Preparing release versions of data

Working data often contain temporary variables, intermediate processing outputs or exploratory analysis artefacts. Preparing release versions involves creating clean, well-structured copies that are intended for sharing.

In practice, this may include:

• separating raw, processed and final release files
• removing temporary or redundant variables
• standardising missing value codes and labels
• ensuring consistent variable naming and formatting
• ensuring consistent file naming and formatting
• providing appropriate metadata alongside data
• checking files for consistent and effective anonymisation
• checking that file contents match collection-level documentation.

Maintaining a clear distinction between internal working files and external release versions helps avoid confusion and supports reproducibility.