L.R.R. Hood writes:
That survey is not one I'd trust, at all.
Phone surveys are rarely reliable.
Um you do realize that virtually all scientific polling in the United States done by Gallup, CNN, USA Today, Zogby, etc. are done by telephone in sufficient numbers to produce a margin of error of 3 to 5%? What manner of polling would you consider reliable?
Having gone through a scientific curriculum at school, and studied statistics, I can tell you that the main factor is whether the sample is random from the point of view of the entire population being estimated. When it is, and when the questions do not suggest an answer, for huge populations, it does not take many data points to produce a result which mimics closely what you would get by asking everyone in the entire population.