Who is the Client?

LiteraSeed is a browser-based application that helps people communicate their medical symptoms to their doctors. 

LiteraSeed was founded after one of the co-founders’ young cousin tragically passed away due to language-based miscommunication. 

What did they want?

To improve LiteraSeed’s user experience for people with low-literacy in the English language. 

What did they think of the current product?

Previous Usability Test Reports

The clients were kind enough to hand us notes from previous usability walkthroughs. These users were refugees with low-literacy in English. We created an affinity map from these notes to identify patterns. The key insight we gleaned from this exercise was that the users were glad to have this survey as a framework to communicate with their doctors but also needed more from it. The survey was helpful, but in some cases it didn’t get to the point that the patients wanted to get across.

What did they need?

User Interviews

Through our interviews we saw a pattern- all three users were familiar with medical terms in a non-English language, but struggled with those terms in the English language, making them feel frustrated. We used WhatsApp audio calls to interview 3 recent refugees. LiteraSeed’s users are a diverse group- refugees, non-native English speakers, people with low digital and language literacy, the elderly and people with disabilities.

Cultural Understanding

Here’s an example of a breakdown in communication that one user shared with us. User 1 told the interpreter that he thought he had the jiggers (fleas). The interpreter’s notes spelled it out as chiggers (a mite) Small difference, one might think. Jiggers are fleas that can burrow underneath a victim’s skin. Chiggers are almost entirely harmless. One can only imagine what the doctor’s diagnosis/ prescription must’ve been.

Focusing on some key pain points

While the user research piece of this project can be its own multi-year project spanning cultures, sub-cultures, languages and regions; we had to focus on identifying actionable objectives for this sprint. The first step was to develop a persona to help us evaluate if our solutions were meeting goals, and if they were

The Persona

 For the purpose of this sprint, our persona was a 30-something woman, who speaks and reads broken English, but is fluent in another language. She is educated in this other language, and understands the importance of accurate communication with her doctor to get the right diagnosis. She runs a small business and sells handmade soap on Instagram, Facebook and accepts orders via WhatsApp. She prefers using Arabic as her primary language on all of these apps. She describes her experience at the doctor’s office as necessary, but boring and frustrating. However, she feels a sense of accomplishment once the doctor’s visits are concluded.

Survey Design- 20 Questions can’t always lead you to the right answer

Heuristic Evaluation showed that the survey provided very little user control and freedom. User interviews pointed out the fact that users thought that the survey didn’t provide for a space to add context to their concerns. Our hypothesis was that if questions were asked in a better way, users’ answers would be more useful to the care providers. We looked at competitors to see how they were tackling surveys and found that there were 2 key factors that the product missed-

1. User control and freedom- the ability to choose which questions to answer first.

2. Free-form questions- Leaving space for users to describe what they were feeling in their own words. 

Patient Engagement & Patient Experience

“Patient engagement refers to a patient’s willingness and ability to interact with a provider, while the patient experience refers to how the healthcare provider met (or failed to meet) the patient’s expectations throughout the healthcare interaction. ” Skedulo

While looking for ways to make this product more engaging, we looked at Duolingo. Duolingo overlaps with this product in two key areas- engaging experience, as well as a foreign language. We really liked the interactions that the app uses.

Culture v/s Language

Culture is the sum of values, traditions, food, art and people over generations. Language is a way of expressing oneself to other human beings. As time changes, culture evolves, but language may take a while to catch up. Can we then, expect a native speaker who is a living embodiment of culture to speak the same language as someone that learnt it in a different time, in a different space? 

“The difference between the right word and the almost right word is really a large matter — it’s the difference between lightning and a lightning bug” Mark Twain

We looked at Skype Translator and Google Translate, however, neither had the ability to detect cultural granularity. Upon discussion with the software engineer on the team, we found that adding a voice-to-text feature was infeasible at the moment.

Scalability

For this design sprint, we were only looking at the OB/GYN survey, meant for pregnant people. However, we wanted to provide the client with a design kit that they could use for other surveys. We were looking at providing colors, font styles, iconography and typical layouts for the phone version. 

TLDR

To summarize, the three main components of our problem statement were re-thinking the survey design, making it engaging for the users, overcoming the cultural barrier and providing scalability.

 While we had a handle on other aspects of the problem, we were still looking for resources to solve for low-literate users.  Our client Aziza, had a repository of research in this area going back a few years. She recommended a paper titled “Designing Mobile Interfaces for Novice and Low-Literacy Users”

The paper conducted usability tests on a banking app in low-literate users. Their design recommendations included these-

1. Provide local language support, both in text and audio.

2. Minimize hierarchical structures.

3. Avoid requiring non-numeric text input.

4. Avoid menus that require scrolling.

It was interesting to note that the apps we used for C & C analyses used most if not all of these suggestions.

Design Studio

In the ideation phase, we started sketching out ideas and decided to conduct a design studio with the stakeholders. This was probably the best decision our team took. The studio saw a bunch of interesting ideas that we debated.

What made the cut, and what didn’t?

We took the ideas from the design studio, fleshed them out a little more, and then sat down with the software engineer for a feasibility check. Based on time, available technological, and human resources, we trimmed the features down.

Survey Design

Easy Questions First-

The heuristic evaluation showed that the current flows lacked user control and freedom.

“This is the essence of intuitive heuristics: when faced with a difficult question, we often answer an easier one instead, usually without noticing the substitution,” Daniel Kahneman, Professor of Psychology.  

Freedom, but structured

The product has a linear structure, with little hierarchy. It was tempting to add hierarchy like traditional apps, however, we needed to maintain a balance since research showed that too much hierarchy did not work for users with low literacy. 

A card sort showed that the overall categories worked for most parts, but the questions within each of these categories needed some reorganization. Group similar things together- We used this principle to group quantitative questions together, questions about the baby separate from questions about the mom and so on.

Adding Cultural Nuance

Transliteration, not translation

The persona we were designing for had basic English vocabulary. Adding more context to the questions and to the symptoms themselves and simplifying medical jargon using everyday English would help the user understand and answer more accurately. The root of this idea came from our design studio. The engineer on the team had the vision to add video clips to improve the explanation.