One application path to restricted federal data, carrying every agency’s rules.
ResearchDataGov is where researchers find and apply for restricted data from 16 U.S. federal agencies. I took over the design in September 2022 and worked on the access pathway: dataset selection, the application and the reviewer side.
- My time
- Sep 2022 – Jan 2024
- My role
- Product designer, ICPSR
- Client
- Federal government data agencies
- Team
- ICPSR internal project team

What I covered
Overall role Product designer Steps 2–4
Before my part The partner agencies set its purpose, its scope and some of the details. ICPSR began the project before I joined ICPSR.
Each agency brought its own requirements to the same application.
Dataset rules, required documents, review periods and provisioning paths all differed by agency. I mapped how a researcher moves from discovery to selection to application, and where each rule enters that path.
I ran multiple rounds of usability testing. The sessions, findings and screens are confidential.
One path, with each rule shown only when it applies.
I designed ResearchDataGov as one shared service, so applicants and reviewers worked from the same application data.
| Researcher step | Agency rule | Review checkpoint | System state |
|---|---|---|---|
| Select datasetApplicant chooses data with restricted access. | Dataset termsRules appear only when relevant. | EligibilityRequired documents and identity checks. | DraftRequirements are still editable. |
| Submit applicationResearch purpose and team details lock. | Agency pathDifferent agencies route differently. | Reviewer queueShared data supports decisions. | Under reviewApplicant sees progress, not internal noise. |
- 01One flowDiscovery, dataset selection, application, review and provisioning in one end-to-end access path.
- 02Many rulesAgency- and dataset-specific requirements, without showing every rule at once.
- 03Two sidesApplicant and reviewer views built on the same application data and status logic.
- 01Versioned required documents.
- 02Review rules set per dataset.
- 03Clear status changes across applicant and reviewer views.
The design was still in exploration, and the screens kept drifting.
Completed the screens and the design specs, with detailed workflow scenarios.
Developer-ready specs, and requirements I completed with the PM.
ICPSR’s team built the MVP from my specs.
I worked with the team through implementation. In September 2023 a second product designer joined, and the project lead team, made up of the PM, the tech lead and the two designers, set the MVP feature list, the work schedule and the delivery schedule. She took over the rest when I left in January 2024.
- Sep 2022I take over a design still in exploration
- Sep 2023A second designer joins; the lead team sets MVP scope and schedule
- Jan 2024Handoff; she takes over the rest
After my part Live at researchdatagov.org. It is now hosted by another organization.