(Originally posted on MySpace)
A survey arrives by snail mail. I'm politely requested by an accompanying letter to complete it and send it in. A winner will be drawn from the respondents, the prize is a gift certificate.
Nice.
But the survey is anonymous. There is nothing on the form or the postage free return envelope to indicate who sent it in. So how do you draw a winner?

I send an e-mail to the intern who set up the survey, and ask him how he thinks he's going to do that.
Says he: "I tried to make things as easy as possible for the respondents and so hope to get enough response. So the gift certificate will end up with someone who actually sent in the survey. Unfortunately I cannot be certain about that".

Soooo... the winner of the gift certificate is drawn from the people who were sent the survey, not those who sent it back. Completing or not completing it does not influence the chance of winning. Frankly I fail to see how that would stimulate anyone to participate.

Not only is it simply untrue that a winner will be drawn from the respondents, the chance that the whole setup will actually work borders on zero... except for those who don't show any brain activity.
Would you want to use survey results gathered solely from that group?

I should win that gift certificate just for figuring it all out!