topic Re: ANOVA of Dose-Response Curve in Discussions

ANOVA of Dose-Response Curve

chanson — Sat, 10 Jun 2023 23:28:17 GMT

Hello,

I have some basic questions.

I need to create a dose response curve for 3 drugs and 2 controls. How can I use 2-factor ANOVA to compare the drugs and controls? Do I make one of the drugs the dependent variable and a control and another drug as the independent variables? I'm very new to this stuff so I would appreciate some help.

Thanks!

Re: ANOVA of Dose-Response Curve

Thierry_S — Mon, 05 Apr 2021 12:43:24 GMT

Hi,
I think that I provided some direction relevant to this question in your other post:

Re: ANOVA of Dose-Response Curve

Mark_Bailey — Mon, 05 Apr 2021 16:42:39 GMT

What do you mean by "compare?"

Also, you have two directions to answer the questions once you decide what comparison makes sense. The first approach is the extract a scalar value, like ED50, You won't be able to apply statistical inference, though, unless you have more than one independent curve for each sample (drug or control). Fit Curve as @Thierry_S suggested provides methods to directly compare parameter estimates or test equivalence. The second approach is to use the entire curve. The Functional Data Explorer will convert each curve into a set of 'functional principle components' that represent the salient features of the curves and can be used as a new response variable. You can them use it with the drug and control categorical variable to compare.