Every study of Internet filters — software that keeps kids away from porn and other inappropriate Web sites — has shown that they are flawed, but a new study commissioned by the Kaiser Family Foundation and published by the Journal of the American Medical Association points out that their effectiveness depends how they are configured.
Setting the filters at their least restrictive level can block most porn sites and still provide access to the vast majority of health sites. But if the filters are set on more restrictive levels, they can block up to 25 percent of health related sites.
The study, "See No Evil: How Internet Filters Affect the Search for Online Health Information," appears in the Dec. 11 edition of The Journal of the American Medical Association and is summarized at www.kff.org.
We tend to see the world in black and white. Either something is good or it's bad. But life is usually a bit more complicated.
Congress thinks that Internet filters are a good idea. A couple of years ago it passed legislation requiring filtering for schools and libraries that receive federal funds. Peacefire.org thinks they're bad — a form of censorship. The American Civil Liberties Union can't seem to make up its mind. The group argued in federal court that filters are a good alternative to legislation that would restrict access to certain types of Web sites, yet also published a report condemning filters as "censorware."
Well, it turns out that filters — like a lot of other tools — are good or bad depending on how they are used.
The Kaiser Family Foundation commissioned a scientific study to determine whether the types of filtering programs used by schools and libraries (and also many parents) can effectively block pornography without keeping kids from accessing health information.
Restricting access to health sites is perhaps the most glaring example of what is called "over blocking," the tendency of filtering programs to keep kids away from sites that they ought to be able to access. In previous studies an overwhelming majority of 15 to 17 year olds said that they have used the Web to look up health information including searches on such topics such as pregnancy, birth control, HIV/AIDS, sexually transmitted diseases and issues regarding drug and alcohol use. When set at their most restrictive levels, filters can block up to 25 percent of health sites, including those operated by the Center for Disease Control and other government agencies.
Another concern is under-blocking — failing to keep kids away from pornography and other objectionable sites. Filters that under-block can give parents and educators a false sense of security. The study found that filters set to their most restrictive levels block 91 percent of pornographic sites. At the least restrictive levels they blocked 87 percent. In other words, by lowering the settings, you increase the amount of porn by only 4 percent while dramatically decreasing the over-blocking of health related sites.
The key is how the person in charge configures the filtering software. All of the major filtering programs have user settings that allow the teacher, librarian, system administrator, parent or other adult in charge to specify just how vigorous the filtering process should be. Should it, for example, block only Web sites that are pornographic or should it also ban sites that include partial or full nudity? What about "hate" sites or sites that encourage the use of tobacco, alcohol or illegal drugs?
The researches found that filters — when set to their least restrictive settings — block only about 1.4 percent of health sites. However, if they are configured at their most restrictive settings, they can block nearly "one in four" health sites (24 percent).
Hardly any sites dealing with breast cancer were blocked by filters set at the least restrictive levels but 6.9 percent of such sites were blocked at the most restrictive levels. The study found serious over-blocking problems with sites dealing with sexual health: Filters set at the most restrictive levels blocked "nearly one in four sites on sexually transmitted diseases (23 percent), one in three on pregnancy (32 percent) or birth control (35 percent), and one in two on condoms (55 percent) or safe sex (50 percent)."
How the filters respond to health related inquiries not only depends on
how they are configured but what terms are being sought. Searches that
include terms like "safe sex" and "condoms" are more likely to be
blocked than those for "birth control" or "herpes."
Trying to block sites that encourage the use of alcohol, tobacco and illegal drugs can also block sites that help prevent the use of those substances, according the Kaiser Family Foundation vice president, Vicky Rideout.
What all this means is that filters can be reasonably effective and reasonably unrestrictive but they can also get in the way of a young person's search for important — potentially life saving — health information. Parents, educators and librarians need to put some thought into how they are configured, especially in terms of filtering out non-pornographic sites such as ones that encourage hate or the use of illegal substances.
Rather than using the filters to block everything that may be objectionable, it might make more sense to simply block pornography. That would comply with federal regulations (the Children's Internet Protection Act) and eliminate the most egregious problems. The other types of problem sites — such as drug, alcohol, tobacco or hate sites — can be dealt with in other ways. In fact, it might make for some good lesson plans, teaching kids to debunk mistruths and understand the role of propaganda.
As always, it's important to understand that filters are not a substitute for adult supervision and educating students about the appropriate use of technology.
Filters are a bit like neckties. When used properly, they can make you look respectable. If you make them too tight they'll strangle you.
By Larry Magid