The Trump Impeachment Thread
Post Reply
 
Thread Rating:
  • 1 Votes - 5 Average
  • 1
  • 2
  • 3
  • 4
  • 5
16-07-2017, 09:08 PM (This post was last modified: 16-07-2017 09:17 PM by GirlyMan.)
RE: The Trump Impeachment Thread
(16-07-2017 08:28 PM)Bucky Ball Wrote:  
(16-07-2017 01:59 PM)Lord Dark Helmet Wrote:  Stupid polls.

Wait what ?

I've already explained to him several times that he does not understand how basic statistical sampling works. His response, if I recall correctly, was "I don't need fucking statistics. It's basic math." All I can conclude at this point is he doesn't understand how basic math works either, let alone basic statistics.

What he should be considering is under what conditions would a properly conducted scientific poll with rigorous sampling methods result in a sample population with 12% more Democrats than Republicans? It's possible, but extremely unlikely, that this sample is more than 2 standard deviations away from the mean. The more polls in fairly close agreement there are, the more unlikely it is. It is most certainly not the case that the professionals who do this for a living intentionally biased their sample. Professional scientific communities police themselves.

#sigh
Find all posts by this user
Like Post Quote this message in a reply
[+] 8 users Like GirlyMan's post
16-07-2017, 09:32 PM
RE: The Trump Impeachment Thread
(16-07-2017 09:08 PM)GirlyMan Wrote:  It is most certainly not the case that the professionals who do this for a living intentionally biased their sample.
Nate Silver includes a "House effect" weighting in his rating of poll organizations; while the biases that show up may not be overtly intentional, bias is notoriously difficult to wholly damp. But you're certainly correct that professional organizations would be shooting their feet off if they deliberately distorted their product; it'd kill their credibility and shorten their client list.
Find all posts by this user
Like Post Quote this message in a reply
16-07-2017, 11:12 PM
RE: The Trump Impeachment Thread
(16-07-2017 09:32 PM)Airportkid Wrote:  ...
Nate Silver
...

You guys have been duped. Five Thirty-eight has its thumb on the scale of justice.

Well, actually, more than a thumb!

quran 5:38 Wrote:As for the thief, both male and female, cut off their hands. It is the reward of their own deeds, an exemplary punishment from Allah. Allah is Mighty, Wise.

Find all posts by this user
Like Post Quote this message in a reply
16-07-2017, 11:22 PM
RE: The Trump Impeachment Thread
(16-07-2017 09:08 PM)GirlyMan Wrote:  
(16-07-2017 08:28 PM)Bucky Ball Wrote:  Wait what ?

I've already explained to him several times that he does not understand how basic statistical sampling works. His response, if I recall correctly, was "I don't need fucking statistics. It's basic math." All I can conclude at this point is he doesn't understand how basic math works either, let alone basic statistics.

What he should be considering is under what conditions would a properly conducted scientific poll with rigorous sampling methods result in a sample population with 12% more Democrats than Republicans? It's possible, but extremely unlikely, that this sample is more than 2 standard deviations away from the mean. The more polls in fairly close agreement there are, the more unlikely it is. It is most certainly not the case that the professionals who do this for a living intentionally biased their sample. Professional scientific communities police themselves.

If they were intending to deceive, reporting that their sample is 12% more democrats is a bit of a strange way of doing so. Also I hate when people do that - report it as "12% more". They should rather say that it's 56% democrat, 44% republican. Or maybe they did and Lord "Clear as Mud" Dark Helmet reported it as "12% more"?

We'll love you just the way you are
If you're perfect -- Alanis Morissette
(06-02-2014 03:47 PM)Momsurroundedbyboys Wrote:  And I'm giving myself a conclusion again from all the facepalming.
Find all posts by this user
Like Post Quote this message in a reply
[+] 1 user Likes morondog's post
16-07-2017, 11:45 PM
RE: The Trump Impeachment Thread
(16-07-2017 11:22 PM)morondog Wrote:  If they were intending to deceive, reporting that their sample is 12% more democrats is a bit of a strange way of doing so. Also I hate when people do that - report it as "12% more". They should rather say that it's 56% democrat, 44% republican. Or maybe they did and Lord "Clear as Mud" Dark Helmet reported it as "12% more"?

Plus, let's not forget that Hillary had nearly 3 million more votes. Trump did not win a majority of the electorate, and yet LDH is dumbfounded by Trump's approval numbers? That is breathtaking in its inanity.

[Image: E3WvRwZ.gif]
Find all posts by this user
Like Post Quote this message in a reply
[+] 1 user Likes EvolutionKills's post
17-07-2017, 01:34 AM
RE: The Trump Impeachment Thread
(16-07-2017 11:45 PM)EvolutionKills Wrote:  
(16-07-2017 11:22 PM)morondog Wrote:  If they were intending to deceive, reporting that their sample is 12% more democrats is a bit of a strange way of doing so. Also I hate when people do that - report it as "12% more". They should rather say that it's 56% democrat, 44% republican. Or maybe they did and Lord "Clear as Mud" Dark Helmet reported it as "12% more"?

Plus, let's not forget that Hillary had nearly 3 million more votes. Trump did not win a majority of the electorate, and yet LDH is dumbfounded by Trump's approval numbers? That is breathtaking in its inanity.

Let's say, hypothetically, that democrats vs republicans are not evenly 50/50 in the population. Then *any* honestly randomly selected survey should most probably pick up different numbers of democrats vs republicans. Of course because it's randomly selected there's a chance that it won't reflect the underlying population, but the whole point of multiple polls and polling large numbers of people is to try to correct for that bias. There *are* plenty of ways to cheat - for example you can select a sample of 100 "randomly chosen" people whom you know, or you can keep selecting samples until you get a sample whose views average to the conclusion you want to support, and only report that one. You can also use a small sample size and only report percentages, and hope that no one asks you for the actual numbers. Percentages are one of the best ways to cheat because you can take them relative to something stupid. For example, if your poll support for Trump increased from 36% to 37% you could report that as a 3% increase in Trump support, by using 1/37 as your increase, rather than the more honest 1% increase.

We'll love you just the way you are
If you're perfect -- Alanis Morissette
(06-02-2014 03:47 PM)Momsurroundedbyboys Wrote:  And I'm giving myself a conclusion again from all the facepalming.
Find all posts by this user
Like Post Quote this message in a reply
[+] 2 users Like morondog's post
17-07-2017, 11:28 PM
RE: The Trump Impeachment Thread
Oh my. 41% already. And Mueller's just getting started. Only 24% supported Nixon's impeachment at the start of his pee tape scandal.

#sigh
Find all posts by this user
Like Post Quote this message in a reply
[+] 1 user Likes GirlyMan's post
18-07-2017, 04:57 AM
RE: The Trump Impeachment Thread
(17-07-2017 11:28 PM)GirlyMan Wrote:  Oh my. 41% already. And Mueller's just getting started. Only 24% supported Nixon's impeachment at the start of his pee tape scandal.

Keep in mind however, that was in the era before the internet's incessant echo-chamber, and before the propaganda mill that is Faux News...




[Image: E3WvRwZ.gif]
Find all posts by this user
Like Post Quote this message in a reply
19-07-2017, 05:51 PM
RE: The Trump Impeachment Thread
This is an interesting article from Newsweek.

http://www.newsweek.com/trump-white-hous...ton-638448

"Trump was never a master builder, entrepreneur or deal-maker. He is good at branding, and at selling that brand: Make America Great Again, Build the Wall, Lock Her Up. Liddle Marco, Lyin’ Ted, Crooked Hillary. But the branding that worked so well during the campaign works a lot less well when your customers are legislators who not only have to have a grasp of policy but to explain their votes to constituents back home."

On his incessant tweeting and golfing....

"That would all be fine if both tweeting and golfing were steam valves for a president otherwise furiously engaged in the greatness-restoring project he promised his supporters. That is not, however, the portrait that emerges from well-sourced White House reports."

Shakespeare's Comedy of Errors.... on Donald J. Trump:

He is deformed, crooked, old, and sere,
Ill-fac’d, worse bodied, shapeless every where;
Vicious, ungentle, foolish, blunt, unkind,
Stigmatical in making, worse in mind.
Find all posts by this user
Like Post Quote this message in a reply
[+] 5 users Like dancefortwo's post
19-07-2017, 07:40 PM (This post was last modified: 19-07-2017 08:30 PM by Bucky Ball.)
RE: The Trump Impeachment Thread
So this is about to spiral down to a new level of scary. Trump said today that it was "unfair to the president" for Sessions to recuse himself, (*as if* the Narcissist-in-Chief never heard the word "ethics"), and that he might fire Mueller if he goes after his family finances, (which is what this has been all about all along, ... money, and loans from oligarchs and money laundering). He threatened a Republican senator sitting next to him today, to vote as Dumpy wanted, or lose his job. He thinks he's still on TV, giving orders on The Apprentice. His astounding unfitness for office, revels itself more and more every day.

Insufferable know-it-all.Einstein God has a plan for us. Please stop screwing it up with your prayers.
Find all posts by this user
Like Post Quote this message in a reply
[+] 4 users Like Bucky Ball's post
Post Reply
Forum Jump: