I can't follow the notation here. You might want to look at some other
reference on simple linear regression if you find text book difficult to follow.
Look here.
RonL
Okay, this chapter is wearing me out. I am not getting anything of all this. Let's start from a basic thing.
It says "compute SSr and SSE and (s_e)^2 and the coefficient of determination, given the following statisics computed form a random sample of pairs od X and Y observations (let's take just one)
Sum from i=1 to n of (y_i - y)^2 = 100,000
r^2 = 0.50
n= 52
No clue no clue..
Is there anybody who can explain me the simple regression. I cannot see the point of all this. Please. Thank you.
Okay thing are getting a very little clearer.
Let us take again that excercice:
"compute SSr and SSE and (s_e)^2 and the coefficient of determination, given the following statisics computed form a random sample of pairs od X and Y observations (let's take just one)
Sum from i=1 to n of (y_i - y)^2 = 100,000
r^2 = 0.50
n= 52
Okay. So. I understood more or less what it is asking. SSR sum square regression; SSE sum square errors. (S_e)^2 is = SSE/(n-2) the variance.
Okay. (S_e)^2 I get with that formula. So I need to know SSE. SSE I get through this formula SSE= Sum of [y_i - (b_o + b_1*x_i) ]^2.
And SSR= (b_1)^2* Sum of (x_i - x_bar)^2 (by definition)
Okay SSR I can know subtracting SSE from 100,000.
But this means I need to know b_1. But b_1= r*s_y/s_x ( b_1 is the slope coefficient estimator)
R I have it. But s_y and s_x (the standard deviations) I don't. The text doe not give them. So how do I get them?
Thanks for your help.