> The 71% is of people is of people who a) had phones and b) answered the > question. While most people in the US do have phones, I'd bet there's a > correlation between NOT having a phone and income, and I know there's a > correlation between income and voting. There are also correlations > between ANSWERING the phone and income, and so on.
Interestingly, before the "no-call" list went into effect, we rarely answered the phone. Due to the large volume of telemarketing calls here in Florida, our answering machine screened all incoming messages. Now, we answer the phone because it may actually be someone on the line who is not a sales representative. I think this may improve accuracy in political poll results in the upcoming campaigns. Lots of people who heretofore gave up on answering the phone may be available and willing to answer polling queries. It is possible, telephone polling may become more realistic given proper sampling techniques. . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
