eLIFE / Social Network Based Digital Support for Individuals with Autism

DURATION/ Aug, 2018 - May, 2019

DELIVERABLES / Concept Prototype, Internal & Client Presentation, Research Documentation

METHODS/ Ethnography, Concept Probe, Questionnaire, Concept & Usability Testing

CLIENT/ Emory Autism Center

PEOPLE / Charlotte Lou (Design Researcher & Product Designer), Rosa Arriaga (Product Manager)

OUTCOME / Handed over for user validation & In proceeding of CSCW 2019 ( Download )



Autism has been a cause I care deeply about (see my research proposal from my study in cognitive psychology). During my time at Georgia Tech, I had the opportunity to join Dr. Rosa Arriaga to further their work on a trusted stranger network (TSN) concept for autism.

In the following 9 months, I worked as a solo design researcher and designer in partnership with our pro bono client Emory Autism Center to validate the concept in context.

Rectangle 376f.png

The challenges that come with autism are prevalent and severe.

Autism is a spectrum disorder that is mainly characterized by social impairment, which exhibited itself in having a small social circle, over-reliance on immediate caretakers, difficulty in socializing, and a large variance in capabilities across the spectrum and  individuals.

Prevalence: 1 in 59 in the US affected by the condition

Severity: Main impairment is social in nature, and social communication affects most aspects of life.


Resources are here, but not enough

Structured in-person learning

Class sessions that teach social and independent skills ranging from social etiquette to healthy eating.The network of people, including teachers and volunteers, that facilitates the practice of those skills.

But in-person training is limited by a usually once-in-a-week learning schedule and in-person only network contact.

Trusted Stranger Network (TSN) 

They are specialized social network that allows the individuals to crowdsource information or advice based on specific needsThey also support the individuals to expand the individuals’ social network.

But the utility of TSN is limited to those who are already comfortable interacting with technology. Plus, qualified volunteers are hard to recruit and incentivize. 


Client saw the potential of a hybrid solution to support continuous learning.

A Mutually Facilitative Solution

The in-person learning program supplies the existing pedagogical and social resources in the form of class content and volunteer network.

The digital network enables the individuals to continuously access and reinforce learning in their daily life.


Sounds Exciting! But the challenge and this hybrid solution need to be validated in context.

2-3 – 1.png

Research Overview

Q: How can a trusted stranger network be leveraged to support continuous learning of social communication skills of high functioning adults with autism in the context of eCenter?


In the initial meeting, the eCenter stakeholders, product manager, and I converged on the following RESEARCH STRATEGY.

▶︎ Vision

A trusted-network based solution to support the learning transfer from classroom to adults’ day-to-day life context. 

▶︎ Strategy

Design Ethnography/volunteer immersion to get an insider’s perspective on the day-to-day needs and challenges of the adults. 

I then decided on the RESEARCH TACTICS, given the strategy as well as resources and constraints.

▶︎ Tactics

OBJECTIVE 1/ To understand the context (the support network & the learning curriculum) >>> Ethnography

OBJECTIVE 2 / To validate the problem (the autism condition & the learning context) >>> Ethnography

OBJECTIVE 3 / To validate the solution (that the TSN concept is ecologically feasible and effective >>> Concept Probe

OBJECTIVE 4 / To gather specific requirements for the TSN solution (user & stakeholders ) >>> Questionnaire

▶︎ Users

Adults with high functioning autism at Emory Autism Center aged 24 - 35, currently enrolled in structured social and independence learning Groups.


Data Collection


Ethnography / Objective 1 &2: Understand Context & Validate Problem

I carried out the ethnography wearing two hats - a design researcher and a fellow volunteer. I was able to make the most of this double perspective to proactively generate more contextual insights about user needs as I interacted with the adults through

  • Participatory Observation (to fully observe behaviors)

  • Contextual Interview (to clarify & deepen understanding of observed behaviors)


The Support Network

  • Primary support network mainly consists of family members, teachers, therapist and volunteers.

  • Teaching staff is overburdened

  • Volunteers are the primary source of peer interaction for most individuals.

  • Bigger life context is not mapped.


The Learning Curriculum

  • Mainly consists of small group sessions and some individual sessions.

  • Topics range from independent living to social communication skills, with social engagement being the main purpose of learning.

  • Volunteers are usually paired up with an adult to (1) aid learning and (2) serve as conversation partners.

Screen Shot 2019-09-26 at 7.51.02 AM.png

Data Synthesis

I organized the data collected into 4 themes that shed light on both the user and contextual challenges:

  • Individual difference

  • Technology use

  • Current learning

  • Current teaching

The Master Board copy.jpg

▶︎ Results

The Autism Condition

The Learning Context


▶︎ Insights

  • Design should accommodate for a wide range of capabilities.

  • Design should better distribute the workload - lesson burden of teacher & better engage users and volunteers.

  • Technology does not seem to be an integral part of the structured learning.

▶︎ Outcomes

  • Network and learning contexts mapped out - Objective 1 ✔︎

  • User needs & contextual problem validated - Objective 2 ✔︎

  • Additional insights identified

▶︎ Decision

  • Need to validate the feasibility of a technological solution by digging deeper to understand

    • IF and HOW users use technology/social network

    • How stakeholders prefer to position this technology.



2. Concept Probe/ Objective 3: Validate solution

I conducted the session in the style of a class, with 4 participants, volunteers and teachers present - to directly gauge if message-based interaction matches users’ cognitive, linguistic and technological capability in order to validate the TSN concept.

  • To test for technological capability, I decided to have each participant verbally share if/how often they use social media app and what social media app they use.

  • To test for cognitive and linguistic capability, I decided to use a paper & pen Q&A session.

▶︎ Results

  • 2 out of 4 participants breezed through both tasks; the other 2 expressed that they have never used social media.

  • The latter 2 participants managed to complete both tasks with much guidance from teachers and interns on grammar and context setting.

  • Facebook and Youtube styled interaction are most familiar to and preferred by users.

▶︎ Insights

  • The solution is technologically accessible to users, though guidance/scaffolding is needed.

  • Scaffolding is needed with grammar and setting context.

  • The solution should leverage user’s familiarity with SNSs such as Facebook and Youtube.

▶︎ Outcome/Decision

  • YES to concept - Objective 3 ✔︎



3. Questionnaire/ Objective 4: Clarify Needs

I sent out a Google Form questionnaire to all adults at eCenter. I heard back from 9, from different groups and different capability levels.

  • Understand the generalizability of the insights so far across a broader user base at the e-Center.

  • Understand specific needs with regard to social media - Why they use social media for, What type of interaction they prefer, Who they interact with, Where they use social media on, How they prefer to use it.

▶︎ Results


▶︎ Insights

  • Mobile as a pilot platform

  • The solution should accommodate both the passive and active styles of interaction

  • Key Stakeholders Requirements > The solution should engage the in-person trusted volunteer in ways that extend the in-person social learning

▶︎ Outcome

  • Specific Needs collected - Objective 4 ✔︎

▶︎ Decision

  • Time to put all the pieces together!


Data Synthesis

Data Mapping

Data Mapping

I did a manual data mapping to:

  • Distill insights by triangulating and cross-validating data from different sources and/or different stages in the research process.

  • Document the how decisions are made and supported by data at different levels, from links to raw data, to summary, insights, and design recommendations.

Persona Spectrum

Persona Spectrum

I created two personas to capture the two most prominent behavioral bracket.

  • The Consumer - less engaged passive interactionist.

  • The Creator - more engaged active interactionist.

Screen Shot 2019-09-18 at 2.21.33 PM.png
Screen Shot 2019-09-18 at 2.21.41 PM.png

Design Guidelines

The overarching goal of the platform is to extend and augment in-person learning. Three specific guidelines emerged:

1. Create scaffolding to ensure equitable access across all capability levels.

This guideline stems from the observation that linguistic, cognitive, social, and technological capabilities vary greatly among individuals with ASD.

For example, some need support browsing video content, yet others comfortably share their stories of autism on social media.

2. Channel appropriate learning to appropriate sources or audience.

This guideline stems from the observation that the learning of social and independence skills and the practice of these skills usually happen through different channels and with different audiences.

For example, individuals learn how to apply deodorant as part of a personal hygiene session led by a teacher; they practice maintaining a conversation through informal chats with volunteers.

3. Create incentives for individuals with ASD to use the system.

This guideline stems from the observation that most individuals are passionate about one particular topic and are excited about spending time with volunteers, both of which can be leveraged to incentivize learning.

For example, one individual has a keen interest in anime, so a volunteer explains how to engage in small talk in the context of an anime conference.

Rectangle 380.png

The Design Process

I went through feature prioritization, feature-context mapping, information architecture design, to arrive at the final prototype eLIFE.

Mini Journey
Information Architecture

The eLIFE Prototype

The prototype consists of four key features and additional scaffolding features, each answering one or more design guidelines.Together, they mirror the two parts of in-person learning - the learning of knowledge and skills, and the social practice of these skills.

Main Navigation

The learning of knowledge and skills is supported by class learning content and social sharing content - MyHome page and My Topics page.

The social practice of the knowledge and skills is realized through two network-based features - My people page and My Nudges page.

Additional Features

Two scaffolding features are incorporated to accommodate the variability in user capabilities- Informational Banner and Volunteer Review page

Additional Features includes Email Notification to reach those who don’t use a smart phone as well as a standard SNS user profile page


Concept Testing


To validate the concept, I evaluated eLIFE on 2 metrics - concept effectiveness and high level usability. As capability is the most critical factor in defining possible user behaviors in interacting with the system, I conducted the session using two capability-based scenarios.

Either scenario corresponds to the needs and capabilities of the Consumer vs. the Creator personas. The Consumer scenario tests more passive interactions such as receiving nudges & browsing content; the Creator scenario test more active interactions such as posting a comment and chatting with volunteers.

▶︎ Participants

  • 8 proxy users (HCI professionals; neurotypicals)


  • Pan-session: Think aloud protocol

  • Concept Effectiveness: Design Guidelines (please see the design guideline section above)

  • High level usability: After Scenario Questionnaire (ASQ) to test ease of use, time spent on task, and sufficiency of scaffolding.


Testing Feedback

Out of a 7-point Likert scale, eLIFE received overall high ratings on both metrics. The overall concept effectiveness is 6.188/7. The overall usability rating for Consumer type interaction is 6.375/7; the Overall usability rating for Creator type interaction is 6.542/7


Concept Effectiveness

High Level Usability


The effectiveness of the digital tool relies largely on quality of in-person resources.

“It's nice to bring all people in the network into one place. This kind of centralized and specialized group is helpful.”

“Quality content is important for the success of the support tool”

“Notifications can also be about guiding users to do something in app (eg. like "Talk to this person")”

Accessibility needs to be more explicitly considered.

“Nuggets of info for each topic is helpful.”

“Have a overview of functionalities and features during on boarding would be nice”

“Not sure what the toggle means; using text ‘follow topic’ may be more clear and more accessible? There are accessibility guidelines specifically for autism.”

Incentivization is critical to the success of the system.

“Feels like checking off a nudge feature seems like motivation and is satisfying”

“Maybe incorporate some kind of reward or gamification system to keep users motivated?”

“Would be nice to introduce a social reward mechanism to notify them for gratitude of their contribution.”

Stakeholder buy-in

After the presentation, the client was very happy with the features and functionalities that the artifact represents as well as the user-centered process that enables genuine user insights. They decided to move forward with our design group to further refine the eLIFE system with more in-depth research and design iterations. The insights and the artifacts have been documented and shared.




Real Content, Real prototype, Real experience

Despite the project being an early concept validation, I chose to go with a high fidelity first prototype for design exploration when doing client presentation. It turned out to be the right decision!

The higher fidelity prototype makes the experience realistic enough for users and stakeholders (who are not from a technical background) to easily understand the functionalities and how they will readily fit into the teaching and learning context of eCenter. It also helped drive conversations that result in richer stakeholder feedback.

Speaking the language of the audience

If connecting with the users (in all senses) is key to successful research, then to speak the users' language is the first step to create that understanding.

Many autistic users I worked with had trouble articulating their thoughts and feelings and find it difficult to stay motivated throughout their learning. Getting to know them as people through immersion helped me replace research jargon with an encouraging volunteer/therapist talk that they’re more familiar with.

Rectangle 376f.png