Digital Pulse - Ch 2 - Sec 3 - See No Evil: How Internet Filters Affect the Search for Online Health Information

Summary

The intent of the study was to evaluate how Internet filters employed in schools and libraries affect the ability of young individuals to locate non-pornographic health information on the web. It systematically measured the effectiveness of six different filtering products at blocking pornography and the rate at which health sites were also blocked. The impact of the filtering systems was delineated not by individual product, but by the blocking configurations used during set up. Configurations were grouped into three categories: Least Restrictive (pornography only); Intermediate (other ‘inappropriate' topics such as nudity, drugs, hate material), as based on one state-wide school network, and; Most Restrictive (tobacco, swimsuits, dating), a setting based on any category that might plausibly be blocked by a school or library. A total of 3,053 health sites and 516 pornography sites were tested against the filtering products. These sites were identified using unfiltered searches through six major search engines.

Key Points

1. On average the least restrictive settings blocked only 1.4% of health sites while intermediate settings blocked 5%. However, at the most restrictive settings, 24% of health sites were blocked.
2. Some health topics were much more likely to be blocked than others. At the least restrictive, setting filters blocked about 10% of sites tied to searches with the words “condoms,” “safe sex,” and “gay.” At the intermediate and most restrictive, substantial amounts of health oriented sites were blocked (25%+).
3. The more restrictive settings did not substantially increase the proportion of pornographic sites blocked.
4. The more restrictive settings did substantially increase the proportion of non-pornographic sites blocked, especially on topics concerned with sexual health.
5. Accidental exposure to pornography through the course of health information searches does not appear to be a substantial problem.
6. Variances existed in the different filtering products and the proportion of health sites blocked across all setting levels.
7. Many youth oriented health websites are being blocked by one or more of theses filtering systems. Websites oriented towards sexual health were even more likely to be blocked. 33% of safe sex sites were blocked by at least one of the filters at the least restrictive setting while 91% were blocked by at least one of the filters at the most restrictive setting.

The report concludes that the adverse impact on the ability of young people to access health information is primarily influenced by how the filters are configured by the schools or libraries installing them. If filters are configured to more restrictive levels, there is a major impact on the access to health information, especially that pertaining to sexual health. Yet, the more restrictive blocking configurations are only marginally more effective at blocking pornography websites. Thus the trade-off for a minor increase in the prevention of pornography is high in terms of the loss of health information. The report suggests that organizations attempting to provide health information need to be aware of these impacts and that those groups attempting to disseminate sexual health information could have their efforts seriously impeded by the overzealous use of filters. The report suggest that the application of filters at more restrictive settings should result from a “conscious and informed policy decision” with an awareness of the effects that it will have on the search for health information rather than a summary decision to set the filters at the highest levels.

Source: Richardson, Caroline & Resnick, Paul. “See No Evil: How Internet Filters Affect the Search for Online Health Information”, Pub. # 3294 (Henry J. Kaiser Family Foundation, Washington, D.C.) December 2002.