Fully Efficient Joint Fractional Imputation for Incomplete Bivariate Ordinal Responses

  title={Fully Efficient Joint Fractional Imputation for Incomplete Bivariate Ordinal Responses},
  author={Xichen She and Changbao Wu},
  journal={Statistica Sinica},
We propose a fully efficient joint fractional imputation method for handling bivariate ordinal responses with missing observations. We show that the method is ideally suited for bivariate ordinal responses to create a single imputed data file and provides valid and efficient inferences for the joint and marginal probabilities, association measures, as well as regression analysis. Asymptotic properties of estimators based on the joint fractionally imputed data set are developed and their… 

Figures and Tables from this paper

Efficient Utilization of Missing Data in Cost-Sensitive Learning

Experimental results on UCI data sets demonstrated the advantages of the proposed Date-drive Incremental imputation Model, DIM, compared to the comparison methods, in terms of prediction accuracy and classification accuracy.



Fractional Imputation in Survey Sampling: A Comparative Review

The empirical performance of FI is demonstrated with respect to multiple imputation using a pseudo finite population generated from a sample from the Monthly Retail Trade Survey conducted by the US Census Bureau.

Parametric fractional imputation for missing data analysis

Parametric fractional imputation is proposed as a general tool for missing data analysis. Using fractional weights, the observed likelihood can be approximated by the weighted mean of the imputed

Fractional Imputation for Ordinal and Mixed-type Responses with Missing Observations

This thesis addresses two essential aspects of large scale public-use data files involving ordinal and mixed-type responses with missing observations: (i) the creation of single complete data sets

GEE for Multinomial Responses Using a Local Odds Ratios Parameterization

In simulations, the parameter estimators with the proposed GEE method for a marginal cumulative probit model appear to be less biased and more efficient than those with the independence "working" model, especially for studies having time-varying covariates and strong correlation.

A multivariate technique for multiply imputing missing values using a sequence of regression models

This article describes and evaluates a procedure for imputing missing values for a relatively complex data structure when the data are missing at random. The imputations are obtained by fitting a

Fractional hot deck imputation

To compensate for item nonresponse, hot deck imputation procedures replace missing values with values that occur in the sample. Fractional hot deck imputation replaces each missing observation with a

Jackknife variance estimation with survey data under hot deck imputation

SUMMARY Hot deck imputation is commonly employed for item nonresponse in sample surveys. It is also a common practice to treat the imputed values as if they are true values, and then compute the

Generalized Estimating Equation

The generalized estimating equation (GEE) uses a quasi-likelihood approach for analyzing data with correlated outcomes. This is an extension of GLM and uses quasi-likelihood method for cluster or

MICE: Multivariate Imputation by Chained Equations in R

Mice adds new functionality for imputing multilevel data, automatic predictor selection, data handling, post-processing imputed values, specialized pooling routines, model selection tools, and diagnostic graphs.

Missing Data, Imputation, and the Bootstrap

Three main topics are discussed: bootstrap methods for missing data, these methods' relationship to the theory of multiple imputation, and computationally efficient ways of executing them.