I know there’s a certain sentiment out there that bad data is better than no data, but I’m not really sure why one would think that since a lot of the polling data available is quite bad. To get a flavor, consider (via Marc Ambinder), Mark Blumenthal pointing out the wildly varying notions of who’s likely to show up in Iowa:
July we have seen 12 public polls released in Iowa by 9 different organizations, and each appears to define and sample the likely caucus-goer universe differently. To the extent that pollsters have revealed the details, their snapshots of the electorate are poles apart, to say nothing of the candidates that those voters support. A month ago, for example, I found the percentage of first-time caucus-goers reported on four different polls of Democrats varying from 3% to 43%, with Edwards doing worse (and Clinton better) as the percentage of newcomers increased.
So, okay, in the aggregate this data is better than no data. It tells us something real. Namely, that public opinion in Iowa is fairly closely divided and that the turnout volume will have a significant impact on the outcome and that the general shape of that outcome is that the more first-timers who show up, the better for Clinton and the worse for Edwards. So it would be good for the press to report results like that and, indeed, the particular branch of the press known as This Blog You’re Reading Right Now will endeavor to report public opinion news in this manner — i.e., an informative one.
It’s striking, however, that most of the organizations who actually sponsor the polls clearly aren’t especially interested in providing their readers with accurate information. Instead, their idea is that if they do a poll, that will generate proprietary information granting them an “exclusive” story on the poll’s results. Thus the results of your firm’s poll should be heavily covered and not placed into the wider context of other poll results and the vagaries of polling methodology.