This article appeared on the front page of the Altoona Mirror on 4

This article appeared on the front page of a newspaper on 4-16-2001. It was kind of embarrassing (I didn't say most of the things that I was quoted as saying), but the article does demonstrate that I've been working with Census data since at least 2001.

Don't be too hard on the reporter. The US Census is probably the most respected and reliable data source in the world, and the temptation to "break a story" on a US Census error was overwhelming to him.

On the next section, you'll see that I said "I found that there was a 97% error ..." That was one quote that I actually did say, but he neglected to mention the full quote. The 97% error rate only applied to a single small fragment (a PUMA, but in the interview we decided to call it a "community") of New York City. Nationwide, the error was small. For those of you who are wondering why I would contact a reporter over a small error ... see the bottom of this page.

Why such a fuss over a small error?

I didn't contact the reporter over the error. I contacted him to demonstrate the power of MAST, and I gave him a tabulation that used some of MAST's capabilities. To make the tabulation interesting, I structured it to reveal the Indian error, naively expecting the reporter to describe what a great tool MAST was. However, the error then became the story, and MAST's superiority was barely mentioned. In hindsight, any reporter is going to be much more excited breaking a story about an error in the census than yet another technological innovation.