Quote:
Originally posted by RatedG²
Actually he's right. 1K people for something like this isn't the best sample size. It's not bad but it's not great
And likely voters is a problem too.
|
For a national poll, 1k people gives a result with a margin of error of around 3% which is decent enough and is actually pretty standard.
Having likely voters is actually done for more representativeness because only interrogating people who are sure to go vote can skew the result (these people are more politically involved, older, more educated...).
