RCMAR Workshop: March 20, 2026

Date: March 20, 2026

Time: 10:00 AM-3:00 PM

Location: Teams

Sponsor: Texas Resource Center on Minority Aging Research

Host: UTMB Department of Biostatistics and Data Science

Cost: FREE!

This workshop focuses on methods for handling survey weights and missing data in survey data analysis. The program introduces complex sampling designs and the use of weights under various conditions, modern imputation techniques, methods for handling non-random missingness, and Bayesian approaches to missing data problems. Examples throughout the workshop will use nationally representative datasets (e.g., NHANES and HRS). Participants will gain practical skills to better understand and apply these methods in their own survey-based research.

Workshop Schedule

Session 1: Introduction to Survey Data Analysis (10:00 - 11:00 AM)

• Overview of complex survey designs (strata, clusters, weights)
• How to use sampling weights
• Common pitfalls in analyzing survey data

Instructor: Xiaoying Yu, PhD, Associate Professor

Session 2: Imputation for Missing Data (11:00 AM - 12:00 PM)

• Types of missing data (MCAR, MAR, MNAR)
• Single imputation (regression imputation, hot-deck imputation)
• Introduction to multiple imputation and key steps in practice

Instructor: Jeong Hoon Jang, PhD, Assistant Professor

Lunch break: 12:00 - 1:00 PM

Session 3: Handling Missing Not at Random (MNAR) (1:00 - 2:00 PM)

• Why MNAR is challenging in practice
• Sensitivity analysis for missing data assumptions
• Conceptual approaches to MNAR modeling (selection model, pattern-mixture model)

Instructor: Jeong Hoon Jang, PhD, Assistant Professor

Session 4: Bayesian Methods in Missing Data Analysis (2:00 - 3:00 PM)

• Bayesian framework with hierarchical modeling of missing data and model parameters
• Iterative methods for Bayesian multiple imputation and inference using the EM algorithm and Gibbs sampling
• Strategies for handling MNAR in the Bayesian framework
• Sensitivity analysis with respect to (1) prior distribution (2) missing data assumptions

Instructor: Moumita Chakraborty, PhD, Assistant Professor