Actually it is not that, oversampling is the deliberate sampling of a subset of the population out of proportion to their true represntation in the population to make sure they are represented.
Other than the deliberate/accidental aspect, how is that different than what I said? Bottom line: You’ve got a higher proportion of X in your sample population than exists in the entire population.
The difference is that in the analysis you can correct for a deliberate over proportion.
If you run a poll of the man on the street in a Maryland suburb and end up with a population of 100 whiltes and 100 blacks, and then treat this as an ubiased sample, then you are in trouble (since you have an over represention of blacks).
If I wanted an unbiased sample I would find say 175 whites and 25 blacks to use as a representative sample. But if I wanted to then say something about the black population I would only have 25 samples far too few to make any claims about black attitudes as a group.
However, I can actively decide to survey 100 white people and 100 black people but when reporting total effects, weight the black results less than the white results to make my results unbiased.
[Slow Clap]
This is it. This has always* been *it. They just haven’t the balls to say so.
[/Slow Clap]
This is all absolutely true.
It’s relatively easy to correct for issues like this. It’s not that hard to estimate the proportions of whites and blacks (and others) in the population. It gets a lot harder when you need to estimate the proportions of many categories at once, especially when they change over time. This takes a fair amount of expertise. Survey companies use proprietary sampling methodologies that are pretty much never transparent, so it can be quite difficult to determine whether a survey actually is well-designed. And even surveys that are well-designed are very often just wrong.
Well, I guess no difference except for the one difference.
You can’t possibly actually think these two cases are equivalent, so what are you up to? Seriously.
Except it’s true, both in letter and spirit.
No he didn’t. I was watching – he didn’t.
You know, the Tea Partiers aren’t denying that they said these things. They’re proud of them. The fact that they said them is a matter of public record.
Why do you guys feel compelled to try to pretty up and spin what the Tea Partiers loudly and proudly proclaim? And why do you insist they don’t speak for other conservatives, but rush in to defend and explain them? If they don’t speak for you, why are you speaking for them?
This.