ROAR Data Flow Diagram (DFD)

Definitions

  • Item: A single stimulus in an assessment.
  • Trial: A single stimulus/response pair.
  • Run: A collection of trials within a single assessment.
  • Assessment: A collection of items designed to assess a facet of the participant's reading ability.
  • Assignment: A list of assessments chosen for a participant to complete.
  • Administration: A collection of assignments assigned to an organization.
  • Organization: A collection of participants, participant administrators, and research administrators.
  • Participant: A user with the ability to access the assessments within an assigned administration.
  • Researcher: A user with the ability to access the participant demographics and assignment run data for specified organizations.
  • Participant Administrator: A user with the ability to enroll participants, generate assignment specifications for specified administrations, and access completion statistics and scores for specified administrations.
  • Research Administrator: A user with the ability to enroll participants, create Participant Administrator accounts, and design assessment and assignment specifications.

System Components (Vendor-Agnostic)

  • Admin Database: system-of-record for identifiable operational data (PII/admin/org/rosters, etc.).
  • Assessment Database: system-of-record for de-identified assessment response/run/trial data and derived scores.
  • Identity Provider & Access Management: authentication and authorization for ROAR users.
  • Authorization & Data Validation Layer: enforcement of access control and request/data validation.
  • Backend Compute Services: backend business logic (serverless functions and/or containerized services), including cross-store synchronization and data-intensive operations.
  • Object Storage: bucket-based object store for assets/corpora and participant artifacts like audio/video recordings.

Level 0 DFD (Context)

Level 1 DFD (User-Facing Processes)

Level 2 DFD (System-Level Processes)

Level 2 is visually dense. To make it maintainable we represent it here as:

  • A shared system component map, then
  • Process-focused sub-diagrams aligned to the Level 2 numbering.

Level 2 — Shared System Components

2.1–2.2 Authenticate + Consent (Gatekeeping)

2.3–2.4 View Assigned Assessments + Take Assessments

2.6–2.7 View Scores + View Administration Progress

2.8–2.9 Enroll Participants + Create Assignments (with Backend Sync)

2.10–2.12 Organizations + View Users for Organization

2.13 Create Participant Administrator

2.14 Create Assessment + Assets/Corpora

2.16 View Raw Run and Trial Data (Researcher CLI)

Notes for Maintainers

  • The "Admin Database" vs "Assessment Database" split is intentional: it supports PII separation and de-identified analytics/scoring.
  • "Backend Compute Services" covers both event-driven sync and batch-ish operations (and can include vendor-agnostic ELT steps, provided they run inside your environment and don't create new external data recipients).
  • "Object Storage" is used for assets/corpora and can be extended for participant artifacts (e.g., audio/video) if that's in scope.