I know that the general formula for multiple regression is:
Y' = b1X1 + b2X2 + ... + bkXk + A
But I have no idea how to compute the coefficients for each factor (b) or A.
I need to be able to do this by hand!
An explination would be brilliant, thank you.
May 12th 2008, 06:32 AM
For starters, and you probably know this: Think of A as the intercept (because it is), and b as slope. It is still a y=mx + b equation.
If you are given an r or R^2 to play with, that would be helpful. To be more a bit more clear: The correlation coefficient relates directly to the regression line
Y = a +bX for any 2 variables, whereby b=r(Sy/Sx). This all extends into multiple regression where you just have to use some summation notation. Namely, r = 1/n-1 Sigma (X-Xhat / Sx) (Y - Yhat / Sy)... Sorry that this is kind of a mess; I am extremely busy/ill today. If someone else could latex up and or expand/clarify this further, as always go for it. Do you have anything that is more specific... like perhaps a problem or an assignment with instructions?