Cautionary Alerts for Skeptics

Abstract

Skeptical analysis requires an awareness of the weaknesses existing in the evidence that may be offered in the evaluation of a hypothesis. That is, cautionary alerts are needed when examining various types of evidence. Four of these ‘alerts’ discussed in this chapter are reliance on correlation analysis, use of phenomenological descriptions, ignoring differential time scales, and inadequate field tests. Correlation between two variables does not necessarily denote a causal relationship, it only indicates the variables may be responding in a similar way possibly to a third, outside variable. A phenomenological description offers an overview of a process but does not attempt to elucidate some of the fundamental mechanisms that account for the behavior of the process. Accounting for the relevant time scale in examining a process can be quite important. For example, accumulation of evidence on an instantaneous or short-timescale process should only be applied with considerable caution to descriptions of behavior at higher levels of hierarchy occurring over longer time periods. Finally, experimental evidence for successful plant improvement, especially in regards to increasing crop yields, should be accepted only if the standard rigors of field experimentation are met.