Abstract
In medicine and chemistry, immunoassays are often used to measure substance concentration. These tests use an S-shaped standard curve to map the observed optical responses to the underlying concentration. The enzyme-linked immunosorbent assay is one such test that is commonly used to measure antibody concentration in vaccine and infectious disease research. The enzyme-linked immunosorbent assay and other immunoassays usually involve a series of doubling or tripling dilutions of the test samples so that some of the diluted samples fall within the near-linear range in the center of the standard curve. The dilution that falls within or is nearest to the center of the near-linear range may then be selected for statistical analysis. This common practice of using one dilution does not fully use the information from multiple dilutions and reduces accuracy. We describe a recently proposed weighted-average estimation approach for analyzing multiple-dilution data (Cheung et al. 2015, Journal of Immunological Methods 417: 115–123), and we present the new