Post-Term Use and Replication Requirements

What is a “Proprietary Dataset”?

The datasets on Dewey are considered proprietary or restricted. This means:

• Data is licensed, not sold
• Users receive access only while an active subscription is maintained
• Raw data cannot be transferred, published, or redistributed
• Users do not gain ownership of the raw data

In short: You may work with the data during your license, but cannot share or keep it beyond the allowed terms.

Post-term use: General rule

Once a subscription ends:

• ❌ You must delete or return all raw data
• ✅ You may keep derived work (models, statistics, summaries) as long as it does not allow reconstruction of the raw data.

However, there is one important exception:

Exception: Project submitted for publication (revise & resubmit allowance)

If a research project has been formally submitted for peer review (e.g., to a journal or conference), you may:

• ✅ Retain a post-term copy of the specific dataset(s) used for the submission, but solely for the purpose of making “revise and resubmit”(R&R) adjustments.
• ❌ You may not continue using the data for new projects or analyses outside of that submission.
• ❌ You may not share or publish the raw data at any point (before, during, or after submission).

When does this exception expire?

The permitted use expires upon the earlier date of the following:

• ✅ Two years after the last day of the Term, OR
• ✅ Accepted for publication, OR
• ✅ Fully withdrawn from consideration

Once expired, users must permanently delete the raw dataset.

Raw data can never be published or shared

Even during peer review or R&R:

❌ Raw data must not be shared directly with journals, reviewers, collaborators, or external parties.

Instead, the proper method is:

➡️ Share a small sample of any aggregated or inferred statistics that you developed for your research (no more than 10% of the resulting dataset)

➡️ Share your notebook or codebook for how you produced the aggregations or statistics from the raw data

➡️ Direct them to Dewey if they need access to the raw data. Take note of the dataset's DOI.

This ensures:

• Compliance with Dewey's terms of conditions and the terms of our data partners
• Consistent data handling and versioning
• Protection of the intellectual property of our data partners
• A standard replication path

Suggested language for publications or submissions:

“The data used in this study is licensed, proprietary data from Dewey Data. To request access, contact Dewey Data directly.”

What data can be shared? (Aggregations & derived outputs)

We support open research, and many derived results can be shared as long as they do not reveal raw data.

✅ Allowed:

• Aggregated statistics (averages, medians, rates)
• Grouped or binned data where no individual records are visible
• Trends or patterns over time/categories
• Model outputs, coefficients, or trained parameters
• Visualizations (charts, maps, plots) based on aggregated data
• Summaries that cannot be reverse-engineered

❌ Not allowed:

• Row-level or record-level exports
• Small breakdowns that expose individual entries
• “Samples” of raw data, even if anonymized (unless explicitly approved)
• Any format that could allow reconstruction of the core proprietary value

Rule of Thumb:

If someone could rebuild or closely approximate the raw dataset from what you shared, or if it replaces the need for someone else to access the raw data to conduct new research, it’s too granular.

After publication

Once a paper is accepted:

• ✅ You may publish a subset of the aggregated and derivative results as part of your findings
• ❌ You must delete any retained raw data (including the R&R copy)
• ❌ You may not provide raw data in supplementary materials
• ✅ You may provide code or processing logic, as long as it does not include, expose, or reconstruct the raw data

Replication pathway (for transparency and reproducibility)

To enable replication while protecting proprietary data:

1. The authors publish a subset of the aggregated results and methodology.
2. Code can be provided (without raw data included).
3. Reviewers or replicators are directed to Dewey Data to obtain access.
4. Dewey provides the same dataset under appropriate licensing.
5. Replication is performed within Dewey’s permitted-use framework.

This keeps research transparent and compliant.

Summary

Final Thoughts

These guidelines balance:

• Responsible data stewardship
• Licensing commitments
• Research transparency and reproducibility
• Protection of proprietary value

If you're unsure whether a particular use or sharing scenario is allowed, just ask—we’re here to help.