Drawing inferences from outliers and exceptions

Abstract All too often, exceptional data values are dismissed, under the assumption that the data analyst’s primary responsibility is to determine the general behavior of the data collection. If we assume that the outlier is a valid measurement, representing a true system outcome, then we must try our best to understand its properties. This chapter explores the assertion that valid outliers are never exceptions to the general rules of a system. To the contrary, outliers are the exceptions upon which the general rules of a system are based. In this chapter, we will provide examples wherein outliers have led observant biologists to important biomedical advances. Examples will include a discussion of cancers having remarkably short latencies, lung cancers occurring in adolescents, rare diseases as sentinels for common diseases, discovery of a hole in the ozone layer, and acral melanomas occurring in African-Americans.