[Norseman]

From the poll results: McCaskill holds a 58% to 34% lead over Akin among very liberal voters

Could someone from MO please explain to me how McCaskill loses 34% of “very liberal” voters? How is that possible? I would expect a number more along the lines of 90% to 10%. Akin, for example, got 81% vs. 5% among “very conservative” voters.

[mrsmith]

“explain to me how McCaskill loses 34% of “very liberal” voters?”

I can’t and I’m not from Mo.

But polls often show Obama and other liberals getting 10%plus of the votes of Tea Party supporters. That’s another head scratcher!

[beenaround]

Because the term “very liberal” or “very Conservative” are subjective terms and what a person in Mississippi interprets as very liberal for them my be simply middle of the road elsewhere.

[Uncle Chip]

Could someone from MO please explain to me how McCaskill loses 34% of “very liberal” voters? How is that possible? I would expect a number more along the lines of 90% to 10%. Akin, for example, got 81% vs. 5% among “very conservative” voters.

Your questions are valid ones and I have no answer for them except that McCasket's negatives are that high, Missourians associate Obamacare with McCasket, and Missourians voted 75% to 25% to overturn Obamacare in Missouri and that included a lot of Democrats in that 75%.

[WILLIALAL]

“Could someone from MO please explain to me how McCaskill loses 34% of “very liberal” voters? How is that possible?”

I don’t think it is possible, that is her strongest base. That number needs to be explained.

[JediJones]

Could someone from MO please explain to me how McCaskill loses 34% of “very liberal” voters?

You really need a large sample in a poll to be able to get demographic breakdowns or cross-references that have any reliability. You can't just take any poll and try to divide it up into smaller polls. That's because the sample of "very liberal voters" is much smaller than the overall sample, and too small to be statistically accurate to draw any conclusions about how "very liberal voters" feel. That doesn't mean there's anything wrong with the overall poll. But you need to have about 1,000 of any group of people to start getting reliable poll results, so this poll is good for "voters" but not for "very liberal voters."