WANG INTERNET AND SAMPLING ROSS IN CHINESE MSM AIDS Education and Prevention, 14(5), 361 366, 2002 2002 The Guilford Press DIFFERENCES BETWEEN CHAT ROOM AND E-MAIL SAMPLING APPROACHES IN CHINESE MEN WHO HAVE SEX WITH MEN Quanyi Wang and Michael W. Ross In a study to determine sampling differences between Internet sites, we obtained data on 353 men who have sex with men in Chinese gay chat rooms and through e-mail web sites. Respondents were approached by the investigator and agreed to fill out an anonymous questionnaire on their Internet use and sexual activity. All materials and contacts were in Chinese characters. Data indicated that there were few differences between the chat room and Internet samples, but that those using e-mail appear to be more isolated, more homosexually-identified (rather than bisexual), have more experience with casual partners on a number of sexual activities, and were less likely to carry condoms and to have safe sex. E-mail respondents were more likely to want to discuss HIV/AIDS prevention on a web site or other site. These data suggest that the two recruiting methods are largely comparable in respondent characteristics, but that e-mail respondents are likely to be more isolated and at higher HIV risk than chat room participants. There has been a recent focus on the Internet as a potentially promising medium for disseminating HIV/AIDS preventive information (DeGuzman & Ross, 1999). Despite the enthusiasm, there have been few methodological studies to determine the population that might be targeted on the Internet and how such a population may differ from those who are accessible by traditional intervention approaches. There has been little research on the differences within Internet populations of men who have sex with men (MSM), for example those accessed from different web sites and chat rooms or via e-mail. Nor has there been much research on populations outside North American and western European Internet-using groups. Those few methodological studies that have compared Internet-using MSM with traditionally acquired non-internet derived samples have found that they are more likely to be younger, more bisexually or heterosexually identified, more rural or from smaller towns or cities, and to have lower education. Ross, Tikkanen and Månsson s (2000) Swedish study showed that Internet-using MSM were younger, more likely to Quanyi Wang is with the Institute of Basic Medical Sciences, Chinese Academy of Medical Science and Peking Union Medical College, Beijing 100005, China. Michael W. Ross is with the World Health Organization Center for Health Promotion and Prevention Research, School of Public Health, University of Texas, Houston. Address correspondence to Dr. Michael Ross, World Health Organization Center for Health Promotion and Prevention Research, School of Public Health, University of Texas, PO Box 20036, Houston, TX 77225; E-mail: mross@sph.uth.tmc.edu 361
362 WANG AND ROSS live in a rural area or in small towns or cities, bisexual, isolated from the gay subculture, and living with parents or a girlfriend than MSM who did not use the Internet. Of particular importance, a third of the MSM using the Internet wanted the opportunity to talk with an expert about HIV/AIDS. These data are surprising given the anecdotal speculation by Binik, Mah, and Kiesler (1999) that subjects recruited using the Internet for sexually related studies will be younger, richer, better educated, and more open than they might otherwise be. A similar smaller study in the United States by Rhodes, DiClemente, Cecil, Hergenrather, & Yee (2002) produced similar findings, with the exception that Internet users were more likely to be older. Mustanski (2002) concluded that further methodological research on Internet samples is needed, particularly controlling for possible confounders in sampling approaches. In China, the concept of homosexuality as defined in Western societies is relatively alien. Chou (2000) reported that there is a reluctance to take on discrete sexual identities, and that a more appropriate concept is that of tongzhi, or comrades. In China, historically, individuals were and still are defined primarily by their family ties and role, and the concept that someone might be defined by what they did or liked sexually is unfamiliar. This disjunction between doing and being makes HIV/sexually transmitted disease (STD) prevention more difficult because many MSM will not identify as being at risk. Although sexual risk behaviors are common among MSM in China (Zhang, Lu, Li, & Hu, 2000), very limited information about MSM have been reported, even in Chinese. We are not aware of any published Internet intervention studies in Asia generally or China specifically. However, Chou (2000) stated that the Internet in China is a highly effective way to break through the oppressive boundaries created by the tight family-kin structure and the social control system of work units and residential arrangements (p. 134). He further noted that modern technology, including e-mail and Internet, have been appropriated by tongzhi to build up an unprecedented space of virtual communities. In China, the hidden nature of MSM and the level of stigmatization of such behavior has been documented (see, e.g., Rosenthal, 2002). The Internet provides an exceptional medium for an intervention to target the more isolated MSM who have access to the Internet, either on a personal basis or through Internet cafés or other connections. According to the Chinese National Network Information Center, by the end of 2001, there were 12.54 million computers connected to the Internet, 33.70 million Chinese accessed the Internet more than 1 hour per week, of whom 60% were males. Internet users increase about 50% per year, and it is estimated that there will be 200 million Chinese Internet users by 2005. The combination of stigmatization of homosexuality, lack of a coherent government HIV/AIDS prevention effort, and growing access to the Internet, make China an ideal context for HIV prevention and education in MSM. METHODS SUBJECTS AND SURVEY PROCEDURES Three hundred and fifty-three Chinese gay men were recruited from seven gay web sites by Internet outreach with informed consent in August to November 2001, including 53 participants recruited from chat rooms and 300 participants recruited by responding to their e-mail. These web sites are seven most popular ones of 500 gay sites in Mainland China. Gay-related news, galleries, communities, and chat rooms are offered by these sites, and the target populations are gay men in mainland China since Mandarin and simplified Chinese characters are used at these web sites. They
INTERNET SAMPLING IN CHINESE MSM 363 also provide basic knowledge about HIV prevention, but they don t have any services for HIV intervention. As with other gay sites in mainland China, the seven sites are nonprofit and porn-free. CHAT ROOM APPROACH Chat rooms of the seven gay web sites are open to all visitors without any limitations on the topics. Chatters can talk by typing with others in two ways: publicly (i.e., all chatters in the chat room can see the words he types) or privately (i.e., only he and the person he is talking with can see the words he types). Considering most Internet users chat with others in the evening and at during the middle of the night, researchers accessed the chat rooms between 8 p.m. to 2 a.m. (Beijing time). In the chat rooms, researchers introduced the survey to all chatters publicly and asked them participate in the survey repeatedly. The significance of the survey was explained to the public, and confidentiality and anonymity were stressed by researchers. Enquiry regarding the survey was answered publicly or individually, depending on whether the request was public or private, to the person who asked the question. Chat room men who agreed to participate the survey told their e-mail addresses to researchers individually, then researchers e-mailed them the informed consent letter. Eighty-two questionnaires were e-mailed to gay men who responded to the informed consent e-mail, 59 questionnaires were e-mailed back by participants, and 53 of them were identified valid by the researchers: six questionnaires with more than one third missing variables were excluded from the sample. Questionnaires covered demographics, location, sexual behavior, HIV risk behavior, HIV/AIDS knowledge, Internet use, HIV/AIDS information needs, and condom carriage and use. A copy is available from the first author on request. E-MAIL APPROACH An enquiry and informed consent letter was sent to members of the selected gay web sites, who opened their e-mail address to all visitors of these web sites. Four hundred and fifty-two questionnaires were sent to participants who responded to the informed consent letter. Of the 319 returned questionnaires, 300 were identified valid, and 19 incomplete questionnaires were identified invalid by researchers. MEASURES AND DATA ANALYSIS An 85-item questionnaire was used to collect information about demographic characteristics, use of the Internet and other AIDS resources, HIV knowledge and attitudes, sexual risk behaviors, and STD and HIV status. The questionnaire was written in simplified Chinese characters, the official language used in mainland China. Data were analyzed using SPSS (Version 11, 2001). The age of the two groups was compared by independent t test; categorical variables were compared by chi-square tests. Because overlapping of the two groups might underestimate the differences, significance levels in this study were set at.01 and.10 instead of.01 and.05. RESULTS Participants in this study were generally young and well educated city residents. Demographic characteristics of the two groups are listed in Table 1. Participants of the two groups were approximately the same age and had similar educational backgrounds and marital status. More than 70% of participants received information
364 WANG AND ROSS TABLE 1. Demographic Characteristics of the Participants (%) Variable Chat Room E-mail Age 24.9 ± 4.4 24.4 ± 5.5 Education (year) 14.3 ± 2.0 13.8 ± 2.6 Marital status Married 17.0 12.7 Single 73.5 74.2 Others 9.5 13.1 Location City 90.6 89.6 Town 7.5 5.4 Rural area 1.9 5.0 about AIDS from gay web sites, and 81.2% of them thought the Internet was the best way to promote HIV prevention for gay men. Close to 70% of participants reported physical sexual experience with men known from the Internet, 10.8% reported having had an STD infection, and 0.9% of them reported being HIV seropositive. Fifteen percent of participants paid for sex with men and 17.7 % of them provided sex for money. One third of participants had sexual experience with women. Participants of the two groups had similar answers about use of AIDS resources and HIV knowledge and attitudes. However, chat room men were less likely to put themselves at a higher risk for HIV/STD infection in casual sex, and a lower proportion of participants recruited from chat rooms practiced unprotected sex with casual sex partners. (Table 2). Other differences of the chat room men compared with e-mail-recruited respondents are also presented in Table 2. These data indicate that chat room users, compared with those contacted by e-mail, were more likely to consider themselves bisexual, were more likely to have had sex with women, used the Internet several times a day (compared with several times a week for e-mail users), and were less likely to want to discuss AIDS and AIDS prevention at a web site or online club. The chat room men were more likely to have ever paid for sex and to carry condoms if sex was possible. DISCUSSION These data must be interpreted with the caution as they are based on a relatively small sample and come from a country where there are still political dangers in admitting to being an MSM, as well as a degree of censorship and risk in making such an admission on the Internet. Thus, their generalizability to other Asian or Western countries is unclear. Nevertheless, they are also of interest because they suggest that there are few major differences between Chinese men recruited on the Internet in chat rooms as compared with e-mail contacts. Understanding such differences is important given that it is important to understand the biases that may occur in recruiting research respondents from one site or the other. Although it is possible that there may have been overlap between the recruitment approaches, if this were the case, this would underestimate the differences between groups. The key difference between chat room and e-mail recruitment is that chat rooms are essentially a real-time group process, whereas e-mail is a delayed individual pro-
INTERNET SAMPLING IN CHINESE MSM 365 TABLE 2. Differences Between MSM Samples Contacted Through Chat Room and E mail (%) Variable Chat Room E mail Sexual identity Homosexual 54.7 72.2 Bisexual 45.3 26.8* Have you had sex with women? (yes) 49.0 31.8** Internet use Few times/day 30.2 22.5 Daily 13.2 21.1 Few times/week 39.6 46.0 Weekly 5.7 7.0 < once/week 11.3 3.4* Would you like to visit a website for AIDS prevention if someone introduced one to you? (yes) 75.5 87.3* Would you like to discuss AIDS at online clubs? (yes) 67.9 86.2** With casual sex partners, what sort of sex have you practiced? Active oral sex without condoms (no) 15.7 27.6* Receptive oral sex without condoms (no) 11.8 24.9 Anal receptive sex without condoms (no) 43.1 53.5* Anal insertive sex without condoms (no) 37.3 50.9* Would you carry a condom with you if sex is possible? (yes) 92.2 77.2** Have you ever paid for sex? (yes) 21.6 12.9* *p <.10 >.01. **p <.01. cess. Not only may the differences in context lead to differences in the respondents, but there may also be differences in the people who use chat rooms as compared to e-mail users. Thus, biases in recruitment from different sites may occur which have important implications for the reach and targeting of interventions. These data suggest that differences between chat room and e-mail recruitment in this Chinese population are relatively small, given that the questionnaire comprised 85 items and that only 12% of these showed significant differences. The differences found, however, were consistent. Chat room users were less likely to see themselves as homosexual and more likely to have had sex with women. They were, however, also less likely to want Internet-based HIV/AIDS information and more likely to use the chat rooms more frequently (several times a day) than e-mail users. However, there was a significant overlap between groups, with nearly eight out of ten e-mail users also chatting with other MSM. Sexually, the chat room respondents were less likely to practice unprotected casual sex and more likely to carry condoms than the e-mail group. These data suggest that chat rooms are at present likely to better reach the more bisexual MSM, but it is the more homosexual MSM who most want HIV/AIDS prevention information. Condom carriage, as an index of prevention, was highest among the chat room participants, further indicating that prevention among e-mail-using participants may be a priority. This finding has an important implication for HIV prevention in Chinese MSM. In China, there are about 500 free gay sites, and each popular site has thousands of members. If there were to be an effective health information distribution program targeted at Chinese MSM by e-mail, it could substantially reduce the risk for HIV infection in Chinese MSM. It may also move a significant number of participants toward contemplation of reducing HIV risks. In summary, these data show that for Chinese MSM there are few significant differences between those recruited in chat rooms compared with e-mail recruitment,
366 WANG AND ROSS possibly due to overlap, but that those using e-mail appear to be more isolated, more homosexually identified, and to have more experience with casual partners on a number of sexual activities, as well as lower condom carriage. Internet interventions to reduce HIV and STD transmission in China could usefully target both modalities of Internet use. REFERENCES Binik, Y., Mah, K., & Kiesler, S. (1999). Ethical issues in conducting sex research on the Internet. Journal of Sex Research, 36, 82-90. Chou, W. S. (2000). Tongzhi: Politics of same-sex eroticism in Chinese societies. New York: Haworth Press. DeGuzman, M., & Ross, M. (1999). Assessing the application of AIDS-related counseling and education on the Internet. Patient Education and Counseling, 36, 209-228. Mustanski, B. (2002). Getting wired: Exploiting the Internet for the collection of valid sexuality data. Journal of Sex Research, 38, 292-301. Rhodes, S. D., DiClemente, R. J., Cecil, H., Hergenrather, K. C., & Yee, L. J. (2002). Risk among men who have sex with men in the United States: A comparison of an Internet sample and a conventional outreach sample. AIDS Education and Prevention, 14, 41-50. Rosenthal, E. (2002). Gays in China step out, with one foot in the closet. The New York Times, April 12, A1& A10. Ross, M., Tikkanen, R., & Månsson, S. (2000). Differences between Internet samples and conventional samples of men who have sex with men: Implications for research and HIV interventions. Social Science and Medicine, 51, 749-758. SPSS for Windows (2001). Version 11. Chicago, Il: SPSS, Inc. Zhang, B., Liu, D., Li, X., & Hu, T. (2000). A survey of men who have sex with men: Mainland China. American Journal of Public Health, 90, 1949-1950.