Linus Van Pelt
Sr. Member
Posts: 2,145
|
 |
« on: August 23, 2014, 07:48:50 PM » |
|
This sentence accompanies other Marist polls, as well; it's not a typo.
The issue is that in households with landlines, women disproportionately pick up the family phone, and old people pick up their phone more too. So a sample based just on who picks up the phone will be biased towards older females. They try to counteract this by asking for the youngest male. It's not that they only poll younger males; they still conduct the poll if the youngest male is out or whatever. So there are plenty of older women in the poll.
Nobody's sample exactly matches the population, which is why pollsters weight the sample by demographics. But they have these schemes to try to make the initial sample as representative as they can get, to minimize the amount of weighting required.
|
