I'm actually quite surprised they would bother with this, unless they intend to completely ignore the results. For this to be valid, somebody would have to receive all the calls, record information about each caller (including whether they were residents of California and their eligibility to vote), and filter out duplicate callers or fraudulent attempts to influence the results (by "flooding").
And even then the results are still going to be useless, because this is what's called a self-selected sample. Self-selected samples are always biased towards respondents with strong opinions, and specifically strong negative opinions. Even without a mobilized anti-gay marriage base, a survey like this is almost guaranteed to "show" strong resistance to gay marriage by the voters of California.
And that may very well be the point. I find it hard to believe that no one in the governor's mansion knows how to do a proper poll, or is unaware of the likely bias inherent in this kind of poll.
Comments 2
Reply
And even then the results are still going to be useless, because this is what's called a self-selected sample. Self-selected samples are always biased towards respondents with strong opinions, and specifically strong negative opinions. Even without a mobilized anti-gay marriage base, a survey like this is almost guaranteed to "show" strong resistance to gay marriage by the voters of California.
And that may very well be the point. I find it hard to believe that no one in the governor's mansion knows how to do a proper poll, or is unaware of the likely bias inherent in this kind of poll.
Reply
Leave a comment