Project info
Work package
- Work
Sustainability threat
- Spillovers
Challenge
- Reconfiguring-roles-and-relationships
Study info
Related studies according to this researcher
Memes to an End: Charting the course of partisan toxicity in a popular political Subreddit
Description of Study
Online political discussions often turn sour once a toxic comment is made, leading to toxic cascades where hostile comments are met with more toxic responses. This can have negative consequences for an individual’s mental health and can lead to hostile perceptions of those holding different opinions. In this paper, we aimed to understand whether different constructive characteristics of a contribution to a discussion can stop a toxic cascade. Using Perspectives API, we quantify toxicity and different dimensions of constructive comments left on a political Reddit forum. Using these dimensions, plus political affiliations of users on this large-scale deliberation/discussions taking place on a political subreddit, we look at 5 comment chains to see whether constructive comments can stop toxicity. We find promising results of being respectful and arguing using reasoning, as well as being curious and showing nuance about certain arguments.
Study research question
How do users talk to each other in these threads, and do certain comment behaviors are able to stop toxicity in online discussions?
What characteristics of a contribution to a toxic discussion from users help stop the toxicity?
Do these changes in comment related behaviors depend on whether commenters share the same political affiliations/opinions?
Collection provenance
- External data
Collection methods
- Text Analysis
Personal data
Yes
External Source
Source description
Data between 1 August 2020 and 31 August 2020 on the PoliticalCompassMemes Subreddit. Data from comments on posts, as well as upvotes and downvotes are used. In addition, user flairs (political affiliations) and time day the comments are made.
File formats
- Json
- .txt
Data types
- Structured
Languages
- English
Coverage start
Coverage end
01/08/2020
31/08/2020
Spatial coverage
Global
Collection period start
01/08/2020
Collection period end
31/08/2020
Variables
Unit
Unit description
Sample size
Sampling method
Communities
Users of the specified Subreddit. Each observation is a chain of 5 comments and the derived/related variables per comment/user who left the comment.
5600 5-comment chains
API Access and text analysis
Hypothesis
Theory
Does engaging with toxic comments with replies containing bridging attributes lead to lower toxicity in subsequent interactions?
Conflict Theory, SIDE, Bridging Characteristics
Variable type
Variable name
Variable description
Dependent variable
Avg Toxicity (4-5)
The average toxicity of comments 4 and 5 in the 5 comment chain.
Independent variable
Toxicity
A rude, disrespectful, or unreasonable comment that is likely to make people leave a discussion. Calculated for each comment in the chain (thus 5 per "observation").
Independent variable
Affinity
References shared interests, motivations or outlooks between the comment author and another individual, group or entity. Calculated for each comment in the chain (thus 5 per "observation").
Independent variable
Compassion
Identifies with or shows concern, empathy, or support for the feelings/emotions of others. Calculated for each comment in the chain (thus 5 per "observation").
Independent variable
Curiosity
Attempts to clarify or ask follow-up questions to better understand another person or idea. Calculated for each comment in the chain (thus 5 per "observation").
Independent variable
Nuance
Incorporates multiple points of view to provide a full picture or contribute useful detail and/or context. Calculated for each comment in the chain (thus 5 per "observation").
Independent variable
Personal Story
Includes a personal experience or story as a source of support for the statements made in the comment. Calculated for each comment in the chain (thus 5 per "observation").
Independent variable
Reasoning
Makes specific or well-reasoned points to provide a fuller understanding of the topic without disrespect or provocation. Calculated for each comment in the chain (thus 5 per "observation").
Independent variable
Respect
Shows deference or appreciation to others or acknowledges the validity of another person. Calculated for each comment in the chain (thus 5 per "observation").
Control variable
Score
Comment popularity. Calculated for each comment in the chain (thus 5 per "observation").
Control variable
Is Author
Binary indicator whether commenter also posted the original post. Calculated for each comment in the chain (thus 5 per "observation").
Independent variable
Author Political Affiliation
Where the user identifies on a political compass. Calculated for each author of a comment in the chain (thus 5 per "observation").
Discipline-specific operationalizations
Conflict of interest
Data packages
Publications
Documents
Filename
Description
Date
Ethics
Ethical assessment
Yes
Ethical committee
EC - Sociology Groningen