Examining the Effects of Conversational Chatbots on Changing Conspiracy Beliefs about Science: The Paradox of Interactivity

Wang, Jinping; Tanes-Ehle, Zeynep

doi:10.1080/08838151.2022.2153842

Cited by 3 publications

(1 citation statement)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since conspiracy beliefs are related to mental health problems (Bowes et al, 2023;Imhoff & Lamberty, 2018;Stasielowicz, 2022b), such as paranoia or schizotypy, providing psychotherapy for people with mental health issues might reduce conspiracy beliefs as a secondary outcome. Other underexplored options include vetted mental health apps, chatbots (Wang & Tanes-Ehle, 2022), and communication for family members (Hümmler & Schiesser, 2023). Various ideas from the misinformation literature (Altay, 2022) could also be examined in the context of conspiracy beliefs.…”

Section: Research Gaps and Future Research Avenuesmentioning

confidence: 99%

The effectiveness of interventions addressing conspiracy beliefs: a meta-analysis

Stasielowicz

2024

Preprint

View full text Add to dashboard Cite

Since new implausible conspiracy theories emerge frequently and are thought to have detrimental effects on individuals and society (e.g., vaccine hesitancy, prejudice, and worse interpersonal relationships), developing effective interventions to reduce conspiracy beliefs is urgent. The present Bayesian three-level meta-analysis with 273 effect sizes from 56 independent samples (N = 27,996) provides the first comprehensive overview of factors affecting intervention effectiveness, including characteristics of samples, interventions, and conspiracy beliefs. The findings indicate that interventions tend to be moderately effective (g = 0.16, 95% CR [0.12, 0.20]). However, effect sizes exhibited substantial variability. By adopting evidence-based recommendations, such as tailoring intervention content to the outcome content and avoiding single-item measures of conspiracy beliefs, intervention effectiveness can be enhanced to g = 0.3 or higher. Furthermore, the identified research gaps, such as the scarcity of research on long-term effects, provide valuable insights for further improvements in this research area.

show abstract