How could I handle left censored data (under detection limits) in a variable that I want to use as a regressor in logistic regression? What's is the best approach?
Asked
Active
Viewed 102 times
1
-
Related: [How small a quantity should be added to x to avoid taking the log of zero?](http://stats.stackexchange.com/q/30728/7290) – gung - Reinstate Monica Jul 17 '16 at 13:54
-
I hope these 2 papers might help- https://sph.umich.edu/dioxin/PDF/OsloPosters/poster_%20belowLOD.pdf https://clu-in.org/conf/tio/ltmo/Nondetects_handout.pdf – Arpit Sisodia Jul 18 '16 at 06:28
-
Hi Arpit thanks for your answer. These papers are interesting and will undoubtedly be helpful, but they treat the left censored data as dependent variable in regression and what I meant to ask is how to treat under detection limits variables as independent variables in multiple regression specifically logistic regression – José Luis Jul 18 '16 at 07:43