Applied Survey Data Analysis
Book Details :
Size3.83 MB

Applied Survey Data Analysis

Applied Survey Data Analysis by Steven G. Heeringa, Brady T. West, and Patricia A. Berglund | PDF Free Download.

Applied Survey Data Analysis Contents

  • Applied Survey Data Analysis: Overview
  • Getting to Know the Complex Sample Design
  • Foundations and Techniques for Design-Based Estimation and Inference
  • Preparation for Complex Sample Survey Data Analysis
  • Descriptive Analysis of Continuous Variables
  • Categorical Data Analysis
  • Linear Regression Models
  • Logistic Regression and Generalized Linear Models for Binary Survey Variables
  • Generalized Linear Models for Multinomial, Ordinal, and Count Variables
  • Survival Analysis of Event History Survey Data
  • Multiple Imputation: Methods and Applications for Survey Analysts
  • Advanced Topics in the Analysis of Survey Data

Preface to Applied Survey Data Analysis PDF

This book is written as a guide to the applied statistical analysis and interpretation of survey data.

The motivation for this text lies in years of teaching graduate courses in applied methods for survey data analysis and extensive consultation with social and physical scientists, educators, medical researchers, and public health professionals on best methods for approaching specific analysis questions using survey data.

The general outline for this text is based on the syllabus for a course titled “Analysis of Complex Sample Survey Data” that we have taught for over 10 years in the Joint Program in Survey Methodology (JPSM) based at the University of Maryland (College Park) and in the University of Michigan’s Program in Survey Methodology (MPSM) and Summer Institute in Survey Research Techniques.

Readers may initially find the topical outline and content choices a bit unorthodox, but our instructional experience has shown it to be effective for teaching this complex subject to students and professionals who have a minimum of a two-semester graduate-level course in applied statistics.

The practical, everyday relevance of the chosen topics and the emphasis each receives in this text has also been informed by over 60 years of combined experience in consulting on survey data analysis with research colleagues and students under the auspices of the Survey Methodology Program of the Institute for Social Research (ISR) and the University of Michigan Center for Statistical Consultation and Research (CSCAR).

For example, the emphasis placed on topics as varied as a weighted estimation of population quantities, sampling error calculation models, coding of indicator variables in regression models, and interpretation of results from generalized linear models derives directly from our long-term observation of how often naïve users make critical mistakes in these areas.

This text, like our courses that it will serve, is designed to provide an intermediate-level statistical overview of the analysis of complex sample survey data—emphasizing methods and worked examples while reinforcing the principles and theory that underly those methods.

The intended audience includes graduate students, survey practitioners, and research scientists from the wide array of disciplines that use survey data in their work.

Students and practitioners in the statistical sciences should also find that this text provides a useful framework for integrating their further, more in-depth studies of the theory and methods for survey data analysis.

Balancing theory and application in any text is no simple matter. The distinguished statistician D. R. Cox begins the outline of his view of applied statistical work by stating, “Any simple recommendation along the lines in applications one should do so and so is virtually bound to be wrong in some or, indeed, possibly many contexts.

On the other hand, the descent into yawning vacuous generalities is all too possible” (Cox, 2007). Since the ingredients of each applied survey data analysis problem vary the aims, the sampling design, the available survey variables there is no single set of recipes that each analyst can simply follow without additional thought and evaluation on his or her part.

On the other hand, a text on applied methods should not leave survey analysts alone, fending for themselves, with only abstract theoretical explanations to guide their way through an applied statistical analysis of survey data.

On balance, the discussion in this book will tilt toward proven recipes where theory and practice have demonstrated the value of a specific approach.

In cases where theoretical guidance is less clear, we identify the uncertainty but still aim to provide advice and recommendations based on experience and current thinking on best practices.

The chapters of this book are organized to be read in sequence, each chapter building on material covered in the preceding chapters.

Chapter 1 provides an important context for the remaining chapters, briefly reviewing historical developments and laying out a step-by-step process for approaching a survey analysis problem.

Chapters 2 through 4 will introduce the reader to the fundamental features of complex sample designs and demonstrate how design characteristics such as stratification, clustering, and weighting are easily incorporated into the statistical methods and software for survey estimation and inference.

Treatment of statistical methods for survey data analysis begins in Chapters 5 and 6 with coverage of univariate (i.e., single variable) descriptive and simple bivariate (i.e., two-variable) analyses of continuous and categorical variables.

Chapter 7 presents the linear regression model for continuous dependent variables. Generalized linear regression modeling methods for survey data are treated in Chapters 8 and 9.

Chapter 10 pertains to methods for the event-history analysis of survey data, including models such as the Cox proportional hazards model and discrete-time models.

Chapter 11 introduces methods for handling missing data problems in survey data sets. Finally, the coverage of statistical methods for survey data analysis concludes in Chapter 12 with a discussion of new developments in the area of survey applications of advanced statistical techniques, such as multilevel analysis.

To avoid repetition in the coverage of more general topics such as the recommended steps in a regression analysis or testing hypotheses concerning regression parameters, topics will be introduced as they become relevant to the specific discussion.

For example, the iterative series of steps that we recommend analysts follow in regression modeling of survey data is introduced in Chapter 7 (linear regression models for continuous outcomes),

but the series applies equally to model specification, estimation, evaluation, and inference for generalized linear regression models (Chapters 8 and 9).

By the same token, specific details of the appropriate procedures for each step (e.g., regression model diagnostics) are covered in the chapter on a specific technique.

Readers who use this book primarily as a reference volume will find cross-references to earlier chapters useful in locating important background for discussion of specific analysis topics. There are many quality software choices out there for survey data analysts.

We selected Stata® for all book examples due to its ease of use and flexibility for survey data analysis, but examples have been replicated to the greatest extent possible using the SAS®, SPSS®, IVEware, SUDAAN®, R, WesVar®, and Mplus software packages on the book Web site ( src/smp/asda/).

Appendix A reviews software procedures that are currently available for the analysis of complex sample survey data in these other major software systems.

Examples based on the analysis of major survey data sets are routinely used in this book to demonstrate statistical methods and software applications.

To ensure diversity in sample design and substantive content, example exercises and illustrations are drawn from three major U.S. survey data sets: the 2005–2006 National Health and Nutrition Examination Survey (NHANES); the 2006 Health and Retirement Study (HRS); and the National Comorbidity Survey-Replication (NCS-R). A description of each of these survey data sets is provided in Section 1.3.

A series of practical exercises based on these three data sets are included at the end of each chapter on an analysis topic to provide readers and students with examples enabling practice with using statistical software for applied survey data analysis.

Clear and consistent use of statistical notation is important. Table P.1 provides a summary of the general notational conventions used in this book. Special notation and symbol representation will be defined as needed for discussion of specific topics.

The materials and examples presented in the chapters of this book (which we refer to in subsequent chapters as ASDA) are supplemented through a companion Web site (

This Web site provides survey analysts and instructors with additional resources in the following areas: links to new publications and an updated bibliography for the survey analysis topics covered in Chapters 5–12;

links to sites for example survey data sets; replication of the command setups and output for the analysis examples in the SAS, SUDAAN, R, SPSS, and Mplus software systems; answers to frequently asked questions (FAQs);

short technical reports related to special topics in applied survey data analysis; and reviews of statistical software system updates and any resulting changes to the software commands or output for the analysis examples.

In closing, we must certainly acknowledge the many individuals who contributed directly or indirectly to the production of this book.

Gail Arnold provided invaluable technical and organizational assistance throughout the production and review of the manuscript. Rod Perkins provided exceptional support in the final stages of manuscript review and preparation.

Deborah Kloska and Lingling Zhang generously gave of their time and statistical expertise to systematically review each chapter as it was prepared. Joe Kazemi and two anonymous reviewers offered helpful comments on earlier versions of the introductory chapters, and SunWoong Kim and Azam Khan also reviewed the more technical material in our chapters for accuracy.

We owe a debt to our many students in the JPSM and MPSM programs who over the years have studied with us we only hope that you learned as much from us as we did from working with you.

As lifelong students ourselves, we owe a debt to our mentors and colleagues who over the years have instilled in us a passion for statistical teaching and consultation: Leslie Kish, Irene Hess, Graham Kalton, Morton Brown, Edward Rothman, and Rod Little.

Finally, we wish to thank the support staff at Chapman Hall/CRC Press, especially Rob Calver and Sarah Morris, for their continued guidance.

Download Applied Survey Data Analysis in PDF Format For Free.