I'm trying to compare the effect of an IV on my DV, where my data are separated into 3 groups. Specifically, in one group there is a significant correlation between IV and DV, but in the other two groups this is much weaker. I would like to test this effect and was thinking that multiple linear regression seems to be the appropriate method.
I have followed the instructions detailed here:
to compare regression coefficients among 3 groups in SPSS and they work perfectly for my data set, however, if I understand it correctly the very last step tests the null hypothesis; Ho: B1 = B2 = B3, which doesn't seem to be what I want, because I want to know if B1 = B2 or B2 = B3
Thanks for any help,