multiple regression- how to calculate b?

#1
If I'm given the means, standard deviations, and correlation coefficients of two independent variables and one dependent variable, how do I calculate the unstandardized coefficient? (b)?
(the entire problem involves finding the bs, the a, and writing the regression equation... getting a and the equation is easy, but I can't do that without getting the bs...)
I tried taking (ry1-ry2*rx1x2)/(1-rx1x2^2) but I think that's giving me the Beta, the standardized coefficient, right? :confused:
 

Dragan

Super Moderator
#2
I tried taking (ry1-ry2*rx1x2)/(1-rx1x2^2) but I think that's giving me the Beta, the standardized coefficient, right? :confused:
Yes, that is correct, it would be Beta1.

All you then need to do is multiply Beta1 by the ratio of the standard deviations (Sy/Sx1)...i.e. b1 = Beta1*(Sy/Sx1).

Similarly, b2 = Beta2*(Sy/Sx2).