[NMusers] Pred-corrected VPC with bounded data

From: Matts Kågedal <mattskagedal_at_gmail.com>
Date: Thu, 7 Sep 2017 10:38:10 -0700

Dear modelers,

Working on a model where the observations are bound between 0 and 100.
Using logit transformation in the model to account for this (I dont
transform the data).

Would like to assess the model based on a VPC.
Since dosing is per the physicians decision based on the observed response,
I would like to generate a pred-corrected VPC using PsN.

When reading the PsN userguide, it says that "*The predcorr feature will
not work well if there is an upper bound, e.g. when the dependent variable
is logit transformed."*

Any thoughts on how to solve this would be appreciated. Would ideally like
to avoid having to simulated the physicians dose decisions.

Matts Kagedal

Received on Thu Sep 07 2017 - 13:38:10 EDT

This archive was generated by hypermail 2.3.0 : Fri Sep 27 2019 - 16:56:20 EDT