Examine data from different angles using diagnostic reports. Determine the extent of missing data and any extreme values with a case-by-case overview.
Use a multiple imputation model to understand patterns and replace values; it helps you choose the most suitable method. Use linear regression or expectation maximization algorithms among others.
Display missing data for all cases and variables. Determine differences between missing and non-missing groups. Assess how missing data of a variable relates to missing data of another.