, I., & Madden, T. (1986). Prediction of goal-directed behavior: Attitudes, intentions and perceived behavioral Control. Journal of Experimental Social Psychology, 22(4), 453–474. Anckar, B., & Walden, P. (2000). Destination Maui? An exploratory assessment of the efficacy of self-booking in travel. Electronic Markets, 10(2), 110–119. Bonn, M., Furr, H., & Susskind, A. (1999). Predicting a behavioral profile for pleasure travelers on the basis of Internet use segmentation. Journal of Travel Research, 37(4), 333–340. Buhalis, D. (2003). eTourism: Information technology for strategic tourism management. Harlow: Financial Times/Prentice Hall. Chau, P., Cole, M., Massey, A., Montoya-Weiss, M., & O’Keefe, R. (2002). Cultural differences in the online behavior of consumers. Communications of the ACM, 45(10), 138–143. China Internet Network Information Centre. (1998). The 2nd survey report. Retrieved July 11, 2004, from www.cnnic.net.cn China Internet Network Information Centre. (2004). The 13th survey report. Retrieved July 1, 2004, from www.cnnic.net.cn China Internet Network Information Centre. (2006). The annual survey report. Retrieved July 11, 2007, from www.cnnic.net.cn Chow, W. (1988). Open policy and tourism between Guangdong and Hong Kong. Annals of Tourism Research, 15(2), 205–218. Choy, D. (1984). Comment on Richter’s political implications of Chinese tourism policy. Annals of Tourism Research, 11(4), 618–621. Christou, E., & Kassianidis, P. (2002). Consumer’s perceptions and adoption of online buying for travel products. Journal of Travel & Tourism Marketing, 12(4), 93–107. Chu, Y. (1994). Evaluation of sightseeing areas in China. Annals of Tourism Research, 21(4), 837–839. Citrin, A. V., Sprott, D. E., Silverman, S. N., & Stem, D. (2000). Adoption of Internet shopping: The role of consumer innovativeness. Industrial Management & Data Systems, 100(7), 294–300. Davis, F. (1986). A technology acceptance model for empirical testing new end-user information systems: Theory and results. Unpublished doctoral dissertation, Massachusetts Institute of Technology, Cambridge, MA. Efendioglu, A., & Yip, V. (2004). Chinese culture and e-commerce: An exploratory study. Interacting with Computers, 16(1), 45–62. Field, A. (2000). Discovering statistics using SPSS for Windows: Advanced techniques for Windows. London: Sage. Journal of China Tourism Research 187 Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior. London: AddisonWesley. Furr, H., Bonn, M., & Hausman, A. (2002). A generational and geographical analysis of Internet travel-service usage. Tourism Analysis, 6(2), 139–147. Gefen, D., & Straub, D. (1997). Gender differences in the perception and use of e-mail: An extension to the Technology Acceptance Model. MIS Quarterly, 21(4), 389–400. Hsee, C. K., & Weber, E. U. (1999). Cross-national differences in risk preference and lay predictions. Journal of Behavioral Decision Making, 12(2), 165–179. Igbaria, M., Schiffman, S., & Wieckowshi, T. (1994). The respective roles of perceived usefulness and perceived fun in the acceptance of microcomputer technology. Behavior and Information Technology, 13(6), 349–361. Li, L. (2000). Geographical consciousness and tourism experience. Annals of Tourism Research, 27(4), 863–883. Luk, S., Fullgrabe, L., & Li, S. (1999). Managing direct selling activities in China: A cultural explanation. Journal of Business Research, 45(3), 257–266. Ma, X., Buhalis, D., & Song, H. (2003). ICTs and Internet adoption in China’s tourism industry. International Journal of Information Management, 23(6), 451–467. Marcussen, C. (2004). Trends in European internet distribution - Of travel and tourism services. Retrieved June 24, 2004, from www.crt.dk Mathieson, K. (1991). Predicting user intentions: Comparing the technology acceptance model with the theory of planned behavior. Information Systems Research, 2(3), 173–191. Merrilees, B., & Miller, D. (1999). Direct selling in the West and East: The relative roles of product and relationship (Guanxi) drivers. Journal of Business Research, 45(3), 267–272. Morrison, A., Jing, S., O’Leary, J., & Cai, L. (2001). Predicting usage of the Internet for travel bookings: An exploratory study. Information Technology and Tourism, 4(1), 15–30. National Bureau of Statistics of the People’s Republic of China. (2004). Communique on major ´ figures of the 2000 population census (No. 2). Retrieved April 20, 2004, from http://www. stats.gov.cn Norusis, M. J. (1993). SPSS for Windows: Advanced statistics, release 6.0. Chicago, IL: SPSS ˇ Inc. Oudiette, V. (1990). International tourism in China. Annals of Tourism Research, 17(1), 123– 132. Park, C., & Jun, J. (2003). A cross-cultural comparison of Internet buying behavior: Effects of Internet usage, perceived risks, and innovativeness. International Marketing Review, 20 (5), 534–553. Richter, L. (1983). Political implications of Chinese tourism policy. Annals of Tourism Research, 10(3), 395–413. Rogers, E. M. (1995). Diffusion of innovations. New York: Free Press. Scandinavian Tourism Board. (2002). China outbound travel market report: Research report. Stockholm: Author. Sexton, R. S., Johnson, R. A., & Hignite, M. A. (2002). Predicting Internet/e-commerce use. Internet Research: Electronic Networking Applications & Policy, 112(5), 402–410. Sheldon, P. J. (1997). Tourism information technology. Oxford, UK: CAB International. Shih, H. (2004). An empirical study on predicting user acceptance of e-shopping on the Web. Information & Management, 41(3), 351–368. Sofield, T., & Li, F. (1998). Tourism development and cultural policies in China. Annals of Tourism Research, 25(2), 362–392. SPSS FAQ. (2004). UCLA: Academic Technology Services, Statistical Consulting Group. Retrieved August 15, 2004, from http://www.ats.ucla.edu/stat/spss/faq/alpha.html Straub, D. (1994). The effect of culture on IT diffusion: E-mail and fax in Japan and the U.S. Information Systems Research, 5(1), 23–47. Szajna, B. (1996). Empirical evaluation of the revised technology acceptance model. Management Science, 42(1), 85–92. 188 Li Li and Dimitrios Buhalis United Nations World Tourism Organization. (2000). Tourism 2020 vision – A new forecast. Madrid: Author. Vijayasarathy, L. (2004). Predicting consumer intentions to use on-line shopping: The case for an augmented technology acceptance model. Information & Management, 41(6), 747–762. Wang, S., & Ap, J. (2003). Tourism marketing in the People’s Republic of China. In A. Lew, L. Yu, J. Ap, & G. Zhang (Eds.), Tourism in China (pp. 217–235). New York: Haworth Hospitality Press. Weber, K., & Roehl, W. (1999). Profiling people searching for and purchasing travel products on the World Wide Web. Journal of Travel Research, 37(3), 291–298. Wong, Y., & Tam, J. (2000). Mapping relationships in China: Guanxi dynamic approach. Journal of Business and Industrial Marketing, 15(1), 57–70. Xiao, H. (1997). Tourism and leisure in China: A tale of two cities. Annals of Tourism Research, 24(2), 357–370. Xiao, H. (2003). Leisure in China. In A. Lew, L. Yu, J. Ap, & G. Zhang (Eds.), Tourism in China (pp. 263–275). New York: Haworth Hospitality Press. Zhang, L. (2002). China tourism Internet: An analysis of survival & future trends. In G. Zhang, X. Wei, & D. Liu (Eds.), Green book of China’s tourism 2001–2003 (pp. 249–265). Beijing: Social Sciences Documentation Publishing House.