Step 1 – page validity and alt text

Of the 127 sites tested in this survey, there was a marked improvement in the automated testing of page validity and application of ‘alt text’ to images. Three sites achieved a perfect score in this set of tests.

A small number of sites had regressed since the last survey. However, this is probably caused by the marked increase in size of those sites. The rate of sites regressing was less than the rate of sites progressing.

Step 2 – visual checks

Of the 127 sites tested, there was no discernible change in the results of visual checks. This is because the conditions for testing are limited to the presence of specific functions on a website, for example server-side image maps. These features tend to be uncommon.

The most common features checked were:

• the association of data cells to header cells in data tables, where present
• pages were usable without JavaScript.

Step 2 – checking for key documents in accessible format

There has been a small improvement in the ratio of sites that provide key documents in an accessible format.

In the 2005 survey, not all sites were checked for accessible key documents as some sites had already failed to proceed to that test. However, of those that were checked, 52% of sites provided key documents in an accessible format.

In the current survey, all 127 sites were checked. Key documents were provided in accessible formats on 54% of these sites.

Step 3 – checking for accessibility barriers

Government websites showed an increase (17%) in technical compliance with Web Guidelines. There were 54% sufficiently compliant websites in 2006 (68 sites out of 127) compared with 37% in 2005 (56 sites out of 150).

The top 30 compliant websites were selected for testing of accessibility barriers. These sites were assessed using the same methodology as in the 2005 survey.

There was no difference in the presence of accessibility barriers for the websites tested between the two surveys.

Step 4 – user testing

The user testers were asked to complete five simple tasks to gauge the accessibility of the top 30 compliant websites. Four tasks were generic and applied to all the sites tested. The fifth task was specific to each site.

The generic tasks were:

• Find the contact details of the organisation and copy the URL of the contact details page.
• Does the website have a Purpose Statement? Type no if not found and copy and paste in the URL if found.
• Does the website have an Accessibility Statement or access keys page? Type no if not found and copy and paste in the URL if found.
• Find and respond to the feedback form on the site.

Some examples of the site specific tasks are:

• Beehive: who is the Minister of Racing?
• Medicines and Medical Devices Safety Authority: name a condition that can be treated by a Pharmacist only medicine.
• Te Kete Ipurangi: what is the web address of the website that provides information on Board of Trustees elections in 2007 and when are results declared?
• Land Information New Zealand: if you object to a revaluation of your property, who will your objection be heard by?
• Employment Relations Service: how many days bereavement leave are you entitled to if your mother or father dies?

The testers recorded how long it took them to complete each task (measured in minutes). Each task was rated on how easy or hard they found it to complete.

Direct comparison of specific user testing between the surveys is affected by the inclusion of Deaf testers in this survey, other changes in the user testers, variances in the site-specific task questions, and changes in the sites. However, average satisfaction scoring on task completion by the user testers between the surveys was similar.