# Thread: R square and F

1. ## R square and F

May someone tell me which one is the correct formula between the 2?

F= Mean Model Sum Square / Mean Residual Sum Square = MSM/MSR= (SSM/k-1) / (SSR/N-k)

or

F= (SSM/k)/(SSR/N-k-1)

from which

F= (R^2 /k-1) / ((1-R^2)/N-k )

or

F= (R^2/k) / ((1-R^2)/N-k-1)

where k is the numer of parameters to be estimated to compute the conditional mean
N is the sample size

Thank you so much!

2. Originally Posted by 0123
May someone tell me which one is the correct formula between the 2?

F= Mean Model Sum Square / Mean Residual Sum Square = MSM/MSR= (SSM/k-1) / (SSR/N-k)

or

F= (SSM/k)/(SSR/N-k-1)

from which

F= (R^2 /k-1) / ((1-R^2)/N-k )

or

F= (R^2/k) / ((1-R^2)/N-k-1)

where k is the numer of parameters to be estimated to compute the conditional mean
N is the sample size

Thank you so much!
FAQ: Why are R-squared and F so large for models without a constant?

3. okay so what are the degrees of freedom of the residual and the df of the model?

4. I mean I really cannot come up with which are the respective denominators if "k" or "k-1" and if "N-k" or "N-k-1" because my professor once said in a way and at another lecture it said the other way and in his slides he keeps confusing the two.. which one is right? I have an exam in a couple of days and I have no clue how to find the answer to this.

Thank you for your help.