RE: Centering (was Re: Missing covariates)

From: Kenneth G. Kowalski Date: July 05, 2001 technical Source: cognigencorp.com
From: "KOWALSKI, KENNETH G. [PHR/1825]" <kenneth.g.kowalski@pharmacia.com> Subject: RE: Centering (was Re: Missing covariates) Date: Thu, 5 Jul 2001 09:48:17 -0500 Alan/NMUSERS: Who really cares if INTERCEPT(no centering) has a different standard error than INTERCEPT(centering)? They are not estimating the same thing! Note that INTERCEPT(no centering)+SLOPE(no centering)*CENTEREDVALUE is estimating the same thing as INTERCEPT(centering). If you calculatedthe standard error for INTERCEPT(no centering)+SLOPE(no centering)*CENTEREDVALUE from the covariance matrix of the estimates of the thetas you will get the same standard error as that reported for INTERCEPT(centering). The really important issue is the slope for the covariate effect. If a global minimum is achieved then SLOPE(no centering) and SLOPE(centering) will have the same estimates and standard errors because they are estimating the same thing. Again, the benefits of centering are purely numerical not statistical. I have no problems with not centering if the model is numerically stable and you achieve the global minimum. This takes the question of how far off from the center of your data to do the centering to the extreme (i.e., centered about zero). Obviously, the answer depends on the numerical stability of the model that you are fitting to your data. Since one doesn't always know at the onset how stable the model is going to be, it is good practice to do centering. Whether one uses some standard value (such as Nick suggests) or the mean or median doesn't really matter provided they all achieve the global minimum...pick the one that is most convenient to you. If achieving of the global miniminum is very sensitive to the choice of centering then you may need to look at the model more closely anyway...perhaps it is overparameterized. I agree with Nick that it is convenient to choose standards. Nevertheless, I probably wouldn't use an age of 40 if the range of ages in the study were between 55 and 75...not that there is anything wrong with that (to quote Jerry Seinfeld) if a global minimum is achieved. I would probably use something like 60 or 65 regardless of what the mean or median of the distribution of ages was in the study. If the model was so sensitive to whether I used 60 or 65 versus the mean or the median then I would worry about the appropriateness of the model more so than what value I should use for centering. Ken
Jul 02, 2001 Nick Holford Centering (was Re: Missing covariates)
Jul 02, 2001 William Bachman RE: Centering (was Re: Missing covariates)
Jul 02, 2001 Kenneth G. Kowalski RE: Centering (was Re: Missing covariates)
Jul 02, 2001 Lewis B. Sheiner Centering (was Re: Missing covariates)
Jul 03, 2001 Jogarao Gobburu Re: Centering (was Re: Missing covariates)
Jul 03, 2001 Alan Xiao Re: Centering (was Re: Missing covariates)
Jul 03, 2001 Nick Holford Re: Centering (was Re: Missing covariates)
Jul 03, 2001 Alan Xiao Re: Centering (was Re: Missing covariates)
Jul 03, 2001 Lewis B. Sheiner Re: Centering (was Re: Missing covariates)
Jul 03, 2001 Alan Xiao Re: Centering (was Re: Missing covariates)
Jul 03, 2001 Diane Mould Re: Centering (was Re: Missing covariates)
Jul 04, 2001 Nick Holford Re: Centering (was Re: Missing covariates)
Jul 04, 2001 Alan Xiao Re: Centering (was Re: Missing covariates)
Jul 04, 2001 Diane Mould Re: Centering (was Re: Missing covariates)
Jul 05, 2001 Nick Holford Re: Centering (was Re: Missing covariates)
Jul 05, 2001 Stephen Duffull RE: Centering (was Re: Missing covariates)
Jul 05, 2001 Nick Holford Re: Centering (was Re: Missing covariates)
Jul 05, 2001 Leon Aarons 70kg neonates
Jul 05, 2001 Nick Holford Re: 70kg neonates
Jul 05, 2001 Peter Bonate Centering
Jul 05, 2001 Alan Xiao Re: Centering (was Re: Missing covariates)
Jul 05, 2001 Leonid Gibiansky RE: Centering (was Re: Missing covariates)
Jul 05, 2001 Kenneth G. Kowalski RE: Centering (was Re: Missing covariates)
Jul 05, 2001 William Bachman RE: Centering (was Re: Missing covariates)
Jul 05, 2001 Diane Mould Re: Centering (was Re: Missing covariates)
Jul 05, 2001 Alan Xiao Re: Centering (was Re: Missing covariates)
Jul 05, 2001 Alan Xiao Question 2 about prediction and covariates
Jul 06, 2001 Matt Hutmacher RE: Centering (was Re: Missing covariates)
Jul 09, 2001 Vladimir Piotrovskij RE: Centering (Impact on SE)
Jul 09, 2001 Alan Xiao Re: Centering (was Re: Missing covariates)
Jul 09, 2001 Kenneth G. Kowalski RE: Centering (Impact on SE)
Jul 09, 2001 Vladimir Piotrovskij RE: Centering (Impact on SE)
Jul 09, 2001 Smith Brian P RE: Centering (Impact on SE)
Jul 09, 2001 Matt Hutmacher RE: Centering (was Re: Missing covariates)
Jul 12, 2001 Juan Jose Perez Ruixo RE: Centering (was Re: Missing covariates)
Jul 12, 2001 Juan Jose Perez Ruixo RE: Centering (was Re: Missing covariates)
Jul 12, 2001 Matt Hutmacher RE: Centering (was Re: Missing covariates)
Jul 12, 2001 Alan Xiao Re: Centering (was Re: Missing covariates)
Jul 30, 2001 Juan Jose Perez Ruixo Re: Centering (was Re: Missing covariates)
Jul 30, 2001 Alan Xiao Re: Centering (was Re: Missing covariates)
Jul 30, 2001 Leonid Gibiansky RE: Centering (was Re: Missing covariates)