Wednesday, July 13, 2016

More Thoughts on Polling Data - UPDATED

Following on to  yesterday's bashing of the Reuters/Ipsos poll and the recent decision by RealClearPolitics.com to include this clearly fake, clear outlier in its daily polling average, we have these results this morning:

  • A national McClatchy poll showing the race at Hillary 42, Trump 39, right in line with every other recent national poll - well, other than Reuters/Ipsos;
  • State polls from Quinnipiac that show the following results:  Ohio - Trump and Clinton tied; Florida - Trump +3; Pennsylvania - Trump +2; 
  • An Iowa poll from Monmouth that has the race at Trump +2;
  • A Nevada poll from Monmouth that has it Hillary +4;
  • Another Florida poll from JMC Analytics that has it Trump +5;
And where is this morning's latest update on Reuters/Ipsos?  Hillary 46, Trump 33.

As I write this, RCP has yet to include this latest clearly fake Reuters/Ipsos result in its daily average.  If they do, well, shame on them, and it's time to stop relying on RCP for anything real.

Have a great day.

UPDATE:  Now at mid-day, we have new results from the YouGov/Economist national tracking poll - another tracking poll which RCP chooses not to include in its average for mysterious reasons - that shows the race at Hillary 40, Trump 37.

At this point, the data is overwhelming that this is a 2-3 point race at this moment in time, yet Reuters/Ipsos comes up with a 13 point margin for Hillary.

Why is the Reuters/Ipsos national tracking poll the only such tracking poll that RCP chooses to include in its average?  This is a question that needs to be answered.

2 comments:

  1. Amazing to me when a polling group just abandons any sense of actually trying to get the numbers correct, in favor of trying to push an agenda.

    ReplyDelete
  2. It's hard to know what the agenda really is with that poll. I guess they could be attempting to lend credence to the "Trump can't win" meme that the #NeverTrump crowd is using to try to steal the nomination from him. But at this point, R/I is so far outside of any sense of reality that no one is really going to take it seriously. I agree that it is really odd that they continue to intentionally use such a skewed sample after all this time.

    ReplyDelete