09-19-2016 10:28 AM
I want to do a sample size calculation for a two proportion equivalence test to show two drugs are not different. According to this link,
SAS does not implement this in proc power.
Does anyone know a work around? Do I need to implement it myself?
09-19-2016 11:12 AM
The usage note you refer to is from 2013, and there have been improvements to PROC POWER. In the SAS/STAT14.1 documentation located here:
there is substantial material related to two sample equivalence and inferiority/superiority testing.
09-19-2016 11:13 AM
I think this paper will give you the answer:
Equivalence and Noninferiority Testing Using SAS/STAT® Software
John Castelloe and Donna Watts, SAS Institute Inc.
The whole paper (23 pages) is worth reading but you can directly go to
PROPORTIONS, RISK DIFFERENCES, AND RELATIVE RISKS (page 9)
Power and Sample Size Analysis for Proportions, Risk Differences, and Relative Risks (page 11)
09-20-2016 12:08 AM
Or you could data simulation: http://blogs.sas.com/content/iml/2013/06/05/simulation-power-curve.html http://blogs.sas.com/content/iml/2013/05/30/simulation-power.html