Home Research Reports Researchers Affiliates Participants Page

PITTSBURGH -- Is the average American family devouring pornographic materials on the Internet?

HomeNet, a new study at Carnegie Mellon University of how ordinary families use the Internet concludes that families aren't big consumers of sexual information on computer networks. Nor do they seem likely to become avid readers.

At the start of the HomeNet field trial in February, 1995, a panel of 50 families in the Pittsburgh area received Macintosh computers, a full Internet connection and Internet services including electronic mail and a World Wide Web browser. The families were also encouraged to explore electronic newsgroups, or discussion groups, on hundreds of topics where anyone can read and post messages to others.

Detailed electronic audit trails were collected by the researchers to understand how the 150 individuals in these families used the Internet. (Subjects all signed consent forms, and the study went through internal CMU human-subjects review. No results are reported that could in any way attribute specific behaviors to specific individuals in the study.)

So, what do ordinary people do with the Internet, and with Usenet newsgroups in particular, when they're there?

• Our population of "regular people" here in Pittsburgh is far less interested in using the net as a way to access sexually-oriented material than the recently publicized Rimm study would seem to predict. And their interest is mostly transient: Most people who do, in fact, look at sexually oriented newsgroups do so only once or twice (over a period of months). Those who have looked at any particular sexually oriented newsgroup more than twice constitute less than 1/5th the sample population, and are mostly adult males and teenagers. And even for these people, their usage of sexually oriented groups is a relatively small portion of their overall activity with newsgroups.
• Newsgroup usage results turn out to be highly sensitive to the sampling technique used. The HomeNet researchers sampled every 10 minutes. But if a one-week sample interval is used instead, it *appears* that sexually-oriented newsgroups and other occasionally-browsed newsgroups are twice as popular as they really are. The more you sample, the more fine-grained your results are, and the more fine-grained your results are, the less important sexually oriented newsgroups become.
• Mostly, people browse newsgroups specific to their interests. Because there are many such specific newsgroups and diverse interests, few such groups show up in the "top 40". But in the aggregate, they far outrank the sexually-oriented groups in popularity. And local groups which allowed users to exchange information relevant to their day to day lives (e.g. "where's the easiest place in Pittsburgh to take the driver's license exam?") were by far the most popular.
• "Lurking": among HomeNet users who both follow (have looked at 3 times or more) Usenet newsgroups and post to them, the median ratio of groups posted on to groups followed is 1:2 (i.e., among people who both read and post, people tend to post to about half as many groups as they read). If we include people who have never posted in the calculation, then the ratio drops to 1:10 (and in case you're wondering, only three HomeNet users have ever posted to a sexually oriented newsgroup).

HomeNet Contact Information:

Vicki Lundmark,
Social and Decision Sciences
5000 Forbes Avenue
Pittsburgh, PA 15213
412-268-7505 (voice) lundmark@andrew.cmu.edu