Mathematical interpretation of targeted volatilome data

Abstract Regardless of the purpose, design, sampling methods, or analysis procedure implemented for a breath study, the ultimate goal is to answer questions about human exposures, health state, or biochemistry. This requires a thorough understanding of the collected data and metadata, including the distributions, confidence in identification, analytical errors and detection limits, human variability, and the ultimate context within which data are considered. This chapter discusses how to prepare a statistically defensible interpretation of breath measurement data. The chapter discussion focuses on mathematical description of distributions, lognormality, comparisons, repeat measures, outliers, and data sorting.