Designing for Usability

Steve Krug explains how we can create better Websites if we just stop wasting time on useless debates within teams and engage in ‘usability testing’ early on in the development process. His book provides the antidote.

He distinguishes between designers and developers in what they consider to be a ‘good’ website design. Designers prefer pleasant layouts, whereas developers enjoy a good amount of features. This is also compared to the duality between commercial culture and craft culture. Which is a question of aesthetics versus usability. Ideally, we would want to have both.

Krug notes that there is no such thing as an “average user”: every user behaves differently and the more you observe and test users the more it becomes clear that users are all different from one another. “All Web use is basically idiosyncratic. […] Good design […] takes this complexity into account.” (p.128) The recurrent mistake is to assume that users alike or even that users will think like you(!). To create a site that ‘works’ Web teams should test the usability of their interfaces by observing how people understand the concept and purpose behind the site and how easily they manage to complete specific tasks. This is why considering the context in which users experience a site is very important as well. Usability testing highlights what works and what doesn’t work with a site. Often it is through testing that the most obvious and important flaws are brought to designers and developers’ attentions, as it reveals user intentions, motivations, perceptions, and responses, which give insights into whether people can use the site.

Moreover, Krug emphasizes the need for ‘usability testing’ in the early stages of designing a site, as it will help teams get rid of underlying issues once and for all (and it saves time!). For him, ‘Testing is an iterative process. […] You make something, test it, fix it, and test it again.’ (p.135) In other words, KEEP TESTING until you are left with a site that users can actually make something of.

The author then prescribes a set of prescriptions for doing your own testing when on a low budget. He proposes that: testers should avoid divulging a site’s content to testees, make sure they think out loud, document both the screen-actions and the testees (using a camera and a screen recorder), ‘usability testings’ can be held once a month, in an office room or any room with a computer, 3 or 4 candidates who have basic knowledge of the Web are a good number for every test made during the development process of a site, offering stipends ($50 to $100 per user) shows that you care about their opinions, and that this procedure can be made in a morning and debriefed with the team during lunch hour. It’s not rocket science!

However, he points out that while ‘usability testing’ is important, there are problems users will encounter that will not be essential to focus on (such as ‘why doesn’t this site have “…” service?’) and other problems that will need immediate fixing. Also, it’s important to design for you audience and others, for experts and beginners, which is why clarity of information (i.e. clarity of the wording used throughout the site) becomes significant when ‘usability testing’ is made on virtually ‘any’ user. He names 2 types responses to focus on while testing: “Get it” testing (do users get it?) and Key task testing (were they able to perform a task? and how well did they do?) (p.144).

What often happens once a few determined problems have been fixed is that new (or hidden) problems come to the foreground, which explains why testing must be made often.

Although I found the text a bit redundant, I must concur that it will be difficult to forget his advice: “Recruit loosely and grade on a curve.” (p.139)

After having been asked “What do you think is most important about interface?” Don Norman responded that interfaces are NOT the place to start, but instead explained that design happens during ideation (or conceptual phase), and that design is a collaboration of different fields of knowledge working together to create sensitive and effective tools: tools that respond to user needs.

Norman stressed the importance of understanding the logic behind the act of designing, which includes considerations for the human factor. In this interview he sets ground rules for ‘good design’ and focuses on the idea that design should be inherently ‘humane’; which is to say that technological tools need to be thought of with a user-centered approach. “Cultivate Sensitivity to Design” explains how experimenting (observing, testing, etc.) helps designers reach a place of “empathy” for the user and simplifies complex products or systems in accordance to audience feedback. So, Norman proposes a shift in priorities. The priority is to provide appropriate tools for people that consider the whole system, or the ‘big picture.’ Design should be task-specific.

He suggests that human-centeredness is attainable in design through collaborative work. Involving people from diverse specializations (cognitive scientists, industrial designers, anthropologists, psychologists, sociologists, etc.) to collaborate in the ideation process can create designs that understand user needs and expectations, and responds to user habits.

In short, having the knowledge background and perfecting tools through prototyping, testing, and experimenting, are primers to any good design creation. In fact, Norman recommends that we invent a new programme of study wherein “each member [would be] trained in design, cognitive science, and programming.”

What I retained from this interview is that: to be a designer is to have the necessary knowledge to create meaningful tools for society. ‘Good design’ requires social-research methodologies and involves a back-and-forth dialogue between users and designers, testing and prototyping (re-designing). ‘Good design’ is also design that understands human beings; that is, design that is sensitive and that follows natural human behaviour and understands both needs and expectations. To be a designer today is to be able to open up to various fields of knowledge and adapting design to changed contexts and audiences.

In “Interacting with Paper,” Bill Buxton proposes paper prototyping which he calls ‘paper interfaces’ as an alternative and a better solution to testing user experiences before starting the design process. The author suggests that paper holds powerful affordances that represent systems of control that users respond to and experience, both in terms of a sketch’s control over users’ actions (control commands) as well as its faculty to lead users to focus on the ‘experience’ and ‘usability’ of an interface rather than on its aesthetic feel or design. This ‘low-tech’ method is useful for finding flaws and understanding the mechanism of an interface during the ideation process.

Buxton walks the reader through the process of sketching user experiences. With paper prototyping, interfaces are rough sketches that explain the ideation phase which helps test-users provide appropriate feedback with no misunderstanding as to whether form follows function or the reverse. Buxton introduces 3 different instances of user testing: paper prototyping (static), paper prototyping (dynamic, here testers replace the next sketch responding to user action), and on screen sketches (also dynamic, using Director, Flash, etc.).

Paper prototyping can be used for: user presentation of concept, informal testing, and usability testing. Buxton insists on the importance of understanding the What, How, When, Where, and Why of sketching user experiences for: ease of use, forgiveness of errors, and aesthetics.

Since a single design can often be a barrier to what a user experiences and how he decides to respond, and to draw the line between design and usability engineering, Buxton offers pointers when testing for incremental improvement:

– know that you “may not be able to trust [user] ratings”;
– present users with a low-cost (non-drafted) interface;
– provide and explore alternative layouts of a same interface;
– test a group of users for all those sketches and other groups for individual sketches, then compare the results; and
– get the design right and the right design.

“What is Usability?”

Patrick Jordan defines the concept of ‘usability’ in regards to product development and interface design. Based on the International Standards Organisation’s categories of “effectiveness, efficiency and satisfaction,” the author determines ways in which usability can be quantified. ‘Effectiveness’ is the capacity for a product to generate (or enable) output; ‘Efficiency’ is measured as the level of effort invested in completing an action or task (for example, the author categorizes usability errors as distinguished between a ‘slip’ and a ‘mistake’ in user performance and experience. For him, a slip is when a user accidentally performs the wrong action which is readily corrected by the user, whereas a mistake is when a user thinks he is doing the right thing (intuitive action), but is unable to perform his task. Errors can be minor, major, fatal, or catastrophic. Designing for Usability is to design products that ideally have very little errors or none.); and Satisfaction, as a separate level of usability, is understood as how comfortable or ‘at ease’ one feels when using a product, which underlies what has been referred as “user-friendliness” (subjective values). Those three considerations are important in guiding the design and features of affirmative usability.

In noting that ‘usability’ is now critical in determining whether or not a product is adopted or rejected, the author emphasizes audience ‘expectation’. Hence, the importance of user-centeredness in design. For the author, ‘usability’ is when effectiveness, efficiency and satisfaction are at their highest peak, and when those measures, then, determine the ease of use and “interaction between a product, user and task.” The author highlights key measures of user-centeredness; such as designing for: experience (first-time or frequent user experience determine a user’s ability to make connections between different tasks of a given product, or between different product of a same kind), domain knowledge (interface fluency or technological knowledge), cultural background (when design is tailored to respond to cultural behaviour and physical characteristics –such as, colour references and ethnic features– then design is able to make product features more intuitive or explicit), disability (considers users with special needs and aims at creating products that include and accommodate universal use –such as disabled toilets, etc.), and age and gender (takes into account generational differences in technological abilities and attitude in general).

Jordan also lists measurable aspects of design when thinking of user-performance: ‘Guessability’, ‘Learnability’, ‘Experienced User Performance’ (EUP), ‘System Potential’, and ‘Re-Usability’. ‘Guessability’ refers to how easily a task is understood at first-hand (which the author calls a cost). ‘Learnability’ refers to how easily it is for a user to ‘learn’ how to use a product’s features, which measures levels of user-performance (recognition, memorability, etc.). EUP (similar to ‘Learnability’) refers to the need for products to provide higher levels of performance for more skilled or experienced users. ‘System Potential’ refers to the need for products to allow access to a product’s maximum potential. Finally, ‘Re-Usability’ is concerned with users who might leave a system for a certain amount to time. Re-Usability measures the easiness with which a product’s tools are recalled.

“Principles of Usable Design.”

Here, the author presents “10 principles of usable design” that describe the different elements in design that affect usability: ‘Consistency’, ‘Compatibility’, ‘Consideration of User Resources’, ‘Feedback’, ‘Error Prevention and Recovery’, ‘User Control’, ‘Visual Clarity’, ‘Prioritisation of Functionality and Information’, ‘Appropriate Transfer of Technology’, and ‘Explicitness’.

‘Consistency’ enables users to generalize task requirements across different products or different tasks (“Inconsistencies are likely to lead to errors.”). ‘Compatibility’ is the potential of design to mean user understanding, expectation, and ability. ‘Consideration of User Resources’ guides design with regards to the bodily-language of hands, eyes, ears, etc. by using sound, visual cues, and tactile functions. ‘Feedback’ refers to systems that indicate an action has been made. ‘Error Prevention and Recovery’ facilitate the recovery from user errors by walking people through an exit path or by giving users the possibility to go back in time (i.e. ‘undo’). ‘User Control’ is enhanced by providing ‘adjustable’ or customizable features which in turn emphasize user-centeredness. ‘Visual Clarity’ considers scale, location and distance (i.e. buttons are labeled, interfaces as at proximal distances and located in convenient and readily accessible ways). ‘Prioritisation of Functionality and Information’ focuses on the hierarchy of information of tools (IA): this measure become subject to errors when products contain clustered information or features. ‘Appropriate Transfer of Technology’ takes into account the surrounding environment (context) of transferring or crossing different systems to enhance one or the other with regards to safety measures. Finally, ‘Explicitness’ coincides with Don Norman’s concept of ‘affordances’ wherein product displays provide perceptible clues as to what they are used for and how they operate.

This text offers a range of considerations for designing for usability which entails designing for user experience, user performance and safety, in terms of effectiveness, efficiency, and satisfaction. It helps designers focus on user-centernedness when thinking of usability and provides them with a vocabulary for investigating and measuring the success of their outputs.