Mind the Gap: An Introduction to Practical Strategies for Handling Missing Data in R

About

Hosted by the Office of Statistics for Doctoral Research, this workshop introduces doctoral students to the challenges that missing data can present in research and practical strategies for addressing them using R.

Missing data are a common and often unavoidable challenge in real-world research. Participants may skip survey questions, miss study visits, or have incomplete clinical assessments. How these missing data are handled can affect statistical power, precision, and – more importantly – the validity of study findings.

Choosing an appropriate approach requires more than simply deleting incomplete observations or filling in missing values. Researchers need to understand the extent and patterns of missingness, consider why data may be missing, and recognize the assumptions underlying different analytic strategies.

This introductory, practical workshop provides a concepts-first approach to understanding and handling missing data, with examples implemented in R. Participants will learn about missing-data mechanisms – including missing completely at random (MCAR), missing at random (MAR), and missing not at random (MNAR) – and use graphical and descriptive methods to explore missingness in their data. The workshop will examine common approaches such as complete-case analysis and simple imputation, discuss their limitations, and introduce multiple imputation as a principled approach for handling missing data under appropriate assumptions.

Emphasis will be placed on developing a practical workflow for evaluating missing data and selecting a scientifically defensible analytic strategy rather than applying a one-size-fits-all solution.

Learning Objectives

By the end of the workshop, participants will be able to:

• Describe common missing-data mechanisms — Distinguish among missing completely at random (MCAR), missing at random (MAR), and missing not at random (MNAR), and explain why these assumptions matter for analysis. • Explore and visualize missing data in R — Summarize the extent and patterns of missingness and use graphical tools to identify where missing data occur. • Compare approaches for handling missing data — Describe the assumptions, strengths, and limitations of complete-case analysis, simple imputation methods, and multiple imputation. • Apply a practical strategy for handling missing data in R — Use the study context, observed missingness patterns, and analytic objectives to select and implement an appropriate missing-data approach.