See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/330894522

A Qualitative Analysis of Internet Trolling

Article in Cyberpsychology, Behavior, and Social Networking · February 2019

DOI: 10.1089/cyber.2018.0210

CITATIONS READS
65 2,254

2 authors:

Evita March Jessica Marrington

Federation University University of Southern Queensland
68 PUBLICATIONS 2,049 CITATIONS 14 PUBLICATIONS 926 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Evita March on 22 August 2024.

The user has requested enhancement of the downloaded file.

 CYBERPSYCHOLOGY, BEHAVIOR, AND SOCIAL NETWORKING
Volume 22, Number 3, 2019
ª Mary Ann Liebert, Inc.
DOI: 10.1089/cyber.2018.0210

A Qualitative Analysis of Internet Trolling

Evita March, PhD,1 and Jessica Marrington, PhD2

Abstract

Internet trolling is receiving increasing research attention and exploration; however, disagreement and con-
fusion surround definitions of the behavior. In the current study, 379 participants (60 percent women) completed
Downloaded by University Of Queensland from www.liebertpub.com at 01/15/20. For personal use only.

an online questionnaire providing qualitative responses to the following: How do you define Internet trolling?
What kind of behaviors constitutes Internet trolling? Does Internet trolling differ from Internet cyberbullying?
Have you ever been trolled online, and if so how did it feel? Word frequency analyses indicated that Internet
trolling is most commonly characterized as an abusive aggressive behavior. Responses also highlight the
subjective nature of humor in trolling depending on whether an individual has trolled. Interestingly, the groups
that indicated trolling as a ‘‘bullying’’ behavior were the groups who had never been trolled. Results of the
current study highlight the need to differentiate between ‘‘kudos’’ trolling and Cyber Abuse.

Keywords: cyberbullying, trolling, qualitative research, antisocial online behavior

Introduction by some researchers to represent only one form of trolling:

‘‘Flame’’ trolling.11 Trolling has changed from provoking
others for mutual enjoyment and entertainment (termed
I nternet trolling (hereon referred to as trolling) has
received increasing scholarly attention in recent years.
One of the earliest definitions described this behavior as
‘‘kudos’’ trolling) to an abusive behavior that is not in-
tended to be humorous (flame trolling).11
luring others online, commonly on discussion forums, into Confusion in the definition appears in empirical articles
pointless and time-consuming activities.1 Trolling is now exploring trolling. One study12 that analyzed tweets refer-
recognized as an antisocial online behavior,2 defined as a encing the disappearance of Madeleine McCann provided
deliberate attempt by an individual to create conflict and examples of tweets from ‘‘anti-McCann’’ members; however,
distress by communicating inflammatory, provocative, and it was not clear which of the tweets (or if all) were considered
menacing comments to their victim.3 examples of trolling. Research exploring trolling typically
Trolling is characterized as malicious in intent and aims to conceptualizes trolling as ‘‘flame’’ trolling, and this definition
disrupt, aggravate, and lure victims into unproductive con- subsequently informs the behaviors chosen for analysis. In one
versation/argumentation,4 with the power to cause harm and study,13 although the author differentiated between the dif-
distress to victims.5–7 Trolling is associated with a number of ferent forms of trolling, the social media post chosen as a
serious outcomes for victims, with increased risk of suicidal representation of trolling was the following YouTube post
ideation and self-harm.4 Concerningly, a recent online poll about footage of an earthquake in Christchurch, New Zealand:
reported that over one-third of ‘‘Millennials’’ have trolled TO THE FAMILIES OF THE EARTHQUAKE VICTIMS:
online, with topics such as politics and religion sparking the You can rest assured I and the rest of the world are pleased your
largest prevalence of trolling behavior.8 piece of s*** family member is dead and rotting in the ground,
Although trolling is commonly considered as antisocial and we laugh at your suffering and think it’s pathetic you are upset
malicious,2 disagreement and confusion exist regarding the because your family member was an insignificant worm who
definition and defining behaviors,9 with no single fixed mean- was s*** while they were alive and now they are dead squashed
ing.4,10 In one of the first studies to conceptualize trolling,6 four filthy s*** rotting in the ground. Especially those two filthy
distinct properties of this behavior were established: Aggres- babies that were squared REST IN PISS YOU F******
sion (verbally attacking others), deception (employed through RODENT PIECES OF S*** (username Annieberkowitz8){
the use of Internet anonymity), disruption (of social proceed- {
Certain terms were redacted with the authors’ permission in
ings), and success (achieved when the ‘‘troll’’ receives their order to be respectful of all cultures while adequately relaying the
desired response).6 This type of trolling is, however, considered tone of the passage.

1
Federation University Australia, School of Health Science and Life Sciences, Berwick Campus, Australia.
2
School of Psychology and Counselling, University of Southern Queensland, Ipswich Campus, Australia.

192
 INTERNET TROLLING ANALYSIS 193

Another study14 investigating online behaviors toward Participants provided demographic information and
people of different social status presented participants with qualitative answers to the following questions: How do you
two responses intended to reflect trolling: ‘‘No matter how define Internet trolling? What kind of behaviors constitute
hard you try, you’ll never get a good grade coz you’re an Internet trolling? Does Internet trolling differ from Internet
idiot!’’ and ‘‘Which lecturer did you sleep with to get that cyberbullying? Have you ever been trolled online, and if so
grade?’’ Although intended to represent trolling, content how did it feel?
validity is questionable as both responses are antisocial, and Data were entered in NVIVO for word frequency analyses.
neither represent mutual enjoyment and entertainment (i.e., The questions were coded in NVIVO as Definition of In-
‘‘kudos’’ trolling11). ternet Trolling (Question 1), Behaviors of Internet Trolling
In a study2 exploring the effect exposure to trolling had on (Question 2), Internet Trolling and Cyberbullying (Question
subsequent perpetration of trolling, the authors examined 3), and Effects of Internet Trolling (Question 4).
what they considered examples of trolling on social media,
including outright swearing (e.g., what a dumb c***), per-
sonal attacks (e.g., you’re an idiot and one of the things that’s Data analysis
wrong with this country), veiled insults (e.g., Hilary isn’t half As stated above, the aim of the current study was to
the man Bernie is lol), sarcasm (e.g., you sound very white, qualitatively explore how the general public defined and
and very male, must be nice), and off topic statements. Re- characterized trolling. The nature of this study was less ex-
sults showed that negative mood and exposure to previous
Downloaded by University Of Queensland from www.liebertpub.com at 01/15/20. For personal use only.

ploratory than standard inductive thematic analysis, as pre-

negative comments increased the likelihood of the individual vious ‘‘themes’’ of Internet Trolling have already been
to troll. However, it could be argued that not all examples established (i.e., deception, aggression, disruption, and suc-
represented trolling. Considering the differentiation of flame cess6). These themes are reflected in the various definitions
and kudos trolling,11 only two examples (outright swearing of Internet trolling: Luring others online, commonly on dis-
and personal attacks) represented flame trolling, and none cussion forums, into pointless and time-consuming activi-
represented kudos trolling. ties1; Common trolling behavior includes starting aggressive
It seems most definitions of ‘trolls’ and trolling best align arguments and posting inflammatory malicious messages in
with flame trolling, for example, ‘‘Troll–the word itself de- online comment sections to deliberately provoke, disrupt,
fines everything about bullying or harassing someone.’’14 and upset others5; and deception performed in multi-party
Furthermore, examples chosen as representations of trolling interactions, which is conducive to (humorous) entertain-
often reflect flame, not kudos, trolling. Perhaps more con- ment of self and/or other participants, at the expense of the
cerning, some researchers have chosen examples (e.g., sar- deceived target. Considering the inconsistencies regarding
casm) that do not meet the definition of flame, or kudos, definitions of trolling, we believed it worthwhile to initially
trolling. analyze the data with a word frequency analysis to hopefully
Clearly, there is a need for a more unified definition of both confirm and highlight discrepancies in established
trolling and trolling behavior. Effective management of this themes and definitions of Internet trolling.
behavior is dependent upon this unified and consistent defi-
nition.4 Considering the change in meaning of trolling11 and
the current discrepancy about what behaviors actually con- Findings
stitute trolling, further research is warranted. Thus, the aim
of the current study was to qualitatively explore how the Word clouds of the most common words were produced
general public defined and characterized trolling. for 379 qualitative responses to Questions 1, 2, and 4 (Fig. 1).
Word frequency analyses were conducted on these ques-
tions. Table 1 shows the top 5 most frequent words for each
Method of these questions.
Participants Table 2 includes selected examples of responses consis-
tent with the common words for each question.
The sample comprised 379 (40 percent men, 60 percent
women; Mage = 31.16 years, SDage = 11.24) individuals from
the community. Of participants, 47 percent reported being Definition of Trolling and Behaviors of Trolling
trolled by others online, and 25 percent reported trolling
others online. Of men, 54 percent reported being trolled by In addition to the qualitative responses in Table 2, there
others online, and 39 percent reported trolling others online. were also definitions of trolling that did not address the
Of women, 41 percent reported being trolled by others on- themes noted above.
line, and 15 percent reported trolling others online. Essentially tricking people into believing crap you say (of-
fensive or not). Making inflammatory comments and baiting
Data collection people into arguments just for fun; maybe giving bad advice
or something (22, male). Intentionally setting someone up for
Participants accessed an online questionnaire through a a joke or fall to make fun of them (22, female). Being a dick
link advertised on social media (e.g., Facebook and In- (22, male)
stagram). The study was advertised as an anonymous con-
For behaviors that constitute trolling, there were also an-
fidential questionnaire investigating characteristics of online
swers that did not align with the commonly identified words
behavior. Due to the content of the questionnaire, partici-
in Table 2.
pants were provided with contact information for counseling
services if required. Boredom and a desire to get a laugh (21, male)
 194 MARCH AND MARRINGTON
Downloaded by University Of Queensland from www.liebertpub.com at 01/15/20. For personal use only.

FIG. 1. Word Clouds of most frequently used words to define (clockwise from left): Definition of Internet Trolling,
Behavior of Internet Trolling, and Effects of Internet Trolling.
Finally, some behaviors described did not represent either
Table 1. Top 5 Frequent Words for Definition flame or kudos trolling.
of Internet Trolling, Behaviors of Internet
Trolling, and Effects of Internet Trolling Stalker tendencies, obsessive, controlling compulsive (34,
female)
How do What behaviors
you define constitute Internet Effects of Summary
Internet Trolling? Trolling? Internet Trolling Responses to the definition and behaviors of trolling in-
Count Count Count dicate that general attitudes toward trolling best meet more
contemporary definitions, where this behavior is a deliberate
Top 5 words attempt to provoke others by creating conflict and distress.3
Provoke 88 Comments 114 Feel 59 Based on the responses, it appears that trolling has consid-
Deliberate 84 Arguments 61 Annoyed 55 erably deviated from its initial definition as luring others
Reaction 76 Bullying 49 Effect 44 online into pointless and time-consuming activities1 into an
Comments 72 Harassment 48 Upset 25 intentional malicious behavior with the specific aims of
Upset 52 Insulting 43 Angry 24
disrupting and aggravating.4
Count represents the number of times the word appeared in the Several responses indicated the possibility for trolling to
379 qualitative responses. be nonoffensive, deviating from the notion that this behavior
 INTERNET TROLLING ANALYSIS 195

Table 2. Selected Responses for Definition of Internet Trolling, Behaviors of Internet Trolling,
and Effects of Internet Trolling
Question Response
How do you define Posting material on the Internet that is inflammatory, with the deliberate intention of
Internet trolling? provoking an emotional response, dividing people on a contentious subject, or
interrupting an otherwise constructive discourse (28, male).
A person who deliberately comments on social media to provoke anger and upsetting
emotions in people by saying nasty things, commenting inappropriately in conversations,
and generally being annoying online (37, female).
Deliberate use of online services or social media to offend, humiliate, bully, or provide
negative feedback or comments to other persons or parties (44, female).
What kind of behaviors Nasty comments, inappropriate comments, slagging off people (37, female).
constitutes Internet Gaslighting or starting arguments, fooling others to perform actions, being a general
trolling? nuisance (21, female).
Any form of harassment, insults, bullying, threatening, spying, stalking. Use of excessive
foul language in online discussions (23, male).
Have you ever been trolled Most of the times it makes me angry when people are that stupid. Sometimes people have
Downloaded by University Of Queensland from www.liebertpub.com at 01/15/20. For personal use only.

online, and if so how taken it too far calling me names and it’s made me upset even though they don’t know me
did it feel? personally (21, female).
Yes, at first annoyed. But have since come to react mostly with eye-rolling. In the long run,
I’d say the event hasn’t had an effect on me (18, male).
It made me feel awful. I felt angry and sad and also sick to my stomach for several weeks
(48, female).

is inherently malevolent; however, trolling does not have one Trolling and Cyberbullying
single fixed meaning.4,10 The current findings do suggest that To obtain metrics regarding whether participants consid-
some definitions of trolling best meet the criteria for ‘‘kudos ered trolling to differ from cyberbullying, we coded responses
trolling’’ (i.e., provoking others for mutual enjoyment and that begin with either yes or no. After running frequencies on
entertainment).11 Still, considering the most common words only yes/no responses, we found that 62 percent of partici-
to define trolling and behaviors, it seems that trolling is most pants indicated that yes, trolling did differ from cyberbully-
commonly seen as an aggressive abusive behavior (i.e., ing. Below are selected examples of qualitative responses that
flame trolling11). discuss the differences between cyberbullying and trolling.
Effects of Trolling Cyberbullying is assaultive in nature, whereas Trolling sim-
ply seeks to ‘‘get a rise’’ out of somebody (36, male). Internet
In addition to the qualitative responses in Table 2 re- trolling does differ from Cyberbullying, although both are
garding the effects of trolling, participants also commonly behaviors meant to invoke emotional distress. Cyberbullying
noted that they tended to ignore these interactions. is often personal and draws on personal emotions. Internet
I am not particularly affected by internet trolls and prefer to trolling is often anonymous and short lived while Cyberbul-
just ignore them (25, female). Yes. I ignore it generally. lying is enduring and personal (23, male). In my opinion,
Maybe it had the effect of slightly annoying me but only Internet Trolling is a tool that can be used to entertain, edu-
momentarily (39, male). cate, or embarrass. Using a parodied tone, the recipient has
been given a chance to see how ridiculous their stance appears
Furthermore, some participants noted some degree of from the other’s perspective. It can, however, be used to cy-
amusement at being trolled. berbully, in that it can be ceaseless, can evolve into stalking,
I felt pretty OK with it. Yeah I mean, it made me laugh I can contain an intent simply to insult, rather than enlighten or
guess. (18, male). I don’t mind being trolled if it’s funny in a debate. Internet Trolling is not Cyberbullying, but Cyber-
creative way. I’ll laugh as well (21, male). Yes, in a playful bullying can happen by way of internet trolling (25, female).
way (but I didn’t realize until later). First I was mad—then I Also commonly noted by participants was the relationship
laughed about it (35, female).
between trolling and Cyberbullying, with many participants
Summary noting trolling as a form of Cyberbullying.

The top five words to describe how someone felt after they Cyberbullying has a larger scope. I would say that trolling can
be a form of cyberbullying (39, male). I think [Internet
had been the victim of trolling included: Feel, annoyed, effect, Trolling is] a form of cyberbullying (32, female).
upset, and angry. These were considered to be emotive words.
This supports the premise that trolling causes victims harm Furthermore, trolling was often noted as the ‘‘less harm-
and distress.5–7 This negative emotional effect associated with ful’’ version of cyberbullying.
experiencing trolling is reflected in previous research report-
Though trolling does share themes and similarities with Cy-
ing a negative association between experiencing trolling and berbullying, comparing the two would be to compare a lesser
self-esteem.10 The emotive effect of experiencing trolling has or petty crime with full-blown major crime (18, male). Cy-
also been recognized by trollers themselves, who reported that berbullying is directly intended to make the targeted person or
victims most often become angry and respond similarly.9 people receive a negative consequence. Internet trolling is
 196 MARCH AND MARRINGTON

typically done in the name of ‘‘fun’’ (18, male). Cyberbul- cent) than to troll (18.7 percent). Qualitative responses were
lying is inherently malevolent. And targeted at specific users, then explored for these four groups, and the five most fre-
often known in real space. Trolling is more often than not for quent words for each category are shown in Table 3.
purely entertainment with occasional external reasons (ideo- Of particular interest are the responses of individuals who
logical). Targets are typically communities as opposed to
have been trolled but do not troll compared to individuals
individuals (24, male).
who have been trolled and who do troll. In particular, for the
Summary group that have trolled, behaviors include ‘‘joke’’ and effects
Although previous research has described trolling as a form included ‘‘laugh’’ and ‘‘funny’’; meanwhile, for the group
of cyberbullying,15,16 there is evidence to suggest differences in that have not trolled, behaviors included ‘‘offensive’’ and
these behaviors and the perpetrators of these behaviors. Results ‘‘insulting’’ and effects included ‘‘angry’’ and ‘‘upset.’’
of the current study indicate that the majority of participants These responses highlight the subjective nature of humor in
believe that trolling differs from cyberbullying. There were, trolling. For individuals who have both trolled and been
however, many responses that described trolling as a form trolled, they may view trolling more in line with the defini-
(even a less harmful form) of cyberbullying. Considering that tion of humorous, ‘‘kudos’’ trolling. Interestingly, the groups
‘‘bullying’’ was one of the top words used to describe trolling, that indicated trolling as a ‘‘bullying’’ behavior were the
the harassing and bullying nature of the troll cannot be ignored. groups who had never been trolled.
Downloaded by University Of Queensland from www.liebertpub.com at 01/15/20. For personal use only.

The Relationship Between Trolling Overall Conclusions

and Being Trolled
Based on results of this study, the general population
To explore the relationship between being trolled and considers trolling an insulting, bullying, and deliberate be-
trolling, a chi-square test was performed with a significant havior, designed to provoke a reaction which has significant
relationship found, v2 = 59.61 (1), p < 0.001, with a large emotive effects. In short, behavior that is synonymous with
effect size of u = 0.40. Individuals who had previously been the definition of flame trolling. Interestingly, current con-
trolled were more likely to have trolled (81.3 percent) than ceptions of trolling better fit the online behavior of flaming,
not trolled (34.8 percent), and individuals who had not pre- which is defined as, ‘‘an electronic mail [.] or message
viously been trolled were more likely not to troll (65.2 per- intended to insult, provoke, or rebuke.’’1

Table 3. Trolling and Being Trolled: Top 5 Frequent Words for Definition of Internet Trolling,
Behaviors of Internet Trolling, and Effects of Internet Trolling
How do you define Internet Trolling? What behaviors constitute Internet Trolling? Effects of Internet Trolling
Count Count Count
Have never been trolled, have never trolled (n = 180)
Top 5 words
Comments 48 Comments 70
Deliberately 44 Arguments 41
Annoy 43 Harassing 38
Provoking 41 Bullying 35
Upset 34 Annoy 21
Have never been trolled, have trolled (n = 17)
Top 5 words
Annoy 15 Arguments 15
Deliberately 13 Cyberbullying 13
Cyberbullying 13 Annoy 12
Arguments 12 Opinions 12
Joke 12 Funny 11
Have been trolled, have never trolled (n = 96)
Top 5 words
Provoke 32 Comments 34 Feel 42
Annoy 31 Annoying 25 Annoying 36
Comments 26 Feel 23 Effect 22
Intent 25 Insulting 21 Upset 22
Reaction 25 Offensive 19 Angry 20
Have been trolled, have trolled (n = 74)
Top 5 words
Reaction 36 Comments 25 Annoyed 35
Deliberately 30 Arguments 24 Feel 31
Annoy 26 Joke 20 Effect 24
Provoking 24 Reaction 19 Laugh 16
Response 20 Annoy 18 Funny 14
Count represents the number of times the word appeared in the corresponding n qualitative responses.
 INTERNET TROLLING ANALYSIS 197

In fact, the only participants to note the ‘‘funny’’ nature of 3. Buckels EE, Trapnell PD, Paulhus DL. Trolls just want to have
trolling were those who perpetrated this behavior. Do those fun. Personality and Individual Differences 2014; 67:97–102.
perpetrating this behavior believe that they are engaging in 4. Coles B, West M. Trolling the trolls: Online forum users
kudos trolling, where there is mutual enjoyment? Or is it more constructions of the nature and properties of trolling.
likely that they are enjoying the potential harm others experi- Computers in Human Behaviour 2016; 60:233–244.
ence at the expense of their behavior? Research on predictors of 5. Craker N, March E. The dark side of Facebook: The
trolling that have found that Internet trolls have high scores of Dark Tetrad, negative social potency, and trolling be-
every day sadism3,17 suggests the latter. Future research could haviours. Personality and Individual Differences 2016;
further explore these variations in responses by conducting a 102:79–84.
thematic analysis with a hypothetico-deductive approach. 6. Hardaker C. Trolling in asynchronous computer-mediated
communication: from user discussions to academic defini-
To better inform research, differentiation in trolling def-
tions. Journal of Politeness Research 2010; 6:215–242.
initions requires further exploration.11 Future research could
7. March E, Grieve R, Marrington J, et al. Trolling on Tinder
also endeavor to explore sociocultural differences in con- (and other dating apps): examining the role of the Dark
ceptualization of trolling, such as cultural variances in def- Tetrad and impulsivity. Personality and Individual Differ-
initions. Results of the current study suggest significant ences 2017; 110:139–143.
deviation from the initial description of trolling as a message 8. Gammon J. (2014) Over a quarter of Americans have made
that is ‘‘intentionally incorrect but not overly controver- malicious online comments. https://today.yougov.com/
sial.’’1 While it might be empowering for Internet users to news/2014/10/20/over-quarter-americans-admit-malicious-
Downloaded by University Of Queensland from www.liebertpub.com at 01/15/20. For personal use only.

picture these trolls as the ugly creature living under the online-comm (accessed August 1, 2018).
bridge, such an image potentially downplays the seriousness 9. Cook C, Schaafsma J, Antheunis M. Under the bridge: an
of their Internet crimes. Social media posts that have pre- in-depth examination of online trolling in the gaming
viously been included in empirical literature as examples context. New Media and Society 2017; 20:3323–3340.
of trolling (e.g., social media posts of Annieberkowitz813) 10. Thacker S, Griffiths M. An exploratory study of trolling in
are incorrectly, and dangerously, referred to as trolling. It is online video gaming. International Journal of Cyber Be-
time that these behaviors are recognized not as forms of havior, Psychology, and Learning 2012; 2:17–33.
trolling (or even flame trolling), but as e-bile18 and Cyber 11. Bishop J. Representations of ‘trolls’ in mass media com-
Abuse. Furthermore, perhaps the adage of not feeding the munication: a review of media-texts and moral panics re-
trolls was best related to kudos trolling. When it comes to lating to ‘trolling’. International Journal of Web Based
e-bile and Cyber Abuse, these actions should be treated as Communities 2014; 10:7–24.
the cybercrimes they are. 12. Synott J, Coulias A, Ioannou M. Online trolling: The case
of Madeleine McCann. Computers in Human Behavior
Notes 2017; 71:70–78.
13. McCosker A. Trolling as provocation: YouTube’s agonistic
a. An initial question asked participants if they were fa- publics. Convergence: The International Journal of Re-
miliar with the online behavior of Internet trolling. Thirty search into New Media Technologies 2014; 20:201–217.
participants (7.33 percent) indicated no and were subse- 14. Lopes B, Yu H. Who do you troll and why: an investigation into
quently excluded from further participation. These 30 par- the relationship between the Dark Trad personalities and online
ticipants were redirected to the end of the questionnaire and trolling behaviours towards popular and less popular Facebook
thanked for their participation. profiles. Computers in Human Behavior 2017; 77:69–76.
b. An additional question asked ‘‘Have you ever trolled 15. Griffiths M. Adolescent trolling in online environments: a
someone online, and if so who did you troll and why’’; brief overview. Education and Health 2014; 3:85–87.
however, responses to this question were minimal and 16. Hong F, Cheng K. Correlation between university students’
therefore did not achieve saturation. online trolling behaviour and online trolling victimization
c. Percentages are based on row totals. forms, current conditions, and personality traits. Telematics
d. We would like to thank an anonymous reviewer for this and Informatics 2018; 35:397–405.
suggestion for future research. 17. Sest N, March E. Constructing the cyber-troll: psychopa-
thy, sadism, and empathy. Personality and Individual Dif-
Author Disclosure Statement ferences 2017; 119:69–72.
18. Jane E. ‘‘You’re a ugly, whorish, slut’’: understanding e-
No competing financial interests exist. bile. Feminist Media Studies 2014; 14:531–546.
References Address correspondence to:
1. Herring S, Job-Sluder K, Scheckler, et al. Searching for Dr. Evita March
safety online: managing ‘‘trolling’’ in a feminist forum. Federation University Australia
The Information Society 2002; 18:371–384. School of Health Science and Life Sciences
2. Cheng J, Bernstein M, Danescu-Niculescu-Mizil C, et al. Berwick Campus, VIC 3806
(2017) Anyone can become a troll: causes of trolling behav- Australia
iour in online discussions. https://arxiv.org/pdf/1702.01119
.pdf (accessed August 1, 2018). E-mail: [email protected]

View publication stats