The woolf plot is more reliable than the scatchard plot in analyzing data from hormone receptor assays
In analysing results from hormone receptor assays, many laboratories present their data in the form of a Scatchard plot and fit a line to the data points by least squares regression analysis to calculate the number of binding sites for the hormone receptor. However, this technique is not reliable when there are outliers present in the data, a feature which is common in receptor assays and which assumes increasing importance when computers are used routinely to analyse assay results. In this study, the Scatchard plot, double reciprocal plot and Woolf plot, were examined as ways for graphically representing data obtained from hormone receptor assays. Toe each of these plots three line-fitting procedures were applied. These were least-squares regression, an unweighted robust regression and a weighted robust regression procedure. Irrespective of the regression technique used, all three plots gave the same answer for the concentration of binding sites when the data were well-behaved. However, in the presence of up to three outlying points, the Scatchard plot, particularly with least squares regression analysis, performed poorly. It is concluded that a more reliable way of representing binding data in these assays is by a Woolf plot and that subsequent line fitting should be by unweighted robust regression analysis.