Google DTorial: An interactive tutorial framework for blind users in a Web 2.0 world - Manual (Page 5)

•Audio Based Content - Because users rely entirely on audio cues,
developers should ensure that content is presented in such a way to
be fully understood through audio only and existing screen readers.
•Mitigate AJAX - Due to the complications created by screen
reader interaction with Web 2.0 applications, tutorial systems must
address this new complexity (for their own interaction model and
the model of the target application).
5 DTorial and Incremental Design/Evaluation
In response to the complexity of learning new applications and the elaborate Web 2.0
+ screen reader interaction techniques, we developed an initial design for embedded
Web-based help systems for VIUs. In order to develop and evaluate our design, we
built the DTorial system. By evaluating, testing and incrementally improving DTorial,
we are able to analyze user behavior, reaction, and resulting design changes to create
a higher-level interaction model.
5.1 DTorial
The DTorial system is a fully functional embedded tutorial, built using JavaScript.
We used Grease Monkey to implement DTorial on top of Google’s Web 2.0 Gmail,
chosen because email is ubiquitous and is crucial to most individuals. Participants in
our initial interviews said that email is empowering and central to their
communications at work and at home. We used a set of JavaScript libraries called
AxsJAX [10] to facilitate audio feedback and audio alerts.
DTorial works by moving a user around a Web page while simultaneously
providing the user with audio feedback and describing the new content, what to
expect, and how to interact. In other words, DTorial teaches the user how to interact
with the computer while using the tool they are learning. As a result, the user is easily
able to apply the skills learned without switching environments.
For example, consider a fictional VIU, Alice. When Alice wishes to understand the
concept of the Folders List in Gmail (Inbox, Drafts, etc), she locates the Drop Down
List on Gmail’s webpage entitled “Tutorials.” Alice selects the tutorial entitled
“Folders” and then presses the Go Button to launch her content. The web browser’s
focus is automatically moved, via JavaScript, from her current element (e.g. Go
Button) to the Folder List in Gmail. Simultaneously, tutorial text explaining Folders
is injected into the DOM, at the top of the Folder list, within the now focused region.
While having no visual impact to the layout of the page (using CSS markup), screen
reader users like Alice can now access the entire body of injected text, including
navigation by headings, lists, and other HTML elements. Alice can immediately
explore the tutorial and test out her new knowledge without having to leave the
instructional guide, because the guide AND environment, are side-by-side. DTorial
further blurs the line between instruction and execution by ensuring a mechanism for
re-reading the instructions at any time. It should be noted that the method for injecting