[FredZarguna]

The USC poll is problematic because it involves a fixed sample of respondents who are randomly sampled initially, but who understand they're part of a sampling group. This means they know they're going to be resampled and their answers are no longer the same as a person receiving a cold call at random. They typically pay closer attention to the news than a randomly sampled population as a result.

Remember also that increasing the size of the sample only decreases the MOE as the square root of the number sampled. Sampling 5000 adults only reduces the MOE from ±3 to ±1.34, but increases the cost by a factor of 5.

And, remember the biggest error is not in the MOE, which is well known, but in determining how to decide which of your respondents are going to actually vote.

[LS]

Agree but it sounds like this poll, then, better captures true “likely” voters than any other. That’s the key. We’ll see.

Since 1996, almost every cycle ONE poll was pretty good at really capturing the race. One year it was Zogby, another year Rasmussen, another 538. Maybe this is USC’s year.