September 19, 2001 12:38pm
Internet Research Methodology of Online Sexual Activities Study Proves Reliable
Source: Mary Ann Liebert, Inc.
by: Company Press Release
(LARCHMONT, NY) -- Studies of internet users who become addicted to Online Sexual Activity (known to researchers as OSA) have been criticized in the past because online survey methodology was thought unscientific.
But a major new study on the subject, which compares two groups of respondents accessing the popular MSNBC.com online news site, indicates that such online research is fundamentally valid, as reported in Volume 4, Number 4, of CyberPsychology & Behavior: The Impact of the Internet, Multi-Media & Virtual Reality on Behavior & Society, a bimonthly peer-reviewed journal published by Mary Ann Liebert, Inc. (www.liebertpub.cpb).
``Overcoming Methodological Concerns in the Investigation of Online Sexual Activities'' was written by Al Cooper, Ph.D.; Coralie Scherer, Ph.D.; and Robin M. Mathy, M.A. Dr. Cooper is training coordinator for counseling & psychological services at Cowell Health Services, Stanford University, Stanford, CA, and director of the San Jose Marital & Sexuality Centre, for which Dr. Scherer is a senior staff psychologist. Robin Mathy, the methodologist and statistician for the project, is affiliated with the Universities of Oxford and Cambridge in England and the University of Minnesota--Twin Cities. Their original paper, one in a projected series, examines the reliability of internet research and its implications for the study of OSA. To access the paper free online or view a free sample issue of the journal, visit www.liebertpub.com/cpb.
Approximately 20 percent of all internet users engage in some kind of online sexual activity (OSA), says the report, including seeking information or advice about sexual health, romance, and relationships; online chatting with a sexual focus; viewing erotic acts; buying erotic materials; or arranging for erotic encounters offline.
``Internet sexuality is of keen interest to sex researchers because of its relatively sudden appearance and rapid expansion in our social consciousness,'' said Dr. Cooper. ``A significant subset of people access the internet for sex, and a smaller subset of users visit X-rated sites excessively, reflecting a mental health problem that can have serious adverse consequences. Research in this area has attempted to get at the magnitude of the problem and to assess its medical and psychological significance. But because previous research has been done online, methods of collecting data have been called into question.''
The new study is based on responses from more than 38,000 MSNBC users to the 2000 Online Cybersex Survey, the largest internet study of online sexuality to date and one of the largest studies of any type on any aspect of sexuality. A convenience sample of 33,391 volunteer participants was used as a control group and a selected random sample of 7,544 subjects as an experimental group in the study. Participation in the convenience sample was available to anyone accessing the MSNBC website in June 2000; participation in the selected random sample was limited to every one-thousandth visitor to the site in the same time period. The survey was a 76-item questionnaire in two parts: 16 questions to assess demographic variables, and 59 questions about use of the internet for sexual pursuits.
Scientists have expressed reservations about the reliability, generalizability, and heterogeneity as well as the impact of self-selection in previous online research. Respondents can easily lie online, compromising the integrity of sample demographics; ephemerality and anonymity make it difficult to get a truly randomized, unbiased study population; internet users' demographics do not reflect the general population, the same time zones or even seasons; and differences between carefully selected respondents and self-motivated volunteers may also skew results. Although all these concerns may apply equally to offline and traditional surveys, related questions had until now remained largely unanswered and in need of empirical investigation.
Comparison of the two sample groups was a first and extremely important first step needed to demonstrate that the internet is a viable tool for social science research, provided that rigorous sampling designs are used, said Robin Mathy. ``We demonstrated that selected random sampling was able effectively to weed out the disproportionately high number of respondents with problematic use of the internet that is ordinarily found in a convenience sample drawn from the internet,'' she said. ``Overall, there were more similarities than differences in our two samples, which suggests that individuals who are self-motivated to participate in a study are not less reliable than those who elect to respond to a systematically administered request to participate. However, further research is needed to compare the reliability of this second group with those who declined to participate.''
Mathy points out that this study lays the groundwork for block sampling and stratified random sampling via the internet to obtain more representative samples, just as professional polling organizations do when conducting large telephone surveys now. Such sampling will yield increasingly reliable results as the demographics of internet users and the general population continue to merge.
``With each passing year, the internet is more closely approximating the demographics of the general population,'' she said. ``Because conducting sexuality research online has the distinct advantages of being able quickly to mount a study, access a large number of potential subjects, and efficiently and economically collect, manage, analyze and store data, it warrants the effort to integrate traditional methodologies with internet technology. The internet also allows recruitment of populations that might otherwise be difficult to study, such as sexually disenfranchised minorities.''
``Hard scientists complain about measurement techniques in the social sciences in that human variables are difficult to control,'' said Dr. Scherer. ``It's not easy to manipulate one variable at a time in studies like this, because people are so complex. But we've done something pretty significant and taken a major step forward. We don't claim that we have a sample representative of all MSNBC users, internet users, or the general population. What we've attempted to do in this study is to apply as rigorous a methodology as possible to a very free-wheeling social and technological phenomenon. Although we can't control all aspects of self-reported responses on the internet, it is as natural a medium for research as any other, with many time- and cost-saving advantages over traditional means of data collection. True randomization is not possible, but we included items to screen out extreme responses and eliminate multiple submissions to get as clean a sample as it's possible to get, which gives us a clear snapshot of people's behavior at a particular moment.''
``Studies like these will lead the way in establishing internet research as a viable and scientifically sound method for generating important findings in this and other areas of research,'' said Mark D. Wiederhold, M.D., Ph.D., editor in chief of CyberPsychology & Behavior.
Detailed reports of the survey's results are available at www.msnbc.com/news/596354.asp and www.msnbc.com/news/596355.asp.
CyberPsychology & Behavior: The Impact of the Internet, Multi-Media & Virtual Reality on Behavior & Society is a bimonthly peer-reviewed journal that explores the psychological and social issues surrounding the internet and inter-active technologies, including demographics of internet users, long-distance learning, social isolation, neuropsychological effects of multimedia, and internet addiction. It is indexed in EMBASE/Excerpta Medica.
Mary Ann Liebert, Inc., is a privately held, fully integrated media company known for establishing authoritative peer-reviewed journals in new and promising areas of science and biomedical research, including Telemedicine and e-Health Journal, Journal of Women's Health & Gender-Based Medicine, the forthcoming Journal of Men's Health, and the book Cybereducation: The Future of Long Distance Learning. Its biotechnology trade magazine, Genetic Engineering News (GEN), was the first in its field and is today the industry's most widely read publication worldwide. A complete list of its 60 journals, books, and newsletters is available at www.liebertpub.com.