Another fake Democrat poll exposed

Arminius said:
Another provably fake poll. This time, it's Quinnipac. They had to over-sample registered Democrats by 10% to manufacture their new claim about the child-molesting Alzheimer's patient now having "double-digit lead." They sampled 36% registered Democrats and 26% registered Republicans to make this happen.

This is why you ALWAYS click on the methodology and see how they got their numbers. This is the same kind of outright lie that caused people to think Hillary had a 92% chance of beating President Trump.

View attachment 15384

Biden Holds Double-Digit Lead Over Trump, New Poll Shows

Imagine the headlines conservatives could fabricate by wildly over-sampling registered Republicans like this:

-Majority of Americans want Obama administration officials prosecuted for treason

-Those who want to finish building the wall have a "double-digit lead" over those who do not

-American people overwhelmingly want power-abusing Democrat judges impeached and purged from federal judiciary

:laugh:
Click to expand...

This survey uses statistical weighting procedures to account for deviations in the survey sample from known population characteristics, which helps correct for differential survey participation and random variation in samples. The overall adult sample is weighted to recent Census data using a sample balancing procedure to match the demographic makeup of the population by region, gender, age, education and race.When including thedesign effect, the margin of sampling error for this study of registered voters is +/-3.3percentage points.
Click to expand...
.
 
giphy.gif
 
it's really not that hard to understand you fucking dummies.

It's called weighting.

Lets say you know the population is 500 democrats, 500 republicans.

You sample 50 people where 30 are democrat and 20 are republican. i.e. 60% democrat vs. 40% republican.

BUT - we know the population data is exactly 50/50, so what do we do if we are not a retard on JPP that doesn't understand basic stats? :thinking:

We weight the responses accordingly. Thus every democrat answer is worth .8x and every republican answer is worth 1.25x and
This brings the sample into line with the known population

I'm pretty sure I learned this stuff in 4th grade.
 
Cohort said:
it's really not that hard to understand you fucking dummies.

It's called weighting.

Lets say you know the population is 500 democrats, 500 republicans.

You sample 50 people where 30 are democrat and 20 are republican. i.e. 60% democrat vs. 40% republican.

BUT - we know the population data is exactly 50/50, so what do we do if we are not a retard on JPP that doesn't understand basic stats? :thinking:

We weight the responses accordingly. Thus every democrat answer is worth .8x and every republican answer is worth 1.25x and
This brings the sample into line with the known population

I'm pretty sure I learned this stuff in 4th grade.
Click to expand...

Don't expect retards to understand it.
 
Grokmaster said:
Quinipiac's OWN METHODOLGY/DEMOGRAPHICS reveal the BIAS. READ THE LINK, LIAR.


PARTY IDENTIFICATION Republican 26% Democrat 36 Independent 31 Other/DK/NA 7




MAKEUP SOME MORE BULLSHIT.
Click to expand...

LOL, the sample is the sample. The breakdown is very likely correct. The margin of error is very small on the poll. It's like saying they oversampled people that don't like Trump. No, there are just lots of people that don't like Trump.
 
Cohort said:
it's really not that hard to understand you fucking dummies.

It's called weighting.

Lets say you know the population is 500 democrats, 500 republicans.

You sample 50 people where 30 are democrat and 20 are republican. i.e. 60% democrat vs. 40% republican.

BUT - we know the population data is exactly 50/50, so what do we do if we are not a retard on JPP that doesn't understand basic stats? :thinking:

We weight the responses accordingly. Thus every democrat answer is worth .8x and every republican answer is worth 1.25x and
This brings the sample into line with the known population

I'm pretty sure I learned this stuff in 4th grade.
Click to expand...

I would say 'yes and no' to this. For demographics like age, gender, race, etc, there is a relatively fixed baseline. For party affiliation, that baseline changes and often changes dramatically over time. So in cases like this, they pollster may very well go with whatever breakdown the sample indicates. I know there is some kind of algorithm and likely a baseline they are working from as a starting point, but I think that type of demographic is less likely to be weighted to a static number.
 
Arminius said:
Another provably fake poll. This time, it's Quinnipac. They had to over-sample registered Democrats by 10% to manufacture their new claim about the child-molesting Alzheimer's patient now having "double-digit lead." They sampled 36% registered Democrats and 26% registered Republicans to make this happen.

This is why you ALWAYS click on the methodology and see how they got their numbers. This is the same kind of outright lie that caused people to think Hillary had a 92% chance of beating President Trump.

View attachment 15384

Biden Holds Double-Digit Lead Over Trump, New Poll Shows

Imagine the headlines conservatives could fabricate by wildly over-sampling registered Republicans like this:

-Majority of Americans want Obama administration officials prosecuted for treason

-Those who want to finish building the wall have a "double-digit lead" over those who do not

-American people overwhelmingly want power-abusing Democrat judges impeached and purged from federal judiciary

:laugh:
Click to expand...
Indeed.
 
MSM and their think tanks are direct branches of the DNC ministry of propaganda and need to be regulated as in kind campaign contributors.
 
Cohort said:
This survey uses statistical weighting procedures to account for deviations in the survey sample from known population characteristics, which helps correct for differential survey participation and random variation in samples. The overall adult sample is weighted to recent Census data using a sample balancing procedure to match the demographic makeup of the population by region, gender, age, education and race.When including thedesign effect, the margin of sampling error for this study of registered voters is +/-3.3percentage points.
Click to expand...

So, they assure us that their wildly biased sampling in favor of Democrats has nothing to do with the wildly disproportionate results, which never end up matching reality?

nyt hillary election.jpg

How comforting. :laugh:
 
Concart said:
LOL, the sample is the sample. The breakdown is very likely correct. The margin of error is very small on the poll. It's like saying they oversampled people that don't like Trump. No, there are just lots of people that don't like Trump.
Click to expand...

Hence the results never matching reality. Like all the profound excitement we're supposed to believe exists for a child-molesting racist crook and Alzheimer's patient like Biden. The left has already proven itself to have zero enthusiasm for Biden and to be willing to take unprecedented steps to frame, sabotage, lie about, and overthrow Trump at any cost. Why would their notoriously wrong polling be the one thing they suddenly decide to have an iota of integrity about?

:bs:
 
Cohort said:
it's important to note that that 92-8 number (which was wrong) was not a poll, it was an extrapolation of data.
Click to expand...

Was it wrong? Things that have an 8% chance of happening happen once out of every 12 times.

I am not saying it was right, but there is no evidence it was wrong.
 
Walt said:
Was it wrong? Things that have an 8% chance of happening happen once out of every 12 times.

I am not saying it was right, but there is no evidence it was wrong.
Click to expand...

I would say it was wrong because it vastly overestimated clintons chances, most metrics never had her anywhere near that high. silver had her only at like 68%

to say someone is 92% to win when they are predicated to have a +2 margin nationally is just totally foolish
 
Arminius said:


Exactly. They're all unscientific bullshit. So why do Democrats depend on them so heavily to misinform people? If your ideas can compete on a level playing field, why resort to such gutter trash nonsense all the time?

:thinking:
Click to expand...


Why do you think only Democrats follow polls? Go back and read the posts here where righties are saying trump has his highest ratings ever, according to the polls (that they want to believe.)
 
Walt said:
Was it wrong? Things that have an 8% chance of happening happen once out of every 12 times.

I am not saying it was right, but there is no evidence it was wrong.
Click to expand...

I think 538 had more reasonable odds (74% Clinton). You have hit on the stupidity of the argument being made about polling in 2016, In fact, the final polls were fairly accurate, predicting a Clinton win by 3.3%. She actually won the popular vote by 2.1%. So when a member of the cult uses some four year old pundit predictions to cast doubt on all polling, it makes me laugh over the abject stupidity of the argument.
 
You must log in or register to reply here.
Back
Top