Trump Leads in ABC poll but Clinton is closing the gap

Discussion in 'Politics, Religion, Social Issues' started by Mac'nCheese, Aug 31, 2016.

  1. Mac'nCheese macrumors 68030

    Mac'nCheese

    Joined:
    Feb 9, 2010
    #1
  2. thermodynamic Suspended

    thermodynamic

    Joined:
    May 3, 2009
    Location:
    USA
    #2
    Yay, more than half the people in each poll* dislikes the candidate, how great America must look right now given all that dislike!

    * How does 56% of the 1,020 adults polled accurately extrapolate as meaning those 1,020 somehow represent the 200,000,000 pool of potential voters?
     
  3. aaronvan Suspended

    aaronvan

    Joined:
    Dec 21, 2011
    Location:
    República Cascadia
    #3
    For statistical analysis of a 200 million population, N = 1020 is perfectly adequate sample size.
     
  4. 1458279 Suspended

    1458279

    Joined:
    May 1, 2010
    Location:
    California
    #4
    It's stats and could be valid, but really most Americans don't think alike and how they poll really matters.

    The more divided the group or the more different opinions can throw things off. One of the problems now is that they don't work well with a mobile tech model.
     
  5. Plutonius macrumors 603

    Plutonius

    Joined:
    Feb 22, 2003
    Location:
    New Hampshire
    #5
  6. 1458279 Suspended

    1458279

    Joined:
    May 1, 2010
    Location:
    California
    #6
    There's a several decades long pattern with Hillary's popularity. She just doesn't do well with people, she never has. Bill was a natural, Hillary is not.

    The more she's out in the public, the worse she does. When she was SOS and we didn't hear much about her, she was popular. When she speaks and gets in front of people, her number go down.

    What's interesting is that the change in the number is the last 3 weeks. Looks like the emails and the FBI lie comment and Powell are starting to take hold.

    Going after the FBI, Powell and lying about lying shouldn't sit well with anyone.
     
  7. steve knight macrumors 68020

    steve knight

    Joined:
    Jan 28, 2009
    #7
    one poll found trump less popular then the middle seat on airplanes.
     
  8. ActionableMango macrumors 604

    ActionableMango

    Joined:
    Sep 21, 2010
    #8
    Yeah, that's pretty much how I feel about the election.

    [​IMG]
     
  9. samcraig macrumors P6

    Joined:
    Jun 22, 2009
    #9
    I would say that describes Trump.
     
  10. 1458279 Suspended

    1458279

    Joined:
    May 1, 2010
    Location:
    California
    #10
    Trump was SOS? I never knew he held a government job. Hmmm... They were doing polls on Trump back in the 90's? Musta missed that. Was he questioned by the FBI back in the 90's too?
     
  11. samcraig macrumors P6

    Joined:
    Jun 22, 2009
    #11
    No. The more he opens his mouth, the more repulsive he becomes. Case in point, tonight's TV appearance
     
  12. MadeTheSwitch macrumors 6502a

    MadeTheSwitch

    Joined:
    Apr 20, 2009
    #12
    It's bad tonight. Real bad. I got through about 45 minutes of it and finally had to mute it as he was literally giving me a headache. Literally. All he was doing was shouting at the audience and camera. I dare say he has never looked LESS presidential.

    But thank you Mr. Trump. Thank you for once again looking like an assclown. Thank you for once again alienating the Latino vote and doubling down on the policies that gave you trouble to begin with. Thank you for undoing whatever little good you might have done this afternoon. You know things aren't working when my mom (a Republican) calls me instead of watching his speech. LOL
     
  13. A.Goldberg macrumors 68000

    A.Goldberg

    Joined:
    Jan 31, 2015
    Location:
    Boston
    #13
    It probably doesn't, but if you look at any surveys for voting they generally don't have more than like 3,000 responds. Many of the voting have well under 1,000 subjects.

    I think what matters more is the population they are drawing from. If all 1,020 people are from Dallas, TX, the results would probably be very different than pooling only San Franciso, CA. Who knows how diverse this poll information sourced from.

    And actually the more important is do these surveyed beliefs actually have an impact on the voters and to what degree. A lot of people don't trust Hillary, but they'd rather see her as POTUS.

    That's why I believe it's best to look at sites that do a meta-analysis of sorts pooling survey results from numerous sources to yield a more realistic estimate.

    The media throws out statistics all the time that are flawed and/or have misleading conclusions.
     
  14. thekev macrumors 604

    thekev

    Joined:
    Aug 5, 2010
    #14
    It really depends on how well the sample is randomized. If any of those 200,000,000 voters are equally likely to be within that 1020, you may be able to approximate sentiment pretty well. Unfortunately it's nearly impossible to eliminate sample bias, especially in this kind of thing.
     
  15. 1458279 Suspended

    1458279

    Joined:
    May 1, 2010
    Location:
    California
    #15
    Aside from the problem that once you pick one, you've changed the odds for every pick after that, but that's another story. They solved that with the Draft (IIRC) they had to add someone back in or something.

    The real issue is how the reach each one of those selected. If you have a list of 200M people and each had a number assigned and you picked a random number, how would you contact that person? IIRC, one large polling company used to "call back" several times until they gave up trying for a given person. It became more and more expensive to actually make contact with each person selected because of several devices and blocking calls.

    They've tried other methods of contact, but haven't been able to get the same results as before when most used land lines and you didn't have call blocking. With all the world becoming spam, more and more people tune others out.

    Younger voters tune out more and repeat tries aren't very effective. This was the reason one large company left. This is also why the polls were off more in mid-terms as the turn out was lower.

    Some do focus groups and actually talk to people in depth, but that's costly.

    http://www.usnews.com/news/the-repo...lic-opinion-polls-are-increasingly-inaccurate

     
  16. thekev macrumors 604

    thekev

    Joined:
    Aug 5, 2010
    #16
    That was actually my point. Randomness isn't just picking a number. I said it's random if each of those 200M people are equally likely to be represented within the sample. That holds regardless of who is easier to reach and method of contact.

    I have a call blocking app on my phone just for that. Beyond just the data collection, surveys are expensive to write, edit, and evaluate. I've looked at some of the questions asked. In most (possibly all) cases I wouldn't be able to come up with that list over the course of a few hours.

    There is certainly going to be some trend away from direct sampling methods in favor of inferential ones. I've been meaning to look at one of twitter's open sourced packages, because I'm curious what they're doing. It looks like they're using shape space variant in their anomaly detection, which I find interesting.

    I haven't tried to gauge how this applies to Trump, but I get your skepticism. Skepticism is good.
     
  17. 1458279 Suspended

    1458279

    Joined:
    May 1, 2010
    Location:
    California
    #17
    I don't know much about the inferential ones, but I've heard they are using google search results. I don't know how they can do that because you don't really know why someone is searching for something and searches can be from bots.

    There's a trend that whoever is thought to be leading at the time, gets extra support because people want to pick the winner.

    Another issue is that many people aren't vocal about who they support. I've never had any signs or stickers for anyone running for office as it tends to offend people and it's not worth the hassle.

    I think these polls now will be worse than the last ones and they weren't good. The biggest deal is that people want a new direction.
     
  18. samcraig macrumors P6

    Joined:
    Jun 22, 2009
    #18
    You missed the point of my post yet again. Also they have rules about trolling.
     
  19. 1458279 Suspended

    1458279

    Joined:
    May 1, 2010
    Location:
    California
    #19
    Your post:
    "No. The more he opens his mouth, the more repulsive he becomes. Case in point, tonight's TV appearance"

    The whole thing was seen as bold and very presidential by news on both sides and the polls went into dead heat.

    The point of you post was yet another "AutoBastTrump", it had no substance. Let's break down the post in question:
    "No" <- shows disagreement, ok let's see the support of the disagreement, shall we?
    "The more he opens his mouth, the more repulsive he becomes" <- completely void of any examples and contradicts the new polls and news reports from both sides.
    "Case in point, tonight's TV appearance" <- support of the "autobash" sites one of the boldest and best moves Trump has made yet.

    Even the generic "The more he opens his mouth, the more repulsive he becomes" just means:
     

    Attached Files:

  20. WarHeadz macrumors 6502a

    WarHeadz

    Joined:
    Aug 30, 2015
    Location:
    Los Angeles
  21. MadeTheSwitch macrumors 6502a

    MadeTheSwitch

    Joined:
    Apr 20, 2009
    #22
    How do you know? Polling that you say is unreliable? Source please.
     
  22. Mac'nCheese thread starter macrumors 68030

    Mac'nCheese

    Joined:
    Feb 9, 2010
    #23
    People always say that and then they vote for he same old people and parties anyway. Its like people who are always looking for the newest coolest place to eat out and then just order the chicken parm. Jesus Christ, why did I have to drive a half hour for you to just order the chicken parm again!?!?
     
  23. MadeTheSwitch macrumors 6502a

    MadeTheSwitch

    Joined:
    Apr 20, 2009
    #24
    Because maybe their chicken parm is better? LOL
     
  24. 1458279 Suspended

    1458279

    Joined:
    May 1, 2010
    Location:
    California
    #25
    LOL, I can see that.
     

Share This Page