Emotion-Driven Belief Updating: How Feelings Shape Evidence Evaluation Across Political and Non-Political Topics

• In Event: Reimagining Epistemic Cognition for a Changing World: Knowing in an Age of Too Much Information

Abstract

Abstract
With increasing digital access and selective engagement, it's unclear how students update beliefs when exposed to credible, especially politically charged, evidence. Emotions may influence how such information is processed. This study examines how epistemic emotions affect alternative information-seeking and belief updating in response to contentious and non-contentious socio-political texts.
Objectives
We investigate how undergraduates’ epistemic emotions shape their decisions to seek alternative information and how both influence belief updating. Three research questions guide the study:
• RQ1: Which epistemic emotions predict alternative information-seeking?
• RQ2: Do emotions and information-seeking predict belief updating toward evidence?
• RQ3: Does topic polarization moderate the impact of seeking alternative information?
Theoretical Framework
Epistemic emotions—such as curiosity, confusion, or frustration—are key to learning and reasoning, especially when confronting belief-challenging content (Muis et al., 2015, 2021). While belief updating is often considered rational, emotions may influence decision-making processes, such as information-seeking behaviors, and may bias reasoning (Sommer et al., 2024). However, research has not examined how they dyanmic interplay between emotions, information-seeking, and belief updating, which is critical in today’s complex information landscape.
Methods
Participants & Procedure
We collected 1,732 observations from 225 undergraduates (26% female, 15% underrepresented minorities, 40% first-generation students) from a Southern California university. Participants read 12 published research summaries—8 on polarizing topics (e.g., abortion), and 4 on non-polarizing ones (e.g., artificial sweeteners). Each text showed empirical evidence favoring a proposition (e.g., poverty programs do help the marginalized). After rating their emotions, students chose whether to learn more. If “yes,” they selected one of two fictitious follow-ups—one supporting, one opposing the claim. Beliefs were rated again post-exposure.
Measures
• Belief Updating: Difference between posterior and prior belief (0–100 scale), in the direction of the evidence.
• Information-Seeking: Categorized as (a) Disinterested (no info), (b) Biased Choosing (aligned with prior belief), (c) Alternative Choosing (opposed to prior belief).
• Epistemic Emotions: Assessed using a 7-item short-form scale (Pekrun et al., 2017), rated 1–5.
Data Analysis
We used a Bayesian mixed-effects multinomial logistic regression (RQ1) with alternative choosing as the reference category, Bayesian linear mixed-effects models for belief updating (RQ2), and interaction models to assess the moderation of topic polarization (RQ3).
Results
• RQ1: Curiosity and enjoyment reduced disengagement but did not decrease biased seeking beyond individual traits (e.g., political orientation, reasoning skill). Other emotions were non-significant. (See Figure 2.)
• RQ2: Compared to alternative choosing, both biased and no information-seeking were linked to reduced belief updating (~6% decrease in updating, respectively). Emotions did not directly predict belief change. (See Figures 3 and 4.)
• RQ3: The positive impact of alternative information-seeking on belief updating was stronger for politically charged topics. (See Figure 5.)
Scholarly Significance
Epistemic emotions shape how students engage with challenging information. While curiosity and enjoyment promote engagement, they don’t guarantee balanced exploration. Alternative information-seeking supports greater belief updating, especially on controversial issues. Cultivating productive epistemic emotions and encouraging engagement with opposing views may strengthen scientific evidence receptivity and epistemic cognition.

Author

• Gabe Avakian Orona, Boston College