Behavioral Prototype

This assignment will give you experience in using behavioral prototyping to explore a user interaction scenario. As discussed in class, this technique can be very effective in testing design assumptions for HCI applications when the actual technology is either not available or too expensive to develop during the design phase of a project.


VOCA is the ‘Voice Operated Cooking Assistant’, which is similar to Amazon Alexa but designed for cooking. It uses voice commands to receive assistance through cooking various recipes. In the scenario, the user asks VOCA to find recipe of (Nutella Cookies), VOCA reads recipe as user assembles ingredients and follows along.

To evaluate the effective of VOCA, we asked the following questions:

  • Are voice commands clear?
  • Can participant easily hear VOCA “speaking”?
  • Are VOCA’s controls intuitive?
  • Can users successfully and correctly follow a recipe through VOCA?
Setting up VOCA

Testing was implemented by:

  • Hooking up to a bluetooth speaker and hiding it under a long tube which resembled the Amazon Alexa
  • Wrapped the long tube in gold-colored paper for a futuristic and modern feel
  • Calling and leaving one phone in the room so that the operator could listen in and respond based on the user commands (live feedback)


For analysis, in addition to our own personal feedback, a critique session was run in class where we presented our video and prototype to fellow classmates. All of this feedback can be seen in the lists below.

What Worked Well:

  • The VOCA audio system was loud enough for participant to hear
  • VOCA was very responsive — user felt like there was no wait time between their commands and VOCA’s response

What Needed Improvement

  • Better preparation with the phone call method by reducing background noise and fixing noise feedback
  • The bluetooth speaker was oddly shaped and the gold VOCA cylinder didn’t stand well on its own
  • Provide additional visual and/or auditory feedback so that the prototype feels more realistic and engaging (i.e. blinking lights or beeping noises)
  • Better and more explicit explanations on how VOCA commands work
  • Julie said: consider pre-recording Neha’s voice instead of having it record-real time. Although it requires more work, it would’ve prevented voice errors. It looked like Brian noticed Neha said “minion” instead of “minutes.”
  • Julie said: consider reading the whole instruction before starting the step-by-step tasks so that users know what to expect in advance. This prevents situations where the user doesn’t realize they have to let their dough cool for 2 hours which may may prevent them from wanting to continue with the recipe.
  • Julie said: Consider shortening recipe instructions, especially for recipes with longer steps. For example, how would you shorten: “Place egg rolls into heated oil and fry, turning occasionally, until golden brown.”?

Conclusion About Design

Given that the participant was able to successfully and confidently make delicious Chewy Nutella Cookies using VOCA’s step-by-step instructions, we think that the design our prototype VOCA was largely effective. It was also fun being behind the curtain and to think about how other VUI’s may have been tested.

Like what you read? Give Alison Chiu a round of applause.

From a quick cheer to a standing ovation, clap to show how much you enjoyed this story.