From this Politico article:
Within the past two weeks, private Democratic polling has shown that the party’s nominee, former Rep. Travis Childers, would start a general election statistically tied with McDaniel. A race against Cochran, who is well-liked by independents and many Democrats, would be difficult to the point of futility.
A few comments:
- It is generally a mistake to ignore polling data.
- The above is not polling data.
We have no idea of the context of the above. For example:
- When (who?) says they were "statistically tied," what is the margin of error? Were the results at the edge of that range, or towards the center?
- What is meant by "within the past two weeks?" Does that mean one poll in that time showed that, or many? How many polls were conducted that show this statistical tie? Were there others conducted that did not show that?
That is not to say that the race, should McDaniel win the runoff, is not closer now than conventional wisdom suggests. However, when I see reference to private poll data without the poll data being released, I cannot help but wonder if there is a reason that the full results were not being released, even if just the topline data. I don't discard the possibility that what is being leaked is accurate, but I also do not take it at face value.
As is often the case with regards to publicly released poll information, more data is needed to help us truly evaluate where things lay.
No comments:
Post a Comment
By and large I am going to rely on Twitter to be the 'comments' section here. You can submit comments, but moderation is enabled, and nearly all of the time I am not even going to check the moderation queue (although in some circumstances, I just might).