HomeLatest ThreadsGreatest ThreadsForums & GroupsMy SubscriptionsMy Posts
DU Home » Latest Threads » Forums & Groups » Main » General Discussion (Forum) » I think Rasmussen distort...

Fri Jul 17, 2020, 09:33 PM

I think Rasmussen distorts 538's

Trump's approval #s. I have been watching these #s since day 1...no one except Rasmussen & affiliates and maybe throw in Harris, polls so high for trump. I know Silver adjusts for past accuracy, but what about frequency? Seems Rasmussen has a poll out constantly. Surely their consistent high approval affects Silver's average?

11 replies, 1162 views

Reply to this thread

Back to top Alert abuse

Always highlight: 10 newest replies | Replies posted after I mark a forum
Replies to this discussion thread
Arrow 11 replies Author Time Post
Reply I think Rasmussen distorts 538's (Original post)
Laura PourMeADrink Jul 2020 OP
Loki Liesmith Jul 2020 #1
Laura PourMeADrink Jul 2020 #4
Loki Liesmith Jul 2020 #9
paleotn Jul 2020 #2
Laura PourMeADrink Jul 2020 #6
Make7 Jul 2020 #3
Laura PourMeADrink Jul 2020 #5
grantcart Jul 2020 #7
BGBD Jul 2020 #8
Laura PourMeADrink Jul 2020 #10
BGBD Jul 2020 #11

Response to Laura PourMeADrink (Original post)

Fri Jul 17, 2020, 09:44 PM

1. They are weighted less into the average

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Loki Liesmith (Reply #1)

Fri Jul 17, 2020, 09:58 PM

4. But isn't that due to o past inaccuracies?

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Laura PourMeADrink (Reply #4)

Sat Jul 18, 2020, 10:20 AM

9. Also they don't adopt all Nate's best practices

They all figure into the pollster rating which helps determine the proper weighting.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Laura PourMeADrink (Original post)

Fri Jul 17, 2020, 09:57 PM

2. Always has.....

and I've always wondered why the keep including a group that obviously focuses on over 65's with land lines. The distortion can't be weighted out unless it's factored at 0.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to paleotn (Reply #2)

Fri Jul 17, 2020, 10:03 PM

6. But aren't people with land lines over 65 hating on him too?

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Laura PourMeADrink (Original post)

Fri Jul 17, 2020, 09:57 PM

3. 538 ranks the pollsters they use. Rasmussen gets a C+.

They adjust the approval numbers based on the rankings. Rasmussen's reported 45% / 53% becomes 40% / 54% after adjustment.

538's overall calculated approval is 40.2% / 55.7% currently.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Make7 (Reply #3)

Fri Jul 17, 2020, 10:01 PM

5. Yes..but talking frequency of their polls. Even with dings

It's consistently adding high approval #s

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Laura PourMeADrink (Reply #5)

Fri Jul 17, 2020, 10:14 PM

7. No because the weight also declines rapidly by date after a few days they

Have a lot less impact.

The real value of any of these pollsters whether PPP or Rasmussen is not the number but the trend


In other words if both Rasmussen and PPP (and the others) all show Trump consistently trending down over a 10 week period, say 7 % , then they are confirming each other even if one is at 40 and the other is 44.

If Rasmussen shows a decline over time using the same turnout projections then his results are as helpful as the others, in some ways more.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Laura PourMeADrink (Original post)

Fri Jul 17, 2020, 10:21 PM

8. They don't just weigh..

 

they also adjust, so a Ras poll at 48 goes into the average at like 41.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to BGBD (Reply #8)

Sat Jul 18, 2020, 08:41 PM

10. Right but nevertheless frequent Rasmussen poll #s

Go into the avg. Even if they come in 41.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Laura PourMeADrink (Reply #10)

Sat Jul 18, 2020, 08:43 PM

11. They do

 

but even if you took them out, the change to the average would be minor.

Reply to this post

Back to top Alert abuse Link here Permalink

Reply to this thread