Missing Data Then and Now - Online Course

Statistical Horizons is celebrating 20 years of advancing research methods—and we couldn’t have done it without you.

Since 2005, we’ve helped thousands of researchers across academia, government, and industry sharpen their skills and apply cutting-edge tools to real-world data. To mark our 20th anniversary, we’re offering a special thank you to our community: a free, 3-hour seminar presented by our Founder and President, Paul Allison.

Widely regarded as a leading authority on missing data, Dr. Allison will revisit one of the most important topics in applied research as part of our Distinguished Speaker Series. Whether you are a longtime fan or just discovering Statistical Horizons, we invite you to join us for this free event.

ABSTRACT

In 2001, Paul Allison published Missing Data as part of Sage’s acclaimed “little green book” series. Cited over 10,000 times, the book focused primarily on two emerging methods for handling missing data: maximum likelihood (ML) and multiple imputation (MI). Much has changed since then—and this seminar is designed to bring you up to date.

In the first hour, Professor Allison will revisit the state of the field as it stood in 2001. After a brief historical overview, he will introduce the core, enduring principles behind ML and MI. Particular attention will be given to the foundational assumptions of that era: missing at random and multivariate normality—what they mean and why they matter.

The second hour will provide an accessible survey of the most important developments from the past 24 years, including:

• • MICE methods for categorical and other non-normal data.
  • New insights and methods for determining the number of imputations needed.
  • Factorization techniques for handling categorical variables with ML.
  • Extensions of MI and ML to multilevel data.
  • Substantive model compatibility in MI.
  • The EMB algorithm for MI.
  • Improved handling of categorical variables under multivariate normality.

In the third hour, Professor Allison will turn to the fully Bayesian approach to missing data—a topic not covered in his original book. While Bayesian methods for missing data have long existed, recent advances in software have made them far more accessible to applied researchers. Today, fully Bayesian tools are not only easy to use but also remarkably flexible and powerful.

Even if your preference is for multiple imputation, Bayesian methods offer a nearly ideal way to generate high-quality imputations across a wide range of models.

Join us for a comprehensive update on missing data methods—what has changed, what has endured, and what you need to know now.

This Distinguished Speaker Series seminar will consist of three hours of lecture and Q&A, held live* via the free video-conferencing software Zoom.

*The video recording of the seminar will be made available to registrants within 24 hours and will be accessible for four weeks thereafter. That means that you can watch all of the class content and discussion even if you cannot participate synchronously.

Closed captioning is available for all live and recorded sessions. Captions can be translated to a variety of languages including Spanish, Korean, and Italian. For more information, click here.

For an in-depth understanding of Missing Data, consider Professor Allison’s 4-week, on-demand seminar, Missing Data Using R. You’ll learn both the theory and the practical application of two modern methods for handling missing data: maximum likelihood (ML) and multiple imputation (MI).