R + ztable: what are lcl and ucl? (logistic regression) -
ztable produces nice zebra (alternating striped) tables, useful presenting dataframes , model objects (lm, glm, etc). it's not clear me lcl , ucl are. documentation says
‘ztable()’ shows odds ratio(or) , 95% confidence interval
but googling indicates lcl/ucl lower/upper control limits, seem confidence intervals constructed +/- 3 standard deviations mean.
most confusingly, lcl/ucl constructed ztable don't contain coefficient estimate:
ztable(glm(factor(am) ~ disp, family = binomial(link = "logit"), data = mtcars))
so lcl/ucl mean?
from examining source code ztable.glm
(thanks @user2957945), answer seems simply:
‘ztable()’ shows odds ratio(or) , 95% confidence interval (lcl, ucl) of odds ratio
the odds ratio exponentiated coefficient (which logit or log odds), , lcl/ucl obtained exponentiating confidence interval of coefficient.
Comments
Post a Comment