Regression to compare mean differences?

#1
Hi everyone,

I am doing a project with two different groups (clinical and control). I want to compare their scores on a couple of variables (ex: body image disturbance), while controlling for age and BMI.

My committee member recommended I do a regression, and noted that it would allow me to control for age and BMI and then see if they in fact differed on my variable.

A different person recommended I do an ANCOVA to look at the same thing.

Can anyone explain how a regression would allow me to see if they differ? Is there any advantage to using one over the other (regression or ANCOVA)? Thank you!!
 

Karabiner

TS Contributor
#2
Regression (with 2 interval scaled predictors, a dummy for Group, and possibly
the 2 interaction terms) and ANCOVA (with Group as a factor, 2 interval scaled
covariates, and possibly the interaction terms) will produce the same results.

With kind regards

K.