« SUSA Polls Suggest a Swing State Bounce for Kerry | Main | General Election Cattle Call, August 4 »

Wednesday, August 04, 2004

A Quick Note on Rasmussen

Posted by DavidNYC

Rasmussen has by far the most annoying - and quite possibly silly - polling methodology out there. In fact, I'd almost say Zogby's "interactive" polling is prima facie more sound. Rasmussen polls over the course of a month which, as someone cleverly observed, is not a snapshot but rather a daguerreotype. I'll quickly mention the swing state results he released today, but since the vast majority of the polling was done before the convention, I think they are even more useless this month than in the past.

�Ģ Missouri: 50-46-4 Bush (May: 44-43-13 Bush)
�Ģ North Carolina: 50-45-6 Bush (June: 49-42-8 Bush)
�Ģ Virginia: 49-46-5 Bush (June: 48-45-6 Bush)

I just want to re-iterate: I'm giving these polls short shrift not because they show Kerry losing, but because I have deep reservations about Rasmussen's methodology. If a more traditional pollster showed results like these for NC and VA, I'd actually be quite happy.

(Thanks to glibfidget.)

Posted at 08:30 AM in Missouri, North Carolina, Virginia | Technorati

Comments

I had no idea that they conducted their polls in over a whole month. So much could change in the course of that month. Pretty useless.

Posted by: Inkan1969 at August 4, 2004 03:59 PM | Permalink | Edit Comment | Delete Comment

For what it's worth, I see today in Rasmussen that they have Kerry with only a 1% point lead (48% to 47%) in Pennsylvania. This is the closest I've seen the race in that state since I can't remember when. I sure hope you're right that they are less valid than the polls showing Kerry leads of 10% to 15% there. Honestly, I haven't even considered Pennsylvania to be a swing state--I still view it as definitely going for Kerry in November.

Posted by: Pepe at August 8, 2004 07:43 PM | Permalink | Edit Comment | Delete Comment