The problem with that approach is that the gap between 20 and 21 percent encompasses far more people -- when you are doing demographic-type studies, thus, you might be over or under reporting your study results by literally 1000's (or 100,000's) depending upon your population base. For example...