Reader "clemTNT" writes in a comment below:

I came accross this yesterday: Ten Reasons Why You Should Ignore Exit Polls

1. Exit polls have a much larger intrinsic margin for error than regular polls. This is because of what are known as cluster sampling techniques. Exit polls are not conducted at all precincts, but only at some fraction thereof. Although these precincts are selected at random and are supposed to be reflective of their states as a whole, this introduces another opportunity for error to occur (say, for instance, that a particular precinct has been canvassed especially heavily by one of the campaigns). This makes the margins for error somewhere between 50-90% higher than they would be for comparable telephone surveys.

Does this relate to what we did in class yesterday?

http://www.fivethirtyeight.com/2008/11/ten-reasons-why-you-should-ignore-exit.html

Yes, I would say that this article certainly does relate to confidence intervals, especially CIs for proportions, which we discussed on Thursday of this week.

The best reason for discounting exit polls (I would not say ignoring) is that participants are self-selecting. Someone wants to ask you as you leave the poll, Whom did you vote for? Many people will avoid the pollster on sight.

However, contrary to the article, precincts are not chosen at random: they compose a stratified sample. Our textbook says, "Researchers obtain stratified samples by dividing the population into groups (called strata [which means layers]) according to some characteristic that is important to the study, the sampling from each group." So precincts are chosen to try to replicate the overall vote with regard to, for example, socioeconomic status, ethnicity, etc.

This is typical for many kinds of polls and tests, such as new-product roll-outs and audience surveys. "Will it play in Peoria?" is not just an old vaudeville joke: products really are tested there.