Survey bias is unavoidable
I wonder what would happen if I put a survey on the Napster web site and asked visitors if people who illegally download music should serve jail time? Given the self-selecting nature of web surveys, would anyone be surprised if an overwhelming majority of Napster "voters" were against legal prosecution?
Speaking as a 25 year marketing vet and researcher, I would say that web surveys have their place and they can be done in a way to remove bias. But the UL survey was not that kind of research -- for a legislator to project the results to the voting population as a whole would be irresponsible.
All we can gather from the survey is that Union Leader readers -- generally more conservative than NH voters as a whole -- voted a certain way. They DO NOT represent anyone but themselves any more than Napster voters represent their point of view and not the public at large.
Sky
|