Lecture缺失值处理策略.pptVIP

The depression trial * * 5. Part II: Multiple imputation * Data set with missing values Result Completed set * * General principles * Informal justification * The algorithm * Pooling information * Hypothesis testing * * MI in practice * MI in practice A simulation-based approach to missing data 1. Generate M 1 plausible versions of . Complete Cases ^ ^ ^ ^ ^ ^ = imputation for Mth dataset 2. Analyze each of the M datasets by standard complete-data methods. 3. Combine the results across the M datasets (M =3-5 is usually OK). * MI in practice... Step 1 Generate M 1 plausib