Nonresponse in Household Interview Surveys E-Book

Robert M. Groves

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Serie: Wiley Series in Survey Methodology
• Sprache: Englisch

Seitenzahl: 675

Veröffentlichungsjahr: 2012

Beschreibung

A comprehensive framework for both reduction of nonresponse andpostsurvey adjustment for nonresponse This book provides guidance and support for survey statisticianswho need to develop models for postsurvey adjustment fornonresponse, and for survey designers and practitioners attemptingto reduce unit nonresponse in household interview surveys. Itpresents the results of an eight-year research program that hasassembled an unprecedented data set on respondents andnonrespondents from several major household surveys in the UnitedStates. Within a comprehensive conceptual framework of influences onnonresponse, the authors investigate every aspect of surveycooperation, from the influences of household characteristics andsocial and environmental factors to the interaction betweeninterviewers and householders and the design of the surveyitself. Nonresponse in Household Interview Surveys: * Provides a theoretical framework for understanding and studyinghousehold survey nonresponse * Empirically explores the individual and combined influences ofseveral factors on nonresponse * Presents chapter introductions, summaries, and discussions onpractical implications to clarify concepts and theories * Supplies extensive references for further study and inquiry Nonresponse in Household Interview Surveys is an important resourcefor professionals and students in survey methodology/researchmethods as well as those who use survey methods or data inbusiness, government, and academia. It addresses issues critical todealing with nonresponse in surveys, reducing nonresponse duringsurvey data collection, and constructing statistical compensationsfor the effects of nonresponse on key survey estimates.

Leseprobe

Contents

Cover

Half Title page

Title page

Copyright page

Preface

Acknowledgments

Chapter One: An Introduction to Survey Participation

1.1 Introduction

1.2 Statistical Impacts of Nonresponse on Survey Estimates

1.3 How Householders Think about Survey Requests

1.4 How Interviewers Think about Survey Participation

1.5 How Survey Design Features Affect Nonresponse

1.6 The Focus of this Book

1.7 Limitations of this Book

1.8 Summary

Chapter Two: A Conceptual Framework for Survey Participation

2.1 Introduction

2.2 Practical Features of Survey Nonresponse Needing Theoretical Explanation

2.3 A Conceptual Structure for Survey Participation

2.4 Implications for Research

2.5 Practical Implications for Survey Implementation

2.6 Summary

Chapter Three: Data Resources for Testing Theories of Survey Participation

3.1 Introduction

3.2 Approaches to Studying Nonresponse

3.3 Qualitative Data from Interviewer Group Discussions

3.4 Decennial Census Match of Survey Records to Census Records

3.5 Documentation of Interaction Between Interviewers and Householders

3.6 Surveys of Interviewers

3.7 Measures of Social and Economic Ecology of Sample Households

3.8 Limitations of the Tests of the Theoretical Perspective

3.9 Summary

Chapter Four: Influences on the Likelihood of Contact

4.1 Introduction

4.2 Social Environmental Indicators of At-Home Patterns

4.3 Household-Level Correlates of Contactability

4.4 Interviewer-Level Correlates of Contactability

4.5 Call-Level Influences on Contacting Sample Households

4.6 Joint Effects of Multiple Levels on Contactability

4.7 Summary

4.8 Practical Implications for Survey Implementation

Chapter Five: Influences of Household Characteristics on Survey Cooperation

5.1 Introduction

5.2 Opportunity Cost Hypotheses

5.3 Exchange Hypotheses

5.4 Social Isolation Hypotheses

5.5 The Concept of Authority and Survey Cooperation

5.6 Joint Effects of Indicators of Social Isolation and Authority

5.7 Other Household-Level Influences on Cooperation

5.8 Multivariate Models of Cooperation Involving Household-Level Predictors

5.9 Summary

5.10 Practical Implications for Survey Implementation

Chapter Six: Social Environmental Influences on Survey Participation

6.1 Introduction

6.2 Trends in Response Rates over Time

6.3 Cross-National Differences in Response Rates on Similar Surveys

6.4 “Natural Experiments” at the Societal Level

6.5 Subnational Variation in Survey Cooperation

6.6 Analysis of Environmental Influences on Cooperation

6.7 Bivariate Relationships of Survey Cooperation and Environmental Factors

6.8 Marginal Effects of Individual Environmental Factors

6.9 Summary

6.10 Practical Implications for Survey Implementation

Chapter Seven: Influences of the Interviewers

7.1 Introduction

7.2 Interviewer Effects on Cooperation

7.3 The Role and Task of Interviewers

7.4 Socio-Demographic Characteristics of Interviewers

7.5 Interviewer Personality

7.6 Interviewer Experience

7.7 Interviewer Attitudes and Expectations Regarding Nonresponse

7.8 Interviewer Behaviors

7.9 Multivariate Models of Interviewer-Level Effects

7.10 Summary

7.11 Practical Implications for Survey Implementation

Chapter Eight: When Interviewers Meet Householders: The Nature of Initial Interactions

8.1 Introduction

8.2 The Initial Interaction from the Householder’s Perspective

8.3 Cues for Judging the Intent of the Interviewer

8.4 interaction from the Interviewer’s Perspective

8.5 Empirical Measurement of Interactions Between Interviewers and Householders

8.6 Nature of the Householder-Interviewer Interaction

8.7 Summary

Chapter Nine: Influences of Householder—Interviewer Interactions on Survey Cooperation

9.1 Introduction

9.2 Tailoring

9.3 Maintaining Interaction

9.4 Useful Concepts Related to Tailoring

9.5 Past Research on Interviewer-Householder Interaction Affecting Cooperation

9.6 Predicting the Outcome of Contacts Using Characteristics of the Interaction

9.7 Effects of Interviewer-Householder Interaction on the Final Disposition of Sample Households

9.8 Summary

Chapter Ten: How Survey Design Features Affect Participation

10.1 Introduction

10.2 The Balance of Cost, Timeliness, Measurement, and Survey Errors

10.3 Survey Design Features Affecting Likelihood of Contact of Sample Households

10.4 Survey Design Features Affecting Cooperation

10.5 Summary

Chapter Eleven: Practical Survey Design Acknowledging Nonresponse

11.1 Introduction

11.2 Selection of Sampling Frames

11.3 Choice of Mode of Data Collection

11.4 Design of Measurement Instruments

11.5 Selection and Training of Interviewers

11.6 Call Attempts on Sample Units

11.7 The First-Contact Protocol

11.8 Efforts at Nonresponse Reduction after the First Contact

11.9 Postsurvey Adjustments for Unit Nonresponse

11.10 Summary

Reference

Index

Nonresponse in Household Interview Surveys

WILEY SERIES IN PROBABILITY AND STATISTICS

ESTABLISHED BY WALTER A. SHEWHART AND SAMUEL S. WlLKS

EditorsVic Barnett, Ralph A. Bradley, Noel A. C. Cressie, Nicholas I. Fisher, Iain M. Johnstone, J. B. Kadane, David G. Kendall, David W. Scott, Bernard W. Silverman, Adrian F. M. Smith, Jozef L. Teugels; J. Stuart Hunter, Emeritus

Probability and Statistics Section

*ANDERSON · The Statistical Analysis of Time Series

ARNOLD, BALAKRISHNAN, and NAGARAJA · A First Course in Order Statistics

BACCELLI, COHEN, OLSDER, and QUADRAT · Synchronization and Linearity: An Algebra for Discrete Event Systems

BASILEVSKY · Statistical Factor Analysis and Related Methods: Theory and Applications

BERNARDO and SMITH · Bayesian Statistical Concepts and Theory

BILLINGSLEY · Convergence of Probability Measures

BOROVKOV · Asymptotic Methods in Queuing Theory

BRANDT, FRANKEN, and LISEK · Stationary Stochastic Models

CAINES · Linear Stochastic Systems

CAIROLI and DALANG · Sequential Stochastic Optimization

CONSTANTINE · Combinatorial Theory and Statistical Design

COVER and THOMAS · Elements of Information Theory

CSÖRGÖ and HORVÁTH · Weighted Approximations in Probability Statistics

CSÖRGÖ and HORVÁTH · Limit Theorems in Change Point Analysis

DETTE and STUDDEN · The Theory of Canonical Moments with Applications in Statistics, Probability, and Analysis

*DOOB · Stochastic Processes

DRYDEN and MARDIA · Statistical Analysis of Shape

DUPUIS and ELLIS · A Weak Convergence Approach to the Theory of Large Deviations

ETHIER and KURTZ · Markov Processes: Characterization and Convergence

FELLER · An Introduction to Probability Theory and Its Applications, Volume 1, Third Edition, Revised; Volume II, Second Edition

FULLER · Introduction to Statistical Time Series, Second Edition

FULLER · Measurement Error Models

GELFAND and SMITH · Bayesian Computation

GHOSH, MUKHOPADHYAY, and SEN · Sequential Estimation

GIFI · Nonlinear Multivariate Analysis

GUTTORP · Statistical Inference for Branching Processes

HALL · Introduction to the Theory of Coverage Processes

HAMPEL · Robust Statistics: The Approach Based on Influence Functions

HANNAN and DEISTLER · The Statistical Theory of Linear Systems

HUBER · Robust Statistics

IMAN and CONOVER · A Modern Approach to Statistics

JUREK and MASON · Operator-Limit Distributions in Probability Theory

KASS and VOS · Geometrical Foundations of Asymptotic Inference

KAUFMAN and ROUSSEEUW · Finding Groups in Data: An Introduction to Cluster Analysis

KELLY · Probability, Statistics, and Optimization

LINDVALL · Lectures on the Coupling Method

McFADDEN · Management of Data in Clinical Trials

MANTON, WOODBURY, and TOLLEY · Statistical Applications Using Fuzzy Sets

MARDIA and JUPP · Statistics of Directional Data, Second Edition

MORGENTHALER and TUKEY · Configural Polysampling: A Route to Practical Robustness

MUIRHEAD · Aspects of Multivariate Statistical Theory

OLIVER and SMITH · Influence Diagrams, Belief Nets and Decision Analysis

*PARZEN · Modern Probability Theory and Its Applications

PRESS · Bayesian Statistics: Principles, Models, and Applications

PUKELSHEIM · Optimal Experimental Design

RAO · Asymptotic Theory of Statistical Inference

RAO · Linear Statistical Inference and Its Applications, Second Edition

*RAO and SHANBHAG · Choquet-Deny Type Functional Equations with Applications to Stochastic Models

ROBERTSON, WRIGHT, and DYKSTRA · Order Restricted Statistical Inference

ROGERS and WILLIAMS · Diffusions, Markov Processes, and Martingales, Volume I: Foundations, Second Edition; Volume II: îto Calculus

RUBINSTEIN and SHAPIRO · Discrete Event Systems: Sensitivity Analysis and Stochastic Optimization by the Score Function Method

RUZSA and SZEKELY · Algebraic Probability Theory

SCHEFFE · The Analysis of Variance

SEBER · Linear Regression Analysis

SEBER · Multivariate Observations

SEBER and WILD · Nonlinear Regression

SERFLING · Approximation Theorems of Mathematical Statistics

SHORACK and WELLNER · Empirical Processes with Applications to Statistics

SMALL and McLEISH · Hilbert Space Methods in Probability and Statistical Inference

STAPLETON · Linear Statistical Models

STAUDTE and SHEATHER · Robust Estimation and Testing

STOYANOV · Counterexamples in Probability

TANAKA · Time Series Analysis: Nonstationary and Noninvertible Distribution Theory

THOMPSON and SEBER · Adaptive Sampling

WELSH · Aspects of Statistical Inference

WHITTAKER · Graphical Models in Applied Multivariate Statistics

YANG · The Construction Theory of Denumerable Markov Processes

Applied Probability and Statistics Section

ABRAHAM and LEDOLTER · Statistical Methods for Forecasting

AGRESTI · Analysis of Ordinal Categorical Data

AGRESTI · Categorical Data Analysis

ANDERSON, AUQUIER, HAUCK, OAKES, VANDAELE, and WEISBERG · Statistical Methods for Comparative Studies

ARMITAGE and DAVID (editors) · Advances in Biometry

*ARTHANARI and DODGE · Mathematical Programming in Statistics

ASMUSSEN · Applied Probability and Queues

*BAILEY · The Elements of Stochastic Processes with Applications to the Natural Sciences

BARNETT and LEWIS · Outliers in Statistical Data, Third Edition

BARTHOLOMEW, FORBES, and McLEAN · Statistical Techniques for Manpower Planning, Second Edition

BATES and WATTS · Nonlinear Regression Analysis and Its Applications

BECHHOFER, SANTNER, and GOLDSMAN · Design and Analysis of Experiments for Statistical Selection, Screening, and Multiple Comparisons

BELSLEY · Conditioning Diagnostics: Collinearity and Weak Data in Regression

BELSLEY, KUH, and WELSCH · Regression Diagnostics: Identifying Influential Data and Sources of Collinearity

BHAT · Elements of Applied Stochastic Processes, Second Edition

BHATTACHARYA and WAYMIRE · Stochastic Processes with Applications

BIRKES and DODGE · Alternative Methods of Regression

BLOOMFIELD · Fourier Analysis of Time Series: An Introduction

BOLLEN · Structural Equations with Latent Variables

BOULEAU · Numerical Methods for Stochastic Processes

BOX · Bayesian Inference in Statistical Analysis

BOX and DRAPER · Empirical Model-Building and Response Surfaces

BOX and DRAPER · Evolutionary Operation: A Statistical Method for Process Improvement

BUCKLEW · Large Deviation Techniques in Decision, Simulation, and Estimation

BUNKE and BUNKE · Nonlinear Regression, Functional Relations and Robust Methods: Statistical Methods of Model Building

CHATTERJEE and HADI · Sensitivity Analysis in Linear Regression

CHOW and LIU · Design and Analysis of Clinical Trials

CLARKE and DISNEY · Probability and Random Processes: A First Course with Applications, Second Edition

*COCHRAN and COX · Experimental Designs, Second Edition

CONOVER · Practical Nonparametric Statistics, Second Edition

CORNELL · Experiments with Mixtures, Designs, Models, and the Analysis of Mixture Data, Second Edition

*COX · Planning of Experiments

CRESSIE · Statistics for Spatial Data, Revised Edition

DANIEL · Applications of Statistics to Industrial Experimentation

DANIEL · Biostatistics: A Foundation for Analysis in the Health Sciences, Sixth Edition

DAVID · Order Statistics, Second Edition

*DEGROOT, FIENBERG, and KADANE · Statistics and the Law

DODGE · Alternative Methods of Regression

DOWDY and WEARDEN · Statistics for Research, Second Edition

DUNN and CLARK · Applied Statistics: Analysis of Variance and Regression, Second Edition

ELANDT-JOHNSON and JOHNSON · Survival Models and Data Analysis

EVANS, PEACOCK, and HASTINGS · Statistical Distributions, Second Edition

FLEISS · The Design and Analysis of Clinical Experiments

FLEISS · Statistical Methods for Rates and Proportions, Second Edition

FLEMING and HARRINGTON · Counting Processes and Survival Analysis

GALLANT · Nonlinear Statistical Models

GLASSERMAN and YAO · Monotone Structure in Discrete-Event Systems

GNANADESIKAN · Methods for Statistical Data Analysis of Multivariate Observations, Second Edition

GOLDSTEIN and LEWIS · Assessment: Problems, Development, and Statistical Issues

GREENWOOD and NIKULIN · A Guide to Chi-Squared Testing

*HAHN · Statistical Models in Engineering

HAHN and MEEKER · Statistical Intervals: A Guide for Practitioners

HAND · Construction and Assessment of Classification Rules

HAND · Discrimination and Classification

HEIBERGER · Computation for the Analysis of Designed Experiments

HINKELMAN and KEMPTHORNE: · Design and Analysis of Experiments, Volume 1: Introduction to Experimental Design

HOAGLIN, MOSTELLER, and TUKEY · Exploratory Approach to Analysis of Variance

HOAGLIN, MOSTELLER, and TUKEY · Exploring Data Tables, Trends and Shapes

HOAGLIN, MOSTELLER, and TUKEY · Understanding Robust and Exploratory Data Analysis

HOCHBERG and TAMHANE · Multiple Comparison Procedures

HOCKING · Methods and Applications of Linear Models: Regression and the Analysis of Variables

HOGG and KLUGMAN · Loss Distributions

HOLLANDER and WOLFE · Nonparametric Statistical Methods

HOSMER and LEMESHOW · Applied Logistic Regression

HØYLAND and RAUSAND · System Reliability Theory: Models and Statistical Methods

HUBERTY · Applied Discriminant Analysis

JACKSON · A User’s Guide to Principle Components

JOHN · Statistical Methods in Engineering and Quality Assurance

JOHNSON · Multivariate Statistical Simulation

JOHNSON and KOTZ · Distributions in Statistics Continuous Multivariate Distributions

JOHNSON, KOTZ, and BALAKRISHNAN · Continuous Univariate Distributions, Volume 1, Second Edition

JOHNSON, KOTZ, and BALAKRISHNAN · Continuous Univariate Distributions, Volume 2, Second Edition

JOHNSON, KOTZ, and BALAKRISHNAN · Discrete Multivariate Distributions

JOHNSON, KOTZ, and KEMP · Univariate Discrete Distributions, Second Edition

JUREKOVÁ and SEN · Robust Statistical Procedures: Aymptotics and Interrelations

KADANE · Bayesian Methods and Ethics in a Clinical Trial Design

KADANE AND SCHUM · A Probabilistic Analysis of the Sacco and Vanzetti Evidence

KALBFLEISCH and PRENTICE · The Statistical Analysis of Failure Time Data

KELLY · Reversability and Stochastic Networks

KHURI, MATHEW, and SINHA · Statistical Tests for Mixed Linear Models

KLUGMAN, PANJER, and WILLMOT · Loss Models: From Data to Decisions

KLUGMAN, PANJER, and WILLMOT · Solutions Manual to Accompany Loss Models: From Data to Decisions

KOVALENKO, KUZNETZOV, and PEGG · Mathematical Theory of Reliability of Time-Dependent Systems with Practical Applications

LAD · Operational Subjective Statistical Methods: A Mathematical, Philosophical, and Historical Introduction

LANGE, RYAN, BILLARD, BRILLINGER, CONQUEST, and GREENHOUSE · Case Studies in Biometry

LAWLESS · Statistical Models and Methods for Lifetime Data

LEE · Statistical Methods for Survival Data Analysis, Second Edition

LEPAGE and BILLARD · Exploring the Limits of Bootstrap

LINHART and ZUCCHINI · Model Selection

LITTLE and RUBIN · Statistical Analysis with Missing Data

MAGNUS and NEUDECKER · Matrix Differential Calculus with Applications in Statistics and Econometrics

MALLER and ZHOU · Survival Analysis with Long Term Survivors

MANN, SCHAFER, and SINGPURWALLA · Methods for Statistical Analysis of Reliability and Life Data

McLACHLAN and KRISHNAN · The EM Algorithm and Extensions

McLACHLAN · Discriminant Analysis and Statistical Pattern Recognition

McNEIL · Epidemiological Research Methods

MILLER · Survival Analysis

MONTGOMERY and PECK · Introduction to Linear Regression Analysis, Second Edition

MYERS and MONTGOMERY · Response Surface Methodology: Process and Product in Optimization Using Designed Experiments

NELSON · Accelerated Testing, Statistical Models, Test Plans, and Data Analyses

NELSON · Applied Life Data Analysis

OCHI · Applied Probability and Stochastic Processes in Engineering and Physical Sciences

OKABE, BOOTS, and SUGIHARA · Spatial Tesselations: Concepts and Applications of Voronoi Diagrams

PANKRATZ · Forecasting with Dynamic Regression Models

PANKRATZ · Forecasting with Univariate Box-Jenkins Models: Concepts and Cases

PIANTADOSI · Clinical Trials: A Methodologic Perspective

PORT · Theoretical Probability for Applications

PUTERMAN · Markov Decision Processes: Discrete Stochastic Dynamic Programming

RACHEV · Probability Metrics and the Stability of Stochastic Models

RÉNYI · A Diary on Information Theory

RIPLEY · Spatial Statistics

RIPLEY · Stochastic Simulation

ROUSSEEUW and LEROY · Robust Regression and Outlier Detection

RUBIN · Multiple Imputation for Nonresponse in Surveys

RUBINSTEIN · Simulation and the Monte Carlo Method

RUBINSTEIN and MELAMED · Modern Simulation and Modeling

RYAN · Statistical Methods for Quality Improvement

SCHUSS · Theory and Applications of Stochastic Differential Equations

SCOTT · Multivariate Density Estimation: Theory, Practice, and Visualization

*SEARLE · Linear Models

SEARLE · Linear Models for Unbalanced Data

SEARLE, CASELLA, and McCULLOCH · Variance Components

STOYAN, KENDALL, and MECKE · Stochastic Geometry and Its Applications, Second Edition

STOYAN and STOYAN · Fractals, Random Shapes and Point Fields: Methods of Geometrical Statistics

THOMPSON · Empirical Model Building

THOMPSON · Sampling

TIJMS · Stochastic Modeling and Analysis: A Computational Approach

TIJMS · Stochastic Models: An Algorithmic Approach

TITTERINGTON, SMITH, and MAKOV · Statistical Analysis of Finite Mixture Distributions

UPTON and FINGLETON · Spatial Data Analysis by Example, Volume 1: Point Pattern and Quantitative Data

UPTON and FINGLETON · Spatial Data Analysis by Example, Volume II: Categorical and Directional Data

VAN RIJCKEVORSEL and DE LEEUW · Component and Correspondence Analysis

WEISBERG · Applied Linear Regression, Second Edition

WESTFALL and YOUNG · Resampling-Based Multiple Testing: Examples and Methods for p-Value Adjustment

WHITTLE · Systems in Stochastic Equilibrium

WOODING · Planning Pharmaceutical Clinical Trials: Basic Statistical Principles

WOOLSON · Statistical Methods for the Analysis of Biomedical Data

*ZELLNER · An Introduction to Bayesian Inference in Econometrics

Texts and References Section

AGRESTI · An Introduction to Categorical Data Analysis

ANDERSON · An Introduction to Multivariate Statistical Analysis, Second Edition

ANDERSON and LOYNES · The Teaching of Practical Statistics

ARMITAGE and COLTON · Encyclopedia of Biostatistics: Volumes 1 to 6 with Index

BARTOSZYNSKI and NIEWIADOMSKA-BUGAJ · Probability and Statistical Inference

BERRY, CHALONER, and GEWEKE · Bayesian Analysis in Statistics and Econometrics: Essays in Honor of Arnold Zellner

BHATTACHARYA and JOHNSON · Statistical Concepts and Methods

BILLINGSLEY · Probability and Measure, Second Edition

BOX · R. A. Fisher, the Life of a Scientist

BOX, HUNTER, and HUNTER · Statistics for Experimenters: An Introduction to Design, Data Analysis, and Model Building

BOX and LUCEÑO · Statistical Control by Monitoring and Feedback Adjustment

BROWN and HOLLANDER · Statistics: A Biomedical Introduction

CHATTERJEE and PRICE · Regression Analysis by Example, Second Edition

COOK and WEISBERG · An Introduction to Regression Graphics

COX · A Handbook of Introductory Statistical Methods

DILLON and GOLDSTEIN · Multivariate Analysis: Methods and Applications

DODGE and ROMIG · Sampling Inspection Tables, Second Edition

DRAPER and SMITH · Applied Regression Analysis, Third Edition

DUDEWICZ and MISHRA · Modern Mathematical Statistics

DUNN · Basic Statistics: A Primer for the Biomedical Sciences, Second Edition

FISHER and VAN BELLE · Biostatistics: A Methodology for the Health Sciences

FREEMAN and SMITH · Aspects of Uncertainty: A Tribute to D. V. Lindley

GROSS and HARRIS · Fundamentals of Queueing Theory, Third Edition

HALD · A History of Probability and Statistics and their Applications Before 1750

HALD · A History of Mathematical Statistics from 1750 to 1930

HELLER · MACSYMA for Statisticians

HOEL · Introduction to Mathematical Statistics, Fifth Edition

JOHNSON and BALAKRISHNAN · Advances in the Theory and Practice of Statistics: A Volume in Honor of Samuel Kotz

JOHNSON and KOTZ (editors) · Leading Personalities in Statistical Sciences: From the Seventeenth Century to the Present

JUDGE, GRIFFITHS, HILL, LÜTKEPOHL, and LEE · The Theory and Practice of Econometrics, Second Edition

KHURI · Advanced Calculus with Applications in Statistics

KOTZ and JOHNSON (editors) · Encyclopedia of Statistical Sciences: Volumes 1 to 9 wtih Index

KOTZ and JOHNSON (editors) · Encyclopedia of Statistical Sciences: Supplement Volume

KOTZ, REED, and BANKS (editors) · Encyclopedia of Statistical Sciences: Update Volume 1

KOTZ, REED, and BANKS (editors) · Encyclopedia of Statistical Sciences: Update Volume 2

LAMPERTI · Probability: A Survey of the Mathematical Theory, Second Edition

LARSON · Introduction to Probability Theory and Statistical Inference, Third Edition

LE · Applied Survival Analysis

MALLOWS · Design, Data, and Analysis by Some Friends of Cuthbert Daniel

MARDIA · The Art of Statistical Science: A Tribute to G. S. Watson

MASON, GUNST, and HESS · Statistical Design and Analysis of Experiments with Applications to Engineering and Science

MURRAY · X-STAT 2.0 Statistical Experimentation, Design Data Analysis, and Nonlinear Optimization

PURI, VILAPLANA, and WERTZ · New Perspectives in Theoretical and Applied Statistics

RENCHER · Methods of Multivariate Analysis

RENCHER · Multivariate Statistical Inference with Applications

ROSS · Introduction to Probability and Statistics for Engineers and Scientists

ROHATGI · An Introduction to Probability Theory and Mathematical Statistics

RYAN · Modern Regression Methods

SCHOTT · Matrix Analysis for Statistics

SEARLE · Matrix Algebra Useful for Statistics

STYAN · The Collected Papers of T. W. Anderson: 1943–1985

TIERNEY · LISP-STAT: An Object-Oriented Environment for Statistical Computing and Dynamic Graphics

WONNACOTT and WONNACOTT · Econometrics, Second Edition

WILEY SERIES IN PROBABILITY AND STATISTICS

ESTABLISHED BY WALTER A. SHEWHART AND SAMUEL S. WILKS

EditorsRobert M. Groves, Graham Kalton, J. N. K. Rao, Norbert Schwarz, Christopher Skinner

Survey Methodology Section

BIEMER, GROVES, LYBERG, MATHIOWETZ, and SUDMAN · Measurement Errors in Surveys

COCHRAN · Sampling Techniques, Third Edition

COX, BINDER, CHINNAPPA, CHRISTIANSON, COLLEDGE, and KOTT (editors) · Business Survey Methods

*DEMING · Sample Design in Business Research

DILLMAN · Mail and Telephone Surveys: The Total Design Method

GROVES and COUPER · Nonresponse in Household Interview Surveys

GROVES · Survey Errors and Survey Costs

GROVES, BIEMER, LYBERG, MASSEY, NICHOLLS, and WAKSBERG · Telephone Survey Methodology

*HANSEN, HURWITZ, and MADOW · Sample Survey Methods and Theory, Volume 1: Methods and Applications

*HANSEN, HURWITZ, and MADOW · Sample Survey Methods and Theory, Volume II: Theory

KASPRZYK, DUNCAN, KALTON, and SINGH · Panel Surveys

KISH · Statistical Design for Research

*KISH · Survey Sampling

LESSLER and KALSBEEK · Nonsampling Error in Surveys

LEVY and LEMESHOW · Sampling of Populations: Methods and Applications

LYBERG, BIEMER, COLLINS, de LEEUW, DIPPO, SCHWARZ, TREWIN (editors) · Survey Measurement and Process Quality

SKINNER, HOLT, and SMITH · Analysis of Complex Surveys

*Now available in a lower priced paperback edition in the Wiley Classics Library.

Copyright © 1998 by John Wiley & Sons, Inc.

All rights reserved. Published simultaneously in Canada.

No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923, [phone], fax [phone]. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 605 Third Avenue, New York, NY 10158-0012, [phone], fax [phone], E-Mail: [email protected].

Library of Congress Cataloging in Publication Data:

Groves, Robert M.Nonresponse in household interview surveys / Robert M. Groves, Mick P. Couper.p. cm. — (Wiley series in probability and statistics. Survey methodology section)“Wiley-Interscience publication.”Includes bibliographical references and index.ISBN 0-471-18245-1 (cloth : alk. paper)1. Household surveys. I. Couper, Mick. II. Title. III. Series.HB849.49.G757 199897-39223CIP

Preface

This book was written out of frustration. Its genesis came in 1986–1988 when a review of the then extant research literature of survey nonrepsonse yielded few answers to the question, “How important is nonresponse to surveys?”

In teaching courses in survey methodology, it was common for us to emphasize that once a probability sample had been drawn, full measurement of the sample was crucial for proper inference to apply. Bright students would sometimes question, “How do we know when nonresponse implies error and when it doesn’t? Is it cheaper and more effective to reduce nonresponse error by decreasing nonresponse rates or by adjusting for it post hoc? Is it more important to reduce nonresponse due to noncontact or nonresponse due to refusals? Why, after all, do people choose not to cooperate with survey requests?” We felt unprepared for such questions and, indeed, grew to believe that the lack of answers was a pervasive weakness in the field, not just a result of our ignorance.

Gathering information post hoc about nonrespondents from diverse surveys, which formed one of the central databases of this book, was an attempt to address a critical weakness in the area—the lack of common information about nonresponse across several surveys. (This was an idea stolen from Kemsley, who in 1971, mounted such a study in Great Britain.) Around 1988, the major U.S. federal household surveys were beginning to consider redesign efforts related to incorporating new population distribution data from the 1990 decennial census. We approached Maria Gonzalez, of the Statistical Policy Office of the Office of Management and Budget, who was leading an interagency group developing those redesign research plans. Our idea was to draw samples of nonrespondents and respondents from household surveys conducted about the time of the decennial census, and match their records to the decennial census data. We would thus have at our disposal all variables on the census form to describe the nonrespondents.

This was an idea whose time had clearly not come to the interagency group. We received tough criticism on what practical lessons would be learned; how would those surveys be improved because of the work, and so on. Maria should be credited with quietly listening to the criticism, but forcefully arguing the merits of our case to the survey sponsors. We dedicate this book to her memory.

What appear to the reader as 11 chapters of theory and analysis are based on many person-years of effort, during which the perspectives on the conceptual foundations of survey participation evolved. Some history of the project may provide a sense of that process.

We sought to develop a diverse set of surveys to match to the decennial records. Ideally, we wanted to represent all major survey design variations among the matched surveys. However, the match tool was to be a unit’s address, so we were limited to area frame surveys, most often conducted in face-to-face mode. We failed to get cooperation from the commercial surveys we approached. We failed to get extra funds to add some academic surveys to the set.

In the end we established a consortium of funders including the Bureau of the Census, Bureau of Justice Statistics (BJS), Bureau of Labor Statistics (BLS), National Center for Health Statistics (NCHS) and the National Institute on Drug Abuse [NIDA, later called the Substance Abuse and Mental Health Services Administration (SAMSHA)]. Research Triangle Institute and the National Opinion Research Center also provided documentation on surveys sponsored by NIDA and Census, respectively, to facilitate the match and administered questionnaires to their interviewers. At each agency there were key contact people, who facilitated our work. These were William Nicholls, Robert Tortora, and Jay Waite (Census Bureau), Cathryn Dippo and Clyde Tucker (BLS), Michael Rand (BJS), Steve Botman (NCHS), and Joseph Gfroerer (SAMHSA).

Completely independent of this research program, in early 1990, Groves took on a temporary post as an Associate Director at the Bureau of the Census, as the project was nearing its implementation. Couper simultaneously took a post as visiting researcher at Census. This permitted Couper to focus full time on the project between 1990 and 1994.

In 1989, samples were drawn from the Census Bureau surveys, mostly by staff in the Statistical Methods Division of the Bureau, under the direction of Jay Waite. John Paletta coordinated the selection of match cases. After the Census, in 1991–1992, Couper began a commuting life between Washington, DC, and Jeffersonville, Indiana, the vast processing complex for the U.S. Census Bureau. There he worked with a team headed by Judith Petty. The leader of the match team, Maria Darr, collaborated in defining the match methods, training and supervising staff, and implementing quality control procedures. Couper directed the match effort, living out of a suitcase, eating too many meals at the Waffle House in Jeffersonville (whose broken sign read “affle House”). Matching survey and census records was a tedious, slow process, but the care and professionalism of the Jeffersonville staff produced a match data set that we believe is as complete and accurate as possible. Acquiring the completed survey data, cleaning data, merging files, determining weighting schemes, variance estimators, and appropriate modeling techniques took some time after the completion of the match in 1993.

We are both indebted to the executive staff of the Census Bureau, which provided a research cocoon at Suitland, permitting Couper to focus entirely on the research activities at crucial times during the match process, and Groves to join him after ending his stint as associate director in 1992.

However, the work of the decennial match project was not our only focus during the years 1988–1992. Even while the match project was being discussed, two other lines of research were developing. The first was a refinement of conceptual thinking on the process of survey participation. This was partially funded by the Census Bureau and was a collaborative effort with Robert Cialdini, a social psychologist who has made important contributions to understanding helping behavior and compliance. We collaborated in a series of focus groups with interviewers from different organizations, seeking insights from their expertise in gaining the cooperation of persons in surveys. This led to a basic framework of influences on survey participation that forms the structure of this book. Cialdini provided important insights about how survey participation decisions might resemble to other decisions about requests and, more broadly, to attitude change. We are in his debt, especially for the insight one Saturday morning that most decision making in the survey context must be heuristically based, ill-informed by the central features of the respondent’s job in a survey.

When our interest grew concerning the effect of the social environment of survey participation, we joined with Lars Lyberg, our friend and colleague, to organize a set of international workshops on household survey nonresponse, starting in 1990. These gave researchers in different countries a chance to compare notes on survey participation across societies. The workshops have stimulated the replication of nonresponse research across countries. Our own research has benefitted from such replication. We have also learned much from the interactions and enjoyed the camaraderie. We thank the regulars at the meetings, including Lars, Bab Barnes, Sandy Braver, Pam Campanelli, Cathy Dippo, Wim de Heer, Lilli Japec, Seppo Laaksonen, Clyde Tucker, and many others.

The other line of research that arose in 1990 involved chances to test empirically our ideas with new data collection efforts. Through the good graces of our colleague Ron Kessler, we smuggled into the National Comorbidity Survey a set of interviewer observations that permitted key initial tests of our notions of the influence of contact-level interactions. This survey was supported by the National Institutes on Mental Health (Grants MH46376 and MH 49098). Later we recieved support from the National Institute on Aging (Grant RO1 AG31059) to add similar measures to the AHEAD survey, which permitted tests of the ideas on a survey of the elderly. Bill Rodgers and Tom Juster were very supportive of including these in AHEAD. Both of these grants were important to Chapters 8, 9, and Chapter 11 of this text.

After the match project data were available, Trivellore Raghunathan became a collaborator when he joined the Survey Methodology Program at Michigan. He collaborated in translating our ideas and findings into a statisitcal modeling strategy for postsurvey adjustment. Raghu deserves full credit for the two-stage adjustment procedures in Chapter 11.

Audience. We’ve written the book for students of survey methodology: those in school, practicing in the field, and teaching the subject. We assume basic knowledge of survey design, at a level comparable to most initial undergraduate survey methods courses. The statistical models are kept simple deliberately

Those readers with limited time should read Chapters 1 and 2 in order to understand the conceptual framework. Then they should read the summary sections of each chapter, as well as Chapter 11.

Those readers most interested in the practical implications of the work should read the last sections of Chapters 4–9, labeled “Practical Implications for Survey Implementation” as well as Chapters 10 and 11.

In using the book as a text in a course on survey nonresponse we have used Chapters 2 and 4–10.

Collaborators. In addition to those mentioned above, other stimulating colleagues helped shape the research. These include Toni Tremblay and Larry Altmayer at the U.S. Census Bureau, and Joe Parsons, Ashley Bowers, Nancy Clusen, Jeremy Morton, and Steve Hanway at the Joint Program in Survey Methodology. Lorraine McCall was responsible for the interviewer surveys at the Census Bureau. Teresa Parsley Edwards and Rachel Caspar at Research Triangle Institute worked with us on parts of the analysis of the National Household Survey on Drug Abuse. Brian Harris-Kojetin, John Eltinge, Dan Rope, and Clyde Tucker examined various features of nonreponse in the Current Population Survey. Judith Clemens, Darby Miller-Steiger, Stacey Erth, and Sue Ellen Hansen provided assistance at various points during the work, especially on the Michigan Survey Research Center surveys. We appreciate the criticisms of a set of students in a summer course on survey nonresponse in 1994 offered through the SRC Summer Institute in Survey Research Techniques. Finally, the administrative staff of the Joint Program in Survey Methodology, including Jane Rice, Pam Ainsworth, Nichole Ra’uf, Christie Nader, and Heather Campbell, provided help at many crucial points.

We are members of the Survey Methodology Program (SMP) at the University of Michigan’s Institute for Social Research, a research environment that stimulates theoretical questions stemming from applied problems. We thank our SMP colleagues for helping us think through much of the material we present in this book. Jim House, as director of the Survey Research Center, has been a consistent supporter of bringing science to survey methodology and we thank him for being there.

We have profited from critical reviews by Paul Biemer, John Eltinge, Robert Fay, Brian Harris-Kojetin, Lars Lyberg, Nancy Mathiowetz, Beth-Ellen Pennell, Stanley Presser, Eleanor Singer, Seymour Sudman, Roger Tourangeau, and Clyde Tucker. Errors remaining are our responsibility.

We are especially indebted to Northwest Airlines, whose many delayed and cancelled flights between Detroit Metro and Washington National airports permitted long and uninterrupted discussions of the research.

Finally, we thank our editor at Wiley, Steve Quigley, for making the publication process as trouble-free as possible.

ROBERT M. GROVESMICK P. COUPER

Ann Arbor, Michigan

College Park, Maryland

Acknowledgments

We are grateful to various copyright holders for permission to reprint or present adaptations of material previously published. These include the University of Chicago Press, on behalf of the American Association for Public Opinion Research, for adaptation of material from Groves, Cialdini, and Couper (1992) and Couper (1992), appearing in Chapters 2 and 10, and for reprinting a table from Dillman, Gallegos, and Frey (1976), as Table 10.1; the Minister of Industry of Canada, through Statistics Canada for adaptation of Couper and Groves (1992) in Chapter 7; Statistics Sweden, for adaptation of Groves, R.M., and Couper, M.P. (1995) “Theoretical Motivation for Post-Survey Nonresponse Adjustment in Household Surveys,” 11, 1, 93–106, in Chapter 9; and “Contact-Level Influences on Cooperation in Face-to-Face Surveys,” 12, 1, 63–83, in Chapter 8; Kluwer Academic Publishers, for adaptations of Couper, M.P., and Groves, R.M. (1996) “Social Environmental Impacts on Survey Cooperation,” 30, 173–188, in Chapter 6; and Jossey-Bass Publishers for adaptation of Couper and Groves (1996) in Chapter 5.

CHAPTER ONE

An Introduction to Survey Participation

1.1 INTRODUCTION

This is a book about error properties of statistics computed from sample surveys. It is also a book about why people behave the way they do.

When people are asked to participate in sample surveys, they are generally free to accept or reject that request. In this book we try to understand the several influences on their decision. What influence is exerted by the attributes of survey design, the interviewer’s behavior, the prior experiences of the person faced with the request, the interaction between interviewer and householder, and the social environment in which the request is made? In the sense that all the social sciences attempt to understand human thought and behavior, this is a social science question. The interest in this rather narrowly restricted human behavior, however, has its roots in the effect these behaviors have on the precision and accuracy of statistics calculated on the respondent pool resulting in the survey. It is largely because these behaviors affect the quality of sample survey statistics that we study the phenomenon.

This first chapter sets the stage for this study of survey participation and survey nonresponse. It reviews the statistical properties of survey estimates subject to nonresponse, in order to describe the motivation for our study, then introduces key concepts and perspectives on the human behavior that underlies the participation phenomenon. In addition, it introduces the argument that will be made throughout the book—that attempts to increase the rate of participation and attempts to construct statistical adjustment techniques to reduce nonresponse error in survey estimates achieve their best effects when based on sound theories of human behavior.

1.2 STATISTICAL IMPACTS OF NONRESPONSE ON SURVEY ESTIMATES

Sample surveys are often designed to draw inferences about finite populations, by measuring a subset of the population. The classical inferential capabilities of the survey rest on probability sampling from a frame covering all members of the population. A probability sample assigns known, nonzero chances of selection to every member of the population. Typically, large amounts of data from each member of the population are collected in the survey. From these variables, hundreds or thousands of different statistics might be computed, each of which is of interest to the researcher only if it describes well the corresponding population attribute. Some of these statistics describe the population from which the sample was drawn; others stem from using the data to test causal hypotheses about processes measured by the survey variables (e.g., how education and work experience in earlier years affect salary levels).

One example statistic is the sample mean, an estimator of the population mean. This is best described by using some statistical notation, in order to be exact in our meaning. Let one question in the survey be called “Y,” and the answer to that question for a sample member, say the ith member of the population, be designated by Yi. Then we can describe the population mean by

where N is the number of units in the target population. The estimator of the population mean is often

where r is the number of respondents in the sample and wi is the reciprocal of the probability of selection of the ith respondent. (For readers accustomed to equal probability samples, as in a simple random sample, the wi is the same for all cases in the sample and the computation above is equivalent to Σyi/n.)

For each possible sample we could draw, given the sample design, we could express a difference between the full sample mean, n, and the respondent mean, in the following way:

which, with a little manipulation becomes

that is,

This shows that the deviation of the respondent mean from the full sample mean is a function of the nonresponse rate (m/n) and the difference between the respondent and nonrespondent means.

Under this simple expression, what is the expected value of the respondent mean, over all samples that could be drawn given the same sample design? The answer to this question determines the nature of the bias in the respondent mean, where “bias” is taken to mean the difference between the expected value (over all possible samples given a specific design) of a statistic and the statistic computed on the target population. That is, in cases of equal probability samples of fixed size the bias of the respondent mean is approximately

or

where the capital letters denote the population equivalents to the sample values. This shows that the larger the stratum of nonrespondents, the higher the bias of the respondent mean, other things being equal. Similarly, the more distinctive the non-respondents are from the respondents, the larger the bias of the respondent mean.

These two quantities, the nonresponse rate and the differences between respondents and nonrespondents on the variables of interest, are key to the studies reported in this book. Because the literature on survey nonresponse does not directly reflect this fact (an important exception is the work of Lessler and Kalsbeek, 1992), it is important for the reader to understand how this affects nonresponse errors.

Figure 1.1 shows four alternative frequency distributions for respondents and nonrespondents on a hypothetical variable, y, measured on all cases in some target population. The area under the curves is proportional to the size of the two groups, respondents and nonrespondents.

Figure 1.1. Hypothetical frequency distributions of respondents and nonrespondents. (a) High response rate, nonrespondents similar to respondents. (b) Low response rate, nonrespondents similar to respondents.

Figure 1.1. (c) High response rate, nonrespondents different from respondents. (d) Low response rate, nonrespondents different from respondents

To provide another concrete illustration of these situations, assume that the statistic of interest is a proportion, say, the number of adults who intend to save some of their income in the coming month. Figure 1.2 illustrates the level of nonresponse bias possible under various circumstances. In all cases, the survey results in a respondent mean of 0.50; that is, we are led to believe that half of the adults plan to save in the coming month. The x-axis of the figure displays the proportion of nonrespondents who plan to save in the coming month. (This attribute of the sample is not observed.) The figure is designed to illustrate cases in which the nonrespondent proportion is less or equal to the respondent proportion. Thus, the nonrespondent proportions range from 0.50 (the no bias case) to 0.0 (the largest bias case). There are three lines in the figure, corresponding to different nonresponse rates: 5%, 30%, and 50%.

Figure 1.2. Nonresponse bias for a proportion, given a respondent mean of 0.50, various response rates, and various nonresponse means.

The figure gives a sense of how large a nonresponse bias can be for different nonresponse rates. For example, in a survey with a low nonresponse rate, 5%, the highest bias possible is 0.025. That is, if the survey respondent mean is 0.50, then one is assured that the full sample mean lies between 0.475 and 0.525.

In the worst case appearing in Figure 1.2, a survey with a nonresponse rate of 50%, the nonresponse bias can be as large as 0.25. That is, if the respondent mean is 0.50, then the full sample mean lies between 0.25 and 0.75. This is such a large range that it offers very little information about the statistic of interest.

The most important feature of Figure 1.2 is its illustration of the dependence of the nonresponse bias on both response rates and the difference term. The much larger slope of the line describing the nonresponse bias for the survey with a high nonresponse rate shows that high nonresponse rates increase the likelihood of bias even with relatively small differences between respondents and nonrespondents on the survey statistic.

1.2.1 Nonresponse Error on Different Types of Statistics

The discussion above focused on the effect of nonresponse on estimates of the population mean, using the sample mean. This section briefly reviews effects of nonresponse on other popular statistics. We examine the case of an estimate of a population total, the difference of two subclass means, and a regression coefficient.

The Population Total. Estimating the total number of some entity is common in government surveys. For example, most countries use surveys to estimate the total number of unemployed persons, the total number of new jobs created in a month, the total retail sales, the total number of criminal victimizations, etc. Using notation similar to that in Section 1.2, the population total is ΣYi, which is estimated by a simple expansion estimator, Σwiyi, or by a ratio-expansion estimator, X(Σwiyi/Σwixi), where X is some auxiliary variable, correlated with Y, for which target population totals are known. For example, if y were a measure of the number of criminal victimizations experienced by a sample household, and x were a count of households, X would be a count of the total number of households in the country.

For variables that have nonnegative values (such as count variables), simple expansion estimators of totals based only on respondents always underestimate the total. This is because the full sample estimator is

that is,

Hence, the bias in the respondent-based estimator is

It is easy to see, thereby, that the respondent-based total (for variables that have nonnegative values) will always underestimate the full sample total, and thus, in expectation, the full population total.

The Difference of Two Subclass Means. Many statistics of interest from sample surveys estimate the difference between the means of two subpopulations. For example, the Current Population Survey often estimates the difference in the unemployment rate for Black and nonBlack men. The National Health Interview Survey estimates the difference in the mean number of doctor visits in the last 12 months between males and females.

Using the expressions above, and using subscripts 1 and 2 for the two subclasses, we can describe the two respondent means as

These expressions show that each respondent subclass mean is subject to an error that is a function of a nonresponse rate for the subclass and a deviation between respondents and nonrespondents in the subclass. The reader should note that the nonresponse rates for individual subclasses could be higher or lower than the nonresponse rates for the total sample. For example, it is common that nonresponse rates in large urban areas are higher than nonresponse rates in rural areas. If these were the two subclasses, the two nonresponse rates would be quite different.

If we were interested in 1 − 2 as a statistic of interest, the bias in the difference of the two means would be approximately

Many survey analysts are hopeful that the two terms in the bias expression above cancel. That is, the bias in the two subclass means is equal. If one were dealing with two subclasses with equal nonresponse rates that hope is equivalent to a hope that the difference terms are equal to one another. This hope is based on an assumption that nonrespondents will differ from respondents in the same way for both subclasses. That is, if nonrespondents tend to be unemployed versus respondents, on average, this will be true for all subclasses in the sample.

Figure 1.3. Nonresponse bias for a difference of subclass means, for the case of two respondent subclass means (0.5, 0.3) by various response rate combinations, by differences between respondent and nonrespondent means.

The figure shows that when the two nonresponse rates are equal to one another, there is no bias in the difference of the two subclass means. However, when the response rates of the two subclasses are different, large biases can result. Larger biases in the difference of subclass means arise with larger differences in nonresponse rates in the two subclasses (note the higher absolute value of the bias for any given [r − m] value for the case with a 0.05 nonresponse rate in subclass 1 and a 0.5 in subclass 2 than for the other cases).

A Regression Coefficient. Many survey data sets are used by analysts to estimate a wide variety of statistics measuring the relationship between two variables. Linear models testing causal assertions are often estimated on survey data. Imagine, for example, that the analysts were interested in the model

which, using the respondent cases to the survey, would be estimated by

The ordinary least squares estimator of βr1 is

Both the numerator and denominator of this expression are subject to potential nonresponse bias. For example, the bias in the covariance term in the numerator is approximately

This bias expression can be either positive or negative in value. The first term in the expression has a form similar to that of the bias of the respondent mean. It reflects a difference in covariances for the respondents (Srxy) and nonrespondents (Smxy). It is large in absolute value when the nonresponse rate is large. If the two variables are more strongly related in the respondent set than in the nonrespondent, the term has a positive value (that is the regression coefficient tends to be overestimated). The second term has no analogue in the case of the sample mean; it is a function of cross-products of difference terms. It can be either positive or negative depending on these deviations.

As Figure 1.4 illustrates, if the nonrespondent units have distinctive combinations of values on the x and y variables in the estimated equation, then the slope of the regression line can be misestimated. The figure illustrates the case when the pattern of nonrespondent cases (designated by “”) differ from that of respondent cases (designated by “”). The result is that the fitted line on the respondents only has a larger slope than that for the full sample. In this case, the analyst would normally find more support for an hypothesized relationship than would be true for the full sample.

Figure 1.4. Illustration of the effect of unit nonresponse on estimated slope of regression line.

1.2.2 Considering Survey Participation a Stochastic Phenomenon

The discussion above made the assumption that each person (or household) in a target population either is a respondent or a nonrespondent for all possible surveys. That is, it assumes a fixed property for each sample unit regarding the survey request. They will always be a nonrespondent or they will always be a respondent, in all realizations of the survey design.

An alternative view of nonresponse asserts that every sample unit has a probability of being a respondent and a probability of being a nonrespondent. It takes the perspective that each sample survey is but one realization of a survey design. In this case, the survey design contains all the specifications of the research data collection. The design includes the definition of the sampling frame, the sample design, the questionnaire design, choice of mode, hiring, selection, and training regimen for interviewers, data collection period, protocol for contacting sample units, callback rules, refusal conversion rules, and so on. Conditional on all these fixed properties of the sample survey, sample units can make different decisions regarding their participation.

In this view, the notion of a nonresponse rate must be altered. Instead of the nonresponse rate merely being a manifestation of how many nonrespondents were sampled from the sampling frame, we must acknowledge that in each realization of a survey different individuals will be respondents and nonrespondents. In this perspective the nonresponse rate above (m/n) is the result of a set of Bernoulli trials; each sample unit is subject to a “coin flip” to determine whether it is a respondent or nonrespondent on a particular trial. The coins of various sample units may be weighted differently; some will have higher probabilities of participation than others. However, all are involved in a stochastic process of determining their participation in a particular sample survey.

The implications of this perspective on the biases of respondent means, respondent totals, respondent differences of means, and respondent regression coefficients is minor. The more important implication is on the variance properties of unadjusted and adjusted estimates based on respondents.

1.2.3 The Effects of Different Types of Nonresponse

The discussion above considered all sources of nonresponse to be equivalent to one another. However, this book attempts to dissect the process of survey participation into different components. In household surveys it is common to classify outcomes of interview attempts into the following categories: interviews (including complete and partial), refusals, noncontacts, and other noninterviews. The other noninterview category consists of those sample units in which whoever was designated as the respondent is unable to respond, for physical and mental health reasons, for language reasons, or for other reasons that are not a function of reluctance to be interviewed. Various survey design features affect the distribution of nonresponse over these categories. Surveys with very short data collection periods tend to have proportionally more noncontacted sample cases. Surveys with long data collection periods or intensive contact efforts tend to have relatively more refusal cases. Surveys with weak efforts at accommodation of nonEnglish speakers tend to have somewhat more “other noninterviews.” So, too, may surveys of special populations, such as the elderly or immigrants.

If we consider separately the different types of nonresponse, many of the expressions above generalize. For example, the respondent mean can be described as a function of various nonresponse sources, as in

where the subscripts rf, nc, and nio refer to refusals, noncontacts, and other noninterviews, respectively.

This focuses attention on whether when survey designs vary on the composition of their nonresponse (i.e., different proportions of refusals, noncontacts, and other noninterviews), they produce different levels of nonresponse error. Do persons difficult to contact have distinctive values on the survey variables from those easy to contact? Do persons with language, mental, or physical disabilities have distinctive values from others? Are the tendencies for contacted sample cases to sort themselves into either interviews or refusals related to their characteristics on the survey variables?

Consider a practical example of these issues. Imagine conducting a survey of criminal victimization, where respondents are asked to report on their prior experiences as a victim of a personal or household crime. As will be seen in later chapters, some of the physical impediments to contacting a sample household are locked gates, no-trespassing signs, and intercoms. These are also common features that households who have experienced a crime install in their unit. They are preventative measures against criminal victimization. This is a situation in which early contacts in a survey would be likely to have lower victimization rates than late contacts. At any point, the noncontacts will tend to have higher victimization rates than contacted cases.

Now consider the causes of cooperation or refusal with the survey request. Imagine that the survey is described as an effort to gain information about victimization in order to improve policing strategies in the local community. Those for whom such a purpose is highly salient will tend to cooperate. Those for whom such a goal is less relevant will tend to refuse. Thus, refusals might tend to have lower victimization rates than cooperators, among those contacted.

This situation implies that the difference terms move in different directions:

Now let’s add to the situation the typical process of field administration. Initial effort by interviewers is concentrated on contacting each sample unit. This initially reaches those with low victimization rates, who disproportionately then refuse to be interviewed. Initial refusal rates are quite high. As contact rates increase, victims, who are interested in responding, are disproportionately contacted. They disproportionately move into the interviewed pool, increasing the victimization rate among respondents. Alternatively, if efforts at higher response rates are concentrated on the initial refusal cases, through refusal conversion, the interviewed pool will increasingly contain nonvictims, lowering the respondent victimization rate.

This is a case where the final nonresponse error is a function of the balance between the noncontact and the refusal rate. For any given overall response rate, the higher the refusal rate, the more likely the survey will overestimate the population’s victimization rate. For any given overall response rate, the higher the noncontact rate, the more likely the survey will underestimate the rate.

This example illustrates the need to dissect the causes of nonresponse into constituent parts that share relationships with the key survey variables. Considering only the overall response rate ignores the possible counteracting biases of different types of nonresponse. This process of dissection is one of the purposes of this book.

1.2.4 Reducing Nonresponse Rates

There are two traditional reactions to survey nonresponse among practitioners: reducing nonresponse rates and using estimators that include adjustments for nonresponse. As we discuss in more detail in Chapter 10, various survey design features act to reduce specific sources of nonresponse.

There is a well-documented set of techniques to increase the likelihood of contacting sample cases. These include advance contacts by mail or telephone in face-to-face surveys in order to schedule convenient times to visit. They include setting the number of days or weeks in the data collection period so that those households that are rarely at home will nonetheless be contacted. In addition, interviewers are trained to call repeatedly on sample units, seeking contact with the household. As the field period progresses, calls on cases tend to be at different times of day or evening; interviewers may be trained to attempt telephone contact, etc.

There are many design features chosen to reduce refusals as a source of nonresponse. These include the use of advance letters, attempting to communicate that the survey is conducted by an organization with legitimate need for the information. The advance communication sometimes contains a cash or in-kind incentive. The interviewer attempts to make appointments with the sample person at times convenient for them to provide the interview. Repeated attempts to persuade reluctant respondents may involve switches to a different interviewer, persuasion letters, or visits by supervisors—all intended to communicate the importance of cooperating with the survey request.

Finally, the design features to reduce the rate of “other noninterviews” include the use of nonEnglish speaking interviewers, translation of the instruments into various languages, and the use of proxy respondents.

Most of these efforts to reduce nonresponse rates are aimed at different potential causes of nonresponse, not directly different characteristics of nonrespondents. They attack the rate term (m/n) in the expression, not the difference terms, [yr − ym]. This means they exert no direct control over the nonresponse error itself, but only on one term of the error expression.

Since design decisions are made under cost constraints, designs often tend to use the cheapest means possible to reduce the nonresponse rate. Usually, noncontact rates can be reduced most cheaply, merely by making more calls on cases not yet contacted. If at any one point in a field period, the current noncontacts are quite different (on the survey measures) from the current refusals, then it is possible that this strategy would not reduce nonresponse error. That is, if [r − nc] is small, but [r − rf