The latest Marquette Law School poll showed walker leading by 6%. But if you look at the poll, there are strong suggestions it might be significantly oversampling right-wingers (for whatever reason).
The 704 respondents broke down as:
Very Conservative or Conservative…..47%
Moderate…..30%
Very liberal or Liberal…..19%
I don’t know what the actual self-identification rates are for the state, but that certainly appears skewed at first blush.
Perhaps more significantly, respondents voted in 2010 for Walker over Barrett 52-41--whereas the actual 2010 vote (which I believe sucked turnout-wise for Dems) was 52-47 for Walker.
Anyone who looked at that poll should have wondered about that rather large (to me) sample skew before reporting it as “Walker leads…” More like “Walker leads in conservative sample…”(!)
I’m basing this on a quick review of the full poll results here. There are possible explanations for these numbers being accurate, but I haven’t seen any offered yet.
...
To wrap this up, my brother graduated top 4 in his college for sociology.
This was his area of expertise.
I give full credit for this disturbing revelation to him and I have great faith in his familiarity with the subject.
2:18 PM PT: I thought this was incredibly relevant so I thought I should bounce it up here.. Perhaps the problem isn't OVER sampling right wingers but rather UNDER sampling the opposition to walker.
Comment by jamess:
I heard on the Ed Schultz Radio Show on Thursday, that a Caller from Wisconsin:
... got a phone survey phone call, during the time of the Marquette Poll. (wasn't sure if it was them though).
The Pollster asked him if he had been involved in or otherwise "supported the Wisconsin recall effort?"
When he answered "Yes",
the Pollster abruptly ended the Survey --
saying "they were trying to conduct an unbiased poll."
2:19 PM PT: I thought this was incredibly relevant so I thought I should bounce it up here.. Perhaps the problem isn't OVER sampling right wingers but rather UNDER sampling the opposition to walker.
Comment by jamess:
I heard on the Ed Schultz Radio Show on Thursday, that a Caller from Wisconsin:
... got a phone survey phone call, during the time of the Marquette Poll. (wasn't sure if it was them though).
The Pollster asked him if he had been involved in or otherwise "supported the Wisconsin recall effort?"
When he answered "Yes",
the Pollster abruptly ended the Survey --
saying "they were trying to conduct an unbiased poll."