Re: Is it possible to have high R-square and fail F-test
From: KellyBoy (ikgantolin_at_gmail.com)
Date: 03/23/05
- Next message: Carsten Steinhoff: "Re: MLE for weighted AND censored samples"
- Previous message: Networking Guy: "Sum of Random Vectors"
- In reply to: thomas_at_the-ramms.com: "Is it possible to have high R-square and fail F-test"
- Messages sorted by: [ date ] [ thread ]
Date: Wed, 23 Mar 2005 00:50:17 EST
only when n is small enough... that is n<5...
we know that the r-square = Sum of squares regression / total Sum of Squares ( SSR/TSS). that is if r-square is large this means that the Sum of squares of Error is Low thus yield a significant F...
- Next message: Carsten Steinhoff: "Re: MLE for weighted AND censored samples"
- Previous message: Networking Guy: "Sum of Random Vectors"
- In reply to: thomas_at_the-ramms.com: "Is it possible to have high R-square and fail F-test"
- Messages sorted by: [ date ] [ thread ]
