Rassumussen vs Gallup - must be polling on different presidents?

[Laura PourMeADrink]

Rasmussen Daily - Improved since Shutdown I started
Approve - 52
Disapprove - 47

Gallup Weekly - Declined since Shutdown I started
Approve - 37
Disapprove - 59

I know, not precisely apples to apples, but still....

https://news.gallup.com/poll/203207/trump-job-approval-weekly.aspx
http://www.rasmussenreports.com/public_content/politics/trump_administration/prez_track_feb11

[leftynyc]

1. Rass only calls landlines

so you can't possibly take them seriously.

[Laura PourMeADrink]

3. Yup. Just saying trump and right bragging about it

[scarytomcat]

16. Let them be delusional

it will only help dems if rethugs think they are in striking distance when they are really down and out
I'm ready for a knock out in 2020

[ProfessorGAC]

15. They Also Willfully Ignore An Incorrect Stratification

I didn't check this one but I've seen several where the political divide was roughly 40R/40I/20D which has nothing to do with the overall population.

[leftynyc]

18. Holy cow

That's ridiculous. I'm no math whiz but a that's a seriously misappropriation of politics today.

[ProfessorGAC]

21. They Are A...

...bought& paid for wing of the RNC

[Ace Rothstein]

2. Rasmussen had R+1 on the general ballot the day of the midterms.

That's all you need to know about them.

[Laura PourMeADrink]

4. Yup ..just think it's crazy obvious. News should look at all

Polls and report on this outlier. It would drive him bonkers

[Downtown Hound]

5. I'm actually annoyed that 538 still includes Rasmussen in their averages

They should have dumped them years ago.

[Laura PourMeADrink]

12. Good point. Republicans probably use a fake poll to make it look closer to cover their suppression

[kurtcagle]

29. They give Rasmussen a C-

That in turn translates into a much smaller weighting than most other polling firms. Nate can't ignore them completely, but Ras doesn't factor significantly into most of their projections.

[Wounded Bear]

6. IIRC Rasmussen is normally 5-10% pro-Repub/Trump...

[Wellstone ruled]

7. Polling is only as

accurate as the Data Base one uses. And,the most important item is,are you using land lines and cell phones,or are you using land lines only and again same for cell phones.

[Laura PourMeADrink]

11. Do pollsters typically call cell phones? I thought they just applied a cell phone factor?

[Wellstone ruled]

14. My experience was,

we were given a script to read,actually,you practiced this script with a coach or monitor,and assigned a cubicle with a heads up display and head set with a mic,I was lucky enough to use a noise canceling mic and head set. Interesting how el cheapo boiler rooms use a open format with people sitting around tables or counters and all the background noise is transferred to the contacted interviewee.

Again,if it was a Political Poll,usually the Team Lead would screen Phone workers before turning them lose on the call board . The call sheets on Political Polling have so manner variables,depending on the possible desired out comes by who ever cut the check to the Polling Company. The Political Polling I did was off of Registered Voters,both Dem and Rethug. Now,there in lies some interesting issuers,you are calling the last known number on their Registration. Again,your call list might be only Persons who voted in a specific Election,and that could be a recent or some time in the past. In the end,found this whole process to be subjective rather than objective. It is all about the Sentence Structure of the questions you are asking. Notice the use of multiply Guess Questions by Conservative push polling used in 2018. And again,the answer choices were evasive or misleading.

[Awsi Dooger]

8. Rasmussen was 45% then 43% in two late January polls

You are correct it's not exactly apple to apples. The 37% Gallup poll was January 27th. Trump's approval has ticked upward since then so Gallup is still reflecting an old number. They'll likely be quite a bit higher on the next sample, although not close to Rasmussen level.

I don't care about specific polling firms as long as Trump's blended number on 530 stays below 42. He can't win from there unless our nominee implodes. If Trump reaches let's say 44% or 45% on 538, then we're not in great shape because the election day electorate will be several points more kind to him than a polling consensus.

I'm going to keep that 44% in mind throughout 2019 and 2020, without making any subjective rationalizations along the way. It is very easy to subjectively fool yourself. Meanwhile, the math and the fundamentals are not impressed.

Hopefully Trump continues to keep talking and tweeting and being himself, driving his number down into the high 30s, where 538 had it a couple of weeks ago.

[John Fante]

9. Reuters and Economist have him at 38% and 42% respectively and

those polls were conducted after his shit SOTU speech, so don't expect a huge bump from Gallup.

Rasmussen is hot garbage regardless of a few outliers results from last month. They're a joke.

[Laura PourMeADrink]

10. The fluff poll I saw off CBS was taken of watchers only - and they adored

him. Didn't delve into it - but it was something like people had to tell them they were going to watch? or something like that. Otherwise, they'd be all night getting people who say "What? SOTU, I'm not watching that idiot."

[DemocratSinceBirth]

13. Rasmussen is a crappo robo poller.

[Takket]

17. ass-puke-sin adds 15% to anything benefitting a rethug.

Should not even be considered a reliable source of polling. it is embarrassing any reputable media outlets even quote it.

[Laura PourMeADrink]

19. By design to cover suppression efforts

[spanone]

20. Must be polling on another Planet

[Yosemito]

22. All adults vs Likely voters

Rasmussen is the only pollster that uses the stupid "likely voters" screen for approval rating.

[Stinky The Clown]

23. Rasmussen includes a heavy Moron-American sampling

[Laura PourMeADrink]

24. Good one. Can I steal?

[Stinky The Clown]

27. Spread it far and wide!!

[DeminPennswoods]

25. Rassmussen sampled LVs who say they'll vote

in 2020 or something. It was a very deliberately culled sample to give Trump the highest possible approval rating. Gallup is probably a whole lot closer to the truth.

[Butterflylady]

26. You mean different planets.

[dlk]

28. The Poll Results Depend Entirely on the Polling Sample and if it is Statistically Significant

Polls aren't always accurate indicators. The biases of those doing the poll impact the results.