Truace
Subscribe

conversational dynamics in health-related YouTube communities discussing scientific versus pseudoscientific treatments

Source article: Exploring Conversational Dynamics in Scientific and Pseudoscientific Health Communities on YouTube: Process Mining and Network Analysis Study

Background Social media platforms, particularly YouTube (Google LLC), are important sources of health information, but also significant vectors for misinformation and pseudoscience. While many studies analyze the content and sentiment of this information, the dynamic, sequential nature of user interactions, which can shape belief formation and community dynamics, remains poorly understood. Objective This study aimed to explore the applicability of network analysis and process mining techniques for identifying an…

By the Truvace desk1 min read1 sourcePeer-reviewedTRV-2026-1277· cite & verify
Exploring Conversational Dynamics in Scientific and Pseudoscientific Health Communities on YouTube: Process Mining and Network Analysis Study
February 2024 Future Audiences office hours by LWyatt (WMF). CC BY-SA 4.0 · https://creativecommons.org/licenses/by-sa/4.0
Trace impact readingNegative state
P 71The P score combines the specificity and measured human impact of the grounded problem claim with the strength of this Trace’s cited sources.

Both sides are scored from claims and sources, not community votes.

G 65The G score combines the specificity and measured human impact of the grounded gain claim with the strength of this Trace’s cited sources.

In brief

Researchers analyzed 52,412 YouTube comments from health treatment videos posted between 2011 and 2025, classified as scientific or pseudoscientific via an LLM-assisted pipeline, to test whether network analysis and process mining could map how conversations unfold.

Distinguishing mixed-valence, treatment-focused exchanges in scientific threads from affirmation-heavy, socioemotional bonding in pseudoscientific threads matters for misinformation response, but the work is exploratory and limited to sampled public comment corpora, leaving causal effects on beliefs and generalizability uncertain.

Main points

  1. Analyzed 20,387 comments from scientific videos and 32,025 from pseudoscientific videos posted between 2011 and 2025.
  2. Used automated pipeline combining API extraction, large language model-based video classification, and multilingual sentiment and thematic classification.
  3. Scientific corpus showed greater prominence of negative expressions, negative comparisons, and medical treatment and advice requests.
  4. Pseudoscientific corpus concentrated around positive expressions, thanking, compliments, and emoji-only or brief acknowledgments.

The gain

Applying network analysis and process mining to 52,412 YouTube comments distinguished scientific from pseudoscientific health discussions, showing mixed-valence evaluation in scientific threads versus affirmation bonding in pseudoscientific threads.

The problem

YouTube health videos serve as significant vectors for misinformation and pseudoscience, with sequential user interactions that can shape belief formation and community dynamics.

The rundown

The study retrieved public YouTube comment threads from 2011 to 2025 and classified videos as scientific or pseudoscientific using an LLM-based pipeline plus multilingual sentiment and thematic NLP.

Network analysis measured normalized node strength of conversational topics, while process mining modeled temporal sequences of interaction types like thanking, compliments, and advice requests.

Authors conclude public health strategies may need to address affective and community-bonding dimensions of engagement in misinformation communities alongside information provision.

What this doesn’t fix

Exploratory observational design based on sampled corpora with preliminary findings, limiting generalizability beyond the retrieved YouTube comment threads.

Sources

  1. Peer-reviewedJournal of Medical Internet Research2026-10-02

The debate