Cook's D measures the change in the parameter estimates caused by deleting each observation. For linear models,
- Di = [1/(p s2)] (b- b(i))' (X'X) (b- b(i))
is the vector of parameter estimates obtained after deleting the i
Cook (1977) suggests comparing Di to the F distribution with p and n-p degrees of freedom.
For generalized linear models,
where W = Wo
when the full Hessian is used and W = We
when Fisher's scoring method is used.
Cook's D statistics are stored in variables named D_yname for each response variable, where yname is the response variable name.
Copyright © 2007 by SAS Institute Inc., Cary, NC, USA. All rights reserved.