"The AP-Yahoo News poll, conducted by Knowledge Networks, has surveyed about 2,000 people since November to gauge how individuals' views are changing during the presidential campaign. The repeated interviews show the candidates' images have evolved gradually since the fall, with both getting higher favorable and unfavorable marks as additional people form opinions."No the title of this article doesn't show off the reporters or the editors bias at all, does it? I wonder which way they lean .... hmmmm.
okay so they have surveyed 2000 people over the last -- 8 months??? That's 250 people a month .... oh yeah - way to get a real good demographic there to draw your conclusions on. Sorry that sampling is way too small to draw any good conclusions. You would think that the AP would know better. [the article leaves it vague as to whether it is the same 2000 people or not, so I will take it that its not else they would have said otherwise]
This is a faulty study and should be taken with a grain of salt.