Ref's Ruling: The NBC/WSJ's small sample and timing led to an absurd result
The NBC/WSJ poll released on Monday, with its high margin of error of 4.6%, looks like a poll intended to reach a certain result. The eleven-point margin is nowhere close to accurate if the point of the poll is to determine where the popular vote is right now. If the poll is intended to show that Republicans were disgusted by Trump's words on a tape eleven years ago over the weekend when the tape ran wall-to-wall, it achieved that. This kind of poll, often provided to us by the NBC/WSJ unit, serves to throw off poll averages. The poll fails all three prongs of the test and will be weighted down in the averages. If I am wrong and other polls reflect this radical shift, I will remove the weight. Somehow I don't think that will happen. If the poll fails only two prongs of the test, incidentally, no weight would be applied. Follow my new account on Twitter for daily updates on the election!
|
National PollsFAILED: Other national polls +2.5 Clinton, NBC/WSJ poll +9.0 Clinton. Well outside the five-point acceptable range. OhioFAILED: OH +0.7 Clinton, NBC/WSJ poll +9.0 Clinton, well outside the acceptable five-point range.
FloridaFAILED: FL+1.8 Clinton. WSJ/NBC poll +9.0 Clinton, well outside the acceptable five-point range.
|