Think Tank RAND Retracts Recent Study -- Absolutely Nothing To Do With The Bakken

The subject of the story does not interest me. Link here.

What interests me is the fact that one of the most renowned think tanks, RAND Corp, made a huge mistake by data mining web sites without confirming the source or the accuracy, it appears. On the surface, it appears that RAND simply mined internet data, probably using google searches for the most part.

The Rand researchers relied on data posted by CrimeReports.com, which they mistakenly believed included LAPD data. Knopman said Rand was not blaming the website. She said Rand reviewers, digging deeply into the data, only recently discovered that it did not include LAPD reports.

This is an understatement:

He said he could understand how the researchers might have made the mistake, but also said they ought to have noticed the lack of crime reported in Los Angeles as compared to other cities. "I think they should have been able to figure it out by seeing a huge pocket of no crime data," he said.

And then, as one delves deeper, it seems to get worse. RAND used only 20 days of data -- 20 days of data, incredible -- and admitted there was a huge margin of error. It seems that with such little data, and admitting a huge margin of error, this report should never have been released in the first place.

One has concern that RAND was rushing to be politically correct in California.