wow...so in essence from the national poll they made a sub poll based on the where the respondents from the national poll lived??? pretty pathetic...
Yup, and it’s always unreliable to do that, unless you have massive numbers of people in the original poll. You need about 1,000 people in a poll to make the numbers credible at all. You cannot take a credible, 1,000 person poll and get anything credible by looking at small subgroups within it. That is junk polling science of the highest order. This episode lends major credence to the idea that the media are manipulating polls and the reporting on them to favor Obama.