N2H2 "State Secrets" - PR and lying with statistics [part 1]

Censorware company N2H2 recently issued a press release which coincidentally illuminates why I have to quit and can't publish certain research. They proclaim: "N2H2 Reports Number of Pornographic Web Pages Now Tops 260 Million and Growing at an Unprecedented Rate". It's a concoction jam-packed with juicy PR nuggets:

"1,800 Percent Increase from 14 Million Pages in 1998"
"N2H2's database contained 14 million identified pages of pornography in 1998, so the growth to 260 million represents an almost 20-fold increase in just five years"
"There have been recent filtering software studies by eTesting Labs for the Department of Justice and a peer-reviewed study conducted by the Kaiser Family Foundation and published in the Journal of the American Medical Association (JAMA). Both found N2H2 superior to competitors at blocking pornographic Internet content without unduly blocking appropriate content, even in related areas such as breast cancer or sexually transmitted diseases."

This censorware PR is going around the world, from Australia to Japan .

There is absolutely no support for me to rebut it, in terms of 1) pay, 2) press-power, 3) protection (from lawsuits). I can start shouting to the wind and risking a lawsuit ... or not.

First, just on the face of it, how much has the web itself grown since 1998? Certainly much ("1,800 percent"?) If the number of "porn"-pages simply remained a constant fraction overall, it would increase in absolute numbers as the web itself grows. What methodology was used to derive that number of "identified pages of pornography" anyway?

And here's where things get interesting. I know N2H2 has to be playing fast and loose (if anyone doubted just based on general principles), because I know their blacklist does not have 260 million entries, and could not have had 14 million entries in 1998. But I can't publish that in any way that is objectively provable in terms of peer-review, because that would require conveying decryption information. And no organization is backing me as a researcher to give my mere words weight in terms of that "name" reputation-capital.

Moreover, I could point out many, many, flaws in that slippery phrase "classified as pornography". In practice, nobody outside of a few hundred hard-core readers (pun unintended) would hear. So, the upside is zero for me, little in general, and the downside is years of litigation. It's not worth it.

[To be continued in further posts. Next planned: Background historical material regarding censorware companies, lying with statistics, attack and defense]

Seth Finkelstein's Infothought blog (Wikipedia, Google, censorware, and an inside view of net-politics) - Syndicate site (subscribe, RSS)

          Twitter