ABSTRACT
To best understand quantitative assay results, it is helpful to review the numerical development of the assay. Typically, a range of known quantities of the analyte is processed and the assay instrumentation produces numeric results. During the development and distribution of a new assay, a round-robin test is often conducted for which a common set of samples are sent to several laboratories to determine the extent to which measurements agree across different laboratories. Qualitative assays, such as dipstick assays, usually indicate the presence or absence of an analyte. For these assays, sensitivity and specificity are of primary importance. Graphical statistical tools are useful for characterizing the shape of the distribution of an endpoint. These tools include histograms, bar charts, and boxplots. Boxplots convey several important features of a distribution. The statistical test applied to data in any given situation will depend on the experimental design, the hypotheses, and the distribution of the endpoint.
