The Wizard of Oz is a wonderful approach to validate interactive products early in the design phase, basically before any of the interaction is actually working (technically). The Wizard of Oz approach is named after a character from the movie with the same name. In the movie, the main character named Dorothy embarks on a quest to find the Wizard of Oz, to help her find way back home. When she finally finds the wizard, she discovers he is actually a fake. All of his ‘wizardry’ is staged by machines which he operated from behind a curtain. The ‘Wizard of Oz’ method applies this fake wizardry to testing interactions. A ‘wizard’ operates controls to make the illusion of the existence of a fully functional product, or at least the interaction under investigation.
An effective way to demonstrate this principle and to test Wizard of Oz approach in a workshop, is by testing different forms and shapes of computer mice. Assume, for example, that the cup shown in the Fig. is a new kind of mouse. You can test this by having a ‘wizard’ copying with a real mouse each and every movement made with cup. The person moving the cup will ‘experience’ the cup as mouse (even if the control may not be perfect).
To make this a bit more challenging, you can use a wide range of objects. For example a photo camera, sun glasses, a toy fork-lift truck. Workshop participants first have to decide/design how to make the functions of a mouse on the object received. How would you map the functions of a mouse on a sunglasses? Or on a photo camera? Working in groups, they have to define and document the mapping. Also, they have to practice playing the wizard. Second, a random member of the workshop is given the task use the ‘mouse’, to find out how it needs to be operated, then use the mouse to make a specific drawing in a paint application (see Fig. below). The task of the wizard is to make sure the mouse reacts as previously defined.
What you see during these experiments are two things. First, with respect to playing the wizard, everybody immediately grasps the concept and all mouses turn into fully functional prototypes without problems. Second, in terms of mapping functionality, there emerge two approaches. One is the ‘it is just a mouse’ approach; independent of the type of object or it’s shape, the object is just a mouse meaning the user will place his/her hand on it and start moving it like a normal mouse, ignoring the specific features and functionality offered by the object. Second is the ‘it is an object’ approach, where the use of the object are mapped on the functions of the mouse. For example, you wear sunglasses, therefore sunglasses as a mouse are worn and you point the cursor by looking at a location. A camera is for pointing, therefore as a mouse you point and ‘shoot’ (see fig above).
The Fig. on the right shows the instructions of a mapping of a Rubic’s cube, where the ability to turn is mapped on the mouse functions. The test showed that the ability to turn was ignored by the subjects, who instead moved the cube as a whole.
Performing this exercise at the beginning of the workshop sometimes inspires adaptation during the prototyping part in the following days. A student who was tasked with designing an intrusive egg timer (see Fig.) showed a nice example of testing a rich interaction by simple mean. The egg timer was designed to emit an intensive light while producing an intruding sound. To validate the impact and effectiveness a prototype was created to include a lamp with a remote switch, and for the sound a mobile phone with a carefully selected ring tone. Invisible for the user, after the indicated time had passed, Francesca would call the mobile phone, and in the time the connection was established, switch the light on. In combination it created an impressive effect, even for the prepared/experienced user.