SEM instead of SD

[lmanchon]

--Hi,

sometimes people prefer to use SEM instead of SD.
pcr_analyze function return SD (error), so is there a way to obtain SEM ?
Thank you --

[MahShaaban]

Hi @lmanchon
It is not clear to me how to get the SEM with error propagation as obtained in the package.
You could just divide the "error" term by the square root of the number of samples, but again I am not sure if this is the correct way to go about it.

[lmanchon]

--Hi,
i have read in some papers they use SEM instead of SD, this reduces the amplitude of the error bar.
When s is the SD or SEM of the delta-ct's, then s' of the delta-deltas is given as s' = SQRT(s²[treat]+s²[control]).
(the delta-delta-ct is from the two average delta-cts (treated and control, for instance)).
But in your package s is SD and s' too, and there is no formula to convert s' to SEM.
SEM=SD/SQRT(N) with N size of the sample.