Liability
Locked Account.
So they do a random sample, but the folks at the WSJ who helped fund it request that they exclude a certain number of Republican responses?
Hang up when someone says they are a Republican?
One thing at a time, bub.
First, it is time to drop the "IF." They once again managed to conduct a random telephone poll of REGISTERED voters that DID over-sample for Dims:
-- http://msnbcmedia.msn.com/i/MSNBC/S...y_Stories_Teases/August_NBC-WSJ_Int_Sched.pdfQF4a Generally speaking, do you think of yourself as (ROTATE a Democrat, a Republican, an independent, or
something else? (IF "DEMOCRAT" OR "REPUBLICAN," ASK Would you call yourself a strong
(Democrat/Republican) or not a very strong (Democrat/Republican)? (IF "NOT SURE," CODE AS "NOT
VERY STRONG DEMOCRAT/REPUBLICAN.") (IF "INDEPENDENT," ASK Do you think of yourself as
closer to the Republican Party, closer to the Democratic Party, or do you think of yourself as strictly
independent? (IF "NOT SURE," CODE AS "STRICTLY INDEPENDENT.")+
Strong Democrat ............................................... 24
Not very strong Democrat ................................. 10
Independent/lean Democrat .............................. 11
Strictly Independent .......................................... 13
Independent/lean Republican ............................ 12
Not very strong Republican ............................... 9
Strong Republican ............................................. 18
Other (VOL) ....................................................... 3
Not sure ........................................................... 1
+ Results shown reflect responses among registered voters.
SECONDLY, it is possible to take the over-sampling into account when analyzing the results. Was such an effort attempted?
Again, Liability: Please read my question carefully. It's an honest question:
How do you know they over-sampled Democrats?
In order to know this, you would need to know the true, accurate weighting of Dems and Reps in the US. There's only one way to achieve that - a census.
Since no one is going to fund a census, we use sampling. For instance, we could sample a couple thousand random people and get a breakdown to use - we could do that and achieve an MOE of 3% or so, with 95% of samples expected to fall within that MOE
Stop. You sound silly.
There are these things we call VOTER REGISTRATIONS.
For almost any given population, it is fucking easy to know the relative percentage of voters who are registered as Dims, Republicans, Liberals, Conservatives, Socialists, Communist Worker Party sheep, Independent, etc.