On Frequency

摘要

Dr. Louis Anthony (Tony) Cox, Jr. has described two principles each of which he argued proves that it is impossible to satisfactorily define frequency.(2,3) Are the principles correct? Do they prove it impossible to satisfactorily define frequency? Why did Dr. Cox argue that they do? If it is possible to define frequency, how should it be defined? I commented on Dr. Cox's first principle~(2)in a previous letter.~(4) Suppose that: (1) You are presented with a probabilistic distribution. (2) The distribution will completely describe random times-to-events in a set of future experiments. Exposures to those events will start at the beginning of each experiment, then if an event arises exposure to a next event will start immediately and be governed by the same distribution. (3) All event severities (meaning loss values caused by individual events) will equal a known constant.