Validity of results

Abstract

Throughout the previous chapters we have been concerned with the acquisition of data, much of it numerical, and its subsequent manipulation to provide information in whatever form it may be required. During these manipulations the assumption has been made that the analytical result is exact. However, as is only too obvious to any analyst who carries out even duplicate analyses, any replication of an analytical measurement will, in general, give a set of results differing by measurable amounts. It is therefore essential that the results from any measurement procedure can be assessed critically to determine the extent of variation which may be expected from a particular procedure and thereby provide a basis for estimating the likely variability, or uncertainty associated with a specific set of analytical results.