Harvard Harris Poll-Trump at 48%

Alaxlmartin

macrumors regular
Original poster
Feb 20, 2018
184
79
Forty-eight percent of Americans say they approve of President Trump's job in office, according to a new American Barometer survey.

The poll showed 24 percent of respondents said they "strongly approved" of Trump, while another 24 percent said they "somewhat approve" of him.

Fifty-two percent of Americans said they disapproved of the president, with 37 percent saying they "strongly disapprove" of him and 15 percent saying they "somewhat disapprove" of him.

The American Barometer is a daily survey conducted by the HarrisX polling company on behalf of The Hill's new video division, Hill.TV

https://apple.news/AhngOOhIxSqOphtGSJVm-BA
 

jkcerda

macrumors 6502a
Jun 10, 2013
682
38,980
Criminal Mexi Midget

GermanSuplex

macrumors 6502a
Aug 26, 2009
963
9,910
Reminds me of when Trump posted that picture of him in all his glory when some random poll had him at 50% for about one day. His dissaproval number has been consistently higher than his approval numbers. Haven't checked any aggregate polls, but his numbers are usually pretty bad.

*We get it. Hillary was predicted to win and she lost. They have a margin of error for a reason...
 
  • Like
Reactions: Eraserhead

RichardMZhlubb

Contributor
Nov 26, 2010
209
14,801
Washington, DC
The problem wasn’t the polls, it was analysts drawing the wrong conclusions from the polls. That Reuters/Ipsos poll showed Clinton with a 3 percentage point lead nationally, which was very close to the actual vote. And the Newsweek poll had Hillary ahead by 8, but it was from September, before the race tightened after the Comey letter re-opening the FBI investigation. Final polling had Hillary up by around 4 points and she won the popular vote by around 2 points. That was actually a narrower variation between polling and the actual vote than in several recent previous presidential elections. But, polling errors of just a few percentage points in a handful of states were what produced the unexpected result of Trump winning. As I noted, Nate Silver was just about the only analyst who correctly anticipated that possibility, giving Trump about a 30 percent chance of winning. It would have been wrong to have declared Trump the favorite, but it was also wrong to say he had virtually no chance.
 

samcraig

macrumors P6
Jun 22, 2009
16,610
35,008
USA

mac_in_tosh

macrumors 6502
Nov 6, 2016
351
4,553
Earth
Improbable things happen all the time. But they won't happen often if the relevant "test" is repeated many times.
 

Herdfan

macrumors 6502
Apr 11, 2011
267
3,842
National polls are meaningless in a Presidential race. You simply need to poll the swing states and be done. So poll FL, NC, VA, PA, MI, OH, WI and maybe AZ and NV. Every other state is pretty much locked in.
 

chown33

Moderator
Staff member
Aug 9, 2009
8,362
4,346
Pumpkindale
90% is not the same thing as 100%.
There's also a big difference between "90% chance of winning" (which is the headline) and "winning 90% of the vote" (which would be an actual landslide).

The first is a prediction about the probablity of an event occurring. Without knowing how it was arrived at, its basis could be anything from mathematical models to examining entrails.

The second is a prediction about the margin of winning should the event occur. As a prediction, it's contingent on the event occurring, and says nothing at all about the probability of the event itself. Again, it could be based on a mathematical model, or another reading of chicken entrails.

It's entirely possible to have a 90% chance of winning with a concurrent prediction of winning only 55% of the vote. Neither one contradicts the other.

It's also possible to have a 51% chance of winning with a concurrent prediction of winning 90% of the vote. Still not a contradiction, but I'd be raising my eyebrows about the confidence factor for both numbers.

To simplify somewhat, it's like the difference between the odds of a team winning a game, as compared to a prediction about the point spread.

Or in a horse race, it's the difference between the odds of winning, as compared to the distance between the winner and the placing horse. "Win by a nose" does not contradict "10:1 odds".
 
  • Like
Reactions: mgguy

A.Goldberg

macrumors 68020
Jan 31, 2015
2,326
7,638
Boston
90% is not the same thing as 100%.
This is the one thing I don’t understand. Trump had a very slim margin to win, but not a statistical impossibility. If someone told you your mother had a 90% of a cancer treatment working, would you not still be worried about that 10% or 1 in 10 chance of treatment failure.

That said, I suspect the polls may have been slightly flawed due to the social stigma of supporting trump. But regardless, as you mentioned 90% does not mean 100%.
 

s2mikey

macrumors 68020
Sep 23, 2013
2,462
2,520
Upstate, NY
The problem wasn’t the polls, it was analysts drawing the wrong conclusions from the polls. That Reuters/Ipsos poll showed Clinton with a 3 percentage point lead nationally, which was very close to the actual vote. And the Newsweek poll had Hillary ahead by 8, but it was from September, before the race tightened after the Comey letter re-opening the FBI investigation. Final polling had Hillary up by around 4 points and she won the popular vote by around 2 points. That was actually a narrower variation between polling and the actual vote than in several recent previous presidential elections. But, polling errors of just a few percentage points in a handful of states were what produced the unexpected result of Trump winning. As I noted, Nate Silver was just about the only analyst who correctly anticipated that possibility, giving Trump about a 30 percent chance of winning. It would have been wrong to have declared Trump the favorite, but it was also wrong to say he had virtually no chance.
Part of what happend too is that once the Trump voters got out of WORK they went to the polls. Most Hillary voters, in leiu of nothing better to do could hit the voting booths early. :D
 
  • Like
Reactions: Herdfan