Data Management

Good data management keeps records understandable, recoverable, and useful after fieldwork. Define conventions before collecting, review records during curation, and preserve the context needed to interpret each project.

Capture complete data

• Use stable identifiers for projects, sites, events, specimens, and parts.
• Use controlled values for people, methods, preparation types, and other repeated fields.
• Record dates, locations, collectors, methods, and uncertainty as soon as practical.
• Link specimens to their sites, events, personnel, taxonomy, media, and associated data.

Review field records

• Confirm that each specimen has a stable identifier and the correct project.
• Check required values, collection dates, sites, events, collectors, and preparation details.
• Attach media and associated files to the correct record and confirm that they can be opened.
• Record uncertainty or unresolved questions in notes for later review.

Review and clean records

Review records in the field before moving to the next specimen. In the lab, check for duplicate identifiers, missing required values, invalid coordinates, inconsistent dates, and broken media links. Normalize taxon names, personnel, localities, and methods while keeping the original field information available when it is needed to understand a correction.

Use a second person to review important changes. Record the project scope, collection methods, responsible curators, and major corrections so another user can understand how the dataset was created.

Protect sensitive information

Identify sensitive locations, personal information, and restricted collection details before sharing data. Export only what the recipient needs and use secure storage for project archives and backups. Follow institutional policies and applicable collecting, access, and data-sharing requirements.

Choose the right export

• Use Export records for structured data review, analysis, and exchange.
• Use Export project for the daily field backup, and for moving a project between devices. Use ZIP or TAR.GZ so media travels with it, or JSON.GZ for a records-only upload over limited internet. Merge project reads all three back, so this is a recovery path as well as a transfer.
• Use Bundle records for publishing, standards-based exchange, or a reproducible data package.
• Use Backup database weekly, and before any merge or other major change, when you need a restorable copy of the whole NAHPU installation rather than one project. It copies every project directory, every personnel photo, and the entire user configs directory, whether or not a project links to those files, so it is much larger than a project export. In the field, run it occasionally, when battery use is not a concern.

Create a fresh backup before major changes. Use exports to review records in a spreadsheet or another database when useful, and check record counts, identifier uniqueness, relationships, media links, coordinates, dates, and sensitive information before the dataset is published or archived.

See the Export guide for the detailed backup, restore, and archiving workflows.