I’m looking for a clear-minded discussion here of the polling data and media bias as of yet. I’m sure Edwards supporters and Clinton supporters have their own opinions as to what’s going on in the polls, and I’d really appreciate it if you’d add any information that you can.
From what I’ve seen so far, there’s a significant polling bias against Obama, which is skewing the polling results.
In Iowa, the polls averaged a 7.9% skew in favor of Hillary. It was a skew that not only disfavored Obama, but also Edwards, who won over Clinton.
In New Hampshire, the polls showed a skew disfavoring Hillary by roughly 8% while showing some statistical accuracy about Obama’s support (an average of the polls put him at 38%, while the final results put him at 36% and within the margin of error). Here it looks as if at least 6% of Hillary’s support base came from Edwards voters or undecided voters.
There’s something else that’s unusual about the New Hampshire polls: the earlier polls show Hillary with a vast statistical advantage over Obama--as high as a 20% difference only two weeks prior to the elections. There can be no test of the validity of those polls. Why you ask? Things change in any election, and so the final polls are always the ones checked for validity.
In Nevada, it finally seems as if the pollsters have figured out the correct results. The skew is only slightly in favor of Hillary (4% by this graph), and one of the polls shows a 2% advantage for Obama. All of this is well within the margin of error, so it looks like these pollsters did good work. Notable is that the Zogby poll managed a perfect zero score.
Then in South Carolina we see an average of an astonishing 16.8% skew against Obama! Zogby–the pollster usually credited with favoring Obama–is off by 13% disfavoring Obama.
?
This is the data I've used to check this polling bias as of yet. Underneath you'll find a link to pollster's reportcard on pollsters.
http://www.surveyusa.com/index.php/2008/02/04/2008-pollster-report-card-through-020408/If we add up the democratic primary skew so far and average it, we end up with a 5.23% skew against Obama across the board. Obviously, some of the polls are more accurate than others, so the validity of this average to actual results is doubtful. This 5.23% skew could only act as an accurate predictor of the outcome of the race if the cause of this skew turned out to be pollster prejudice–of whatever kind–against Obama. Even then, there are wide variations from one pollster to the next, meaning that applying this average in reverse as an attempt to eliminate the bias would not work (it would not be enough in some cases, and it would be too much in others, depending on the bias).
What this exercise does show is the inaccuracy of the polls, which are biased against Obama.
Any thoughts? More information? Please check my results!