Allahpundit is probably making too much out of all this. Rather than showing anything really substantive about democrats, this Zogby Poll probably merely demonstrates the insufficiently-often-recognized ability of the people commissioning polls to produce the results they desire by how they frame the questions.
07 Sep 2007
Scott D
Having constructed and fielded hundreds of pieces of research over the last 30 years, I can tell you for certain that whoever wrote this question did so in a way that could not possibly have accurately measured what it purports to. Journalists (and I use the term loosely these days) are not much more educated in statistics and research than the general public and tend to blithely accept results at face value without even examining the questions that were asked.
About 25 years ago we asked the following two questions of 100 people:
1. Do you think you are harming your children by allowing them to eat pre-sweetened cereals for breakfast occasionally?
2. Scientists have determined that too much sugar in children’s diets may be harmful to them. Do you think children could be harmed by eating lots of sugar-coated cereals for breakfast?
Needless to say, the “results” were wildly different.
Writing questionaires to get “real” attitudes is something any experienced researcher can do. When you see questions like the one in the Zogby poll, you can be certain that the writer had a result in mind when he wrote the question. I’m a bit surprised that an organization like Zogby would even accept such a piece of research to field.
Please Leave a Comment!