- GiveButtons Facebook Page
- GiveButtons Gift Ideas Facebook Page
- GiveButtons Google+ Page
- The Future of Shopping
- The Science of Gift Giving
Revising the Taxonomy of User Interface Terminology after 22 YearsFebruary 5, 2013
In 1990 I created a taxonomy of user interface terminology. A taxonomy of user interface functions had been developed by my friend Jim Carter a few years earlier, but it focused on user interface functions and I wanted something a little broader that included things like the cognitive engineering background to human computer interaction and user interface design. It’s been 22 years since the taxonomy was published and the taxonomy is getting a bit long in the tooth considering how much the world of computing, devices, and user interfaces has changed in the interim. I looked around and didn’t find a lot of similar taxonomies, although there is one taxonomy from 1992 that is somewhat related (I will review that taxonomy in more detail in a separate post).
Here’s a quick overview of the 1990 taxonomy that I developed. The development of the taxonomy began with the assumption that the relevant topics could be organized into a hierarchical representation. While a tangled network might ultimately be more appropriate the strategy was to start by forcing everything into a tree structure and then create associative links across the branches of the tree a s found necessary.
The first step in developing the hierarchy was the selection of a relevant set of terms that spanned the field of human-computer interaction. In the original (1990) version of the taxonomy, the terms were chosen from three sources: section headings in the handbook o f human-computer interaction (Helander, 1988) ; index terms from a well-known book on user interface design (Shneiderman, 1987), and lecture notes from courses on cognitive engineering and intelligent interfaces that I had taught at the University of Southern California. The terms that were initially selected were further refined based on the comments of various colleagues and reviewers.
A sorting task was then used to organize the selected terms into a hierarchy. The terms were each written on separate 3 x 5 cards . The cards were then shuffled and spread over a large table top. The cards were then grouped together according to which cards appeared to belong together. This was done without any (conscious) preconceived notion of what the organization should be. Once each group was formed it was collected together (tied with a rubber band) and then assigned a label. After all the cards had been placed in groups, the groups themselves were formed into groups. This agglomerative process was continued until all the subgroups (and the terms that they contained) had been collected into a single supergroup representing the root of the tree.
The results of the sorting process were then transcribed onto a large sheet of paper and edited. This editing included revising the names of terms and categories to fit i n with standard usage, and the expansion of terms within a group. This type of expansion was found to be much easier after a basic structure (taxonomy) had been defined. The revised taxonomy was then passed to members of the user interface and hypermedia groups at the Institute of Systems Science, National University of Singapore, for their review. Further revisions then led to the version of the taxonomy that was published in the SIGCHI Bulletin in 1990.
The taxonomy was designed to organize terms in human-computer interaction for use in research and instruction. Thus its aim and character was different from a taxonomy of user-oriented functions that had been developed earlier (Carter, 1986). Carter’s taxonomy was concerned with the standardization of user interface functions or activities. In this taxonomy, we are concerned with the standardization of terminology over the discipline of human-compute r interaction. Our interpretation of human-computer interaction is broadly defined, ranging from cognitive science issues underlying user behaviour to design guidelines and tools.
The Four Main Categories of Terms
The taxonomy that resulted from the process described above had four main branches, which are described in this section. A detailed analysis of each of these branches will be given in the following section.
The four main branches (categories of terms) of the taxonomy are as follows:
1. The Basic Interface Mode
2. Cognitive Engineering
3. User Interface Engineering
To give you a flavor of the taxonomy, here is an example of the main sections for the user interface engineering branch of the taxonomy.
And here are the sub-branches for interaction evaluation (section 3.4 under user interface engineering).
In this taxonomy, the basic interface model is a simple characterization of the major components of all interfaces. In this model, a transaction begins with a (task-related) goal in the mind of the user. This then leads to a user behaviour (such as pointing with a pointing device and similar behaviours ) which occurs in the context defined by the current status of the task and the computer system. The behaviour and the context in which it occurs then jointly define an action that is carried out by the system (such as retrieving a file, or presenting requested information on the screen). Displays consist of information that is presented to the user for it s own sake, i.e, to be read or listened to. Effects are outputs from the system that are designed to assist users in interpreting the system’s actions and updating their mental model of system and task status. Forms are consistent models in which actions, effects, and displays are embedded. They generally conform to the notion of interface metaphors.
The basic interface model contains declarative knowledge about what user interfaces consist of. The second main category is cognitive engineering, which covers the areas of applied cognitive science that are relevant to understanding human-computer interaction.
The user interface engineering category includes subtopics that are relevant to the interface engineering process, without making any commitments as to what the (procedural) details of that process might be. While there have been numerous books and papers on the software engineering process, and on processes of design in general, the user interface design and implementation process was less well defined when the original version of the taxonomy was created in 1990. For instance, the overall process of user interface design was not directly addressed by Hartson and Hix (1989) in their lengthy review of human-computer interface development.
The fourth main category deals with general classes of application to which user interface design is directed. It is difficult to discuss broad issues in user interface design without regard to the type of application being considered. In addition, the type of interface often has a strong influence on the type of user interface that is developed and the general issues that are considered. For instance, user navigation issues tend to predominate in information technology interfaces, while cursor movement and text selection tends to be particularly important in applications such as word processing.
Each of the four main branches of the taxonomy is potentially controversial. For instance, there may be other ways of defining a basic interface model, or other topics in applied cognitive science that should take precedence over the topics selected in the taxonomy. However, while the details of each branch may be debatable, it seems useful to distinguish between the broad issues represented by each branch. The major impact of this taxonomy is that it distinguishes between the basic interface model (the what) and the interface engineering process (the how), and it also recognizes the importance of cognitive engineering as a foundation for human-computer interaction. In addition , the heterogeneity of user interfaces across different type s of application is also recognized.
Now, twenty-two years later, I feel that the taxonomy has withstood the test of time reasonably well. But perhaps it is time to produce a revised version. I don’t know how long this will take me, but your comments would be very welcome.
Leave a Reply