This is quite interesting. While both political party affiliation and marital status are both self-reported, martial status is certainly more an objective criteria.
So to me it would make much more scientific sense to make sure your sample reflects the population in terms of marital status than self-identified political affliation. Surely they stratify the sample by sex, race etc as they should marital status.
Are you sure you read it right. We cant tell since you didnt post it. I have not heard of married couples being referred to by Zogby or any other pollster. Generally what they are referring to is that Dems usually win by a decent margin among single women and lose among married women. White men on the other hand, single or married typically go for the Republican. Single moms? Very likely to be Democratic voters. May or may not be counted as singles by pollsters since they are head of household with children.
I don't know if this is pertinent or not, but up until about 3 weeks ago, I was sent Zogby polls about once a week. I completed and sent them immediately as they became available to me, since I believe the polls can be used to sway, and thought every one of us must be counted!
As of 3 weeks ago, they no longer care for my input. Maybe it is that I am from AZ and our state is not a swing state, but how do they get this "national" number of 3%? Maybe it's a different type of poll. All I know is it seems odd that they tracked my opinion religiously (and it did not waver for Bush), and suddenly I am not getting any email from them whatsoever.
Can anyone shed some light on this?