Making Web Content Accessible

By IGI Global on Jun 24, 2011
IGI Global would like to thank Shalin Hai-Jew for contributing this article outlining methods for making multimedia web content accessible through several mediums. Dr. Hai-Jew's newest publication, Constructing Self-Discovery Learning Spaces Online: Scaffolding and Decision Making Technologies, will become available this Winter. An excellent resource for any library, her edited research volume, Virtual Immersive and 3D Learning Spaces: Emerging Technologies and Trends, is currently available in the IGI Global Bookstore.

To increase the broader usage and shelf life of digital learning objects, it is important to ensure that the digital objects are fully accessible. Given the variety of multimedia and Web-deliverable contents, the definition of "accessibility" varies. One good starting point, though, is the U.S. federal government requirement, as encapsulated in Section 508. The WebAIM Section 508 Checklist translates these requirements into more easily applied specific details.

For many decades now, the Americans with Disabilities Act (ADA) has specified accessible design for websites. Generally, all users of a site should have a clear sense of orientation to the site and have access to the necessary informational value of the contents of the site—no matter if he or she has varied visual, hearing, symbolic processing, or other limitations.

How many perceptual channels does a learning object have to be rendered in? In general, the guidelines seem to suggest that the information has to be available in two. Something that is a visual has to have a text form (usually in the alt-text or "longdesc" element of XHTML) that captures the informational value of an image. Something that is an audio piece of information also has to have a text form. An audio-visual simulation should have a transcript and text descriptors of the action (which is three channels: sight, hearing, and symbolic processing).

These standards are not meant to put a dampener on the types of rich media used in online learning. After all, rich media convey plenty of information, and they often engage learners. However, such accessibility considerations will certainly add an extra layer of steps to the development of such objects.

On-the-Ground Realities for Transcription

Building accessibility into a design early on saves on having to double-back later on, when the work is often more costly to bring up-to-par and when some information may no longer be available from the original subject matter expert (SME). Retrofitting a course for accessibility with proper alt-texting may mean deeper research on the topic for accurate alt-texting or transcription. (If information is cannot be established in a credible way, that data should be left off—rather than conveying inaccurate data. Or, the data would have to be qualified to show the clear limits of verifiability.) Also, knowing what accessibility work will be required will help the development team decide which pieces of learning to render into video. Understanding the real costs of the work may dissuade some faculty members from going full-bore on in capturing all their semester's lectures into video.

Practitioners have long hoped for technology-enabled ways of transcribing speech into text effectively. It seems that those technologies are maybe just a few years away in terms of public consumption, but, then again, that hope has been around for decades. It may be that the real costs currently, and for the near-term, will require human transcriptionists focusing on the nuances of the spoken speech, the tone, the context, and other elements—to render an effective transcript.

The Contents of Effective Transcripts

So what makes for an effective transcript? Is it only an effective verbatim rendering of the spoken speech? Some basic requirements for these transcripts would include the following:

  • The transcript should be accurate in terms of what was spoken—even if the speaking was done in a heavy accent or at a fast clip. Anything that is unintelligible should be represented as such.
  • The transcript should be written in correct sentence format and with correct punctuation and symbology. Transcripts that are in all block caps are not preferred.
  • The spelling should be fully correct even if complex terminology appears in the audio format.
  • Measurements should be represented in all the main conversion formats.
  • All sound effects and music should be described in the transcript—especially if those elements add informational value or mood value.

Beyond what is spoken, other aspects can be highly important to include. If speakers take actions that augment the dialogue or the communications, that should be described. If a presenter puts up an informational image in a video, that should also be included in the transcript. Anything that has informational value should be included.

URLs that have been mentioned should be included and activated in a value-added transcript.

Relevant images should be included in a transcript—with alt texting.

It also helps to have an information structure—through the use of tagging headings, subheadings, body text, and other textual elements in a transcript.

A transcript should be time-stamped every so often, so that users of the transcript can reconnect with a video or simulation at an exact time for highest utility. A transcript is seen as less optimal than synchronized captions, but this type of captioning—where verbatim speech is tied to a particular time in a video—involves another layer of work which is often considered a little too high-cost in a university or college environment. Some high-end grant-funded projects, though, will have to-the-second timed text.

Some Work-Arounds

There are some ways to cut corners without losing quality. Lecturers who stay very close to their slideshows (and who design effective slideshows) may conduct a desktop lecture capture and also make the slideshows available, which means that the audio of the slideshow is text-readable in the properly-designed slideshows. Images that are being uploaded into a database may be submitted with the metadata. The proper metadata (information about information) may serve as the machine-readable alt-texting for users.

Proper transcription enhances learning possibilities for a much wider potential audience. It also enhances understanding of the material for non-native speakers of the language-of-instruction. Further, it's a legal requirement to ensure accessibility of Web-delivered contents.
Browse for more posts in:
EducationApplied E-Learning

No comments Comments

Log in or sign up to comment.
Be the first to comment!