The Regression Equation

#1
I am currently trying to find bo. The formula for it is 1/n(Eyi-b1Exi)=y with a line on it-b1x.

I am having a tough time trying to solve this. Do the two different equations equal one another? Where you you find y with the line on top of it? I would appreciate an example of how to solve this type of equation. Thank you
 

Mean Joe

TS Contributor
#2
I am currently trying to find bo. The formula for it is 1/n(Eyi-b1Exi)=y with a line on it-b1x.

I am having a tough time trying to solve this. Do the two different equations equal one another? Where you you find y with the line on top of it? I would appreciate an example of how to solve this type of equation. Thank you
I'm kind of confused; you have a formula with b1 in it, and you're trying to find b0? Is this not a linear regression?

"y with a line on it" sounds like you're talking about "y bar", which is the average value of y in your sample.
 

Dragan

Super Moderator
#3
I am currently trying to find bo. The formula for it is 1/n(Eyi-b1Exi)=y with a line on it-b1x....

I would appreciate an example of how to solve this type of equation. Thank you


Suppose Y and X have means, standard deviations, and a Pearson correlation (r) as follows:

YBar = 20, XBar = 10, stdY = 2, stdX = 1, and r = 0.50.

As such, b1 = r*(sy/sx) = 1.0 and thus

b0 = YBar - b1*XBar = 10.

The regression equation is: YHat = 10 + 1*X.