"Did the Dog Eat the Data?"
"Strategic Vision is cooking the books. And whoever is doing so is doing a pretty sloppy job."
Blogger Nate Silver crunched over four years of data from the firm's polls -- and says he's discovered the pollster's most-commonly reported numbers end in either a seven or an eight. "Over a sample of more than 5,000 data points, such an outcome occurring by chance alone would be an incredible fluke," Silver argues, "millions to one against." Statisticians (including the IRS) detect fraud by noting when numbers are reported in a non-random distribution - and Silver says that appears to be true for Strategic Vision's poll results.
"[T]his data is not random. It's not close to random."
How were the poll results generated? The polling firm won't release their methodology, argues the American Association for Public Opinion Research, who had issued a reprimand Wednesday saying the firm "repeatedly refused to release essential facts about polls". But the polling firm has responded,
saying they did provide the requested information - in June - according to a Florida news site. They cite a response from Strategic Vision's David E. Johnson, who points out his company isn't even a member of the polling organization.
Johnson's story? After providing the requested information, the organization simply claimed that "it was not done in a timely enough fashion so they were going to issue a press release reprimanding us... At no time did they request additional information as they now allege they had." In fact, Johnson argues, there could be suspicious circumstances around the reprimand, since "several members of this panel that voted to investigate us serve as consultants for one of our competitors and indeed contacted a client several days prior to this being released telling our client that they should switch to the company they consult for."
And his response to Silver's accusations? "[W]e categorically deny them and will refute them.
We have a call into our attorney on this and fully intend to take action that will vindicate us...he has attempted to do severe damage to our reputation and what is he going to do when we disprove him just say I am sorry. That isn't enough at this point."
Blumenthal is urging caution on Silver's report, citing a University of Michigan professor who suggested better questions for the analysis. But the controversy is already having an impact, according to a Thursday announcement from the "Political Wire" service.
"Even though Strategic Vision regularly gives exclusive peeks to Political Wire about their new polls, we're no longer going to report them."