Howard I agree that any statistics that appear in the media (or anywhere else for that matter) are suspect. There is seldom given any real information on how the data was arrived at. The number of respondents or size of the data base (and how many people who were contacted who did not respond) and the validity of the data base (how representative it was) can have an effect on the results. Added to this is how truthful the responses were, and much of the information is highly suspect. Scientic analysis is not always much better when the the data involve statistical analysis, as many of the researchers do not have enough math background to do a properly objective analysis.