Doug, What I mean by my experience" with Rasmussen is that for certain US events (such as elections) or attitudes about an issue (such as unemployment), I compared different US polling bureaus (Gallup, Harris, Nielsen, Zogby, Pew etc.) in terms of what they surveyed and predicted, and what the outcome was.
And generally my impression was, that Rasmussen scored pretty well, i.e. their elections prognosis for example tallied pretty well with the actual result. Admittedly though my sampling wasn't very comprehensive. When statistical information becomes a commodity, the "statistics business" usually degenerates, since "the price of truth is in truth highly variable" - to get some truths, is cheap, to get others, very expensive - which leads to plenty swindling. People also try to make maximal profit out of "cheap and nasty" surveys. When I worked for Statistics New Zealand in a questionnaire design/classification development section 1991-94, they made me do among other things a literature review of scientific studies on response effects (specifically, order effects) in closed survey questions, blah. I covered about a hundred studies in all which showed clearly among other things that attitudinal questions are - if I didn't know already - especially prone to response effects (including question-wording effects and response order effects). See e.g. Sudman & Bradbury's classic textbook "Response effects in surveys". As regards my experience in general, I actually often prefer it if people regard it as worthless - I have more of a chance that way to do what I want to do, without being plagued by parasitic American, Dutch and British scumcunts (including left-of-the-left scumcunts and criminals) trying to exploit me! J. _______________________________________________ pen-l mailing list [email protected] https://lists.csuchico.edu/mailman/listinfo/pen-l
