Theory & Background

Brief theoretical background for understanding MICE.

• Missing Data Mechanisms
  • MCAR: Missing Completely at Random
  • MAR: Missing at Random
  • MNAR: Missing Not at Random
  • Practical Implications
• Multiple Imputation Theory
  • Why Multiple Imputation?
  • The Three-Step Process
  • What is MICE?
  • Key Points
  • When MICE Works
• Method Details
  • PMM: Predictive Mean Matching
  • CART: Classification and Regression Trees
  • Random Forest
  • MIDAS: Distance Aided Substitution
  • Sample: Random Sampling
  • Comparison Summary
• Rubin’s Rules
  • The Problem
  • Basic Formulas
  • Fraction of Missing Information (FMI)
  • How Many Imputations?
  • Usage in mice-py

Overview

This section provides essential theoretical concepts needed to use MICE correctly.

Missing Data Mechanisms

MCAR, MAR, and MNAR definitions

Multiple Imputation Theory

Why multiple imputation and how MICE works

Method Details

Technical overview of each imputation method

Rubin’s Rules

How results are pooled

For detailed mathematical derivations and proofs, see the references.