Missing data
-
Joint Modelling Multiple Imputation is implemented in the R package jomo.
-
This allows for imputation of either continuous or discrete, participant-level or study-level, and systematically or sporadically missing data.
-
A tutorial paper introducing the package has been published.
​
-
Several options are available to perform multiple imputation using the fully conditional specification approach.
-
In particular, the R package mice offers the implementation of one-stage multilevel multiple imputation methods for handling systematically and sporadically missing data in clustered data, and has built-in functions for the imputation of continuous and binary variables.
-
The functionalities of mice have been extended by the R package micemd, which offers the implementation of one-stage and two-stage multilevel imputation methods for continuous, binary and count variables.
-
Other R packages are available for dealing with missing values in clustered data (e.g. hmi or miceadds), but they are generally not dedicated to handle systematically and sporadically missing data jointly.