Behavioral Prototype - Wizard of Oz

For our sixth assignment in HCDE 451: UX Prototyping Techniques, we explored the use of a "Wizard of Oz" prototype to test the validity of a high-tech design solution. The assignment required us to explore an advanced user interaction scenario by making a "fake" version of the product to test UI decisions. Often times, solutions that arise from ideation are difficult to test due to their technology requirements; prototyping high-tech solutions quickly is more difficult than low-tech solutions such as a mobile app (paper prototype) or a general product concept (video prototype).

My group of four HCDE students decided to test motion gestures for the iPhone game Temple Run. In order to do so, we needed to convince users that they were actually controlling an iPhone game with motion gestures. We each played a role in the testing session: a moderator, an observer, a camera-person, and a "wizard" (the one secretly controlling the gestures). We set up an Apple TV to a flat screen and propped an iPhone up with a camera pointed at the participant. We explained to participants that the iPhone was using an advanced motion capture technology to track gestures and map them to touch/tap/swipe controls. In reality, the iPhone (see the upright purple iPhone in the photo to the right) was in sleep mode. Meanwhile, the "wizard" was using the real iPhone to Airplay Temple Run to the Apple TV

IMG_20180212_110739.jpg
IMG_20180212_111209.jpg

Being an avid Temple Runner myself, I played the role of the "wizard" and secretly played the Temple Run game while responding to the participant's movements (pictured to the left). Through a combination of salesmanship, dummy technology, and practice, we successfully convinced all three participants that they were controlling Temple Run with their motion gestures.

In doing so, we were able to receive real feedback about the motion gestures themselves-- without actually having to develop any motion tracking technology!

What we learned

Being our first experience with a "Wizard of Oz" prototype, we were bound to make a few mistakes. Overall, I think we were successful in accomplishing our main goal: to test the validity of our motion gestures. We also did well to set up the testing room and develop a rough script to ensure we covered all of our bases and convinced the participants that the technology really worked! Here's what we could have improved:

  • document the entire process better;

  • record audio on participants for better feedback;

  • have a more rigid script, protocol, and follow-up questions;

  • deliver the "news" of the fake tech in a more approachable way (we often felt very bad for "fooling" participants and we wished we came up with a better plan for revealing the wizardry);

  • test multiple gestures to find the "best" ones (A/B testing);

  • set up the room so the iPhone and Apple TV are closer to prevent lag;

Overall, we had an absolute blast and got some great feedback from users, including a few being upset that they couldn't actually play Temple Run this way! I can't wait to try out my next Wizard of Oz prototype!

Nathan Young1 Comment