Web Site Ergonomics and User Tests Page II

Surface ergonomics depends little on the context, and instead concerns the presentation of information (typography, colors, etc.). The tools applicable in this context are style guides or ergonomic guidelines. These guides give tips for improving the design of user interfaces, and derive essentially from theoretical research and consulting work in the field. The ergonomic guide(4) that we have devised offers recommendations for achieving a good degree of homogeneity across a site, in terms of presentation, as well as behavior and the system's reactions; we often use the term " look and feel." From these recommendations and the particular features of the project, the ergonomic standards (detailed specifications) for the project can be drawn up.

User tests

The ergonomist has a number of specific tools and devices at his disposal: interviews, observations, task description method, expert assessment, and so on. User tests are an integral part of these processes, but are still not implemented as often as they should be. In a user-oriented approach, it seems quite paradoxical that this option is not used as a matter of course.

There are a wide variety of user tests available. To put it broadly, we might say that the user test is a means of qualitative assessment based on individual analysis of target users of a Web site. More precisely, user tests enable us to directly observe the behavior of Internet users as they visit a site, to identify potential problems and to pinpoint the difficulties they encounter. From this data, precise recommendations can be compiled, and used to make improvements to the site.

How many users should be observed?

The number of users that make up our sample groups varies between 6 and 10. One oft-heard criticism is that this is not a very scientific approach... which is a valid remark, but do not forget that we are not trying to obtain statistically meaningful results. The aim is simply to pinpoint the main problems of the site, or to validate choices concerning browsing, categorization or functionality.

As Jacob Nielsen(5) puts it in his Alertbox of March 19, 2000: "The most striking truth of the curve is that zero users give zero insights."

Source: Jacob Nielsen, Why You Only Need to Test With 5 Users, Alertbox dated March 19, 2000

By carrying out tests on a sample of 5 users, it is possible to detect 80% of the ergonomic problems that a Web site can pose.

Here, we shall look at three of the most cost-effective types of test:

• Card sorting
• Scenario-based test.
• Free visit

Card sorting for site categorization

Card sorting is a method used to define or validate the overall categorization of a site. It enables information blocks to be grouped together in line with the user's logic and lets you designate categories or grouped information. The method is simple: it involves showing users a selection of cards, at random, featuring:

• the information provided by the site
• the services/features offered by the site.

Users are requested to explain, spontaneously, how they understand each card and to group them together according to their own logic. Once the cards are grouped together, the users are asked to give a title to each set of cards. After correlating the results from each user, the site categories can be devised. This operation can be reiterated to achieve the required degree of accuracy, to compile the sub-categories and layout of the different aspects of the site. The technique provides us with qualitative data regarding the users' needs, and makes it possible to structure content of categories to fit their logic. It also lets us validate or select the titles of categories, so that they match the target users' requirements as closely as possible.

Go to page: 1  2  3 

(1) Cognition refers to all the mental process involved in our relations with the environment
(2) The Francophone Ergonomics Society (SELF) was set up in 1963 to promote ergonomics in French-speaking countries
(3) Definition adopted by the board of the SELF and approved during the Paris conference in 1988
(4) Ergonomic recommendations, web design guide…
(5) Jakob Nielsen is a leading expert on web ergonomics; his Alertbox column can be read on his site http://www.useit.com/
(6) American telephone company

Printer Friendly Version