How can I prove that COVAR(x,y)/VAR(x) = b? I'm pretty sure the first part equals m, but I don't see how the slope and y-intercept can be the same thing, unless their both zero.

Any nudge in the right direction would be greatly appreciated.

Printable View

- Feb 14th 2011, 07:50 PMthejollywaghow does m = b (or slope=y-intercept
How can I prove that COVAR(x,y)/VAR(x) = b? I'm pretty sure the first part equals m, but I don't see how the slope and y-intercept can be the same thing, unless their both zero.

Any nudge in the right direction would be greatly appreciated. - Feb 14th 2011, 08:02 PMpickslides
My understanding is, for $\displaystyle \displaystyle y=a+bx$

then $\displaystyle \displaystyle a=\bar{y}-b\bar{x}, b= \frac{cov(x,y)}{var(x)}$

So this works if you call the equation $\displaystyle \displaystyle y=mx+c$ , many text books do. - Feb 15th 2011, 07:27 AMthejollywag
Thanks, pickslides!