Testing the Null Hypothesis for a Specified Difference — The Right Way to Test for Parallelism

by Janice D. Callahan, PhD and Nancy C. Sajjadi
Volume 2, Issue 2 (March/April 2003)

Testing for parallelism is a fundamental requirement for assessing the validity of bioassay data used to calculate relative potency. When a test sample and the reference material are diluted and assayed, the assumption is that their dose response curves have an identical shape. The standard statistical approach to demonstrate that two curves are parallel involves testing the null hypothesis for equality. This article explains why that approach is not appropriate and presents an argument for testing the null hypothesis for a specified difference to statistically ascertain parallelism...

Citation:
Callahan JD, Sajjadi NC. Testing the Null Hypothesis for a Specified Difference — The Right Way to Test for Parallelism. BioProcess J, 2003; 2(2): 71-77. https://doi.org/10.12665/J22.Callahan.

Download