The Range of Normal

Abstract

One of the primary aspects of any test is its capacity to distinguish “normal” from “abnormal.” However, the issue of what normal truly is remains complex. There are two categories: correlated normality and isolated normality. Each has numerous advantages and disadvantages, but the latter is much more common for defining normal. The method of establishing limits of normal varies depending on whether correlated or isolated normality is utilized. With correlated normality, the limits are dependent on whether the purpose of the test is discovery, exclusion, or confirmation. Isolated normality limits are frequently based on the consequences of being termed abnormal. It is imperative that normal limits be based on values of an appropriate reference group: the normal group should be healthy equivalents of the individual being tested. Finally, both intrapersonal and menstruational variations that result when multiple tests are performed on the same individual should be accounted for.

Key Words:

clinical trial - normal - statistical analysis