Personal attacks decrease user activity in social networking platforms

doi:10.1016/j.chb.2021.106972

Computers in Human Behavior

Volume 126, January 2022, 106972

https://doi.org/10.1016/j.chb.2021.106972 Get rights and content

Highlights

•
Exploration of the effects of online personal attacks on victims' activity on social media (Reddit).
•
First large-scale (150K users) analysis with a high-precision Artificial Intelligence system not based on self-reported data.
•
Data analysis with classical statistical methods, Bayesian estimation, and model-theoretic analysis.
•
Personal attacks received online significantly decrease victims' online activity.

Abstract

We conduct a large scale data-driven analysis of the effects of online personal attacks on social media user activity. First, we perform a thorough overview of the literature on the influence of social media on user behavior, especially on the impact that negative and aggressive behaviors, such as harassment and cyberbullying, have on users' engagement in online media platforms. The majority of previous research were small-scale self-reported studies, which is their limitation. This motivates our data-driven study. We perform a large-scale analysis of messages from Reddit, a discussion website, for a period of two weeks, involving 182,528 posts or comments to posts by 148,317 users. To efficiently collect and analyze the data we apply a high-precision personal attack detection technology. We analyze the obtained data from three perspectives: (i) classical statistical methods, (ii) Bayesian estimation, and (iii) model-theoretic analysis. The three perspectives agree: personal attacks decrease the victims’ activity. The results can be interpreted as an important signal to social media platforms and policy makers that leaving personal attacks unmoderated is quite likely to disengage the users and in effect depopulate the platform. On the other hand, application of cyberviolence detection technology in combination with various mitigation techniques could improve and strengthen the user community. As more of our lives is taking place online, keeping the virtual space inclusive for all users becomes an important problem which online media platforms need to face.

Section snippets

The influence of social media

Even since the beginning of the social media popularity burst, studies have indicated that participation in SNS influences the users’ well-being and social self-esteem, especially in adolescents. Valkenburg, Peter, and Schouten (2006), in a small initial survey study (881 users) among young users (10–19 yo.) on a popular Dutch SNS called CU2, already noticed that especially (1) the frequency and (2) the tone (positive vs. negative) of the feedback they received had a significant (and

Technology applied for personal attack detection

For the need of this research we define personal attack as any kind of abusive remark made in relation to a person (ad hominem) rather than to the content of the argument expressed by that person in a discussion. The definition of ‘personal attack’ subsumes the use of specific terms which compare other people to animals or objects or making nasty insinuations without providing evidence. Three examples of typical personal attacks are as follows.

•
You are legit mentally retarded homie.
•
Eat a bag of

Large-scale quantitative analysis of impact of personal attacks on reddit user activity

At this point we move on to the presentation of our observational study. It is crucial to emphasize that due to space limitations some technical details and methodological decisions have not been fully explained in the publication, but we did our best to expand on such issues in the online documentation of the study available at https://rfl-urbaniak.github.io/redditAttacks/.

Discussion: potential solutions to personal attacks, verbal aggression and hate speech

A number of studies have investigated potential means to mitigate social network depopulation caused by personal attacks and other forms of online harassment, with almost all of them pointing to the importance of efficient moderation.

One of the strategies that can serve as a potentially effective way to curb verbal abuse is counter-speech. As shown in (Bilewicz et al., 2021; Munger, 2017) it can lead to a significant reduction of the number of racist tweets or attacking comments on Reddit, but

Conclusions and future work

In this paper we introduced our study on the effects of personal attacks on users' engagement and continuation of activity in social media. Despite of the existence of a vast literature on this topic, the previous works were small-scale (up to several hundred subjects), and were based on self-reporting. With the growing importance of social networking services (SNS) in many people's lives, and with the quickly growing number of online harassment and cyberbullying cases world-wide, there is a

CRediT author contribution statement

Rafal Urbaniak: Methodology, Formal analysis, Data Curation, Writing – Original Draft, Writing – Review & Editing, Visualization. Michał Ptaszyński: Conceptualization, Validation, Investigation, Resources, Writing – Original Draft, Writing – Review & Editing. Patrycja Tempska: Conceptualization, Validation, Investigation, Resources, Writing – Original Draft, Writing – Review & Editing, Data Curation, Project administration. Gniewosz Leliwa: Conceptualization, Software, Writing – Original Draft,

References (77)

D.A. Ellis et al.
Do smartphone usage scales predict behavior?
International Journal of Human-Computer Studies
(2019)
Y. Kelly et al.
Social media use and adolescent mental health: Findings from the UK millennium cohort study
EClinicalMedicine
(2018)
C. Lampe et al.
Crowdsourcing civility: A natural experiment examining the effects of distributed moderation in online forums
Government Information Quarterly
(2014)
J. Mullahy
Specification and testing of some modified count data models
Journal of Econometrics
(1986)
M. Näsi et al.
Association between online harassment and exposure to harmful online content: A cross-national comparison between the United States and Finland
Computers in Human Behavior
(2014)
J. Preece et al.
The top five reasons for lurking: Improving community experiences for everyone
Computers in Human Behavior
(2004)
L. Rösner et al.
Dangerous minds? Effects of uncivil online comments on aggressive cognitions, emotions, and behavior
Computers in Human Behavior
(2016)
A.N. Saiphoo et al.
Social networking site use and self-esteem: A meta-analytic review
Personality and Individual Differences
(2020)
M. Weber et al.
Online hate does not stay online–how implicit and explicit attitudes mediate the effect of civil negativity and hate in user comments on prosocial behavior
Computers in Human Behavior
(2020)
H.C. Woods et al.
# sleepyteens: Social media use in adolescence is associated with poor sleep quality, anxiety, depression and low self-esteem
Journal of Adolescence
(2016)

B. Afful et al.

Whatsapp and academic performance among undergraduate students in Ghana: Evidence from the university of cape coast

The Journal of Education for Business

(2020)

H. Akaike

A new look at the statistical model identification

IEEE Transactions on Automatic Control

(1974)

A.M. Alkhalaf et al.

The impact of WhatsApp use on academic achievement among saudi medical students

Medical Teacher

(2018)

E.S. Andersen

Satiation in an evolutionary model of structural economic dynamics

J. Arguello et al.

Talk to me: Foundations for successful individual-group interactions in online communities

C.P. Barlett

Anonymously hurting others online: The effect of anonymity on cyberbullying frequency

Psychology of Popular Media Culture

(2015)

C.P. Barlett et al.

Predicting cyberbullying from anonymity

Psychology of Popular Media Culture

(2016)

J. Barlińska et al.

Ich słowami—obraz pomocy w sytuacjach cyberprzemocy rówieśniczej z perspektywy uczniów

Dziecko krzywdzone. Teoria, badania, praktyka

(2018)

T. Beran et al.

Cyber-harassment: A study of a new method for an old behavior

Journal of Educational Computing Research

(2005)

M. Bilewicz et al.

Hate speech epidemic. the dynamic effects of derogatory language on intergroup relations and political radicalization

Political Psychology

(2020)

M. Bilewicz et al.

Artificial intelligence against hate: Intervention reducing verbal aggression in the social network environment

Aggressive Behavior

(2021)

M. Campbell et al.

Victims' perceptions of traditional and cyberbullying, and the psychosocial correlates of their victimisation

Emotional & Behavioural Difficulties

(2012)

X. Cao et al.

Consequences of cyberbullying and social overload while using SNSs: A study of users' discontinuous usage behavior in SNSs

Information Systems Frontiers

(2019)

M.C. Carey et al.

The influences of participation and moderation on the development of a sense of virtual community

International Journal of Web Based Communities

(2016)

E. Chandrasekharan et al.

You can't stay here: The efficacy of Reddit's 2015 ban examined through hate speech

Y.T. Chua

Understanding radicalization process in online far-right extremist forums using social influence model

(2019)

J.G. Cragg

Some statistical models for limited dependent variables with application to the demand for durable goods

Econometrica

(1971)

E. Cross et al.

Virtual violence II: Progress and challenges in the fight against cyberbullying

(2012)

M. Culpepper

Exploring the relationships of social media usage and symptoms of anxiety and depression in adolescents

(2020)

I.P. Cvijikj et al.

Online engagement factors on Facebook brand pages

Social network analysis and mining

(2013)

W. De Koster et al.

‘stormfront is like a second home to me’ on virtual community formation by right-wing extremists

Information, Communication & Society

(2008)

W.L. Hamilton et al.

Loyalty in online communities

A.B. Hill

The environment and disease: Association or causation?

(1965)

S. Hinduja et al.

Cyberbullying: An exploratory analysis of factors related to offending and victimization

Deviant Behavior

(2008)

J. Howick et al.

The evolution of evidence hierarchies: What can Bradford hill's ‘guidelines for causation’contribute?

Journal of the Royal Society of Medicine

(2009)

A. John et al.

Self-harm, suicidal behaviours, and cyberbullying in children and young people: Systematic review

Journal of Medical Internet Research

(2018)

N. Johnson et al.

Hidden resilience and adaptive dynamics of the global online hate ecology

Nature

(2019)

K. Kircaburun

Self-esteem, daily Internet use and social media addiction as predictors of depression among Turkish adolescents

Journal of Education and Practice

(2016)

Cited by (12)

She is seeking attention: Women's self-objectification and hostility toward peer selfies
2024, Computers in Human Behavior
Although the relationship between self-objectification and selfie-related behaviors has been extensively investigated, limited attention has been given to understanding how women's self-objectification impacts their attitudes toward peer selfies from the observer's perspective. This issue was addressed across four studies (N = 801). Specifically, Studies 1a and 1b (correlational research) revealed a positive association between women's self-objectification and hostile attitudes toward peer selfies. In Studies 2 and 3 (experimental research), a causal relationship was established, such that women with higher levels of state self-objectification were more inclined to attach negative tags to other women's selfies. Notably, perceived attention-seeking motivation plays a mediating role in the process (Studies 1b, 2, and 3). Furthermore, these effects occurred not only in response to selfies taken by highly attractive women but also in response to targets who are moderately attractive or average-looking (Study 3). These findings bridge a significant gap in the field of self-objectification and social interaction on social media.
Why do we not stand up to misinformation? Factors influencing the likelihood of challenging misinformation on social media and the role of demographics
2024, Technology in Society
This study investigates the barriers to challenging others who post misinformation on social media platforms. We conducted a survey amongst U.K. Facebook users (143 (57.2 %) women, 104 (41.6 %) men) to assess the extent to which the barriers to correcting others, as identified in literature across disciplines, apply to correcting misinformation on social media. We also group the barriers into factors and explore demographic differences amongst them. It has been suggested that users are generally hesitant to challenge misinformation. We found that most of our participants (58.8 %) were reluctant to challenge misinformation. We also identified moderating roles of age and gender in the likelihood of challenging misinformation. Older people were more likely to challenge misinformation compared to young adults while, men demonstrated a slightly greater likelihood to challenge compared to women. The 20 barriers influencing the decision to challenge misinformation, were then grouped into four main factors: social concerns, effort/interest considerations, prosocial intents, and content-related factors. We found that, controlling for age and gender, “social concerns” and “effort/interest considerations” have the significant impact on likelihood to challenge. Identified four factors were analysed in terms of demographic differences. Men ranked “effort/interest considerations” higher than women, while women placed higher importance on “content-related factors”. Moreover, older individuals were found to be more resilient to “social concerns”. The influence of educational background was most prominent in ranking “content-related factors”. Our findings provide important insights for the design of future interventions aimed at encouraging the challenging of misinformation on social media platforms, highlighting the need for tailored, demographically sensitive approaches.
Cyberbullying detection for low-resource languages and dialects: Review of the state of the art
2023, Information Processing and Management
The struggle of social media platforms to moderate content in a timely manner, encourages users to abuse such platforms to spread vulgar or abusive language, which, when performed repeatedly becomes cyberbullying — a social problem taking place in virtual environments, yet with real-world consequences, such as depression, withdrawal, or even suicide attempts of its victims. Systems for the automatic detection and mitigation of cyberbullying have been developed but, unfortunately, the vast majority of them are for the English language, with only a handful available for low-resource languages. To estimate the present state of research and recognize the needs for further development, in this paper we present a comprehensive systematic survey of studies done so far for automatic cyberbullying detection in low-resource languages. We analyzed all studies on this topic that were available.We investigated more than seventy published studies on automatic detection of cyberbullying or related language in low-resource languages and dialects that were published between around 2017 and January 2023. There are 23 low-resource languages and dialects covered by this paper, including Bangla, Hindi, Dravidian languages and others. In the survey, we identify some of the research gaps of previous studies, which include the lack of reliable definitions of cyberbullying and its relevant subcategories, biases in the acquisition, and annotation of data. Based on recognizing those research gaps, we provide some suggestions for improving the general research conduct in cyberbullying detection, with a primary focus on low-resource languages. Based on those proposed suggestions, we collect and release a cyberbullying dataset in the Chittagonian dialect of Bangla and propose a number of initial ML solutions trained on that dataset. In addition, pre-trained transformer-based the BanglaBERT model was also attempted. We conclude with additional discussions on ethical issues regarding such studies, highlight how our survey improves on similar surveys done in the past, and discuss the usefulness of recently popular AI-enhanced tools for streamlining such scientific surveys.
FALCoN: Detecting and classifying abusive language in social networks using context features and unlabeled data
2023, Information Processing and Management
Social networks have grown into a widespread form of communication that allows a large number of users to participate in conversations and consume information at any time. The casual nature of social media allows for nonstandard terminology, some of which may be considered rude and derogatory. As a result, a significant portion of social media users is found to express disrespectful language. This problem may intensify in certain developing countries where young children are granted unsupervised access to social media platforms. Furthermore, the sheer amount of social media data generated daily by millions of users makes it impractical for humans to monitor and regulate inappropriate content. If adolescents are exposed to these harmful language patterns without adequate supervision, they may feel obliged to adopt them. In addition, unrestricted aggression in online forums may result in cyberbullying and other dreadful occurrences. While computational linguistics research has addressed the difficulty of detecting abusive dialogues, issues remain unanswered for low-resource languages with little annotated data, leading the majority of supervised techniques to perform poorly. In addition, social media content is often presented in complex, context-rich formats that encourage creative user involvement. Therefore, we propose to improve the performance of abusive language detection and classification in a low-resource setting, using both the abundant unlabeled data and the context features via the co-training protocol that enables two machine learning models, each learning from an orthogonal set of features, to teach each other, resulting in an overall performance improvement. Empirical results reveal that our proposed framework achieves F1 values of 0.922 and 0.827, surpassing the state-of-the-art baselines by 3.32% and 45.85% for binary and fine-grained classification tasks, respectively. In addition to proving the efficacy of co-training in a low-resource situation for abusive language detection and classification tasks, the findings shed light on several opportunities to use unlabeled data and contextual characteristics of social networks in a variety of social computing applications.
Namespotting: Username toxicity and actual toxic behavior on reddit
2022, Computers in Human Behavior
Citation Excerpt :
The data analysis was performed with the assistance of Samurai, a proprietary system developed by Samurai Labs.4 Samurai is a technology which employs a combination of symbolic and statistical approaches to detection of various types of online violence, applied in a number of studies in recent years (Bilewicz et al., 2021; Ptaszynski et al., 2021; Ptaszyński, Leliwa, Piech, & Smywiński-Pohl, 2018; Urbaniak et al., 2022). In the present study we used a relevant selection of Samurai Labs tools.
Without relying on any user reports, we use algorithmic detection and Bayesian statistical methods to analyse two large data streams (329 k users) of Reddit content to study the correlations between username toxicity (of various types, such as offensive or sexually explicit) and their online toxic behavior (personal attacks, sexual harassment among others).
As it turns out, username toxicity (type) is a useful predictor in online profiling. Users with toxic usernames produce more toxic content than their neutral counterparts, with the difference in predicted mean increasing with activity (predicted 1.9 vs. 1.4 toxic comments a week for users with regular activity, and 5.6 vs. 4 for top 5% active users). More users with toxic usernames engage in toxic behavior than among neutral usernames (around 40% vs. 30%). They are also around 2.2 times more likely to have their account suspended by moderators (3.2% vs. 1.5% probability of suspension for regular and 4.5% vs. 2% for top 5% users)—detailed results vary depending on the username toxicity type and toxic behavior type. Thus, username toxicity can be used in the efforts of online communities to predict toxic behavior and to provide more safety to their users.
Do Digital Jobs Need an Image Filter? Factors Contributing to Negative Attitudes
2023, SSRN

View all citing articles on Scopus

^☆: During the course of the study, we have utilized content that is publicly available on Reddit.com and can be accessed via the Reddit API or other similar technologies. This study was not interventional research. Moreover, although Reddit usernames are anonymous and usually do not display any personal information, we have additionally anonymized each one of them. For these reasons, no informed consent was required (following point 8.05 of the Ethical Principles of Psychologists and Code of Conduct of the American Psychological Association). Samurai Labs contributed 3000 PLN to the funding of this research. Samurai Labs prepared the raw data and results of personal attack recognition, and provided the humanpower to manually check which of 540 outliers (even after the initial filtering out of bots) were bots. Some computing power was provided by National Science Centre research grant number 2016/22/E/HS1/00304. The following authors have affiliation and involvement in Samurai Labs: Michał Ptaszyński, Patrycja Tempska, Gniewosz Leliwa, Maciej Brochocki, Michał Wroczyński— they are either co-founders or employees of Samurai Labs. Study design, R code, statistical analyses, their explanation and visualizations are due to Rafal Urbaniak (who has no interest in results going one way or the other, and no incentives related to any specific outcome were involved), theoretical discussion and further editorial work on the manuscript are joint effort. Full.Rmd file with analysis code and anonymized datasets is available at https://rfl-urbaniak.github.io/redditAttacks/.

¹: https://rfl-urbaniak.github.io/ (R. Urbaniak).

²: http://arakilab.media.eng.hokudai.ac.jp/~ptaszynski/ (M. Ptaszyński).

³: https://www.linkedin.com/in/patrycjatempska/ (P. Tempska).

⁴: https://www.linkedin.com/in/leliwa/ (G. Leliwa).

⁵: https://www.linkedin.com/in/maciejbrochocki/ (M. Brochocki).

⁶: https://www.linkedin.com/in/michalwroczynski/ (M. Wroczyński).

View full text

Personal attacks decrease user activity in social networking platforms☆

Highlights

Abstract

Section snippets

The influence of social media

Technology applied for personal attack detection

Large-scale quantitative analysis of impact of personal attacks on reddit user activity

Discussion: potential solutions to personal attacks, verbal aggression and hate speech

Conclusions and future work

CRediT author contribution statement

International Journal of Human-Computer Studies

EClinicalMedicine

Government Information Quarterly

Journal of Econometrics

Computers in Human Behavior

Computers in Human Behavior

Computers in Human Behavior

Personality and Individual Differences

Computers in Human Behavior

Journal of Adolescence

Whatsapp and academic performance among undergraduate students in Ghana: Evidence from the university of cape coast

The Journal of Education for Business

A new look at the statistical model identification

IEEE Transactions on Automatic Control

The impact of WhatsApp use on academic achievement among saudi medical students

Medical Teacher

Satiation in an evolutionary model of structural economic dynamics

Talk to me: Foundations for successful individual-group interactions in online communities

Anonymously hurting others online: The effect of anonymity on cyberbullying frequency

Psychology of Popular Media Culture

Predicting cyberbullying from anonymity

Psychology of Popular Media Culture

Ich słowami—obraz pomocy w sytuacjach cyberprzemocy rówieśniczej z perspektywy uczniów

Dziecko krzywdzone. Teoria, badania, praktyka

Cyber-harassment: A study of a new method for an old behavior

Journal of Educational Computing Research

Hate speech epidemic. the dynamic effects of derogatory language on intergroup relations and political radicalization

Political Psychology

Artificial intelligence against hate: Intervention reducing verbal aggression in the social network environment

Aggressive Behavior

Victims' perceptions of traditional and cyberbullying, and the psychosocial correlates of their victimisation

Emotional & Behavioural Difficulties

Consequences of cyberbullying and social overload while using SNSs: A study of users' discontinuous usage behavior in SNSs

Information Systems Frontiers

The influences of participation and moderation on the development of a sense of virtual community

International Journal of Web Based Communities

You can't stay here: The efficacy of Reddit's 2015 ban examined through hate speech

Understanding radicalization process in online far-right extremist forums using social influence model

Some statistical models for limited dependent variables with application to the demand for durable goods

Econometrica

Virtual violence II: Progress and challenges in the fight against cyberbullying

Exploring the relationships of social media usage and symptoms of anxiety and depression in adolescents

Online engagement factors on Facebook brand pages

Social network analysis and mining

‘stormfront is like a second home to me’ on virtual community formation by right-wing extremists

Information, Communication & Society

Loyalty in online communities

The environment and disease: Association or causation?

Cyberbullying: An exploratory analysis of factors related to offending and victimization

Deviant Behavior

The evolution of evidence hierarchies: What can Bradford hill's ‘guidelines for causation’contribute?

Journal of the Royal Society of Medicine

Self-harm, suicidal behaviours, and cyberbullying in children and young people: Systematic review

Journal of Medical Internet Research

Hidden resilience and adaptive dynamics of the global online hate ecology

Nature

Self-esteem, daily Internet use and social media addiction as predictors of depression among Turkish adolescents

Journal of Education and Practice