ABSTRACT
This chapter describes the methodologies of biostatistics, including both the potential strengths and limitations of different study designs. It discusses the implications, on the selection of statistical tests, of different types of data, including continuous versus categorical and paired versus unpaired, and their distributions, such as normal versus skewed. A major challenge to meteorologists, and medical writers (MWs) and other researchers and communicators, is to not only comprehend but clearly disclose the statistical methods used to estimate outcomes. The chapter also discusses the logic and "mechanics" of statistical analyses, including implications, for choosing statistical tests, of different types and distributions. One of the most influential forms of statistical testing is analysis of variance (ANOVA). ANOVA comprises the ratio of variance due to a treatment or exposure, in the numerator, divided by the variance due to random probability or biological variation, in the denominator.
