LondonJohn
Penultimate Amazing
- Joined
- May 12, 2010
- Messages
- 21,162
Depends what question you think it's addressing. Knox and the case had been discussed at length on the forum. The poll tells you what members of the JREF with an interest in the original thread thought of Knox and her innocence or guilt, and to an extent how they were feeling about the thread as well. Answering that kind of question I don't see that the poll is particularly improper or biased. As for non-randomised self-selected, sure that's the kind of survey it is. What are the randomised non-self selected polls that you guys are referring to? Most people don't know enough to comment one way or the other on the case. Those that do are already a self-selected group.
I agree with your penultimate sentence. The only valid poll (if it were important, which it isn't) would be to pick a sample that was representative of the population (e.g. US/UK city/country), and to ask an unbiased question such as "Do you believe that Amanda Knox, who's currently appealing a conviction for the murder of Meredith Kercher in Perugia, Italy, should be convicted or acquitted by the appeal court?" The available options would be "convicted", "acquitted", "undecided", "don't know enough to give an answer".
Like you, I guess that around 75-80% of people would give the last response. But this would be the only way to gauge public opinion with any sort of accuracy whatsoever. It wouldn't interest me as a worthwhile exercise anyway. I'm reminded of Superhans' response to Jeremy talking about public opinion on the most excellent TV comedy "Peep Show":
"People like Coldplay and voted for the Nazis. You can't trust people, Jeremy."