A GOMS Model for Keyboard Navigation in Web Pages and Web Applications

At the ICCHP conference(July 12-14, 2006; University of Linz, Austria), Martin Schrepp and Patrick Fischer of the University of Mannheim, Germany presented their paper on GOMS models for keyboard navigation in Web pages and Web applications.

The paper discusses efficiency of keyboard navigation compared to mouse navigation. With the Internet, availability of information is easy and quick. However, getting access to all the information is not always easy, especially for disabled users, if Web applications or Web pages do not allow unrestricted keyboard support.

There are many types of non-text Web elements, such as frames, images, buttons and tables. Many Web pages contain numerous hyperlinks spread randomly across the page. If the Web page does not provide adequate keyboard support, disabled users depending on assistive technologies find it very difficult to access these non-text elements. At times, assistive technology products might not even be able to access certain non-text elements.  

Web applications should enable a user to access such non-text elements and hyperlinks efficiently through a keyboard. ‘Efficiently’ here means that it is not adequate that the user can access all elements using the keyboard; they should be able to do so at acceptable speeds. For instance, if a Web page has 50 hyperlinks, and the page does not support shortcuts and access keys, navigation is possible only by the TAB key. In such a case, the time taken to navigate a Web page through a keyboard would be more than twice the time taken using a mouse.

As per the concept of Universal Design, technology should be designed in such a way that products are accessible and usable by everyone easily, including people with disabilities. Currently, however, the Design for All or Universal Design concept is not adopted properly by designers. As per the report on website accessibility by the British Disability Rights Commission, 81% of the websites fail to satisfy even the most basic standards for accessibility as recommended by the World Wide Web Consortium. The study also found that only 19% of websites comply even with the lowest priority checkpoints for accessibility. The study also implies that a task than can be done in one minute using a mouse may require five to 10 minutes if done using a keyboard.

Many users (disabled, standard and experts alike) prefer to handle all applications. including Web browsers, using the keyboard, as it is frequently faster than using a mouse. Providing efficient keyboard support is necessary to increase the usability of the Web software/application/page. For achieving this, first, we need a method to compare mouse and keyboard navigation. Doing this by user tests requires a lot of effort in terms of time, money and resources. This is where the GOMS model comes in. GOMS stands for Goals, Operators, Methods and Selection rules.

The GOMS models allow comparison of different methods to operate a Web page or Web application. These models can also be used to compare keyboard and mouse navigation. According to the GOMS model, the average time taken to perform an action like following a link will be 11.83 seconds using a mouse and 28.56 seconds using the keyboard. If the time taken to work on a Web page using a keyboard should not be twice or even more than twice the time taken using a mouse, then additional keyboard support should be implemented for this page.

The GOMS models can be used to determine whether the amount of keyboard support for a Web page or Web application is sufficient to ensure that there are no unacceptable disadvantages to keyboard users.