Product Designer
UWConversationalSystem3.jpg

UW Prosocial Computing Lab -- Designing a conversational system for data-driven reflection

 

UW Prosocial Computing Lab

Designing a conversational system for data-driven reflection

 

Overview

As a member of the research team, I translated participatory user research data into conversational scripts, designed surveys, and analyzed user engagement and interview data to inform the design of a mobile conversational system. Our research findings will be published and presented at UbiComp 2018.

 

Problem & Opportunity

Fitness trackers (e.g., Fitbit) collect large amounts of user data. However, learning from and acting upon that data remains a challenge for many people.

Design Challenge

How might we help people ultimately act upon the data collected from their fitness trackers?

Solution

We designed a mobile conversational system that helps users reflect upon their physical activity, with the hypothesis that reflection leads to learning and action. We designed 25 conversational scripts that Fitbit users can interact with using SMS/MMS. At the end of a 2-week field study with 33 participants, we found that engagement with the system increased motivation and the adoption of new health behaviors.

My Role

Research assistant to Professor Gary Hsieh and PhD student Rafal Kocielnik

Methods: Survey design, qualitative coding & data analysis

Timeline: March 2017 - September 2017

 

Impact

I'm excited to announce that I'll be second author in a peer-reviewed paper at UbiComp 2018.

 

Sample Dialogues

 Design by Rafal Kocielnik; I provided research support during the design process

Design by Rafal Kocielnik; I provided research support during the design process