.svg)
.jpg)
How might we improve the user experience of a mobile phone survey focused on diet quality?
Pop up usability lab
Mobile phones are transforming how we collect data - faster, cheaper, and far more scalable than traditional in-person surveys.
In Rwanda, the International Institute for Tropical Agriculture (IITA) had brought the global Diet Quality Questionnaire (DQQ) directly to peoples' hands via USSD, a technology that enables simple, text-based interactions on mobile phones without requiring internet access.
Our challenge: Ensure people not only understand the questions but enjoy the experience when answering the survey. Because better experiences mean more user generating better data - and ultimately better opportunities to make data-based governance decisions for food-security.
Over two weeks in Rwanda, we took a hands-on approach - part of the time based at the IITA office in Kigali, and part of the time right inside respondents' homes. Our team blended local expertise and global design thinking: a human-centered design specialist, a local scientist deeply familiar with the tool, and four trained enumerators.
We tracked usability progress live through the platform provider, VIAMO, carefully monitoring data quality and drop-off rates before and after each design improvement.
1. Can respondents easily complete the questionnaire via USSD on their mobile phones?
2. Do survey respondents like to answer questions about their diet using their phones?
Qualitative methods are suited best to discover unkown unkowns - like peoples' perceptions of the task. Through the interviews, we realized that respondents were not aware of the importance of reporting their own consumption exactly for the prior day- they instead reported based on their general diet or the diet of their community.
The survey consists of a series of questions, each one with a short list of food items and the question asking whether the respondents ate these. When asked about whether any food item in a list was consumed, some respondents interpreted the question as asking whether they consumed all food items in the list - We expected this to influence data flaws.
“I cannot answer yes to something I didn’t consume - I ate tomatoes, but no eggplant”
* System errors & time outs in USSD impacted the user experience and frustrated people
* No correction or error prevention possible within the USSD system
* Challenges with comprehensibility of the questions around dietary habits
* Poor legibility of the long, repetitive questions on small screens
* No in-context help for the automated USSD protocol which appeared in English on peoples' phones
We advised to revisit the communication between the platform provider and the telecommunications company to reduce system errors. In addition, we supported IITA to clarify larified instructions on how to complete the survey (e.g.: outlining when to respond “yes”) to reduce data entry errors. The team decided to address the repetitive questioning by separating one framing question ("did you eat one of the following:...") from the many responses/ food options to choose from to avoid repetition
With the redesign, we helped people get through the survey! We made the survey more inclusive and we helped generate better data: Data quality (measured in low deviant responses) increased significantly.
.jpg)
