3.3 Dealing with duplicates, errors, outliers, and missing data

Before running any analysis, you might want to make sure your data is reliable, i.e. as much as possible free from duplicates, errors and inconsistencies. Screening your data looking for these will help you understand the reliability of your data.

Keep in mind that data correction might not always be possible, desirable, or sufficient to increase the reliability of your data.

This section is divided into 4 sub-sections:

• 3.3.1 Duplicates,
• 3.3.2 Errors,
• 3.3.3 Outliers,
• 3.3.4 Missing data.