Once the basic ideas behind user-centered design become clear, some guidelines may seem so obvious as to go without saying. For example, Jakob Nielsen writes, "Error messages should be expressed in plain language (no codes), precisely indicate the problem, and constructively suggest a solution." This advice applies to most types of information in an interface. A novice designer might think, "Of course--why would feedback or instructions be provided in any other way?" Giving users information that's timely and appropriate, however, even for very specific situations, can be harder than it seems.

We can see this in our everyday environments. Consider signs, which often play a comparable role to informational dialogs and error messages in software. Although the state of knowledge concerning correct wording, visual design, and placement of signs has had decades to evolve, almost everyone can think of examples of signs that are inadequate, inappropriate, or even deceptive. One of the classrooms in which I teach offers a vivid example for my students. On a solid wooden door at the front of the room, behind the lectern, a large sign reads, "Fire door. Do not block." The meaning is obvious, except that if you approach the door to see how it works, you find that there is no way to open it--no knob, lever, or metal plate to push on. Further, the hinges are on the inside of the door, so that it opens into the classroom. A bit of thought leads to the realization that the door is for the students in the chemistry lab adjacent; if a fire breaks out in the lab they can escape into the classroom and from there to the corridor and out of the building.

All well and good, but where does that leave students in the classroom? Imagine a fire alarm going off and the smell of smoke in the air. Students rush to the obvious exit but find that there's no way to open the door marked "Fire door", and that pushing on it is not the solution in any case. When I describe this scenario to my students in the classroom, it usually meets with some uneasy laughter, but also an appreciation for the pitfalls of information design. They take a few lessons from this example. First, messages are targeted at specific audiences, and messages must be appropriate for their audience. For the students in the chemistry lab, the sign would be perfectly appropriate if it were visible on the other side of the door, but for the students in the classroom, it is less than useless. Second, less information can sometimes actually improve a message: "Important: Do not block this door" would be sufficient in this case. This avoids drawing attention to the function of the door, functionality that is not aimed at those reading the sign. (Recently the sign was taken down from the door, which now stands blank. Though not an ideal solution it is at least a slight improvement.)

Other examples of poor signs are easy to find. My students write,

"On a country road near my town is a small bridge that goes over a creek. A sign on the bridge reads, "Warning: Bridge floods in heavy rain." I've noticed that when it does rain and the bridge floods over, the sign is submerged under the water. This makes it very hard to know that the bridge is out during a rain storm. Some drivers have even missed the bridge altogether in the rain at night."

"I have made three signs with magic marker for customers at the video store where I work. The first one was a 'heat death' sign made with bright colors and an illustration of fire, with the warning not to leave tapes in the car. The second sign explained how discounts work. Again, I used bright colors and made the important, attention-grabbing parts larger. The third sign, complete with examples, explained the difference between 'widescreen' and 'formatted to fit your TV.'
"I thought my signs were great because they were easy to read quickly and didn't include extra information. Much to my dismay, the owners decided to 'improve' my signs by making them on a computer and printing them out. They didn't make the new signs balanced, with some parts smaller and some larger, bright colors, etc. As a result, I noticed people NOT SEEING the signs any longer. I kept getting the questions that my homemade signs had been answering for them. It's true, I could just be bitter that my contribution went unnoticed and was tossed aside. But I felt the homemade ones really were easier to read than the ones made on the computer, and when the customers obviously aren't reading them anymore, well, that only confirms my belief."

These examples emphasize further lessons about the importance of visual design, layout, and proper placement of information. If relevant information is not available at the right place and time, it can go unnoticed in the course of the natural activities associated with a task.

In general, one of the most important interface design concepts that students can learn is that the usefulness of information is tightly connected with task context. This can be seen clearly in so-called interface wizards, in users are presented with choices individually or in small groups, accompanied by information about the result of possible actions. The same rules that govern the effectiveness of signs, of which we have only mentioned a few, apply to these sequential dialogs. Wizard dialogs can be specialized to different user populations, with complexity hidden from novice users but with more detailed decisions available to expert users (who in any case can use different, more flexible interface facilities to accomplish their tasks.) The information that users see at any point in their decision-making process can be restricted to the current context, with less immediately relevant data delayed until later or made accessible only on request. Because wizards give users a well-focused view of the components in a sequence of choices, visual design problems are much simplified, even if additional constraints on consistency must be taken into account. For user interfaces in other genres, such as fill-in forms and question-answer dialogs, the presentation of information can often be improved if designers are aware of basic rules that make everyday signs effective.