08-08-2015, 12:14 AM
(08-08-2015, 12:01 AM)BmorePat87 Wrote: How is asking "who would you vote for" biased? I understand how some questions can be phrased in a way to lead voters, but that's why we scrutinize professional polling to eliminate that.
Also, you're giving an example of hypothetical poll not being completely representative of the voter turnout population, which has nothing to do with being biased but more to do with not being able to guess the turnout.
If we look at independent polling for 2012, you'll see that none had Romney winning beyond the margin of error:
http://www.realclearpolitics.com/epolls/2012/president/us/general_election_romney_vs_obama-1171.html
However, to get back to what we were actually discussing (real polls versus online polls), I have to imagine that a voter poll from Breitbart or Drudge would have had Romney winning by more than what these independent polls had. Why? Because those who frequent those sites are predominately conservative. Likewise, a DailyKOS poll would have been far more in favor of Obama.
Well the sample size is never 50/50. And what good is any poll with a margin of error over 4%?
A poll from drudge or breitbart would be perfect for a GOP primary.
I'm not saying online polls are the best. Just saying you can find bias in any of these polls.
