The answer is simple. While some on the Left have been gleefully touting the numbers, we've been looking for the two C's: cross-tabs and context. Has this firm or agency polled this race before? If so, are there any observable trends? What methodology does the company use? How big was the sample? What was the demographic breakdown of the sample - by party registration, gender, ethnicity, socioeconomic status, geography, etc?
If a reader has access to any of this information, please feel free to share it with us. Otherwise, there are very few conclusions that can be drawn from this information. If the Left wants to draw sweeping conclusions from this outlying survey that shows their candidate cruising to victory, that's their prerogative by all means. But frankly, any discussion that doesn't highlight the major qualifiers isn't credible.
Then again, there's also this gem from the same poll:
Nearly two-thirds of voters in Colorado (68%), Maine (72%) and Minnesota (65%) oppose the ["Employee Free Choice Act"]. Moreover, voters in Minnesota and Colorado would be less likely to support candidates who support the EFCA. Specifically, a plurality of voters would be less likely to vote for Mark Udall (44%) and Al Franken (41%) if they support this legislation. Moreover, at least 80% of voters in all three states believe that secret ballot elections are the cornerstone of democracy and should be kept for union elections.In a poll that's an outlier showing a big lead for Mark Udall, these measures of Colorado opinion against a Big Labor bill that would take away workers' rights to a secret ballot in the workplace are a bad sign for the Left. Unless they want to argue with a straight face that this polling sample is clearly skewed to the Republican side.
Of course, Udall was a co-sponsor of the EFCA in 2007, even though he had "serious reservations" about the issue (but many union financial contributions).
For more reliable information on trends in the Schaffer-Udall race, we'll wait for the next Rasmussen tracking poll to be released.
2 comments:
Interesting. So you question the methodology of a poll you don't like without knowing the methodology. Have you questioned any of the other polls?
Nice try at sophistry, but go back and read the post. It isn't unreasonable to ask reasonable questions about a survey that shows vastly different results from everything else.
I haven't said the poll is valid or invalid. I question all polls, which individually tell observers a lot less than they do collectively. But some polls have a record of being better than others - Rasmussen has one of the best records, and in this case a history for tracking over time.
Do you know the record of McLaughlin? I don't, but if you want to latch onto its conclusions because it shows results you like, it's your prerogative. You can draw sweeping conclusions from an outlier. But at the cost of any remaining intellectual honesty.
Post a Comment