Party affiliation

[Darkwind]

Go out on the street and ask three people

Does Gallup know you are giving away trade secrets?

 

[Darkwind]

What is your criteria for the quality of the information? What are you planning to use it for?

There is this link which I assume is four years old.
How many registered Democrats are there in the US

I'm trying to figure out why people think that every poll "over samples" Democrats. In order to know if a poll over sampled Democrats, you would need to know what the current breakdown by party affiliation actually is.

wow, really?

The pollster will ask the person being polled how they consider themselves...D, R, I.

 

[8537]

What is your criteria for the quality of the information? What are you planning to use it for?

There is this link which I assume is four years old.
How many registered Democrats are there in the US

I'm trying to figure out why people think that every poll "over samples" Democrats. In order to know if a poll over sampled Democrats, you would need to know what the current breakdown by party affiliation actually is.

wow, really?

The pollster will ask the person being polled how they consider themselves...D, R, I.

Exactly! You are EXACTLY right! Somebody gets it.

The best reasonable way to assess the breakdown of D's, R's and I's is to ask them in a scientific random sampling poll. Precisely!

Really.

 

[Darkwind]

I'm trying to figure out why people think that every poll "over samples" Democrats. In order to know if a poll over sampled Democrats, you would need to know what the current breakdown by party affiliation actually is.

wow, really?

The pollster will ask the person being polled how they consider themselves...D, R, I.

Exactly! You are EXACTLY right! Somebody gets it.

The best reasonable way to assess the breakdown of D's, R's and I's is to ask them in a scientific random sampling poll. Precisely!

so, when you look at the polling data and you see they polled 951 democrats, and 37 independants, and 5 republicans, and the poll skews liberal.......


???

 

[8537]

wow, really?

The pollster will ask the person being polled how they consider themselves...D, R, I.

Exactly! You are EXACTLY right! Somebody gets it.

The best reasonable way to assess the breakdown of D's, R's and I's is to ask them in a scientific random sampling poll. Precisely!

so, when you look at the polling data and you see they polled 951 democrats, and 37 independants, and 5 republicans, and the poll skews liberal.......


???

I've never seen such a poll.

On the other hand, I've seen a couple dozen random sample polls regarding party affiliation in recent months. In almost every case, the results show that democrats outnumber republicans by between about 2 and 12 percentage points - and the results from almost every poll falls within the MoE of the other ones. Polls from Pew, The WSJ, CNN, Fox etc...

In the reality based community, that serves as a very, very strong indication that people who identify as Dems are between 2 and 12 points more common than people who identify as Republican.

 

[emptystep]

According to the 2008 data there are about 20% more Dems, national average. So the 2 to 12 percent is conservative on the average. How well you can trust any of these numbers is the question.

 

[8537]

According to the 2008 data there are about 20% more Dems, national average. So the 2 to 12 percent is conservative on the average. How well you can trust any of these numbers is the question.

Well you have to get the baseline figures from...somewhere...and assuming a 50:50 split is senseless.

So instead, you use the best data source you can find - a series of scientifically administered nationwide random sample polls.