iSIFT

UX Designer & Researcher | Winter & Spring 2021

iSIFT if a fact-checking tool that helps adults find the information they need to determine the credibility of an article in a simple and efficient manner.
award icon
Out of 150+ projects, iSIFT won the 2021 UW iSchool Capstone Award for Social Impact
open laptop and iPhone showing summary of article page, with Sifty mascot to the left

SETTING & Sponsors

iSIFT was completed as a capstone project for the University of Washington’s Information School and was sponsored by the Center for an Informed Public (CIP). The CIP is an organization that seeks to “resist strategic misinformation, promote an informed society, and strengthen democratic discourse.”

Role

UX/UI Designer UX Researcher
UX Designer

TEAM

Judith Wong, Project Manager
Joseph Altamira, Developer
Saasha Mor, Developer

TIMELINE

January - May 2021

TOOLS

Figma
Miro
Adobe Illustrator
process
01 problem

PROBLEM SPACE

Misinformation is everywhere...and it’s dangerous
misinformation
(n.) false or inaccurate information that is spread with or without the intent to deceive
Although misinformation can be spread about a wide range of topics, my team and I decided to focus on misinformation relating to COVID-19 since we were in the midst of a global pandemic.
misinformation cycle; false news leads to blind acceptance, which leads to confusion and distrust, leading to questionable actions, leading to dissemination of false news; cycle repeats
Click image to enlarge

THE PROBLEM

People who avoid fact-checking important articles about COVID-19 are more likely to believe false information, get confused, and make decisions that have potentially negative consequences.

OBJECTIVE

Our task was to design a digital solution that uses the SIFT Method to combat misinformation
What is the SIFT Method?
black down arrow

PROBLEM STATEMENT

With misinformation, media literacy, and COVID-19 in mind, I was able to formulate the following problem statement...
How might we help adults (aged 25-55) on Facebook achieve stronger media literacy so that they can feel more confident when consuming information about COVID-19 and vaccinations?
02 research

RESEARCH METHODS

After formulating some research questions to guide our research, I employed the following research methods to gain insights about our users, discover gaps in the market, and check our assumptions and biases...
Summary of research methods used, the purpose of each method, and key findings from each
Click image to enlarge
My team and I documented, summarized, and categorized our findings in Miro. Check out our Miro Board for a closer look at our research process!

RESEARCH INSIGHTS

Our research revealed key insights about our users and market space...
I was able to identify critical pain points that people experience when faced with misinformation about COVID as well as gaps in the market that we could address. Our research also revealed other key insights that would inform future design decisions.
Summary of iSIFT research insights including pain points, market space, and other insights as well as evidence from user research to support each insight
Click image to enlarge

VALUES & GOALS

I identified 3 key values to keep in mind throughout the ideation process, which translated into 3 design goals
I carefully considered the ethics and values our team wanted to prioritize as well as those of our sponsors and capstone advisors. I documented those values in Miro and identified the 3 key values to focus on...
two people with lightbulbs for head icon
Autonomy
Users must be able to validate and form their own thoughts; do not explicitly tell them what is right/wrong
beige down arrow, large
black magnifying glass icon
SIFT Method
The solution must incorporate the SIFT method in some way
capital building icon
Bias
Exclude partisan bias in our design and reports; must consider how to check fake news and websites
beige down arrow, large
ballot sticking out of ballot box icon
Non-Partisan
The solution will at least try to minimize political bias
person, full body in circle icon
Accessibility
Ensure product design is efficient and accessible to all target users (e.g. visual accessibility features)
beige down arrow, large
analog clock icon
Efficient
The solution should fulfill its purpose in 3-5 minutes
Those values translated into 3 design goals that would address both painpoints and user needs identified during user research as well as our sponsors’ objective for this project.

USER PERSONA

Based on research data, I developed 2 user personas that were representative of our target demographic
While developing these personas, was especially cognizant of various factors that affect vaccine hesitancy and the spread of misinformation such as health concerns due to socioeconomic, racial, and political factors, as well as motivations for sharing news on Facebook.
Click image to enlarge
03 ideation

AFFINITY MAP

Since our sponsors did not have a specific solution in mind, we kicked off our ideation by brainstorming different types of solutions and features that address our users’ needs.
My team and I pitched our top 3 solutions to our sponsors, but faced some pushback. During our meeting with our sponsors, I noted what they found most important:
After much deliberation, we agreed upon a solution...
sticky notes on Miro explaining key features/functionality of solution
Click image to enlarge
I then led another round of ideation to brainstorm and prioritize specific features, which we documented in Miro.
categorized and prioritized iSIFT features on sticky notes in Miro
Click image to enlarge

SITEMAP

I began fleshing out our new solution by mapping out the website component. Reviewing notes from our user interviews and survey results helped me organize and prioritize the content the website would display.

Overall, the website is quite simple...
sitemap for website component of iSIFT in a flow chart from Miro
Click image to enlarge

USER JOURNEY

Since our solution is comprised of 2 components - the chatbot and the website - I drafted a couple of user flows to demonstrate how the 2 components interact.
user journey map for both personas using iSIFT; indicates how they feel and what they do as they use the product to solve problems
Click image to enlarge

STORYBOARDING

I revisited our user personas and created storyboards for each persona to check in with our end users’ goals and envision how our solution can help them meet their goals.
04 design

MEDIUM-FIDELITY PROTOTYPE

After sketching some low-fidelity wireframes, I created a mid-fidelity prototype for both mobile and desktop in preparation for preliminary user testing on concept validation.
screenshot of desktop and mobile view of medium-fidelity prototype of iSIFT
Click image to enlarge

HIGH-FIDELITY PROTOTYPE

After reviewing feedback from both our users and our sponsors, I created a final prototype of iSIFT for mobile and desktop...

Main Flow

iPhone showing messenger chat with Sifty; user sent a message to sifty
01 Facebook user shares article with chatbot, Sifty
open laptop showing facebook and open messenger chat with Sifty; user sent a message to sifty
iPhone showing open messenger chat with sifty; sifty responds to article link sent by user
02 Sifty sends a link to iSIFT website and pre-filled Google search
open laptop showing Facebook and open messenger chat with sifty; sifty responds to article link sent by user
iPhone with screen showing iSIFT summary of article
03 User reads summary of information to determine its validity
open laptop with screen showing iSIFT summary of article
iPhone showing iSIFT 'about page' detailing the SIFT method
04 User can learn more about media literacy if needed
open laptop showing iSIFT 'about page' detailing the SIFT method

Key Features

iSIFT: Resources

FINAL PRESENTATION

Out of 150+ projects, iSIFT was one of 5 projects to win an award. We won the Best Impact Award!
Check out our final presentation! This presentation and the poster I designed for this project can also be found on the University of Washington iSchool’s capstone archive.
05 TEST & DEVELOP
Timeline of design process, highlighting that concept validation happened between mid-fi and hi-fi prototypes and usability testing happened after hi-fi prototype
I had the opportunity to conduct 2 rounds of user testing during the course of this project - one for concept validation and one for usability testing.

CONCEPT VALIDATION & USABILITY TESTING

I used a mid-fidelity prototype to conduct the concept validation tests before moving on to designing a hi-fidelity prototype. Usability testing was done using our fully developed product.

Both tests were semi-structured; I presented a scenario in which participants would use our solution, observed them as they interacted with the prototype, and took note of their feedback.
Summary of concept validation and usability testing; includes number of users, positive and negative feedback, and opportunities for improvement
Click image to enlarge
Overall, the solution was well-received. I adjusted the prototypes based on user feedback from concept validation. In the end, our goals were met; users found iSIFT to be efficient, simple, and useful. I did not make any major changes after usability testing, before concluding the project.

DEVELOPMENT

Our developers created a Facebook page for iSIFT to enable the chatbot interaction. Meanwhile, the website component was coded using React.js and hosted on AWS. More information about technical development is available on GitHub.
06 reflect

reflection

In contrast to working on projects with my peers who are familiar with product development, working with sponsors whose expertise is not in tech was an eye-opening experience. It was challenging at times to interpret our sponsors’ feedback and to build our solution, but by the end of the project, I learned to...

Prioritize inclusive design - Design for one, design for all
See more
black down arrow

Defend my design decisions
See more
black down arrow

next steps

As our project concluded, my team and I prepared a handoff document for another capstone team interested in continuing our work. This document featured all the tasks we would have prioritized if we had more time. These tasks include:

Acknowledgements

Special thanks to Mike Caulfield, Emily Porter, and Koko Koltai for their collaboration and sponsorship for this project. Huge thanks to Sasha Anderson for incredible guidance as an advisor throughout the process. Last but not least, thank you to my incredible teammates for all your hard work!