WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: Procedure of making web accessibility testing.

for

From: Karl Groves
Date: Jun 9, 2009 12:55PM

From an auditing standpoint - and in consideration of the production techniques of modern websites and web based applications - the best approach for thorough, qualitative data is to test representative sets of content based on media type. This is for a few reasons:

First, website interfaces are now often constructed with some sort of template system. The example I like to use most is the CNN website. Going to that site, you'll observe that every page within the 'www.cnn.com' domain has the same header. Imagine for a moment that the 'Powered by Google' image next to the "Search" button had no alt text. Because this is the same exact code calling this image, you only need to report on that missing alt text once. Find it once, fix it once, and the issue is remediated across the whole system. It would be an inefficient use of auditor (and developer, and stakeholder) time to continuously find & report on the same issues. From an auditing standpoint, their efforts would be best reserved for finding new problems, not ones they already know about.

Somewhat related is the idea that developers typically have their own "patterns" to the way they develop things. Generally speaking, a developer who doesn't make one data table correctly won't make any of them correctly. Of course there might be instances where an otherwise good developer will miss something, for the most part if you test a handful of similar items you'll get the qualitative data you need in order to make the necessary remediation recommendations. So, carve out portions of the interface into media types (i.e. "tables", "forms", "images", etc.) and test a few representative samples of each rather than testing every single one.


Karl



----- Original Message -----
From: "Angela Colter" < <EMAIL REMOVED> >
To: "WebAIM Discussion List" < <EMAIL REMOVED> >
Sent: Tuesday, June 9, 2009 10:27:47 AM GMT -05:00 US/Canada Eastern
Subject: Re: [WebAIM] Procedure of making web accessibility testing.

Hi Rakesh,

I recently worked on an accessibility evaluation for a recently redesigned
association web site. Now, the full site has around 60,000 pages; there are
many, many content editors updating different areas of the site.

Because of the limited budget of the project, we reviewed a sample of 15
pages across the site, chosen by the client. While focusing on most popular
pages certainly makes sense, our samples were chosen in an effort to ensure
that the main content areas (each edited by a different group) were
represented. That way, no one could look at our results and think, "well,
that's a problem for THAT group, but not for ours." That was the thinking
behind choosing the sample, anyway.

What we found was there were two different types of accessibility errors:
errors in the design and errors in the content.

The "design" errors (such as text links that were styled so that the
underlines were removed) were located in the page templates, CSS, etc. These
errors would need to be corrected by folks with specialized knowledge or
access to fix these global issues. And because they were global, fixing them
once fixed them everywhere.

The "content" errors (such as poorly chosen or missing alt text) were
usually introduced by the content editors themselves. These are easy enough
to fix once the content editors understand what the errors are, how the
affect accessibility and how to fix them. The problem is, content errors are
very easy to reintroduce.

So by all means, do an accessibility review on a selection of pages. My
suggestion would be to consider your popular pages, critical tasks,
different page types (forms vs. pages, for example), and the different
content editors in your sample.

Where you find global issues in the design, fix them. Where you find errors
introduced by the content editors, focus your accessibility training efforts
to address the types of errors you're content editors are making.

Hope this helps,

Angela Colter

Usability / Accessibility Consultant
angelacolter.com

On Tue, Jun 9, 2009 at 12:18 AM, Rakesh Chowdary Paladugula <
<EMAIL REMOVED> > wrote:

> Dear all,
> I have a doubt from the beginning. I have a existing website with
> 1000 pages to be tested for web accessibility should I test all the
> 1000 pages or some selected pages . If so what are the pages I gave to
> test.
> One more issue is I will have websites under construction which will
> be in local surver. Wave will not take the URL of the staging site. Is
> the procedure used is file upload and copying the source code and
> validate or is there any other procedure.
> Thanks & regards
> Rakesh Chowdary
> Iridiuminteractive Limited
> Changing a face doesn't change anything but facing a change changes
> everything.
>