Conceptual Models & Affordances

April 12, 2009

Notes on a lecture given by co-creativity director Sean A.; based on a CS160 lecture

Conceptual Models

People have preconceived notions about the way things work; you see a doorknob, what do you do with it? You turn it, you don’t push it. You have an idea of how to use a remote control. These conceptual models, called “image schemas” in computational linguistics, are based off of your common experiences in the world. When something conflicts with a person’s conceptual models, they often have a delayed understanding of that thing. A good example of this is in mathematics. People are used to using infix notation, which looks like this:

(1 + 5)

When someone is presented with something like prefix notation, which is sometimes used in computer science, they take much longer to figure out what is going on. Look:

(+ 3 4)

Changing someone’s conceptual models is very difficult. Often times when designers attempt to do something that doesn’t fit someone’s conceptual models their designs don’t do well in the market. As cool as, say, a doorknob that you push in, might sound, it would inevitably  confuse and frustrate users.

Affordances

Affordances make up conceptual models, they are the perceived conceptualizations of how things are supposed to work.

Take these scissors for example:

Let’s assume that you’ve never touched a pair of scissors in your life. What does looking at the scissors tell you about its function? You might notice that it has

• handles, one bigger, one smaller
• sharp edges on the interior of the metal pieces
• a hinge

One’s basic assumptions about the object have to do with these observations. You would probably thing that the handles with holes are for your hand, that the blades are for cutting, and that the hinge shows you how the scissors move.

There’s more to it though, an important things to remember about affordances is that they aren’t simply features of intentional designs; that is, affordances also come from inherent features of an object: is it light, how big is it–is it hand sized?, is it aerodynamic?

In the case of something like a refrigerator, where the mechanisms for setting freezer and fridge temperatures may be difficult for a user to understand, something that designers do–or should do–is abstract complicated controls to fit conceptual models. This is where the importance of interfaces comes into play.

A designer will model their design to function (cool food, for example), and they design use a system image (an interface–controls, etc.) to allow for user interaction. One thing that designers cannot design is the user’s model of how something should work. So we have this relationship given below:

Design Model => System Image <=> User’s Model

The user cannot infer from the system image (always) the design model, or how a thing actually works. Because of this, if we do not have system images that resonate with user models, errors and frustration will happen. That isn’t to say that a user’s model is correct–often times they are wrong, because they are extracted from fragmentary evidence. Here are some important design principles to keep in mind:

Design Principles

1. Make Controls Visible

• Leave the guesswork out of it.

2. Make Sure That Mapping Is Clear

• Visual cues are often better than “elegance.”
• Functionality is often better than “elegance.” Example: The Onion’s Click-Wheel Laptop

3. Be Aware of Transfer Effects

• Previous experience with similar object will be applied to new experiences. Example: why we have the QWERTY layout on computers.

4. Provide Feedback

• You know how you always press the “WALK” button on crosswalks a million times? You probably want to make sure it registers your feedback. Newer crosswalk buttons that make a sound when you press them fix the problem.