Error Profile

An error profile is a structured summary of the accessibility issues detected on a page, typically represented as a numeric vector with one component per checkpoint or rule — counts of violations, binary pass/fail indicators, or failure rates. Error profiles were introduced in web-accessibility sampling research (King et al., 2005; Brajnik, Mulas, and Pitton, 2007) as a way to cluster pages that share similar patterns of issues, so that stratified sampling can pick representative pages from each cluster rather than blindly selecting at random. The same idea underlies modern automated-testing dashboards that group pages by rule fingerprint to help human auditors focus their attention efficiently.