Find the Fake: Combating health misinformation with a chatbot game

Location

Jordan

Date

2022

Link

Introductory video

Link

Working paper

Project type

gamified chatbot, digital public health tool

Problem

In Jordan, health misinformation spreads quickly through messaging apps like WhatsApp. With low digital literacy and high levels of social sharing, false claims about vaccines and public health were gaining traction. The Ministry of Health needed a scalable, accessible solution to help people identify and resist misleading information.

Solution

At the World Bank, I helped develop Find the Fake, a gamified chatbot that delivers a "prebunking" experience through WhatsApp. Prebunking exposes users to false or manipulative content examples, helping them recognize similar real-life patterns. The game uses local language, relatable scenarios, and immediate feedback to teach users how to spot misinformation.

Key Features:

Delivered entirely through WhatsApp, the country’s most widely used messaging platform
Interactive storyline, quizzes, and badge rewards to keep users engaged
Designed to be accessible for users with low digital literacy

Results

We tested the tool with 2,851 social media users through a randomized study.

32% improvement in recognizing misinformation tactics
56% reduction in intention to share false headlines
Balanced examples helped users identify false and accurate information, improving trust without increasing skepticism.

This chatbot game proved a practical, low-cost way to boost public resilience to health misinformation.

My Role

Product design and UX research. Developed and tested multiple chatbot formats with local partners and users
Stakeholder coordination. Worked with Meta, the donor, the Jordan government, and World Bank teams
Impact evaluation. Contributed to study design and co-authored the policy research paper

Find the Fake promotional graphic: a character in a Jordan flag shirt beside the line '40% of people spread misinformation online', with a prompt to message the game to play

Find the Fake game messages: a suspicious health-product message, a yes or no misinformation question, an explanation of the fake-expert tactic, and a badge earned for spotting it

Find the Fake study results: bar charts showing the game raised misinformation-detection accuracy (up to +19%) and cut willingness to share it (down to -57%) versus a control group, n=1,059

Find the Fake results comparing control, game, and infographic conditions: the game improved detection accuracy (+5%) and reduced sharing (-43%), outperforming the infographic, n=2,261

Find the Fake evaluation method: how the study measured accuracy in detecting misinformation and likelihood of sharing, with example Arabic survey items and response scales

Find the Fake study design: participants recruited through Facebook ads, split into a WhatsApp game arm and an infographic arm, then assessed on misinformation detection and sharing

Find the Fake concept diagram of psychological inoculation: learning about and being exposed to misinformation tactics builds the skills to recognize it, producing resistance to misinformation

Find the Fake background: a chart of misinformation beliefs in Jordan (60% believe the virus is man-made) and an estimate that misinformation costs about US$78 billion

Find the Fake findings summary: the game raised detection skills by over 5%, cut intended sharing by over 40%, had greater impact among the unvaccinated, outperformed infographics, and over 80% would recommend it

Source and credits: Find the Fake, World Bank (developed for a program in Jordan). The screenshots and research figures are shown to document work on the project, with attribution to the World Bank. They do not imply endorsement.