UNISTAT - the ultimate Excel statistics add-in

Four-Parameter Logistic Regression

Unistat’s optional Analysis of Bioassays module now features a complete implementation of Four-Parameter Logistic Regression following recommendations of United States Pharmacopoeia (2010) chapters <1032>, <1033> and <1034>.

Full and Reduced models, plate effects, equivalence tests and outlier detection are featured. Potency of multiple test preparations can be determined.

It is possible to estimate effective dose for any percentile (i.e. the so-called EDanything) together with its confidence interval and make predictions for the response variable with their confidence and prediction intervals. 96-well ELISA and unbalanced designs are supported.

For details see 10.5. Four-Parameter Logistic Regression.
Four-Parameter Logistic Regression Full Model
Four-Parameter Logistic Regression Reduced Model
Four-Parameter Logistic Regression Outlier Plot
Four-Parameter Logistic Regression Variance Plot
Four-Parameter Logistic Regression Difference Plot Upper Asymptote
Four-Parameter Logistic Regression Difference Plot Lower Asymptote
Four-Parameter Logistic Regression Difference Hill Slope