Richard Valliant

Biographical Sketch

Richard Valliant is a Research Professor Emeritus at the University of Michigan and the Joint Program for Survey 
Methodology at the University of Maryland. He received his PhD in Biostatistics from Johns Hopkins University in 1983. He has over 40 years of experience in survey sampling, estimation theory, and statistical computing. He was formerly an Associate Director at Westat and a mathematical statistician with the Bureau of Labor Statistics. He has a range of applied experience in survey estimation and sample design on a variety of establishment, institutional, and household surveys. He is also a Fellow of the American Statistical Association, an elected member of the International Statistical Institute, and has been an editor of the Journal of the American Statistical Association, the Journal of Official Statistics, and Survey Methodology.

Research Areas

His research covers a variety of areas, including inference from non-probability samples, replication variance estimation, price index estimation, calibration estimation, use of commercial lists in design of household samples, and regression diagnostics for models fitted with survey data.  He is the co-author of three books and an R package.

Practical Tools for Designing and Weighting Survey Samples (2013)
by R. Valliant, J.A. Dever, and F. Kreuter

The goal of the book Practical Tools for Designing and Weighting Survey Samples (2013) is to put an array of tools at the fingertips of practitioners by explaining approaches long used by survey statisticians. The links below take you to files of R code for the examples and projects in the book. Sadly, there are some typos and other errors in the 1st edition. You can get a list of those, too.

Practical Tools book

Finite Population Sampling and Inference: A Prediction Approach (2000)
by R. Valliant, A.H. Dorfman, and R.M. Royall

The use (explicit or implicit) of models in inferences from survey samples has been common for decades. This book, published in 2000, brings together the major theoretical results for non-Bayesian model-based theory as applied to finite populations. This book remains relevant, especially considering the ongoing concerns about deteriorating response rates in many types of surveys and the need to make inferences from what are essentially non-probability samples.


Survey Weights: A Step-by-step Guide to Calculation (2017, to be published)
by R. Valliant and J.A. Dever

This is a brief, practical guide to computing survey weights using the statistical package Stata. Many examples are given of how to code the calculations in Stata.



Education in Survey Methodology

For information on education in survey methodology and data science, you should visit the sister programs at Michigan and Maryland. Master's, PhD, and Certificate programs are offered through both universities.