Ghost Hunters
Research Project Developed for Facebook
October 2020 - December 2020
I developed a mobile-based game that leverages embedded design techniques to covertly teach players how to identify fake news. Armed with this information, our players can curb their inadvertent spread of fake news on social media.
Research Goal
How can we motivate online news readers to critically evaluate information before sharing to social media?
People rely on social media as a news source, but these platforms popularize misleading and even false information (or “fake news”). To stop fake news’ spread, users need to become more critical thinkers.
Solution Benefits
Ghost Hunters improves players’ ability to evaluate news credibility quickly
It trains users on what elements are the most important to focus on in evaluating news effectively
It teaches users the importance of evaluating news by associating critical thinking with social status
My Role
UX Researcher + Game Designer
I researched user rationale for sharing fake news, then developed and led prototyping methods to test game designs with these users.
I researched and integrated persuasive design principles into our game, to motivate desired user behaviors in an enjoyable context.
Client
Facebook (Academic Project)
Collaborators
Project Timeline
Immersive interviews helped us identify motivation issues with flagging fake news. We explored various concepts to improve motivation, eventually settling on a single game that we rapidly iterated on thanks to parallel prototyping and usability testing.
Understand Fake News Spread
Immersive Interview + Literature Review
Why do well-meaning people share fake news?
From initial literature review, we learned quickly that most social media users don’t share fake news out of malice, but because they think an article is legitimately compelling. What gives?
To understand how users perceived fake and credible articles, we developed a “contextual think-aloud”.
CONTEXTUAL THINK-ALOUD:
Recreate two articles in the Facebook news format, then insert them into an otherwise benign Facebook feed and ask participants their unfiltered thoughts on the articles’ credibility
We learned...
Readers want to share news with provocative perspectives, because they think this makes them appear more intelligent
Readers use peripheral cues as proxies for article credibility (e.g., number of likes / comments, blue “verified” checks) because they’re time poor
Readers are far more likely to share news that aligns with their own perspective, and see it as more credible
Our initial research was right; participants hated the spread of fake news but didn't feel motivated enough to critically assess everything they read
Explore Opportunities
Generative Ideation + Research Presentation
How could we improve users' ability and motivation to think critically about news?
With our new focus area, each of us brainstormed 15 ideas to diverge in our thinking, after which we clustered ideas into groupings and identified three priority groups.
We received feedback from a panel of professors with backgrounds in persuasive design and game design:
IDEA:
FEEDBACK:
Narrative-based detective game: Players find and analyze clues to catch a perpetrator trying to spread misinformation
Disguising the intervention should encourage replayability, but a long narrative could become boring quickly and deter playing
Gamify real-life Facebook experience: Publicly award “Scholar Badges” to Facebook users when they act critically (e.g., click articles before sharing)
The social reward system is powerful, but how do we prevent users from gaming the system or create objective measures of critical thinking?
First-person empathy-building story: Players enter the life of a conservative to better appreciate the validity of dissenting opinions
It’s near-impossible to drive empathy with “the other side of the aisle” today, as political values are intertwined with personal identity
Generative Ideation
How should we combat the spread of fake news?
To explore the best method of execution, we reverted to generative ideation, and stress-tested our ideas against three criteria that our critique had uncovered as critical:
1
How enjoyable is the game to improve engagement and replayability?
2
How well are we educating players on how to evaluate fake news without creating a shallow, potentially misused game?
3
How are we building empathy for those who share fake news, instead of vilifying them?
With these criteria, we agreed that we should focus on a modified version of the detective game for its potential.
ISSUE:
What game concept could both teach users how to combat fake news, while being an enjoyable experience?
GHOST HUNTERS:
Ghost hunters enter a haunted house with a shapeshifting demon in their midst. Ghost hunters find clues to identify the demon, and the demon misleads hunters with fake clues to get them to turn on one another.
Parallel Prototype Ideas
Competitive Analysis
How can persuasive design help us improve the game’s education value?
I turned to other educational games to understand how they balanced enjoyability with teaching moments, and extracted four key principles that underpinned our game’s effectiveness:
Obfuscation: Set the game in a haunted house with news articles about exorcism so players are more engaged and think less about the game’s true intent
System 2 Thinking: Incentivize credibility evaluation and show players how to evaluate quickly, improving their long-term motivation and ability recognition
Social Proof: Decide on which news is unreliable through voting, so critical thinking is seen as socially desirable, even outside the game
Increased Moral Sensitivity: With the demon gameplay, players learn firsthand the power of fake news to motivate combatting it
Parallel Prototyping + Usability Testing + Think-Aloud
Which game flow is the most enjoyable for players?
With so many different ways to execute gameplay, I proposed parallel prototyping to identify the elements that resonated with users so we could consolidate them into a single flow.
We developed three flows with distinct elements to test:
1
Complete the Ritual: Players have to find credible exorcisms and ritual items to defeat the demon, further obfuscating the game’s true intent
2
Use the Guidebook: Players get help in pinpointing exorcism credibility with a guidebook that teaches them what to look for
3
Everyone's a Hunter: Keep things simple. The demon is bot-controlled and players don’t have to navigate rooms to find evidence
ISSUE:
How can we determine the most enjoyable gameplay mechanics and flow for our players?
IDEA:
Parallel prototype several flows with distinct differences to identify which elements resonate most
We learned...
None of our designs did well in think-aloud protocols. We had created games that felt like homework instead of fun distractions, for many reasons:
The guidebook was necessary to evaluate credibility quickly, but digesting it all was far too boring
The articles were long and complex, so users would resort to skimming and guessing about credibility
The overly simple flow was boring and had no replay value, but the complex flows with multiple roles took 2-3 times to fully understand
Users loved debating credibility during voting, but the ritual items felt tacked on and unrewarding to collect
Play-Test a Mobile Game
Evidence-Based Design
How can we ensure our educational game doesn’t feel like a chore?
Our game felt less like a game and more like a lesson. I proposed a few solutions to increase interactivity, enjoyability, and understandability:
CRITIQUE:
IDEA:
I want guidance, but a guidebook is way too much information
A “friendly ghost chatbot” (Juju) can advise ghost hunters on credibility, while Zuzu can help demons figure out how to mask credibility
The articles are too long to read, and blend together in my mind
No article over 100 words, and vary mediums and visuals to keep things fresh and digestible
I want an involved game flow, but that takes too long to understand fully
Replace the old, multi-page tutorial with an interactive tutorial narrated by Juju and Zuzu to introduce both gameplay routes more clearly
I want to focus gameplay on voting and debate of my submitted articles
Create a time limit on article collection to focus on voting, and simulate real-life time constraints when assessing news
Usability Testing + Wizard of Oz (WoZ) Prototyping
How can we test a multiplayer game on Figma?
With so much importance placed on the game’s social element, how could we test a full multiplayer run-through with just a single clickthrough Figma prototype?
We invented a “parallel flows” Wizard of Oz method with two participants at a time that involved:
1
Bring each player into a separate Zoom room to play a flow with a tester adjusting screens based off input
2
Bring both players into a single room with 3 “confederates” that pretended to go through their own flow
3
Discuss and vote as a team amongst all 5 “players” to create a fully immersive experience
We learned...
Our novel prototyping method immersed users and created a highly enjoyable experience:
Players finally understood both routes of gameplay upon starting, thanks to our interactive tutorial.
Players appreciated the time cap to force efficiency and create focus on the voting discussions
The game finally felt enjoyable, with multiple users commenting that they’d be open to replaying a more developed version
Although shortened articles were appreciated, players wanted novel ways to test critical thinking (e.g., riddles, mini-games) so their sole focus wasn’t reading
Final Solution
Fight the Spread of Fake News
Teaches players how to identify fake news with the “Ghost Hunter” role, where the friendly ghost Juju partners with users to critically evaluate the news they come across
Teaches players how fake news is spread with the “demon” role, where the evil ghost Zuzu finds ways to covertly make fake news seem more appealing so players can identify those tactics in real life
Immerse Users in a Fictional World
Combines education with enjoyability by teaching players to critically evaluate news as part of a fictional “ghost hunter” context to drive engagement and obfuscate the game’s educational goals
Teaches with an interactive tutorial that is more enjoyable to follow as users learn the game’s purpose and roles
Collaborate to Assess News Credibility
Vote to eliminate any unreliable news from the group, with Ghost Hunters trying to identify the news submitted by the demon, and the demon trying to present their fake news as reliable
Associates critical thinking with social status as winning players are those that can identify fake news consistently to identify and defeat the demon in the team’s midst
Results & Final Thoughts
I co-developed a multiplayer social deduction game that helps players learn how to read news more critically, while still being an engaging and enjoyable experience
Creating an experience that’s educational and fun
Our final game flow struck a balance between education and enjoyability, motivating and teaching players to care more about criticality in their social media news-reading and news-sharing, while having a genuinely fun social experience with their friends.
It creates moments of delight for players, and improves Facebook’s newsfeed:
Educates players on which news elements are most critical when evaluating credibility
Increases evaluation efficiency for players assessing news, accommodating the speed of social media consumption
Motivates players to be more critical by associating evaluative thinking with social status
Reduces spread of fake news with more thoughtfully shared articles from platform users
Rehabilitates Facebook image with clear and effective method of curbing fake news spread, hosted on Facebook Gaming
Increases platform usage as increased news credibility increases user likelihood of engagement
