Abstract
Modern medical research often includes too many statistics which are misused or misunderstood. This problem has been identified and discussed by several authors, see, e.g., Altman et al. 1983, Editorial 1988, 1992, Morris 1988. In this article I will explain some general reasons for applying statistical principles and give a few hints about how to do this. The interested reader can find detailed information on the interpretation of specific statistical methods in modern textbooks, e.g., Machin and Campbell (1993), Pagano and Gavreau (1993).