There is support here in this poll for the idea that sampling relatively small numbers of people can give an fairly accurate view of the larger group. These stats have been pretty consisitent since the beginning sample of 50-60 people or so. Its cool, cause now I have a little better faith in polls.