Multiple Imputation for Categorical Time Series

Abstract

The mict package provides a method for multiple imputation of categorical time-series data (such as life course or employment status histories) that preserves longitudinal consistency, using a monotonic series of imputations. It allows flexible imputation specifications with a model appropriate to the target variable (mlogit, ologit, etc.). Where transitions in individual units’ data are substantially less frequent than one per period and where missingness tends to be consecutive (as is typical of life course data), mict produces imputations with better longitudinal consistency than mi impute or ice.