Hi John,
You've definitely raised some good points. Before I get into some of them, have you seen the age breakdown of respondents?
3% 18 - 24
7% 25 - 34
17% 35 - 44
23% 45 - 54
24% 55 - 64
25% 65 AND ABOVE
1% REFUSED
With almost 75% of the respondents 45 and over, I think that this is causing a serious skew. When you also add in that it is only 800 people, I have a feeling that their margin of error is probably a lot more than the claimed +-3.5%, at least in the <45 age group.
Originally Posted by
John T. Fishel
Hard to answer with any degree of surety but I would hypothesize that
1. Most polls do not ask the question giving any kind of alternative. (In fact, I haven't seen any others that do so.)
2. This poll poses raises questions of outcome however inelegantly.
3. Other polls focus on the negative.
4. The election not only addressed the war but other issues as well.
None of these are definitive nor even necessarily correct. But they may offer a start at an answer to Tequila's question.
I will freely admit to a bias against most survey data for a lot of the reasons you list above. Still, I think some decent inferences could be made from this poll if there was a decent amount of proper analysis. Hey, I'd even be satisfied with some decent cross tabs stratified by age and gender .
Marc
Bookmarks