0

As a follow up to this question unscale and uncenter glmer parameters, how can I unscale and uncenter glmer standard errors?

Could someone provide a function similar to rescale.coefs but for standard errors? If needed, the question referred to above includes an example.

Community
  • 1
  • 1
bdu
  • 302
  • 1
  • 3
  • 12
  • You should offer a particular data-situation. And since it really does look trivial you should indicate what difficulty you are having with understanding how to "unscale" a value. The scaled vectors will have the necessary information in their attributes. Read `?scale` before responding. – IRTFM Oct 22 '15 at 15:24
  • Thanks. Unfortunately it is not simply a matter of undoing `scale` but of understanding how to "re-adjust" the standard errors of scaled regressors. – bdu Oct 22 '15 at 17:30
  • Then it's really not a well specified coding problem yet, and you should seek appropriate statistical advice. – IRTFM Oct 22 '15 at 18:08
  • Clarification: it is very simple to un-standardize (reverse the centering and scaling of) the standard errors of the parameter estimates. I am just struggling with the the un-standardization of the standard error of the intercept. Thanks. – bdu Oct 29 '15 at 18:07
  • @BondedDust As you suggested, I asked a similar question on [Cross Validated](http://stats.stackexchange.com/questions/179281/determine-the-original-standard-error-of-the-intercept-in-a-regression-estimated) – bdu Oct 29 '15 at 19:59

0 Answers0