These shifts were not outside the margins of error in the Reuters/Ipsos and Echelon Insights surveys, while the Times’s pollsters said that they could not calculate such a margin for their recontacting survey. Nevertheless, the fact that the same shift was recorded across three different surveys lends credence to its validity.
I’ve heard it’s like a 2% change? That’s within the margin of error of most polls.
It’s noise. Random fluctuation.
There’s two interesting things the article points out.
Obviously you can’t go by polls, and there’s a ton of time between now and Nov, but it’s interesting nonetheless.
Only one of them was a recontact poll, which they don’t even report a margin of error for.
That one at least produced some interpretable anecdata, to wit: very few people changed their mind.
It’ll take months of pounding away at it to reach most people, if it does at all.