topic Re: power and sample size calculation in JMP in Discussions

power and sample size calculation in JMP

DFereidouni — Mon, 22 Jan 2024 23:15:58 GMT

I'm trying to determine the sample size using one sample mean. The Std. Deviation from the historical study is 5.8E+8 and I don't know how to use it in that platform as the one sample mean has a Std. Deviation limit of 200,000.

Any idea how to overcome this issue? Is it possible to use CV in the calculator?

Re: power and sample size calculation in JMP

Thierry_S — Tue, 23 Jan 2024 05:19:47 GMT

Hi,

It appears that you have a scale issue. Would it be appropriate to transform the standard deviation to a larger unit (e.g., 1e-6 initial unit)?

Of note, I don't think you can use the CV because the power calculation tool expresses effect size as multiples of the standard deviation.

Let us know if that addresses your issue.

Best,

Re: power and sample size calculation in JMP

MRB3855 — Tue, 23 Jan 2024 13:50:29 GMT

Hi @DFereidouni , to add to @Thierry_S comment: CV is often used if the assumed distribution is lognormal.

But, in your case you can just scale everything as @Thierry_S said and all should work (e.g., divide SD and Difference to Detect by 1,000,000). And, looking at the Power Explorer, it doesn't look like the effect size (Difference to detect) is a multiple of the standard deviation.

Power for One Sample Mean (jmp.com)