Testing the Sonic Interface was a little trickier than testing the visual interface. I’ll be the first to admit I am not used to designing physical objects, and trying to figure out how to get two plastic shells to emit sound was a little tricky. Eventually, I got these two objects to fit speakers inside them that could project enough sound to test the soundboards I’d created with Farrago on users.
The non-human sonic interface tests actually went fine.
Stevi, the human-looking interface, needed some work. I realized that after about three tests. First, you could barely hear the voice recording. I re-recorded using better equipment (the mics and mixer I use for The Future Of, thanks Marcomm department!) and mixed with a lot more vocal treble. The hollow, hard plastic interior was adding a lot more deepness to the already pretty bass-y voice in the first recording, so this more feminine filter was WAY more audible. The script was also altered–instead of telling the user that they were being tested, Stevi became a character in the situation, asking their governor (the user) what choice they should make.
I took Stevi and the non-human interface to playtest Thursday, which was still a little loud. It was hard to get people to interact with Stevi–they did not like this uncanny figure asking them. I managed to get 5 user tests, all of whom chose the Prospect Theory-modeled choice.
Oy vey. I guess sonic interfaces are not particularly persuasive, or at least the ones I built are not. Again, physical objects are not my strong suit, so there may be a much more talented designer out there capable of tricking people into going again their instincts.