Overcoming the harmful effects of confirmation bias, entrenched misinformation, and social conformity

Li Shu, Faculty of Applied Science & Engineering; Matthew Feinberg, Rotman School of Management

Combining the systematic approach of engineering design and prototyping with behavioral-science insights, we address three challenges for pandemic resilience: 1) confirmation bias when people consume information, 2) misinformation that has taken hold in individuals, and 3) social conformity to behave in ways inconsistent with accurate information.

Individuals are subject to confirmation bias when they assimilate information. For example, those who do not (want to) believe that a disease is harmful will seek and preferentially attend to information that minimizes harm. We will develop and test new interventions that help people overcome such confirmation bias.

Misinformation has led many people to hold incorrect views about pandemic-relevant topics. Simply providing the correct information is unlikely to help. Instead, conversations where such people reflect on their reasoning and motivation, have been shown to succeed in changing minds. However, such conversations are time and labor-intensive. We propose developing chatbots to examine whether they can be programmed to change the minds of those with deeply entrenched and inaccurate views.

Finally, even with accurate information, many people struggle to act consistently with that information when those around them do not. We aim to design, prototype, and test technology-based interventions to help people overcome harmful social conformity. For example, video displays of people performing behaviors that are consistent with accurate information, may be optimally situated in a physical environment to create the social proof that may otherwise not exist.

Addressing challenges in how accurate information may be sought, internalized, and translated to behavior will improve pandemic resilience.