personalization and crowdsourcing for misinformation prevention
Problem: Misinformation on social media is rampant, and different people may have different needs and attitudes towards misinformation mitigation strategies like warning tags.
Solution: We assess the viability of crowdsourcing as a means to identify misinformation online and identify factors which may impact a person’s vulnerability and attitudes towards prevention efforts.
I co-founded the interdisciplinary Cognitive Media Lab at UCSD which studies online misinformation.
Study 1: Crowdsourcing to Detect Online Misinformation (CSCW 2022)
Can we use crowdsourcing to detect online misinformation? How do a person’s cognitive, information assessment, and personality traits affect their detection ability?
Design should be sensitive to the needs of different users.
Crowdsourcing platforms can help detect and stop the spread of online misinformation. However, past studies have not deeply examined the individual characteristics—such as cognitive factors and biases—that predict crowdworker accuracy at identifying misinformation.
In our study, participants assessed the truthfulness and sentiment of COVID-19 related tweets as well as answered several surveys on personal characteristics.
Results support the viability of crowdsourcing for assessing misinformation and content sentiment related to ongoing and politically-charged topics like the COVID-19 pandemic, however, alignment with experts depends on cognitive, informational, and dispositional, and personality traits.
This study offers insight into how crowdsourcing can be used for detecting misinformation in practice, using crowd composition as a means for crowd recruitment and filtering.
Study 2: Misinformation Intervention Strategies Based On Individual Differences (CSCW 2025)
How can we shape misinformation prevention strategies to fit groups with different attitudes?
Personalizing misinformation warning tags to the individual characteristics of social media users may enhance mitigation effectiveness. In this study, we used survey to assess how people differ and how those differences predict a person’s attitudes and behaviors toward tags and tagged content.
Results show attitudes towards warning tags and behaviors are influenced by factors such as personality, information processing, trust and credibility dispositions, and cognitive abilities.
We synthesize our results into design insights that can inform the creation of effective and personalized misinformation warning tags and misinformation mitigation strategies more generally.
An example misinformation intervention.
Publications that have come out of this research agenda:
Kaufman, R. A., Broukhim, A., Haupt, M. (2025). WARNING This Contains Misinformation: The Effect of Cognitive Factors, Beliefs, and Personality on Misinformation Warning Tag Attitudes. Proceedings of the 2025 ACM Conference on Computer Supported Cooperative Work (CSCW). PDF
Kaufman, R., Haupt, M., Dow, S. (2022). Who’s In the Crowd Matters: Cognitive Factors and Beliefs Predict Misinformation Assessment Accuracy. Proceedings of the 2022 ACM Conference on Computer Supported Cooperative Work (CSCW). PDF