Web accessibility metrics are crucial for the quantitative evaluation of web sites. We present a new automatic metric called WAB, based on the WAB metric with extensions inspired from the UWEM metric. The first results are encouraging as we have obtained better precision when calculation the accessibility level measured for a site. Related Work The following is a brief summary of the features of the two metrics considered: WAB and UWEM.

We need quantitative evaluations of accessibility [

We focused on two of these metrics: WAB and UWEM. The first one is popular in
accessibility studies [

WAB metric [

WAB score = ?p ?v ???? Nnvv ×Wv ?? ? ? N p

With p = Total pages of a Website, v = Total violations of a Web page, nv = Number of violations, Nv = Number of potential violations, Wv = Weight of violations in inverse proportion to WCAG priority level, and Np = Total number of pages checked.

WAB scores greater than 5.5 denote a web site with serious accessibility problems
[

The UWEM [

One of the outputs is a score computed using the following metric:

F (s) =

Bp1 + Bp2 + ... + Bpn + Np1 + Np2 + ... + Npn +

? Bst t _ on _ site _ level

? N st t _ on _ site _ level

With F(s) = UWEM score for the site s, Bpi = Total number of "fail" results from all tests within page pi and Npi = Total number applications of all tests within page pi. Bearing in mind that some UWEM tests apply to the web site as a whole and not to individual pages, Bst = Total number of "fail" results from all tests on site level, Nst= Total number applications of all tests on site level.

The UWEM metric (Fs) value is a number between 0 and 1. A score of zero denotes that the Web site should not present serious problems of accessibility. 3

WAB* As mentioned before, the WAB metric evaluates 25 checkpoints from the 3 priority levels. However, the tests performed to evaluate the checkpoints are sometimes vague when specifying the way to determine the number of potential violations of each checkpoint. This is owed mostly to the evolution of the technology used in the construction of web sites. Table 1 shows some examples.

On the other hand, the UWEM metric specifies priority 1 and 2 checkpoints of greater precision than the ones used in WAB (as seen in Table 2) when identifying the number of potential violations. However, the downside of this metric is that no priority 3 checkpoint is used.

WAB* considers all 3 priority levels (as WAB), and not only 2.

WAB* uses the same WAB metric (with the same weights).

WAB* has all WAB checkpoints but updated to eliminate deficiencies. For example: a. b.

Not only frames, but iframes are tested.

Not only table, th, td, and frame are tested for relative size and position, but also div, iframe, object, and applet.

Besides, WAB* includes eleven automatic checkpoints from the UWEM metric.

The final result is a metric with 36 fully automatic, evaluable checkpoints. 6 are priority 1, 23 are priority 2, and 7 are priority 3. The complete table is not included for space reasons.

We performed an accessibility study of the European Banking Section (some results
are published in [

We have introduced WAB*, a new fully-automatic metric. It is based on WAB but has a much more severe accessibility evaluation, using also UWEM-inspired checkpoints. It evaluates 36 checkpoints against WAB?s 25, and with more precision. The first results from studies performed using the new metric show it provides more precise markers for accessibility levels than the other two metrics.