Preventing accidental data loss

Data loss most often occurs through operational mistakes rather than technical failures. Common risks include accidental deletion, overwriting working files, unsynchronised copies across devices, and errors in automated processing workflows.

To reduce these risks, research teams should:

• Maintain clear separation between raw, working and derived data.
• Protect master files from accidental editing.
• Use consistent folder structures and naming conventions.
• Avoid manual overwriting of source data during cleaning or processing.
• Test automated scripts on copies of data before applying them to production files.

For projects using automated pipelines or live data feeds, it is good practice to implement safeguards such as read-only source folders, checkpoint outputs and routine integrity checks.

Combining these operational practices with regular backups and version control provides layered protection against data loss during the active research phase.