Semi-Automatic Evaluation

An approach to accessibility evaluation that combines automated checks with human expert judgment. Semi-automatic tools run algorithmic tests to detect issues that can be identified programmatically — such as missing alt text or invalid ARIA attributes — and then guide the evaluator through manual inspection of criteria that require human interpretation, such as whether alt text is meaningful or whether content is logically ordered. This approach recognizes that fully automated testing can only reliably detect a portion of accessibility barriers, while purely manual evaluation is time-consuming and inconsistent across evaluators.