About multiple correction T-test

#1
Hi everyone, I'm newbie in this field.

I was confused on my data in this few days.

I would like to compares the gene expression between "control and treatment"

Then, I have done the welch's t tests without followed by Bonferroni test because I think there are only two groups for comparison and alpha value is same even go through the Bonferroni test.

So, i would like to ask some suggestions from experts, is it correct about my thinking?

Thank you!:)
 

hlsmith

Less is more. Stay pure. Stay poor.
#2
If you are not doing multiple comparisons, then a correction is likely not needed. Though, of note you should set your alpha to your a priori level of type I error acceptance. Meaning, you don't have to commit to 0.05, it should be based on the context and risks you would like to circumnavigate.
 

fed2

Active Member
#4
sounds like maybe micro array? probably should mention maybe multiple comparison is from many genes then, traditionally false detection rate / benjamin hochberg adjustment then.