Accessibility auditors have to choose a method when evaluating accessibility: expert review (a.k.a. conformance testing), user testing, subjective evaluations, barrier walkthrough are some possibilities. However, little is known to date about their relative strengths and weaknesses. Furthermore, what happened for usability evaluation methods is likely to repeat for accessibility: that there is uncertainty about not only pros and cons of methods, but also about criteria to be used to compare them and metrics to measure these criteria. After a quick review and description of methods, the paper illustrates a comparative test of two web accessibility evaluation methods: conformance testing and barrier walkthrough. The comparison aims at determining merits of barrier walkthrough, using conformance testing as a control condition. A comparison framework is outlined, followed by the description of a laboratory experiment with 12 subjects (novice accessibility evaluators), and its results. Significant differences were found in terms of correctness, one of the several metrics used to compare the methods. Reliability also appears to be different.
A comparative test of web accessibility evaluation methods
BRAJNIK, Giorgio
2008-01-01
Abstract
Accessibility auditors have to choose a method when evaluating accessibility: expert review (a.k.a. conformance testing), user testing, subjective evaluations, barrier walkthrough are some possibilities. However, little is known to date about their relative strengths and weaknesses. Furthermore, what happened for usability evaluation methods is likely to repeat for accessibility: that there is uncertainty about not only pros and cons of methods, but also about criteria to be used to compare them and metrics to measure these criteria. After a quick review and description of methods, the paper illustrates a comparative test of two web accessibility evaluation methods: conformance testing and barrier walkthrough. The comparison aims at determining merits of barrier walkthrough, using conformance testing as a control condition. A comparison framework is outlined, followed by the description of a laboratory experiment with 12 subjects (novice accessibility evaluators), and its results. Significant differences were found in terms of correctness, one of the several metrics used to compare the methods. Reliability also appears to be different.File | Dimensione | Formato | |
---|---|---|---|
p113-brajnik.pdf
non disponibili
Tipologia:
Documento in Pre-print
Licenza:
Non pubblico
Dimensione
2.2 MB
Formato
Adobe PDF
|
2.2 MB | Adobe PDF | Visualizza/Apri Richiedi una copia |
p113-brajnik.pdf
non disponibili
Tipologia:
Altro materiale allegato
Licenza:
Non pubblico
Dimensione
2.2 MB
Formato
Adobe PDF
|
2.2 MB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.