ITC Research Computing Support
Missing Values Analysis and Imputation

Missing Values

Agenda

Assumptions

Software Assumptions

Software Documentation Contains Information about How Missing Values Are Handled

Software Actions on Missing Data

Inspecting the Data

Partial Results from PROC UNIVARIATE
Sample Drawn From a Normal Distribution

Partial Results from PROC UNIVARIATE
Data from Non-normal Distribution

Generating Some Test Data in SAS

Partial Data – testdata Data Set

Use a Scatterplot to Inspect the Data Visually

Use a Scatterplot to Inspect the Data Visually

Use a Scatterplot to Inspect the Data Visually

Use a Scatterplot to Inspect the Data Visually

SPSS – Use Interactive Graph

SAS Code

Missing Values in SAS TTEST Procedure

T-Test Results

Types of Missingness - MCAR

Types of Missingness - MAR

Non-ignorable Missingness

Deleting Cases With Missing Values
CCA – Complete Case Analysis

Disadvantages of CCA

Methods Generally Not Recommended

Pairwise Deletion
Available Case Analysis (ACA)

Mean Substitution

"Imputation involves replacing an incomplete..."

   Case – A Comparison of Methods – Truxillo
A Simulation Example Which Shows That Different Methods of Handling Missing Values Lead to Different Results

Expectation-Maximization (EM)

Data for EM and MI
From SAS Help

SAS Code for EM

Slide 34

Advantages of Multiple Imputation

Disadvantages of Multiple Imputation

Implementing a Multiple Imputation Strategy

PROC MI

Missing Data Patterns

Imputation Methods in PROC MI

PROC MIANALYZE

Additional Imputation Methods in PROC MI

Additional Imputation Procedures

Summary

References

"UPCOMING COLLOQUIA"

© 2008 by the Rector and Visitors of the University of Virginia.

The information contained on the University of Virginia’s Department of Information Technology and Communication (ITC) website is provided as a public service with the understanding that ITC makes no representations or warranties, either expressed or implied, concerning the accuracy, completeness, reliability or suitability of the information, including warrantees of title, non-infringement of copyright or patent rights of others. These pages are expected to represent the University of Virginia community and the State of Virginia in a professional manner in accordance with the University of Virginia’s Computing Policies.