True to assume equal variance when testing difference between the means???

wanye

New Member
#1
hi all,

In testing for the difference between the means, is it true that we assume that each of the populations is normally distributed with equal variances? why?
 

JohnM

TS Contributor
#2
I think people get into the "habit" of assuming normality and equal variances because the violations of these assumptions need to be pretty severe before they will lead us to erroneous conclusions.