https://communities.sas.com/t5/Statistical-Procedures/Estimate-Group-Means-for-Significance/m-p/677650#M32551 <P>Hi,</P><P>For a project, I have confirmed the sample size, but am wanting to know what is the average difference that’s needed across the two groups to be considered significantly different. Any idea on where to start with this? I have 618 in each cohort , looking for power at 0.80 and alpha of 0.05</P> Tue, 18 Aug 2020 21:21:50 GMT mekono 2020-08-18T21:21:50Z https://communities.sas.com/t5/Statistical-Procedures/Estimate-Group-Means-for-Significance/m-p/677650#M32551 <P>Hi,</P><P>For a project, I have confirmed the sample size, but am wanting to know what is the average difference that’s needed across the two groups to be considered significantly different. Any idea on where to start with this? I have 618 in each cohort , looking for power at 0.80 and alpha of 0.05</P> Tue, 18 Aug 2020 21:21:50 GMT https://communities.sas.com/t5/Statistical-Procedures/Estimate-Group-Means-for-Significance/m-p/677650#M32551 mekono 2020-08-18T21:21:50Z https://communities.sas.com/t5/Statistical-Procedures/Estimate-Group-Means-for-Significance/m-p/677664#M32552 <P>You also need to estimate (or guess at) the root mean square error or standard error of the mean. Then you can use <A href="https://documentation.sas.com/?cdcId=pgmsascdc&amp;cdcVersion=9.4_3.4&amp;docsetId=statug&amp;docsetTarget=statug_glmpower_toc.htm&amp;locale=en" target="_self">PROC GLMPOWER</A> to determine the size of mean difference that will be statistically significant.</P> <DIV id="tap-translate">&nbsp;</DIV> <DIV id="tap-translate">&nbsp;</DIV> <DIV id="tap-translate">&nbsp;</DIV> Wed, 19 Aug 2020 12:56:21 GMT https://communities.sas.com/t5/Statistical-Procedures/Estimate-Group-Means-for-Significance/m-p/677664#M32552 PaigeMiller 2020-08-19T12:56:21Z