book

Multiple Imputation of Missing Data Using SAS

by Patricia Berglund, Steven G. Heeringa

July 2014

Intermediate to advanced

164 pages

5h 36m

English

SAS Institute

Read now

Unlock full access

About This Book
About The Authors
Acknowledgements
Chapter 1: Introduction to Missing Data and Methods for Analyzing Data with Missing Values
1.1 Introduction1.2 Sources and Patterns of Item Missing Data1.3 Item Missing Data Mechanisms1.4 Review of Strategies to Address Item Missing Data1.4.1 Complete Case Analysis1.4.2 Complete Case Analysis with Weighting Adjustments1.4.3 Full Information Maximum Likelihood1.4.4 Expectation-Maximization Algorithm1.4.5 Single Imputation of Missing Values1.4.6 Multiple Imputation1.5 Outline of Book Chapters1.6 Overview of Analysis Examples
Chapter 2: Introduction to Multiple Imputation Theory and Methods
2.1 The Origins and Properties of Multiple Imputation Methods for Missing Data2.1.1 A Short History of Imputation Methods2.1.2 Why the Multiple Imputation Method?2.1.3 Overview of Multiple Imputation Steps2.2 Step 1—Defining the Imputation Model2.2.1 Choosing the Variables to Include in the Imputation Model2.2.2 Distributional Assumptions for the Imputation Model2.3 Algorithms for the Multiple Imputation of Missing Values2.3.1 General Theory for Multiple Imputation Algorithms2.3.2 Methods for Monotone Missing Data Patterns2.3.3 Methods for Arbitrary Missing Data Patterns2.4 Step 2–Analysis of the MI Completed Data Sets2.5 Step 3–Estimation and Inference for Multiply Imputed Data Sets2.5.1 Multiple Imputation–Estimators and Variances for Descriptive Statistics and Model Parameters2.5.2 Multiple Imputation–Confidence Intervals2.6 MI Procedures for Multivariate Inference2.6.1 Multiple Parameter Hypothesis Tests2.6.2 Tests of Linear Hypotheses2.7 How Many Multiple Imputation Repetitions Are Needed?2.8 Summary
Chapter 3: Preparation for Multiple Imputation
3.1 Planning the Imputation Session3.2 Choosing the Variables to Include in a Multiple Imputation3.3 Amount and Pattern of Missing Data3.4 Types of Variables to Be Imputed3.5 Imputation Methods3.6 Number of Imputations (MI Repetitions)3.7 Overview of Multiple Imputation Procedures3.8 Multiple Imputation Example3.9 Summary
Chapter 4: Multiple Imputation for the Analyzsis of Complex Sample Survey Data 49
4.1 Multiple Imputation and Informative Data Collection Designs4.2 Complex Sample Surveys4.3 Incorporating the Complex Sample Design in the MI Imputation Step4.4 Incorporating the Complex Sample Design in the MI Analysis and Inference Steps4.5 MI Imputation and Analysis for Subpopulations of Complex Sample Design Data Sets4.6 Summary
Chapter 5: Multiple Imputation of Continuous Variables
5.1 Introduction to Multiple Imputation of Continuous Variables5.2 Imputation of Continuous Variables with Arbitrary Missing Data5.3 Imputation of Continuous Variables with Mixed Covariates and a Monotone Missing Data Pattern Using the Regression and Predictive Mean Matching Methods5.3.1 Imputation of Continuous Variables with Mixed Covariates and a Monotone Missing Data Pattern Using the Regression Method5.3.2 Imputation of Continuous Variables with Mixed Covariates and a Monotone Missing Data Pattern Using the Predictive Mean Matching Method5.4 Imputation of Continuous Variables with an Arbitrary Missing Data Pattern and Mixed Covariates Using the FCS Method5.4.1 Imputation of Continuous Variables with an Arbitrary Missing Data Pattern and Mixed Covariates Using the FCS Method5.5 Summary
Chapter 6: Multiple Imputation of Classification Variables
6.1 Introduction to Multiple Imputation of Classification Variables6.2 Imputation of a Classification Variable with a Monotone Missing Data Pattern Using the Logistic Method6.3 Imputation of Classification Variables with an Arbitrary Missing Data Pattern and Mixed Covariates Using the FCS Discriminant Function and the FCS Logistic Regression Method6.4 Imputation of Classification Variables with an Arbitrary Missing Data Pattern and Mixed Covariates: A Comparison of the FCS and MCMC/Monotone Methods6.4.1 Imputation of Classification Variables with Mixed Covariates and an Arbitrary Missing Data Pattern Using the FCS Method6.4.2 Imputation of Classification Variables with Mixed Covariates and an Arbitrary Missing Data Pattern Using the MCMC/Monotone and Monotone Logistic Methods with a Multistep Approach6.5 Summary
Chapter 7: Multiple Imputation Case Studies
7.1 Multiple Imputation Case Studies7.2 Comparative Analysis of HRS 2006 Data Using Complete Case Analysis and Multiple Imputation of Missing Data7.2.1 Exploration of Missing Data7.2.2 Complete Case Analysis Using PROC SURVEYLOGISTIC7.2.3 Multiple Imputation of Missing Data with an Arbitrary Missing Data Pattern Using the FCS Method with Diagnostic Trace Plots7.2.4 Logistic Regression Analysis of Imputed Data Sets Using PROC SURVEYLOGISTIC7.2.5 Use of PROC MIANALYZE with Logistic Regression Output7.2.6 Comparison of Complete Case Analysis and Multiply Imputed Analysis7.3 Imputation and Analysis of Longitudinal Seizure Data7.3.1 Introduction to the Seizure Data7.3.2 Exploratory Analysis of Seizure Data7.3.3 Conversion of Multiple-Record to Single-Record Data7.3.4 Multiple Imputation of Missing Data7.3.5 Conversion Back to Multiple Record Data for Analysis of Imputed Data Sets7.3.6 Regression Analysis of Imputed Data Sets7.4 Summary

Chapter 8: Preparation of Data Sets for PROC MIANALYZE
8.1 Preparation of Data Sets for Use in PROC MIANALYZE8.2 Imputation of Major League Baseball Players’ Salaries8.3.1 PROC GLM Output Data Set for Use in PROC MIANALYZE8.3.2 PROC MIXED Output Data Set for Use in PROC MIANALYZE8.4 Imputation of NCS-R Data8.5 PROC SURVEYPHREG Output Data Set for Use in PROC MIANALYZE8.6 Summary
References
Index

Overview

Find guidance on using SAS for multiple imputation and solving common missing data issues.

Multiple Imputation of Missing Data Using SAS provides both theoretical background and constructive solutions for those working with incomplete data sets in an engaging example-driven format. It offers practical instruction on the use of SAS for multiple imputation and provides numerous examples that use a variety of public release data sets with applications to survey data.

Written for users with an intermediate background in SAS programming and statistics, this book is an excellent resource for anyone seeking guidance on multiple imputation. The authors cover the MI and MIANALYZE procedures in detail, along with other procedures used for analysis of complete data sets. They guide analysts through the multiple imputation process, including evaluation of missing data patterns, choice of an imputation method, execution of the process, and interpretation of results.

Topics discussed include how to deal with missing data problems in a statistically appropriate manner, how to intelligently select an imputation method, how to incorporate the uncertainty introduced by the imputation process, and how to incorporate the complex sample design (if appropriate) through use of the SAS SURVEY procedures.

Discover the theoretical background and see extensive applications of the multiple imputation process in action.

This book is part of the SAS Press program.

Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.

Read now

Unlock full access

More than 5,000 organizations count on O’Reilly

O’Reilly covers everything we've got, with content to help us build a world-class technology community, upgrade the capabilities and competencies of our teams, and improve overall team performance as well as their engagement.

Julian F.

Head of Cybersecurity

I wanted to learn C and C++, but it didn't click for me until I picked up an O'Reilly book. When I went on the O’Reilly platform, I was astonished to find all the books there, plus live events and sandboxes so you could play around with the technology.

Addison B.

Field Engineer

I’ve been on the O’Reilly platform for more than eight years. I use a couple of learning platforms, but I'm on O'Reilly more than anybody else. When you're there, you start learning. I'm never disappointed.

Amir M.

Data Platform Tech Lead

I'm always learning. So when I got on to O'Reilly, I was like a kid in a candy store. There are playlists. There are answers. There's on-demand training. It's worth its weight in gold, in terms of what it allows me to do.

Mark W.

Embedded Software Engineer

Publisher Resources

ISBN: 9781612904528

Cloud Computing

Data Engineering

Data Science

AI & ML

Programming Languages

Software Architecture

IT/Ops

Security

Design

Business

Soft Skills