When we proposed the development of the OneNet authoring tool that would enable non-technical authors to easily create accessible electronic documents, we first performed a usability study to examine the necessity and feasibility of this goal by testing two hypotheses:

  1. Existing authoring tools do not adequately support "typical" authors in producing accessible documents.
  2. An authoring tool designed to intentionally promote accessibility would significantly improve the likelihood that typical authors will produce accessible documents.

The project team developed a practical accessibility scoring metric, based on W3C and Section 508 standards, to measure and compare the accessibility of electronic documents. The team then developed a prototype authoring tool with features designed to actively assist users in addressing accessibility. A series of user and expert tests were performed, comparing the accessibility of test documents created with existing authoring tools to those created using the prototype.

In user testing, twelve non-technical users created a test document with three leading tools (Microsoft Word, Microsoft FrontPage, and Macromedia Contribute) and the prototype. The accessibility of each document was scored using the accessibility metric and compared across tools. Results were significant (a = .05) and dramatic:

  • Microsoft Word, Microsoft FrontPage, Macromedia Contribute 31.2%
  • OneNet Prototype 72.3%

In expert testing, four accessibility experts created the test document using a wider range of authoring tools. Experts evaluated each tool in a "typical use" scenario, in which they mimicked the patterns of use observed in the user tests, and in a "best case" scenario, in which they did everything possible to create an accessible document. Results of "typical use" tests were similar to those found in user testing, showing an average of 38.7% accessibility for existing tools and 77.5% for the prototype. "Best case" tests demonstrated that accessibility was possible with many of these tools, with the best achieving as high as 95%. This observation led to an important additional conclusion:

  • It is not adequate to make accessibility possible, it must be automatic.

Based on the results of this study, the project team is convinced that it is both necessary and possible to develop a tool that would improve the accessibility of typical electronic documents.

Full Paper

OneNet White Paper: Development of an Authoring Tool that Promotes Accessibility (pdf)