Skip to content

openfed/AccessibilityScale

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 

Repository files navigation

AccessibilityScale

A proposal for a more formal Accessibility Scale metric based on the EN 301 549 v3.2.1 standard

Why?

  • Only three conformity levels in the WAD
    • Full
    • Partial
    • Non
  • Small errors immediately lead to “partial conformance”
  • Heavy investment in accessibility will not necessarily be reflected in a change in conformance level

Goals

  • Increase the granularity – more opportunities for noticeable improvement
  • Better vision on the impact of accessibility issues for different use cases
  • Less legalistic and more public-friendly communication
  • Clear and objective approach

Sources

  • Based on EN 301 549 v3.2.1
    • Criteria (list Annex A Table A.1)
    • “Functional performance statement” (Annex B Table B.1) – renamed to “Use Cases” here
    • Weighing based on Table B.2 “Primary” and “Secondary” relationships of criteria vs Use Cases
  • Based on the presence of an Accessibility Statement
  • Evaluations based on in-depth or simplified audits
  • Inspired by the Dutch “Accessibility status” logic and presentation on
    https://www.digitoegankelijk.nl/toegankelijkheidsverklaring/status

In-Depth audit scoring

  • Across 10 use cases, calculate a %-based score based on pass/fail – “Not Applicable” criteria not included
    • Weighed calculation (prototype!!!) per use case
      • Primary relationship: 1
      • Secondary relationship: 0.5
      • No relationship: 0
    • (Weighted sum of all “pass” criteria) / (Weighted Sum of all applicable criteria)
  • Final score is
    • Average of score per use case
    • Except if at least one use case scores below a threshold (25%) – then total score = lowest score

Simplified audit “scoring”

  • Three levels
    • 0 errors detected in simplified audit
    • >=50% of verified criteria failed
    • <50% of verified criteria failed
  • Important distinction with in-depth
    • Simplified audit based on automated checks can establish violations (“fail”)
    • Cannot (for most criteria) establish “pass” since automated testing will miss certain failure cases
    • So the “%” here is something very different from the in-depth scoring

Visualisation of the scale

  • Using letters to indicate conformance level (A+, A,…,F)
  • Visuals inspired on Nutriscore

Conformance levels

  • Level A+: score 100% (in depth audit) and valid accessibility statement
  • Level A: score >98% (in depth audit) and valid accessibility statement
  • Level B: score >=60% (in depth audit) and valid accessibility statement
  • Level C: either
    • score <60% (in depth audit) and valid accessibility statement
    • no failures on simplified audit and valid accessibility statement
  • Level D: =<50% failures on simplified audit and valid accessibility statement
  • Level E: either
    • >50% failures on simplified audit and a valid accessibility statement
    • no audit but a valid accessibility statement
  • Level F: No valid accessibility statement

Visualisation on in-depth audit report

Screenshot of detailed explanation of the conformity level (in Dutch)

  • States the validity/presence of an accessibility statement
  • Provides the aggregate score
  • Shows a vertical visualisation of the scale, in this example “C”, and a sentence stating “the score for this website is lower than 60%”
  • For each use case
    • shows a visual and textual representation of the indiviual scores
    • where a score is below the 25% threshold, the visual is in a different colour and surrounded by a thick outline of the same colour

Some remarks

  • We did excluded the Use Case “4.2.11 Privacy” because it is linked to a very small set of criteria and would have a disproportionate effect on the score
  • We did include Use Case “4.2.9 Minimize Photosensitive Triggers” because violating it has such an immediate and high impact on the affected user
  • Note that all proposed thresholds and weights described need to be tested against more real-world examples before finalisation

About

A proposal for a more formal Accessibility Scale metric based on the EN 301 549 v3.2.1 standard

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published