I agree, the surveys are generally anything but scientific in nature, and I generally cringe while reading them. But if we can make the assumption that the sample population was random and then count on the fact that the survey was simple enough that it would be hard to screw up, then at least...