(Multiple regression) Can I develop an F-test for zero-intercept?

Mar 2016
5
0
somewhere
(Multiple regression) Can we develop an F-test for testing this hypothesis?

RM : H0 : Y = e
FM : H1 : Y=B0 + B1X1 + .... + BpXp + e

(e : error term, r.v)

As I learned, when we want to find the fitted regression equation, we should not neglect intercept part.
It should not be considered zero, and in general case, RM of H0 doesn't touch B0 part.

Anyway this problem is far from general multiple regression analysis case, so I want to know how to solve this problem.
Please give me some help, or a little hint quickly, anybody!
I attached the original problem!

???? 2016-04-19 04.51.14.png???? 2016-04-16 18.26.36.png
 
Last edited:

chiro

MHF Helper
Sep 2012
6,608
1,263
Australia
Hey melodian92.

If you are doing what I'm thinking you are doing (which is using an ANOVA to test B0 = B1 = B2 = .. = Bp = 0) then if the F-statistic is the same as that for the ANOVA then I believe you can.

Can you show us the F-test you are using and what the test statistics are as a function of your data?