sample size-new

[Tommaso Salvitti]

From a published study we have the following information. Here there is only one group (preterm infants, 56 infants) in which the variable Deltamet (mean change in methylation of the SLC6A4 gene) is measured. deltamet (mean (sd) = 0.07 (1.23)

The results of regressions to see the effect of NICU-related stress and deltamet on ATL MPR (anterior lobe brain),is reported below

deltameta	beta	SE	t	p
	-0.48	0.16	-3.06	0.01

I tried to see the sampsi command THIS TIME choosing the case where you have only one group but here too I had difficulty Assuming a power of 80% is an alpha of 0.05 what is the sample size? In other words, how many subjects must be taken if the estimate deriving from the sample must fall within 2 percentage points of the regression value indicated, with a confidence level (1-α) equal to 95%? I don't know if sampsi is the right command . Is anybody can help me? Thanks in advance Tommaso

[Rich Goldstein]

unless you are using quite an old version of Stata (and, as the FAQ says, you should tell us if you are using anything other than the current version), you should be using the -power- command; see

h power