1634 An Exploratory Study of the Design Preferences of U.S. and Chinese Virtual Communities Table 1. Web design features in each cultural dimension Adapted from Chau et al. (2002); Fink and Laupase (2000); Luna et al. (2002); Robbins and Stylianou (2001); Singh (2003) Cultural dimensions Design Features Examples Content Promotion Interactivity Navigation Others High context Design focus is the use of art Politeness and soft sell approach Links to consumer groups, suppliers, distributors; Hierarchical structure Asian Low context Design focus is informative Hard sell DSSURDFK%HQH¿WV highlighted, Search-based structures. Western Individualism Secure communication; Interaction supports individual effort Search engine; Avoidance of registration requirement U.S. Collectivilism Social communication opportunity; Clubs; Interaction supports group activity Links to local Websites China Strong uncertainty avoidance Guided interactions Use of cookies statements; Customer service & local stores; Free trials & downloads; Disclosure of use of cookies N/A Weak uncertainty avoidance Presence of a job list Interactions are free for choice N/A Masculinity Presence of ¿QDQFLDOUHSRUWV Hit or visitor counter; quizzes and games N/A Femininity Presence of social responsibility N/A Short term orientation Presence of corporate history FAQ, Site map/index pages, Search engine U.S. Long term orientation Heavy animated graphics/ video/audio H.K. 1635 An Exploratory Study of the Design Preferences of U.S. and Chinese Virtual Communities and interactivity categories were also employed in Benbunan-Fich’s (2001) study. Combining all these suggestions, categories employed in this study are content, promotion, interactivity, and QDYLJDWLRQDV PRVW RIWKH IHDWXUHV FDQ¿WLQWR these categories. There are also a number of studies that have examined the impact of culture on Web site design (Chau, Cole, Massey, Montoya-Weiss, & O’Keefe, 2002; Fink & Laupase, 2000; Luna, Peracchio, & Juan, 2002; Robbins & Stylianou, 2001; Singh, 2003). Table 1 presents a summary of the Web site design features required to support each of Hofstede’s (2001) cultural dimensions. Virtual Community Tools A variety of tools have been used in virtual com- munities. These tools are all computer-mediated. The software often used in virtual communities can be broadly grouped as asynchronous or syn- chronous tools. Synchronous tools are software supporting communication in real time (Preece, 2000); they can be chat room, audio broadcast and so forth. They are not in an on-going form like asynchronous tools, which do not require users to be available at the same time (Preece, 2000). A good asynchronous example is forum, which allows users to post and respond to posts Synchronous/ Asynchronous Tool Asynchronous Tools Synchronous Tools Web site (the basic is asynchronous, but also with synchronous software embedded frequently). Email Chats Bulletin boards Instant messaging Listserver MUDs Forums MOOs UseNet Virtual environments Virtual postcards and greeting cards Table 2. Synchronous and asynchronous tools adapted from Preece (2000) MUDs: Multiuser Dungeons MOOs: Object Oriented MUDs Table 3. Different tools employed in varied types of virtual communities Adapted from Chaudhury et al. (2001); Churchill and Bly (2000); Ruhleder (2002); Toomey et al. (1998) Communities Industry Workspace Business to Consumers Academic Toomey Churchill Chaudhury, Mallick, & Rao Ruhleder ³<RX´ ³7KH\´ ³,W´ text-based chat rooms, MOOs MUD, 2-D graphical MUD for long distance Email, chat sessions, virtual postcards and greeting card BBS, threaded discussion, forums web display of documents, multimedia contents Audio broadcast, forum, chat rooms, BBS 1636 An Exploratory Study of the Design Preferences of U.S. and Chinese Virtual Communities any time and continuously (Ruhleder, 2002). A Web site can employ both synchronous and asynchronous tools at the same time. Table 2 presents the different kinds of synchronous and asynchronous tools that can be employed in a Web-based environment. Chaudhury et al. (2001) suggested that a virtual community should be developed using differ- ent tools. Table 3 shows tools used in different types of communities. From the table, we can see that asynchronous tools are used for supporting QRQUHODWLRQVKLSW\SHVRIFRPPXQLWLHVLH³LW´ type communities use Web display); on the other hand, communities focused on relationship prefer V\QFKURQRXVWRROVLH³\RX´W\SHFRPPXQLWLHV use chat). In summary, the literature review highlighted three areas of research focus: virtual community types, Web site design, and tools used. There has been little work into cultural preferences in the choice of tools and types of communities preferred. Although many have suggested that Web sites (in general) should be designed to suit different cultural preferences (Chen, 2002; Fink & Laupase, 2000), our understanding of the cul- tural preferences in virtual communities Web site design remain limited. The aim of this study is, therefore, to explore the design preferences of U.S. and Chinese virtual communities in hope of ¿OOLQJWKLVJDSLQRXUNQRZOHGJH METHODOLOGY Content Analysis This study investigates the design preferences of U.S. and Chinese virtual communities. To ascertain these design patterns, content analysis is employed; as it makes it possible to compare a large volume of content over different virtual communities (Neuman, 2000). &RQWHQWDQDO\VLVLV³DWHFKQLTXHIRUJDWKHU- ing and analysing the content of text” (Neuman, 2000); it is a systematic and objective way to analyse message characteristics (Neuendorf, 2002). It has a long history and has been used in PDQ\¿HOGV$FFRUGLQJWR 1HXPDQWKH advantages of employing content analysis are that it enables researchers to compare content across many texts (units) and reveal aspects of the context that are not easy to see. For instance, we may have some vague feelings based on our unsystematic observation, but content analysis can KHOSWRFRQ¿UPLIWKHIHHOLQJVDUHWUXHREMHFWLYHO\ and quantitatively. Content analysis has also appeared in many Web analyses since the 1990s as the Web pages contain graphics, texts, and other rich context that make content analysis applicable (Neuendorf, 2002). Web site features are one of the design preferences studied in this research. A Web site contains multiple Web pages; miscellaneous contexts are included within the page, ranging from user information, news stories, promotion details, and so forth. Content analysis can help to handle this large volume of information properly. By employing content analysis, similarities and differences of the design preferences between virtual communities in the U.S. and China can EHLGHQWL¿HG Coding Scheme An a priori coding scheme was created (see Table 4). It describes all the measures in the study and their references. Measures selected were Web site design features, tools, and types of virtual com- munities. Web site design features are grouped into content, navigation, promotion, emotion, and interactivity. These categories have been suggested in previous studies on Web-usability HYDOXDWLRQJXLGHOLQHV$IHZIHDWXUHVGRQRW¿W into any of these categories are grouped into the ”other” category. A codebook and a corresponding coding form were developed. All measures are fully explained in the codebook so ambiguity of the measures 1637 An Exploratory Study of the Design Preferences of U.S. and Chinese Virtual Communities can be largely reduced (Neuendorf, 2002). A coding form was also developed, based on the codebook, for recording all variables measured. The codebook and coding form were created ZLWKWKHPHDVXUHVGH¿QHGIURPSULRUUHVHDUFK The codebook is presented in Appendix A along with the coding form in Appendix B. Sampling of Virtual Communities Hofstede’s study divides countries into six clusters (Hofstede, 2001). These clusters have been used LQPDQ\RWKHUVWXGLHVLQWKH¿HOG.KDVODYVN\ 1998; Robbins & Stylianou, 2001). It is also noted that the Anglo and Asian clusters have appeared in many studies for comparison of cultural differ- ence (Fink & Laupase, 2000; Singh, 2003). These WZRFOXVWHUVVKRZVLJQL¿FDQWGLIIHUHQFHLQHDFKRI the cultural dimensions (see Table 5). Thus, U.S. and China were selected in the study to represent the Anglo and Asian cluster correspondingly. The virtual communities chosen are also based in these two countries. In order to make sure that virtual communities selected are highly representative of the clusters chosen, only those most popular communities are included in the study. These communities have attracted more visitors and members. A more popular virtual community implies that it is more acceptable and preferred by some users than oth- ers and therefore, advantages of its design can be assumed. All virtual communities were retrieved from Yahoo directory (Directory > Society and Culture > Cultures and Groups > Cyberculture > Virtual Communities). The communities rep- UHVHQWLQJ$QJORFOXVWHUDUHRQWKH<DKRR³0RVW 3RSXODU´OLVW+RZHYHUWKH³0RVW3RSXODU´OLVW is not available in the Yahoo Chinese version; instead, a total of 49 virtual communities are listed. In order to determine the most appropri- DWHFRPPXQLWLHVWKHWUDI¿FUDQNRIHDFKFRP- munity is checked with http://www.alexa.com. Table 4. Coding scheme Measures 'H¿QLWLRQ Reference Web Site Design Features Content The informational and transactional capabilities of a Web site. Agarwal and Venkatesh, 2002 Navigation What allows users to acquire more of the information that they are seeking, and what makes the information easier to be found. Palmer, 2002 Promotion The advertising of a Web site. Agarwal and Venkatesh, 2002 Interactivity Nature of the interaction between users and Websites Benbunan-Fich, 2001 Tool Use of multimedia content Agarwal and Venkatesh, 2002 Community type Types of virtual communities Amstrong & Hagel, 1996, Chaudhury, Mallick, & Rao, 2001 1638 An Exploratory Study of the Design Preferences of U.S. and Chinese Virtual Communities Table 5. Comparison of clusters selected Adapted from Chau et al. (2002); Hofstede (2001); Singh (2003) Table 6. U.S and Chinese virtual communities sampled Dimension Anglo (USA) Index Asian (China) Index Context Low N/A High N/A Power Distance Small 40 Large 68 Collectivism VERSUS Individualism Individualism 91 Collectivism 25 Masculinity VERSUS Femininity Masculinity 62 Femininity 57 Uncertainty avoidance Weak 46 Strong 57 Time Short term orientation 29 Long term orientation 96 The U.S. Communities Yahoo! Groups http://groups.yahoo.com/ Well http://www.well.com/index.html MSN Groups http://groups.msn.com/ Everything2 http://www.everything2.com/ Cybertown http://www.cybertown.com/main_ieframes.html Bolt http://www.bolt.com/ Bianca http://www.bianca.com/ Meet Up http://www.meetup.com/ SmartGroups.com http://www.meetup.com/ Makeoutclub http://www.makeoutclub.com/ The Chinese Communities (Sina) http://www.sina.com.cn/ (NetEase) http://bj.163.com/ (Tianya Club) http://www.tianyaclub.com/ (4) (The 9 City) http://www.the9.com/ http://city.the9.com/ (Yinsha Chat) http://154008.chat.yinsha.com/ - (Ccidnet) http://bbs.ccidnet.com/ (Yuhou) http://www.yuhou.net/ (Seasky) http://www.seasky.cn/ (Joyie) http://ww.joyie.com/ (Exia) http://www.exia.cn/ Alexa.com was founded in 1996, and it provides VHUYLFHVVXFKDVWUDI¿FUDQNVLWHVWDWLVWLFVDQG so forth (http://pages.alexa.com/company/index. html). The top-10 virtual communities were then selected. Table 6 lists the U.S. and Chinese virtual communities sampled. 1639 An Exploratory Study of the Design Preferences of U.S. and Chinese Virtual Communities Coding Process Each community selected is evaluated against each of the variables individually. The detailed coding processes are as follows: • The access date and time of the community, along with the community name, URL, and cultural cluster, must be noted at the begin- ning of each evaluation. • If the Web site designing feature is present, a ³\HV´LVJLYHQRUHOVHD³QR´LVUHFRUGHG • Tools used in each community are recorded. Multiple tools can be checked for one single community. • Types of community are also coded accord - ingly. As one community may provide more than one type of community service, mul- tiple community types may be recorded. All data collected was keyed into a spreadsheet for reporting in the next stage. Data was double checked against the original data on the code form by a second person to ensure accuracy. RESULT AND ANALYSIS Background Information of Virtual Communities Selected The most popular virtual communities were se- lected since they have higher visiting rates and more members, which means that they may be better designed and are more preferred by users. A total of 20 virtual communities were included, with half of them based in the U.S. and the other half based in China. The background information of each community is introduced in Tables 7 and 8, with their history, characteristics, services, and community type(s) included. Virtual Community Types, Tool Used, and Design Preferences Table 9 presents the frequency of each item. For a better visual comparison between the two cluster groups, a bar chart is also included, as shown in Figure 2. Web-Design Features Web-design features can be grouped into six cat- egories, namely content, promotion, interactivity, navigation, emotion, and others. Content: Features in the content category are WKHMREOLVWFRUSRUDWHKLVWRU\¿QDQFLDOUHSRUWV social responsibility; design focus is use of art or informative content. • For both clusters, 3 out of 10 communities have presented job opportunities within the company on their Web sites. No difference is shown. • Eighty percent of the Chinese communities provide information on the companys’ his- tory compared with 50% of U.S. communi- ties providing such information. • Not much difference was shown on the presentation of social responsibility state- ments, with 70% of Chinese communities and 80% of U.S. communities providing information. • The focus of design is different between Chinese and U.S. communities. Seventy percent of Chinese communities choose to use more colour and graphics on their Web sites, while 60% of U.S. communities are more focused on providing information on the Web sites. Promotion: Two types of promotion ap- proaches are evaluated. Seventy percent of the Chinese communities used the soft-sell approach, 1640 An Exploratory Study of the Design Preferences of U.S. and Chinese Virtual Communities Table 7. Background information of U.S. virtual communities Group name Community background Community type Yahoo! Groups Yahoo started in 1994. It creates and manages online email communities. A great variety of groups are available for discussing all sorts of topics. (http://groups.yahoo.com/) ,QWHUHVW³7KH\´ 5HODWLRQVKLS³<RX´ Well Well is a pioneering online community, which has operated for more than seventeen years. It is known for engaging conversation and intelligent debate. More than 260 conferences ranging from technical DQGVSHFL¿FWRDEVWUDFWDQGVXUUHDODUHDYDLODEOHLQWKHFRPPXQLW\http://www.well.com/aboutwell. html ,QWHUHVW³7KH\´ MSN Groups ,QWHUHVWJURXSVDUHFODVVL¿HGLQWRVHYHQWHHQFDWHJRULHVZLWKVXEFDWHJRULHVLQFOXGHG,WLVXVHU RULHQWDWHGDQGVWDWHVDV³FRPPXQLWLHVRISHRSOHDVXQLTXHDV\RX´http://groups.msn.com/Browse. msnw ,QWHUHVW³7KH\´ 5HODWLRQVKLS³<RX´ Everything2 A community that simply aims to catalogue all human knowledge. It shows the interconnections between all people from all places; many ideas are presented. http://www.everything2.com/ ,QWHUHVW³7KH\´ Cybertown )RXQGHGLQ&\EHUWRZQLVWKHODUJHVW'YLUWXDOFRPPXQLW\WKDWKDVHYHUH[LVWHG,WLVDVFL¿ community of the future that enables members to interact and live in the virtual world. http://www. cybertown.com/info/about/business.html Fantasy; Interest & ³7KH\´ Bolt Bolt is one of the most popular Websites in the U.S. Bolt provides a communication platform targeting teenagers. The current member number is about 6 million. Its services include news and features, email, chat, message boards, and etc. http://www.bolt.com/ ,QWHUHVW³7KH\´ Bianca Bianca is an alternative online community; a home of Bianca’s Smut Shack. It provides different interest topics, BBS and chat. ,QWHUHVW³7KH\´ Meet Up Meetup connects people around their own interest topics locally. Events are held for members to join across 45 countries and 604 cities. Its registered members are nearly 1 million, with 2745 topics to choose from. http://www.meetup.com/about/ ,QWHUHVWDQG³7KH\´ SmartGroups. com SmartGroups is powered by Freeserver, a company established in 1998. SmartGroups helps the members to organize the different groups of people in their life. Services provided are email and the Internet Web page. http://www.smartgroups.com/text/en/about.cfm? ,QWHUHVWDQG³7KH\´ Makeoutclub Makeoutclub established in the year 2000. The purpose is to share personal interest, and bring people together. It was initially a personal Web site and has become an online independent music and subculture community. http://www.makeoutclub.com/about.htm ,QWHUHVWDQG³7KH\´ 1641 An Exploratory Study of the Design Preferences of U.S. and Chinese Virtual Communities Table 8. Background information of Chinese virtual communities Group name Community background Community type 新浪 (Sina) Sina started in the year 1997. It is voted the most popular Chinese site. It provides community services for Chinese globally. Its registered users are about 83 million. A total of 693 topic groups are available. http://corp.sina.com.cn/eng/sina_intr_eng.htm ,QWHUHVW³7KH\´ 5HODWLRQVKLS³<RX´ 网易 (NetEase) Started in 1997, NetEase now has about 144 million registered members. It is also the second most popular site in China. Communities are divided based on geographical areas and interest topics. http:// corp.163.com/eng/aboutus/aboutus.html ,QWHUHVW³7KH\´ 5HODWLRQVKLS³<RX´ Fantasy 天涯虚拟社区 (Tianyaclub) The Tianya Club has about 3000 members. It offers nearly 40 interest groups. The main services provided are posting and checking messages, articles and ideas. http://www.tianyaclub.com/ ,QWHUHVW³7KH\´ 第九城市(4) The9City The ‘9’ city was established in 1999 and has become one of the top 15 Chinese Websites, with an average of 400 million pagea viewed per day. It is a virtual city that allows its citizens to interact and exchange ideas together. The number of members has reached 14 million. http://city.the9.com/other/ corp_intro.htm Fantasy 碧聊大观园 Biliao The Biliao chat community is powered by Yansha. The community includes the most popular Chinese on-line chat rooms; with not only text chatting but also voice chatting available. All chat rooms are FDWHJRULVHGLQWRJURXSVDQGIXUWKHUFODVVL¿HGLQWRPDQ\PRUHVXEJURXSV ,QWHUHVW³7KH\´ 塞迪网 Ccidnet Ccidnet, established in 2000, now has 13 million members . The focus of the communities is to provide Information Technology (IT) related Bulletin boards services for members. Other IT information is also made available on-line. ,QWHUHVW³7KH\´ 雨後池塘 Trueme Trueme was established in the year 2000. It provides community services to both individuals and other organisations. It has about 1 million registered members. Each member can apply for a virtual space in the community. http://www.yuhou.net/about.html ,QWHUHVW³7KH\´ 5HODWLRQVKLS³<RX´ Fantasy 逸海情天 Seasky Seasky established in 2002, and has 50 thousand members now. It provides services such as personalised space, and Mud to enable member interactions. http://www.seasky.cn/about/aboutus.htm ,QWHUHVW³7KH\´ 5HODWLRQVKLS³<RX´ Fantasy 乐乐社区 Joyie Joyie was established in 2000. It has more than 1 million registered members. The community enables members to build a virtual home in the virtual world, with other services provided to encourage interactions between the members. http://www.joyie.com/community/about.htm Fantasy 异侠 Exia Exia provides services for members to enjoy on-line games and join different topics of discussion. It also provides spaces for members to upload their serial stories on-line and share them with others. http:// www.exia.cn/ Fantasy 1642 An Exploratory Study of the Design Preferences of U.S. and Chinese Virtual Communities Table 9. Item frequency comparison between Chinese and U.S. communities Categories Features Chinese U.S. Content Job list 30% 30% Corporate history 80% 50% Financial reports 20% 10% Social responsibility 70% 80% Design focus is use of art 70% 40% Design focus is informative content 30% 60% Promotion Soft sell approach 70% 30% Hard sell approach 30% 60% Interactivity FAQ 100% 80% Free choice of interactions 90% 80% Guided choice of interactions 0% 0% Interactions support group activity 80% 60% Interactions support individual effort 60% 70% Social communication 100% 90% Secure communication 10% 40% Clubs 60% 40% Navigation Hierarchical structure 90% 40% Search-based structure 10% 60% Links to local Websites 40% 0% Links to consumer groups 20% 0% Site map/index 70% 20% Others Cookies 20% 70% Registration 20% 10% Hit or visitor counter 100% 10% Quizzes and games 70% 30% Free trials and downloads 50% 40% Customer service 50% 10% Local stores & terminology 50% 0% Search engine 60% 60% Tools Asynchronous tools 40% 32% Synchronous tools 40% 18% Types 7UDQVDFWLRQ³,W´W\SH 0% 0% Fantasy type 60% 10% 5HODWLRQVKLS³<RX´W\SH 50% 40% ,QWHUHVW³7KH\´W\SH 60% 100% 1643 An Exploratory Study of the Design Preferences of U.S. and Chinese Virtual Communities Figure 2. Frequency comparison . of U.S. and Chinese Virtual Communities Table 1. Web design features in each cultural dimension Adapted from Chau et al. (2002); Fink and Laupase (2000); Luna et al. (2002); Robbins and Stylianou. Study of the Design Preferences of U.S. and Chinese Virtual Communities any time and continuously (Ruhleder, 2002). A Web site can employ both synchronous and asynchronous tools at the same time 2000). &RQWHQWDQDOVLVLV³DWHFKQLTXHIRUJDWKHU- ing and analysing the content of text” (Neuman, 2000); it is a systematic and objective way to analyse message characteristics (Neuendorf, 2002). It has a long history and has been