1st Edition
R Companion for Sampling Design and Analysis, Third Edition
The R Companion for Sampling: Design and Analysis, designed to be read alongside Sampling: Design and Analysis, Third Edition by Sharon L. Lohr (SDA; 2022, CRC Press), shows how to use functions in base R and contributed packages to perform calculations for the examples in SDA.
No prior experience with R is needed. Chapter 1 tells you how to obtain R and RStudio, introduces basic features of the R statistical software environment, and helps you get started with analyzing data.
Each subsequent chapter provides step-by-step guidance for working through the data examples in the corresponding chapter of SDA, with code, output, and interpretation. Tips and warnings help you develop good programming practices and avoid common survey data analysis errors.
R features and functions are introduced as they are needed so you can see how each type of sample is selected and analyzed. Each chapter builds on the knowledge developed earlier for simpler designs; after finishing the book, you will know how to use R to select and analyze almost any type of probability sample.
All R code and data sets used in this book are available online to help you develop your skills analyzing survey data from social and public opinion research, public health, crime, education, business, agriculture, and ecology.
- Getting Started
- Simple Probability Samples
- Stratified Sampling
- Ratio and Regression Estimation
- Cluster Sampling with Equal Probabilities
- Sampling with Unequal Probabilities
- Complex Surveys
- Nonresponse
- Variance Estimation in Complex Surveys
- Categorical Data Analysis in Complex Surveys
- Regression with Complex Survey Data
- Additional Topics for Survey Data Analysis
Obtaining the Software
Installing R packages
R Basics
Reading Data into R
Saving Output
Integrating R Output into LATEX Documents
Missing Data
Summary, Tips, and Warnings
Selecting a Simple Random Sample
Computing Statistics from an SRS
Additional Code for Exercises
Summary, Tips, and Warnings
Allocation Methods
Selecting a Stratified Random Sample
Computing Statistics from a Stratified Random Sample
Estimating Proportions from a Stratified Random Sample
Additional Code for Exercises
Summary, Tips, and Warnings
Ratio Estimation
Regression Estimation
Domain Estimation
Poststratification
Ratio Estimation with Stratified Sampling
Model-Based Ratio and Regression Estimation
Summary, Tips, and Warnings
Estimates from One-Stage Cluster Samples
Estimates from Multi-Stage Cluster Samples
Model-Based Design and Analysis for Cluster Samples
Additional Code for Exercises
Summary, Tips, and Warnings
Selecting a Sample with Unequal Probabilities
Sampling With Replacement
Sampling Without Replacement
Selecting a Two-stage Cluster Sample
Computing Estimates from an Unequal-Probability Sample
Estimates from With-Replacement Samples
Estimates from Without-Replacement Samples
Summary, Tips, and Warnings
Selecting a Stratified Two-Stage Sample
Estimating Quantiles
Computing Estimates from Stratified Multistage Samples
Univariate Plots from Complex Surveys
Scatterplots from Complex Surveys
Additional Code for Exercises
Summary, Tips, and Warnings
How R Functions Treat Missing Data
Poststratification and Raking
Imputation
Summary, Tips, and Warnings
Replicate Samples and Random Groups
Constructing Replicate Weights
Balanced Repeated Replication
Jackknife
Bootstrap
Replicate Weights and Nonresponse Adjustments
Using Replicate Weights from a Survey Data File
Summary, Tips, and Warnings
Contingency Tables and Odds Ratios
Chi-Square Tests
Loglinear Models
Summary, Tips, and Warnings
Straight Line Regression in an SRS
Linear Regression for Complex Survey Data
Multiple Linear Regression
Using Regression to Compare Domain Means
Logistic Regression
Additional Resources and Code
Summary, Tips, and Warnings
Two-Phase Sampling
Contents iii
Estimating the Size of a Population
Ratio Estimation of Population Size
Loglinear Models with Multiple Lists
Small Area Estimation
Summary
A Data Set Descriptions
Bibliography
Index
Biography
Yan Lu is Associate Professor of Statistics at the University of New Mexico. Her research interests include survey sampling, mixed models, nonparametric regression, and data mining. Recent publications develop new statistical methods for combining data from multiple surveys, selecting probability samples from massive data streams, and applying nonparametric regression to survey data.
Sharon L. Lohr, the author of Measuring Crime: Behind the Statistics, has published widely about survey sampling and statistical methods for education, public policy, law, and crime. She is a Fellow of the American Statistical Association and an elected member of the International Statistical Institute, and has received the Gertrude M. Cox, Morris Hansen, and Deming Awards. Formerly Dean’s Distinguished Professor of Statistics at Arizona State University and a Vice President at Westat, she is now a statistical consultant and writer.
