# FAIR Self-Assessment for Audiovisual Research Data **Adapted from the [ARDC FAIR Data Self-Assessment Tool](https://ardc.edu.au/resource/fair-data-self-assessment-tool/) for audiovisual masking workflows.** Use this checklist to assess how Findable, Accessible, Interoperable, and Reusable your masked audiovisual dataset is. Score each item and use the tips to improve FAIRness before archiving. --- ## Audience - Researchers preparing masked data for deposit - Data stewards reviewing audiovisual datasets - Ethics boards assessing data sharing plans --- ## F — Findable ### F1. Does your dataset have a persistent identifier? - [ ] **Best**: DOI assigned via DANS, DataverseNL, or Zenodo - [ ] **Good**: Handle or PURL from institutional repository - [ ] **Minimal**: Web URL only - [ ] **None**: No identifier assigned > **Tip**: Request a DOI when depositing in DANS Data Station SSH. Reserve a DOI before publication to include in your paper. ### F2. Is the identifier included in all metadata records? - [ ] Yes — DOI appears in dataset metadata, related publications, and DMP - [ ] No ### F3. How richly is the dataset described? - [ ] **Best**: Standard metadata schema (Dublin Core / DDI) with title, description, creators, subjects, temporal/spatial coverage, masking method used - [ ] **Good**: Structured text with title, description, and key fields - [ ] **Minimal**: Title and brief description only - [ ] **None**: Not described > **Tip for audiovisual data**: Always document the masking method (blur, pixelation, face swap), software version, and which body regions were masked. Use the SYNAPSIS Metadata Schema template. ### F4. Where is the metadata record searchable? - [ ] **Best**: DANS + NARCIS + institutional repository (multiple registries) - [ ] **Good**: Domain repository (DANS Data Station SSH, DataverseNL) - [ ] **Adequate**: Institutional repository only - [ ] **None**: No searchable registry --- ## A — Accessible ### A1. How accessible is the data? - [ ] **Open**: Entire masked dataset publicly accessible - [ ] **Controlled**: Accessible to authorized researchers under conditions - [ ] **Partial**: De-identified subset publicly accessible; original restricted - [ ] **Embargoed**: Available after a set date - [ ] **Metadata only**: Only the description is public - [ ] **None**: No access > **Tip**: For masked videos, the anonymized output can often be shared openly while the original recordings remain restricted. Use tiered access: open for masked, controlled for originals. ### A2. Is the data available online once access is approved? - [ ] **Best**: Downloadable via standard API or web service - [ ] **Good**: File download from repository - [ ] **Adequate**: Available on request (email/form) - [ ] **None**: No online access ### A3. Will the metadata persist even if the data is removed? - [ ] Yes — repository guarantees metadata persistence (e.g., DANS) - [ ] No or unsure > **Tip**: DANS retains metadata records even when datasets are removed, which is important for citation continuity. --- ## I — Interoperable ### I1. What file formats is the data available in? - [ ] **Best**: Open, standard formats — MP4 (H.264), CSV, JSON - [ ] **Good**: Open but less standard — MKV, TSV - [ ] **Adequate**: Proprietary but widely supported — AVI, MOV - [ ] **Poor**: Proprietary, application-specific formats > **Tip for audiovisual data**: Use MP4 (H.264/AAC) for masked videos, CSV for kinematics/pose data, JSON for blendshapes and metadata. Avoid codec-specific containers. ### I2. Do you use standard vocabularies to describe data elements? - [ ] **Best**: Dublin Core / DDI metadata with controlled vocabulary terms - [ ] **Good**: Recognized terms but not from a formal vocabulary - [ ] **Minimal**: Free-text descriptions only - [ ] **None**: Data elements not described ### I3. Are relationships to other datasets documented? - [ ] **Best**: Machine-readable links (JSON-LD, RDF) to source data, publications, or code - [ ] **Good**: URI links in metadata to related resources - [ ] **None**: No links > **Tip**: Link to the masking software version, the original study, and the consent forms used. Use the `relatedIdentifier` field in DataCite metadata. --- ## R — Reusable ### R1. What license is assigned? - [ ] **Best**: Standard license (CC-BY 4.0 or CC-BY-NC 4.0) with machine-readable URL - [ ] **Good**: Standard license referenced in text - [ ] **Adequate**: Custom data use agreement - [ ] **None**: No license specified > **Tip**: CC-BY 4.0 is preferred for publicly shared masked data. For restricted data, use a Data Use Agreement specifying conditions. ### R2. How much provenance information is recorded? - [ ] **Best**: Full processing pipeline documented — software version, parameters, input/output checksums, operator - [ ] **Good**: Masking method and software version recorded in metadata - [ ] **Adequate**: Brief note in abstract/description - [ ] **None**: No provenance > **Tip**: Use the SYNAPSIS Masking Report template to document the full processing provenance. This satisfies both FAIR requirements and ethics board documentation needs. ### R3. Does the dataset follow community standards? - [ ] **Best**: Follows CESSDA/DANS guidelines for SSH data deposit - [ ] **Good**: Follows institutional RDM policy - [ ] **Minimal**: Ad hoc structure - [ ] **None**: No standards followed --- ## Scoring Summary Count your "Best" and "Good" answers: | Category | Best | Good | Adequate | Minimal/None | |----------|------|------|----------|-------------| | **F** (4 items) | | | | | | **A** (3 items) | | | | | | **I** (3 items) | | | | | | **R** (3 items) | | | | | **Interpretation**: - 10+ Best/Good: Your dataset is well-prepared for FAIR archiving - 7-9: Good foundation; address gaps before deposit - <7: Significant improvements needed; consult your data steward --- ## Resources - [ARDC FAIR Self-Assessment Tool](https://ardc.edu.au/resource/fair-data-self-assessment-tool/) — original tool - [CESSDA Data Management Expert Guide](https://dmeg.cessda.eu/) — SSH-specific guidance - [DANS Data Station SSH](https://dans.knaw.nl/en/social-sciences-and-humanities/) — Dutch SSH data repository - [SYNAPSIS Metadata Schema](downloads/templates/metadata-schema.json) — audiovisual masking metadata - [SYNAPSIS Masking Report Template](downloads/templates/masking-report.md) — processing provenance