Smoke and mirrors - Wizard-of-Oz prototyping
In a recent blog post my colleague Marcus G highlighted the importance of making prototypes personalized and realistic. As an example of how to suspend disbelief he mentioned Wizard of Oz prototypes, where you have a human behind the scenes simulating the behavior of an assumed intelligent service. This method has traditionally been used to observe the use and effectiveness of a proposed user interface, although creative ways of including it in less usability-centric activities have been shown. The latter is also primarily how we have used Wizard of Oz in a number of projects when wanting to create stimuli material that serves as a common ground for feedback sessions and concept discussions. Depending on the concept there are of course many different ways to go about in order to present a realistic (but not entirely real) experience, but I will give two examples to illustrate how we have explored the use of the method.
The first example is a user study in which we wanted to understand how people think about and perceive a home network of connected devices. Networks are not only a very abstract area for most people, it is also rather complicated to implement something that actually works – especially if you have a number of conceptual ideas involving both physical and virtual artifacts. What we then did was basically to multiply the traditional Wizard of Oz setup and use separate devices (phones and computers) for controlling the five connected home devices (TV, computer, stereo, phone, central unit) that were part of our scenarios. This could potentially put a large pressure on the wizard (times 5 in this case) behind the one-way mirror. Therefore, we did not let the respondents interact with the artifacts. Instead we used a moderator that demonstrated the interaction sequences, carefully following a script that involved a delicate collaboration with the wizard. The result was a simulation of a holistic user experience that the respondents perceived as very realistic, generating spontaneous, emotional and very interesting reactions.
The second example is a demo of a visual communication and remote collaboration service. In this case the purpose was not to use Wizard of Oz prototypes as part of a user study, but rather as a starting point for a discussion around future strategies. This time, our approach was to create a short play in which interactive visualizations for the service concept was used as props. The actual video communication was solved through 10-meter long cables running to another room! The wizard was also hiding behind the scenes, but still being able to follow the events and time all the visualizations as the play progressed. And to give a realistic impression of a touch-based shared interactive space, we had created a virtual frame in mid-air (using strings not visible on the camera view) for the person behind the scenes. An interesting sidetrack for this example; in a technical organization where managers are used that everything in demos have actually been implemented, Wizard of Oz presentations can be quite entertaining :-)
There are of course a few obvious drawbacks with the method, like the lack of flexibility, the need to follow a linear script, and the difficulties for respondents in user studies to interact directly with the prototypes. Nevertheless, we have found that the immersive presentation of concepts can generate strong affective responses instead of only intellectual ones, and also be a very valuable starting point for discussions. For services, including several interconnected applications, devices, objects, and so on the method is particularly relevant as implementing these concepts as functioning prototypes would take a large amount of resources as well as time. And perhaps requiring all this before you know whether you are on the right track or not. In the two examples above, we could use off-the-shelf tools to create inspiring and realistic experiences of highly advanced services within weeks.