Since I mentioned a couple of polls in the previous post, it would not hurt to talk about the poll that Markos commissioned.  You see, Markos is writing a book about his preconceived notions about the political party he disagrees with.  In order to make sure that his "data" about the opposition is correct he commissioned a poll.  I'll let Steve Pendlebury of AOL News take over the story from here.

Most Republicans think Barack Obama is a socialist and Sarah Palin is more qualified than him to be president. Most either think Obama wasn't born in the U.S. or aren't sure. Nearly 40 percent want him impeached. Those are a few of the results in a Daily Kos poll that's shaking up the blogosphere.

The liberal Web site commissioned Research 2000 to survey more than 2,000 self-identified Republicans last month. Site founder Markos Moulitsas -- the "Kos" in Daily Kos -- called the results "nothing short of startling."

A few more nuggets from the poll:

• 21 percent believe the Association of Community Organizations for Reform Now (ACORN) stole the 2008 election and 55 percent aren't sure.
• 31 percent believe Obama is a racist who hates white people and 33 percent aren't sure.
• 23 percent want their state to secede from the U.S. and 19 percent aren't sure.
• Only 26 percent favor letting openly gay Americans serve in the military.
• 67 percent believe the only way to get to heaven is through Jesus Christ.

On his blog, Politico's Ben Smith said the poll was conducted "more or less with the goal of making Republicans look extreme" and "does a pretty good job of that." Daily Kos responded Wednesday, "Actually, the poll just asked questions of self-identified Republicans. The answers made Republicans look extreme."

Reactions from the left were swift and predictable....

"If this poll is accurate (and Research 2000 has a good reputation for accuracy), the Republican base has an absurdly high percentage of wackos of all stripes, from religious fanatics to creationists to secessionists to conspiracy theorists," said Charles Johnson on Little Green Footballs.

While others were more skeptical in their response...

"One has to wonder if Republicans really believe that Obama stole the election, was born in Kenya and that all non-Christians are going to hell," noted Matt Sussman, Technorati politics channel editor. "Granted, several of the questions are loaded, and Kos himself may be reading too much into the 'undecided' pool, but the results are nonetheless startling.

"What's missing from the research is context, Darren Lenard Hutchinson observed on his blog Dissenting Justice. "It does not provide comparative results for self-identified Democrats and Independents. So, while Kos and many Daily Kos readers believe that the poll proves the extremism among Republicans, it is difficult to make this conclusion without comparative data," he said.

Hutchinson and Moderate Voice's David Adesnik both pointed out that the Kos poll's results on questions about same-sex marriage and gays in the military were far different from those from recent Gallup and Pew surveys.

Mr. Hutchinson is correct in that statement but as Mr. Pendlebury mentions in his article...

NewsBusters' Scott Whitlock slammed MSNBC's David Shuster for using the poll to declare that most Republicans are "birthers" who think Obama isn't really an American citizen.

"So, one poll, by a left-wing website (in conjunction with the firm Research 2000), is enough for MSNBC to assert that 58 percent of GOPers subscribe to a bizarre conspiracy?" Whitlock fumed. "A Rasmussen poll from May of 2007 found that 61 percent of Democrats either believed that George Bush knew about the 9/11 terror attack in advance or aren't sure. Does that mean that 'most Democrats' are Truthers?"

Emphasis mine.  A very valid question.

Jim Geraghty offers some very valid points about criticism on the sampling.

The pollster, Research 2000, is legitimate, but some are asking fair questions about whether the sample skews too Southern, too male, and too old to accurately represent Republicans as a whole.

I'd note that pollsters generally agree that the number of self-identified Republicans shrank since, say, 2004, and while GOP candidates are doing better, you're not seeing a comeback among voters who want to identify as Republicans. In other words, a lot of conservatives left the GOP during Bush's second term, and would see themselves as independent, and not identify as a Republican if called by Research 2000. The folks who answered the phones for that pollster are very, very, very loyal Republicans.

Beyond that, I don't find too many of these findings that surprising or even that troubling.

Another very valid point.

Mr. Pendlebury then goes on to mention the skewed methodology of this poll, which long time readers of this blog know is where I have the tendency to "geek out".  However I am not going to go there today, instead I am going to go right to the horses mouth because while Mr. Pendlebury hits on some very valid methodology flaws, what I was more interested in was how Markos got to the conclusions that he did.  For example, the MSNBC claimed that 58% of the GOP base are "birthers".  Kos asked in question #4 " Do you believe Barack Obama was born in the United States or not?"  Your choices are a simple "yes", "no", or "Not Sure".  42% of the responders said "Yes" they believed that Barack Obama was born in the United States and 22% said "Not Sure".  In order to get the interpretation that 58% of the Republican base "Believed" that Barack Obama was not born in the US, MSNBC (and Kos)  had to add that 22% that answered "Not Sure" to the 36% that answered "NO".  That is simply not an accurate reading of the data.  Many people that I know say that they are unsure on this issue because they simply do not know or care to know the answer.  That does not mean they believe in the "birthers" theory because they simply (say it with me.....) do not know!!!!!

Another one of Markos' gems is the 55% that say that they "don't know" if ACORN "stole" the 2008 election.  Here is a hot tip for you all...the reason that we "don't know" is because there is absolutely, positively no proof that they did or did not.  Oh to be certain, there were some instances were voter fraud happened in some states, but was it part of a concerted conspiracy to defraud the American people - there is no proof to say there was so we simply do not know that for a fact!

One of the things that I noticed about the methodology is that it was set up specifically to get high "undecided" or "don't know" answers.  Because those answers are the ones that are the easiest for the reader to "interpret" to fit his/her own biases.  Take for instance question #4 (again).  That 22% that said that they were not sure could just have easily been added to the 42% that said that they believed the President was born in the US to show "proof positive" that the "birthers" are a minority of Republicans and you would be more correct than David Schuster was in adding it to the 36% that said "No" to show that the majority of Republicans were "birthers".  The first 7 questions in this poll could all be interpreted that way, based on what you did with the "Not Sure" vote.

Some of the questions were indeed designed to try to make Republicans look bad in the eyes of the center.  Take for instance question number 10 - "do you think your state should secede from the union?"  That is one of those questions that normal people look at you and ask "are you high?  What the he** kind of question is that?" if you asked them.

But here is where Markos falls into the same trap both political parties fall into.  He ASSUMES that he knows what the important questions are for Republicans...abortion, gays, God and Creationism.  The only thing he left out was guns....The problem is those kinds of social issues always fall by the wayside at times like this - for the right, the center AND the left.  People don't worry about that kind of thing when they are worried about paying the bills and whether their job will still be there at the end of the year!

In the end, Mr. Pendlebury sums up my opinions of opinion polls quite nicely...

It's human nature for people to interpret any poll in a way that reinforces what they already believe. And if they can't, they simply dismiss the research as being flawed. Perhaps that is the real takeaway from this survey.

Which is why everyone should look into the methodology anytime the hear someone talking about this poll and that poll and realize that, while they are great indicators of trends, they are not the be all and end all.  Especially in ANY poll where the questions are as loaded as this ones were.  It is the wise thing to do.