Seed Grant Report — Truth, lies, and misinformation during cognitive development
Celeste Kidd's lab received a BCNM Faculty Seed Grant for "Truth, lies, and misinformation during cognitive development." Read more about the results of the project below!
The award from BCNM for our proposal “Truth, lies, and misinformation during cognitive development” allowed us to collect empirical data on interventions designed to give children a greater ability to discern truth from falsity in online materials. We conducted two sets of foundational empirical studies on two types of interventions designed to facilitate children’s ability to discern fact from fiction in new media. This work led to two new research insights, communicated in multiple published scientific papers and presentations:
1. Exposure to detectable inaccuracies makes kids more diligent fact-checkers of novel claims
Children are uniquely vulnerable to misinformation because of their nascent knowledge states and tendencies towards overconfidence and social trust; yet psychological research focused on misinformation interventions has with rare exception focused on adults. Further, children are increasingly exposed to misinformation via increased connectivity to the internet and generative AI systems on computers and mobile devices (Common Sense Media, 2023; Magis-Weinberg et al., 2021). Existing efforts to protect children from misinformation typically employ sanitation methods, which seek to carefully curate content for children (e.g., YouTube Kids). Our findings show that this solution could actually worsen the problem by dampening children’s natural skepticism. Instead, our findings suggest that exposure to controlled-but-imperfect information may scaffold children’s epistemic vigilance and critical thinking skills.
In two experiments and one simulation, we show that young children adaptively adjust the amount of evidence they seek to verify new claims based on the amount of prior inaccuracies they’ve encountered. These results hold in a search engine context, underscoring the work’s relevance to children’s digital media literacy.
Evan Orticio presented this work at the Budapest CEU Conference on Cognitive Development (2023), 44th Annual Meeting of the Cognitive Science Society (2023), UC Berkeley Developmental Area Seminar (2022), and UC Berkeley Institute of Human Development (2023). In 2024, Evan plans to present this work at the Misinformation & Belief Science preconference at the Society for Personality and Social Psychology Convention. It was published as a CogSci paper, and is currently under review at Nature Human Behavior.
2. Children can use the confidence of others to form more accurate beliefs
We need to uncover truth with limited time and resources to make good decisions. Humans rely on information from others by dividing cognitive labor and collaborating to maximize our limited resources. Multiple perspectives make it more likely for us to uncover the truth (Bahrami et al., 2010; Surowiecki, 2005), but this comes with a challenge. Because of the inherent diversity of experiences, we will inevitably remember events differently, or come to different conclusions from the same evidence. Humans need strategies to resolve disagreements so that we can make the most of the information available to us.
We conducted a large empirical study to examine how children navigate differences of opinion across development. Our results demonstrate that children are able to distinguish differing confidence judgements and default to adopting the belief of the more confident person by age 5. By age 8, they can integrate the confidence judgments of multiple people in order to form more accurate beliefs with finer granularity.
Carolyn Baer presented this work at the Budapest CEU Conference on Cognitive Development (2023), 44th Annual Meeting of the Cognitive Science Society (2023), University of British Columbia Developmental Brown Bag talk series (2023), and UC Berkeley Institute of Human Development (2022). It has been prepared as a full manuscript which will be submitted to Developmental Science imminently.