The Student Room Group
You're not getting much response - so I'll take a stab at this.

If you add a fixed amount to either the x or y values you will not affect any of Sxx Sxy or Syy because they all subtract the mean.

If you multiply the x values by say 2 you will quadruple the Sxx value, double the Sxy value and not change the Syy value.
Reply 2
Original post by ian.slater
You're not getting much response - so I'll take a stab at this.

If you add a fixed amount to either the x or y values you will not affect any of Sxx Sxy or Syy because they all subtract the mean.

If you multiply the x values by say 2 you will quadruple the Sxx value, double the Sxy value and not change the Syy value.


So it's like variance then; in saying that Var(2X+5)=22Var(X) in the case of Sxx and then for Sxy it's like standard deviation (presumably because the x value isn't squared in an Sxy calculation).

Thanks for responding :smile:
Reply 3
Original post by the.austrian
So it's like variance then; in saying that Var(2X+5)=22Var(X) in the case of Sxx and then for Sxy it's like standard deviation (presumably because the x value isn't squared in an Sxy calculation).

Thanks for responding :smile:


Yes, exactly the same, make sure you square it as you've said :smile:
So if you are multiplying x by 9, would Sxx be multiplied by 81??
Original post by Ellieseftel
So if you are multiplying x by 9, would Sxx be multiplied by 81??


You catch on fast

Latest