The Design of Everyday Things: Full Blog

Reference:
The Design of Everyday Things
Donald A. Norman
1988 Basic Books

Summary:
In The Design of Everyday Things, Norman attempts to impress upon the reader how much the design of a device can lead to, or inhibit, understanding of it.  He argues for physical mapping whenever possible, as arbitrarily mapped controls require specialized knowledge instead of intuitive understanding.  What is more, in many instances a user will blame themselves for their lack of operating knowledge, so typically a designer won't receive negative feedback about poor designs.  As a result, when they do get complaints they often dismiss them as minority views.

When interacting with the world, people rely on cues to accomplish their tasks.  We learn what results from specific actions in specific environments, but when attempting the same action in an unfamiliar environment we tend to mess up.  This is a result of the fallibility of memory, so care must be taken in the design stage to make arbitrary knowledge easy to remember.  When the user has a good mental model, they have a better understanding of a system.  When the designer chooses an unconventional model, a task becomes much more difficult for someone to complete since they don't have the appropriate knowledge.  Ideally, labels should not be needed to explain a device.

In designing a device, the architect should try to reduce the possibility of errors.  Slips are unconsious errors, while mistakes result from reaching an incorrect conclusion.  One way to minimize both is through the use of forcing functions.  These functions behave in such a way that action B can't occur until action A is completed.  While these functions can be a nuisence to a user, when user properly they can prevent disastrous errors.

Convention and aesthetics must be taken into account during design.  When designing a device with many arbitrary mappings, like a keyboard, it should probably be manufactured with the conventional layout even though morre efficient arrangements exist.  It would simply be too costly for everyone to have to relearn how to type.  Aesthetics behave in much the same way.  One design might look better than another, but an ugly functional device is better than a pretty unusable one.

Many things must be considered when designing something.  The designer must consider conceptual models, mapping, constraints, and standardization just to name a few.  The designer should focus on giving the user what they need, because as the functions of a device increase, so does the complexity of it.

Discussion:
While I agree with most of Norman's writing, he tends to get a bit... repetative... with his points.  In many cases it seemed like I was reading a variaton of the same idea over and over and over again.  Still, I suppose he made his opinions clear.  I certainly learned a thing or two from this book.  To me, some of the ideas seemed like simple common sense, but there were others that I hadn't considered before.  In the past, I have written programs and thought that the complexity was getting out of hand.  Now I have a better understanding of methods to prevent that from happening.