Conversational Inoculation to Enhance Resistance to Misinformation

Dániel Szabó¹ ✉️, Chi-Lan Yang², Aku Visuri¹, Jonas Oppenlaender¹, Bharathi Sekar¹, Koji Yatani², Simo Hosio¹

¹ University of Oulu
² The University of Tokyo

Research of Crowd Computing Oulu • More Projects

MindFort System Conditions — The MindFort system implemented three experimental conditions (the treatments). (a) Reading, (b) Writing, and (c) Chatbot condition.

TLDR; Can arguing with a chatbot make you immune to fake news? Yes. We built "MindFort" to inoculate people against misinformation through debate. It worked better than just reading about the facts.

Abstract

Prolific of misinformation is a globally acknowledged problem. Cognitive Inoculation helps build resistance to different forms of persuasion, such as misinformation. We investigate Conversational Inoculation, a method to help people build resistance to misinformation through dynamic conversations with a chatbot. We built a Web-based system to implement the method, and conducted a within-subject user experiment to compare it with two traditional inoculation methods. Our results validate Conversational Inoculation as a viable novel method, and show how it was able to enhance participants' resistance to misinformation. A qualitative analysis of the conversations between participants and the chatbot highlighted adaptability, independence, trust and friction as the main factors affecting Conversational Inoculation. We discuss the opportunities and challenges of using Conversational Inoculation to combat misinformation. Our work contributes a timely investigation and a promising research direction in scalable ways to combat misinformation.

Paper (Coming Soon) Code

Introduction

Misinformation is now a global problem. False claims about nutrition, vaccines, and treatments threaten public health by encouraging unhealthy habits and fostering misconceptions that lead to poor decisions. Cognitive Inoculation Theory (CIT) establishes cognitive inoculation as a systematic method for strengthening resistance to persuasion. We investigate Conversational Inoculation, a paradigm in which participants' resistance to persuasion is developed through structured and interactive conversation with chatbots. We developed MindFort: a Web-based system for Conversational Inoculation using an LLM-powered chatbot.

Method

We evaluated MindFort in an online within-subject experiment with 65 participants to investigate how conversational inoculation compares against reading and writing treatments. The experiment included four topics: Exercise and Mental Health, Binge Drinking, Dental Hygiene, and Protecting Nature.

Study Design — Overview of the experimental procedure and study design.

Results

Our results validate Conversational Inoculation as a viable novel method. The chatbot condition yielded greater resistance to misinformation than Reading or Writing conditions after controlling for baseline susceptibility. The conversational approach was also found to be engaging without sacrificing user motivation.

Inoculation Outcomes — Post-attack certainty change compared between the four conditions.

Conclusion

Our study solidifies Conversational Inoculation as a promising research direction in the fight against misinformation. It demonstrates that conversational agents can effectively facilitate inoculation interventions in ways that traditional methods cannot, offering adaptability and engagement.