multiple imputation.

A missing-data method that creates several plausible completed datasets, analyzes each, and combines estimates with uncertainty.

Use it for

Words that belong together

• multiple imputation for missing data
• impute missing values
• combine imputation estimates

On the page

The researchers used multiple imputation and reported how many completed datasets contributed to the final estimates.

Usage notes

Each imputed dataset reflects uncertainty about missing values; estimates and variance are combined across datasets, rather than treating one filled-in file as certain.