Raw vs Processed Data in Experiment Records: What Both Contribute to Reproducibility

Raw and processed data in experiment records serve different purposes: raw data is the unmodified instrument output, the original observation that supports audit and re-analysis; processed data is the analyzed, summarized, or visualized result that supports conclusions. Both belong in the record, because processed data without raw data is unverifiable, and raw data without processed context is uninterpretable.

Many labs keep processed data in the record and raw data somewhere else, then discover during an audit or a reproducibility check that the raw data is lost, corrupted, or no longer accessible. This guide covers the role of raw vs processed data in experiment records, what each contributes to reproducibility, and how to manage both.

What Each Data Type Contributes

Raw data is the evidence; processed data is the interpretation. Both are needed. An auditor or a future team member should be able to see the processed conclusion and trace it back to the raw data that supports it, or re-analyze the raw data independently if needed.

How Zettalab Supports Data Management

For labs that need raw and processed data linked to records, Zettalab connects ELN-style documentation with team file storage. ZettaNote supports file attachments, and ZettaFile supports organized storage, so a team can keep both data types connected to the experiment record. To manage raw and processed data inside a connected R&D platform, explore Zettalab's cloud-based R&D lab platform.

FAQ

Why keep raw data if I have processed data?

Raw data is the original evidence. Processed data can be checked against it, and re-analysis may reveal insights the original processing missed. For audit and regulatory purposes, raw data is often explicitly required. Processed data alone is a conclusion without its foundation.

When is processed data alone sufficient?

Rarely, and only when the processing is lossless and reversible, or when raw data cannot be preserved for legitimate technical reasons. In general, any conclusion that matters should have its raw data preserved and linked. When in doubt, retain the raw data.

How long should raw data be retained?

For the same period as the experiment record it supports. For regulated work, specific retention periods apply. The retention obligation covers both raw and processed data; do not retain one without the other.

Conclusion

Raw and processed data both belong in experiment records: raw data provides the evidence, processed data provides the interpretation. Keeping both, linked and retained together, is what makes experiment results reproducible and auditable. A connected R&D workspace that supports both data types, such as Zettalab, fits labs whose records must hold up over time. To manage raw and processed data inside a connected lab workspace, explore Zettalab's cloud-based R&D lab platform.