Well, for starters.. all three of those recent posts are based on ONE poll... done by Rasmussen, on Jan. 29-30. It was a poll of 500 people, done only on the phone. It used only land line phones, no cell phones. They did use some online survey forms to help correct for people who don't have land line phones. It was LIKELY VOTERS... and, it has a margin of error of +/- 4.5%.
The poll I linked was done on Jan 30 - Feb 4. It was 1200 people, of which 1082 were REGISTERED VOTERS. Of the 1082 Registered voters, ~ 40% were registered as Republicans, and only THOSE people were asked the primary question.
The poll was done on land lines (77%) and cell phones (23%) and also included some on-line surveys. It has a margin of error of +/3%.
So, no... I wouldn't say the other poll was WRONG? It was just slightly different. The results from both polls are within the margin of error. SO, statistically... they say the same thing.
Personally, I Think a poll of "Likely Voters" is more accurate than "Registered Voters". So, the data on the general election might be better in the Rassmussen poll. However, I don't think either of the polls is accurate on that part. It's just too far out. Grimes will not defeat either Bevin or McConnell.
In my opinion, the WHAS poll is MUCH more accurate on the Republican Primary race. That's why it's such bad news for Bevin. He's still really far behind. The GOOD NEWS for Bevin is, the data indicates that his problem isn't that people don't LIKE him... they just don't KNOW him. That's what campaigns are for... to introduce and sell yourself to the public. But, he's going to need a LOT more money. And, some kind of issue to separate himself from Mitch.
Of course, Mitch is making that very difficult, since... he's the "uber-conservative" during election years. It's an uphill battle for Bevin.
ping