JMP's equivalence test, subcategory of nonlinear fitting with sigmoid curves - logistic curves - logistic 4p using log transformed concentration data, is a fantastic tool for comparison two drugs in bioassay. Result contains four parameters (growth rate, Inflection point, lower asymptote, and upper asymptote).
In some cases, a higher variability with the lower asymptote ratio can be observed due to the large impact of small variations in the ratios of reference and test sample. To overcome this potential drawback, Yang et al. suggest not considering the lower asymptote alone, but together with the upper asymptote as the ratio of upper to lower asymptote (e.g. New Variable=(upper asymptote - Lower asymptote)ref / (upper asymptote - lower asymptote)test).
For detailed information, please see the article,
Comment and Completion: Implementation of Parallelism Testing for Four-Parameter Logistic Model in Bioassays, DOI:10.5731/pdajpst.2015.01060
They suggested two new parameters. First one is the range from lower to upper asymptote and second is the slope (or growth factor) at inflection point.
I think their idea is quite reasonable and hopefully I could apply their concept in JMP.
Thank you.