T. R. Girill
trgirill@acm.org
Technical Literacy Project
February, 2008

Handbook Table of Contents

What is Usability?

Usability is a strange word but an increasingly influential idea. Any product designed for a specific purpose (such as tools, home appliances, or computer software) can be judged for success by its usability. For some people, improving a product's usabilty is a full-time job. In 2007, Barbara Whitaker featured usability professionals in her New York Times employment column (Whitaker, 2007), where she passed along this explanation of their work from Microsoft's Eric Danas:

We bridge the gap between what [any] technology is capable of doing and what users want to achieve.

...[usability is] the extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency, and satisfaction in a specified context of use.

Breadth of Importance

The diversity of professions concerned with usability reveals the breadth of its importance. Among the endorsing organizations and websites are:

• Usability Professionals Association (www.upassoc.org)
  This cross-disciplinary group of "human factors" experts ranges from linguists to librarians and from anthropologists to engineers. The common interest that unites such different people is in empirical evidence about what technology users need and want, along with what design techniques most reliably meet those needs and wants. Besides its website, UPA also supports an annual conference and the Journal of Usability Studies.
• Usability.gov
  This informative website is sponsored by the U.S. Department of Health and Human Services. Perhaps surprisingly, it focuses not on the usability of medical tools or services but rather on how to design government agency websites for effective information delivery.
• World Usability Day (www.worldusabilityday.org)
  The goal of this annual event is "to ensure that the services and products important to life are easier to access and simplier to use." Software developers (and computer science undergraduates) are always major players. But the topical theme in 2007 was actually the usability of health care resources.
• HCI Bib (www.hcibib.org)
  Human-computer interaction (HCI) is a chronic concern for usability professionals because much hardware and software is needlessly hard to use well. This website offers a 39,000-item bibliography, with every citation annotated and categorized, on the physical, psychological, and social aspects of improving computer interfaces.

Usability's Complexity

Usability is complex is a very distinctive way: it is multi-dimensional. Several largely independent (although sometimes casually confused) strands comprise it.

Barbara Mirel devoted a whole book (Mirel, 2004) to explaining why

• usefulness (enabling user-relevant tasks), and
• convenience (easy exercise of available features)

• effectiveness (achieving intended goals, Mirel's usefulness),
• efficiency (performing with economy and grace), and
• satisfaction (meeting user expectations, matching user skills or limitations well).

• learnability (important if new users are common),
• memorabilty (if the gaps between use are often long), or
• adaptability (if the tool or product should handle diverse situations).

Text Usability

Usability is an important property of nonfiction, nonnarrative text, just as it is of consumer products, software, or medical devices. The readers of technical text are also users:

• They have goals or tasks in whose context they consult the text.
• They judge the text (as they would any tool) by how well it helps them achieve those goals or tasks.
  • Do instructions (e.g., for extracting DNA from cheek cells) help get the job done?
  • Do descriptions (e.g., of bone fracture mechanisms) promote better understanding and far-transfer use?

Text usability has always been a background concern for writers and readers of technical prose. The computer boom of the 1970s, however, brought it to the foreground. For example, IBM was so often criticized for its unhelpful computer documentation that the company chartered a group of engineers at its Santa Teresa (California) Laboratory to explore how to improve IBM technical publications. The result was a 27-page internal booklet called Ease of Use (Ease of Use Study Group, 1979), which concluded that:

• IBM software was often useless without adequate explanatory publications.
• Publication "ease of use" in turn had three independent and vital aspects (p. 2). Technical information must be:
  • Easy to understand,
  • Easy to find, and
  • Task sufficient.
• This was best achieved by pointing out these key text features to everyone who worked on IBM documentation and by deploying human editors who revised draft text with these three ease-of-use factors in mind.

Cultivating such technical ease of use might seem obvious until one compares this text design strategy with the advice often found in technical writing books of the same period. For instance, consider Henreitta Tichy's Effective Writing for Engineers, Managers, Scientists (first edition 1966, second edition 1988). This was a very influential work, widely used in professional development courses for chemists or engineers. Yet despite 'effective' in its title, this book never mentions usability (or ease of use) in its detailed index. Tichy's very typical stress was on sentence-level style improvements (grammar and word choice). The book largely ignores "task analysis" and hence improving a technical text by overtly pursuing the three dimensions of usability championed above by the ease-of-use engineers.

Text usability's more recent influence shows in a different professional reference work that appeared a decade after Tichy's second edition, namely Developing Quality Technical Information by Gretchen Hargis (1998). Although this work has a general, comprehensive title, it focuses largely on computer (software) documentation. This is perhaps unsurprising because it was copublished by IBM, along with Prentice Hall. The 1979 ease-of-use themes shape all the topics and examples throughout this book. Indeed, 'ease of use' has replaced 'task sufficiency' as the label for the third essential usability (or quality) feature here. At 311 pages, this treatment is 10 times longer than IBM's pioneering pamphlet, but it stresses quite the same view that usability is the proper measure of success for all technical text.

Text Usability and Teaching

Explicitly introducing students to usability when you teach technical writing helps them in two ways:

• Exposing this as a writing goal is part of "revealing the magic" behind crafting helpful nonfiction nonnarrative text, a teaching strategy more fully explained in the cognitive apprenticeship section later. Students can only draft or revise toward (more) usable prose if they are aware of what this aim means in specific detail.
• Usability talk provides a practical vocabulary for "metacognitive discourse," for students evaluating the adequacy of their own text and that of others. Educator Mike Sharples has noted (Sharples, 1999) that most adults simply lack any vocabulary for discussing their own writing strengths and weaknesses.

This diagram visually reveals the three dimensions, or independent aspects, of basic text usability, in the long "ease of use" tradition sumarized above:

Easy to Understand

This is the most familiar usability aspect by far. Everyone has experienced struggling with technical prose that they couldn't understand (including students with inappropriate textbooks). To be understandable, text must be:

1. Clear.
   Words or phrases that could easily mean different things to different readers (ambiguity) or that don't seem to have any clear meaning (vagueness) make text needlessly confusing for everyone and impossible for ESL readers.
2. Concrete.
   Technical text often discusses abstract topics (sometimes by means of mathematical expressions). But even such passages can be made more understandable with the help of realistic cases, models, analogies, or tangible examples.
3. Well Styled.
   This may seem like a literary issue encroaching on science prose. Actually, however, there are fairly simple heuristics (rules of thumb) for improving the style of technical text (see Bennett and Gorovitz, 1997; for example, prefer shorter Anglo-Saxon words to longer Latin-based ones when possible). And scientists who read English-language articles but who work primarily in another language complain often about the hurdles that careless styling imposes on their understanding (e.g., Montesi and Urdiciain, 2005).

Relevant

Writers sometimes forget that even if a passage is easy to understand, it is of little use to readers for which the content is irrelevant. Usable technical text anticipates the information needs (both scope and level) of the likely audience (a good cake recipe won't help those who want to cook a chicken). To be relevant for readers, text must be:

1. Task Oriented.
   Most people read technical prose with an eye toward some planned future activity (perhaps specific and physical, like operating a fork lift; perhaps general and cognitive, like designing their next research experiment). Text oriented to the prerequisites, demands, and sequences of their likely tasks is more helpful than text that ignores those tasks.
2. Accurate.
   Readers who rely on text for factual information expect tested procedures and reliable scientific claims and relationships.
3. Complete.
   Balance is the key to usable completeness. Omitting needed steps or vital connections makes text unhelpful, even dangerous. Yet smothering those steps or connections in a flood of overwhelming or confusing detail undermines usabilty in another way. Relevant text omits nothing needed without including distracting stray information.

Easy to Find

This is the most overlooked aspect of text usabilty, and novice writers often neglect it. Yet even hightly relevant text that is easy to understand is useless without effective access features. Accessibility really is a third, independent dimension of usability for science prose. Easy to find text has three characteristics (which, when they perform well, often don't call attention to themselves):

1. Organized.
   Text structure is important even in story telling. For nonfiction nonnarrative text the structure may be less familiar to some readers (a problem-solution sequence or a hierarchy, for example), but revealing it is crucial for reader success. All writing involves planning, and for technical text planning the structure is as important as planning the content.
2. Retrievable.
   Good text structure helps readers only if they become aware of it through navigation aids planned and installed by the writer (or by a subsequent editor). To help readers find relevant parts of a long or complex text, writers need to pay attention to:
   • the signals that they provide.
     Parallel, overt lists of related items are one familiar text signal. Hierarchical section headings and subheadings are another. (These never appear in novels because findabilty is not a goal of novelists nor a concern for most novel readers.)
   • the chunk size of their passages.
     Long blocks of undivided content can make retrieving specific answers to specific questions difficult.
   • their access vocabulary.
     Besides the words within the normal text, writers must also consider supporting alternative, synonymous terms in indexes or other look-up aids.
3. Visually Effective.
   Technical text often looks quite different than most prose fiction or even most newspaper stories (Bernhardt, 1986). Including appropriate pictures, drawings (e.g., of equipment), or data graphs is one aspect of making science prose visually effective. Integrating the text and graphics so that they truly complement each other's information delivery (with callouts and captions, for instance) is vital too. And managing data density using visual features to make text data rich without overwhelming readers is a constant challenge (e.g., see Tufte, 1983).

References

Bennett, Jonathan and Gorovitz, Samuel. (1997).

Improving academic writing. Teaching Philosophy, 20(2), 105-120.

Berhardt, Stephen. (1986).

Seeing the text. College Composition and Communication, 37(1), 66-78, reprinted in ACM Journal of Computer Documentation, 16(3), 3-16 (1992).

Ease-of-use Study Group. (1979).

Ease of Use. San Jose, CA: IBM Santa Teresa Laboratory.

Hargis, Gretchen, et al. (1998).

Developing Quality Technical Information. New York: IBM and Prentice Hall.

ISO. (1996).

Ergonomic Requirements for Office Work with Video Terminals. Geneva, Switzerland : International Standards Organization.

Mirel, Barbara. (2004).

Interaction Design for Complex Problem Solving. San Francisco: Morgan Kaufmann.

Montesi, Michela and Urdiciain, Blanca. (2005).

Abstracts: problems classified from the user perspective. Journal of Information Science, 31(8), 515-526.

Sharples, Mike. (1999).

How We Write. London: Routledge.

Tichy, Henrietta. (1966, 1988).

Effective Writing for Engineers, Managers, Scientists. New York: John Wiley.

Tufte, Edward. (1983).

The Visual Display of Quantitative Information. Cheshire, CT: Graphics Press.

Whitaker, Barbara. (2007).

Technology's untanglers: they make it really work. New York Times, July 8, 2007. Available online at: http://www.nytimes.com/2007/07/08/business/yourmoney/08starts.html