Social media commenters’ understanding of the antecedents of teacher-targeted bullying

This article reports on the findings of a small-scale, extant, qualitative social media research study on commenters’ understanding of the antecedents of teacher-targeted bullying. Comments on an article posted by Sarah Sorge (2013) on The Educator’s Room were used as data source. Guided by an ecological model and the attribution theory, the study identified victim and perpetrator attributes, colleagues’ indifference and unprofessionalism, school management’s lack of leadership and failure to address the problem, as well as socio-cultural factors and policy changes as antecedents of teacher-targeted bullying. It is argued that conventional teacher-learner power relations are flawed due to the unsupportive, even antagonistic attitudes of parents, colleagues, society at large, people in leadership positions and policy makers towards the victims of teacher-targeted bullying. It is concluded that, despite ethical dilemmas, the advent of the Internet and social media has created opportunities for researchers to use comments posted on the Internet as a data source to investigate teacher-targeted bullying.


Introduction
Since the publication of research by Pervin and Turner (1998) and Terry (1998) on the bullying of teachers by their learners in the United Kingdom (UK), there has been a growing research interest in the topic (e.g. De Wet, 2012;Emmerová & Kohútová, 2017;Garrett, 2014;Kauppi & Pӧrhӧlä, 2012a). Pӧrhӧlä (2012a:1060) note that, despite this interest, "research on the victimisation of teachers is still scarce, and understanding of the phenomenon is rather limited." Hence there is a need for further research on teachertargeted bullying (TTB).
Two-thirds of adults worldwide used the Internet in 2015. Internet access rates are over 80% in advanced i economies such as Canada, Australia and the United States of America (USA). In emerging economies adult Internet access varies from as high as 72% in Turkey and Russia to as low as 42% in South Africa. Internet access rates are lower in poorer countries, mostly in sub-Saharan Africa, and parts of Asia, including 25% in Ghana, 21% in Tanzania and 15% in Pakistan (Poushter, 2016). Despite the lower levels of access to the Internet compared with the global median (67%), adult Internet users in emerging economies are more likely to use social media in comparison with those in developed countries. Poushter (2016:21) found that [o]nce online, people in emerging and developing nations are hungry for social interaction. Majorities of adult Internet users in almost every emerging and developing nation surveyed say that they use social network sites, such as Facebook and Twitter. Comparatively fewer online adults in advanced economies say they use social networks, though half or more still report using social media in these countries.
The proliferation of Internet access and social media users worldwide provides "new avenues for researchers across multiple disciplines … to collect rich, vast, and networked data" (McCay-Peet & Quan-Haase, 2018:14). Despite the posting of videos on YouTube, remarks on Twitter and comments on anti-bullying and school violence websites on TTB (Jordaan, 2018), no evidence could be found of researchers using social media to investigate TTB (Snelson, 2016). With this article I aim to fill the aforesaid hiatus in TTB research by reporting on the findings from a small-scale social media research project on commenters' ii understanding of the antecedents of TTB. This article not only expands the existing body of knowledge on TTB, but also sheds light on the use of social media as a data source for the investigation of TTB in an advanced economy (USA). I argue that social media may be seen as a viable data source in advanced and emerging economies due to the expansion of Internet access and the use of social media in these countries.
An extensive literature search revealed that researchers predominantly use quantitative research methods (e.g. De Wet & Jacobs, 2006;Kauppi & Pӧrhӧlä, 2012a;Kopencký & Szotkowski, 2017;Özkılıç, 2012;Terry, 1998) to investigate TTB. Despite a broad literature search I was able to identify only three qualitative empirical research papers on TTB (De Wet, 2010, 2012, 2019. Whereas the abovementioned quantitative and qualitative studies investigated TTB from the perspective of the teachers, Chen andJames et al. (2008) conducted quantitative studies among learners. Researchers claim that TTB is a serious, prevailing problem (e.g. De Wet & Jacobs, 2006;Woudstra et al., 2018), yet only two longitudinal studies could be found that shed light on the escalation of TTB: Kõiv's (2015) comparison of the prevalence of TTB over a ten-year period (2003 versus 2013) in Estonia found a statistically significant increase in 11 of the 15 identified forms of TTB. James et al. (2008), on the other hand, found a decrease in TTB in Ireland. In 2003, 28.2% of the learners who took part in their study admitted that they had bullied their teachers. In 2005, 16.3% learners admitted to being guilty of TTB.
Quantitative TTB studies emphasise the serious negative effects of TTB in the private and professional lives of victims (De Wet & Jacobs, 2006;Espelage et al., 2013;Kopencký & Szotkowski, 2017;Woudstra et al., 2018) and the disintegration of teaching and learning at schools where TTB prevails (De Wet & Jacobs, 2006;Pervin & Turner, 1998;Pyhältӧ et al., 2015;Woudstra et al., 2018). Victims of TTB who took part in Pervin and Turner's (1998:5) study claimed, for example, that "the bullying caused them [teachers] to suffer stress and had made them lower their expectations of teaching as a career." In addition to this, a qualitative study by De Wet (2010) found that TTB may result in the breakdown of the parent-teacher relationship due to the fact that teachers hold parents responsible for their children's behaviour, and parents critique teachers' classroom management.
A qualitative study by De Wet (2012) found that there is a scarcity of research on the antecedents of TTB. The quantitative studies that explore the antecedents of TTB emphasise demographic characteristics, such as the gender (Kauppi & Pӧrhӧlä, 2012b;Özkılıç, 2012) and age (Khoury-Kassabri, Astor & Benbenishty, 2009) of the perpetrators, as well as the age (Emmerová & Kohútová, 2017;Pervin & Turner, 1998), gender (De Wet & Jacobs, 2006Özkılıç, 2012) and years of working experience of the victimised teachers (De Wet & Jacobs, 2006;Emmerová & Kohútová, 2017;Özkılıç, 2012;Pervin & Turner, 1998;Terry, 1998). Antecedents of TTB are also explained in terms of the location of the school (Khoury-Kassabri et al., 2009;Pervin & Turner, 1998). Three studies were identified that moved beyond demographics in search of the antecedents of TTB: De Wet's qualitative study (2012) focuses on individual, institutional and broad societal factors underlying TTB; Steffgen and Ewen (2007) investigate the influence of strain and school culture on school violence in Luxembourg; and Khoury-Kassabri et al. (2009) report on the impact of individual and school-related factors, such as school climate and culture (Jewish and Arab learners), on violence against teachers in Israel. With the exception of the publications by Chen and Astor (2009:9) and Khoury-Kassabri et al. (2009), all of the above studies focus on the antecedents of TTB from the perspective of the victims.
The scarcity of qualitative research on TTB results in a one-dimensional understanding of the phenomenon. The reviewed quantitative research studies focus on the different types of bullying, the negative effects of TTB on teachers' private and professional lives, and the impact of demographic variables on bullying. De Wet (2010Wet ( , 2012, on the other hand, tried to move beyond statistics, and reports on the lived experiences of victims of TTB.

Theoretical Framework
To support my data analysis, I drew on two social theories. The first theory was an adaptation of Johnson's (2011) model of workplace bullying. Johnson's (2011:56) model has its origin in Bronfenbrenner's (2005) ecology of human development. Bronfenbrenner (2005) argues that human development is shaped by factors in four nested layers of hierarchical systems. The ecological model that underpins this study also consists of four interrelated systems, namely the microsystem (the bullying learner(s) and the victimised teacher(s)); the mesosystem (the empathic or apathetic demeanour of the immediate workgroup (colleagues) may act as deterrent or incitement for TTB); the exosystem (the attitude and behaviour of school principals and/or management towards victims or perpetrators of TTB may either discourage or provoke TTB); and the macrosystem (social-cultural, political and societal norms and values may have a positive or destructive influence on teacher-learner relationships) (cf. Johnson, 2011). The study was also informed by the attribution theory. Proponents of this theory argue that we constantly want to explain why "people behave in the way they do" and "what is happening around us" (Kauppi & Pӧrhӧlä, 2012a:1061. In their explanation of this theory, Pӧrhӧlä (2012a:1061) write that people make their attributions of themselves and others according to internal and external attributions. Internal attributions are made when "a person's behaviour is considered to be caused by the individual's characteristics or personality." External attributions come about when the behaviour is thought to be the result of "circumstances or the situation" (Kauppi & Pӧrhӧlä, 2012a:1061. A key concept of the attribution theory, namely fundamental attribution error, is important when the antecedents of TTB are discussed. When people interpret others' positive behaviour, they tend to overemphasise the situational causes and underemphasise the internal ones. The opposite is true when people try to explain other people's negative behaviour. The reverse, as explained in the previous two sentences, is true when people explain their own successes and failures (Kauppi & Pӧrhӧlä, 2012a:1061. Pӧrhӧlä (2012a:1062) call this "selfserving bias."

Research Methodology
This article is positioned within an interpretivist research paradigm. The interpretive paradigm is characterised "by a concern for the individual." The core undertaking of the interpretivist researcher is "to understand the subjective world of human experiences" (Cohen, Mansion & Morrison, 2009:21). This article, therefore, focuses on individual commenters' subjective experiences and understanding of TTB, and aims to gain an understanding of the phenomenon through commenters' postings on a website (cf. Henning, Van Rensburg & Smit, 2004:21). A small-scale, extant, qualitative social media research design was followed.
McCay-Peet and Quan-Haase (2018:17) define social media as "web-based services that allow individuals, communities, and organizations to collaborate, connect, interact, and build community by enabling them to create, co-create, modify, share, and engage with user-generated content that is easily accessible." Social media research, that is, research that uses data obtained from social media itself (Social Media Research Group, 2016), can be either largescale studies that aggregate terabytes of information or, as is the case with the current study, small-scale studies investigating the comments of a limited number of Internet users (McCay-Peet & Quan-Haase, 2018). Extant materialthat is, "material developed without the researcher's influence" (Salmons, 2018:182) and where there is no direct contact between the researcher and the individual commenterswas used as data for this qualitative study. Comments on an article posted by Sarah Sorge (2013) on The Educator's Room were used as data for this study. On 6 May 2013, Sorge posted an article, The bullied teacher, and invited the readers of her post to share their experiences and suggest possible solutions for the problem. Over a period of five years (6 May 2013 to 23 May 2018), 66 readers posted comments in response to the article. Only five of the commenters who took part in the Internet conversation had not been victims of TTB.
During August 2018, I did several Internet searches on TTB using phrases such as "teachers being bullied," "bullied teachers," "teachers bullied by pupils" and "teacher-targeted bullying." I also used synonyms such as "educator," "learner" and "student." Even though my Internet searches were specifically aimed at identifying postings regarding TTB on social media websites, my search resulted in the identification of numerous academic studies and newspaper articles available on the Internet. Even though I read many of the identified comments on the topic on social media websites, I cannot claim that I worked through the entirety of social media websites before identifying the comments on the article by Sorge (2013) as my data source. The sheer size of the World Wide Web makes such an undertaking impossible and improbable.
The guiding principles for qualitative content analysis proposed by Henning et al. (2004) were followed to reduce, condense and group the content of the postings on the website. I immersed myself in the comments (data) of the 61 commenters who had either been victims of TTB or made comments on Sorge's (2013) article or their fellow commenters' posts. I coded the data by hand. I followed an open coding system and awarded codes to different "segments or units of meaning" (cf. Henning et al., 2004:105). After that, related codes were inductively merged and categorised. Being familiar with the context, content and theory of bullying research, and having worked through "the corpus of raw data" (Henning et al., 2004:106) several times, I was able to see the relationship between the different categories and identified four broad themes: what TTB is, the antecedents of TTB, the consequences of TTB and possible solutions to the problem. TTB researchers predominately focus on these four themes. As mentioned in the Introduction, TTB research is, with the exception of a few publications by De Wet (2010, 2012, 2019, quantitative in nature. My decision to concentrate on the four themes may have been influenced by a dearth of qualitative research on the topic, and the opportunity that the existing quantitative research on the topic would afford me to advance the trustworthiness of my findings. In this article I focus, in line with the stated aim thereof, on only one of the four broad themes, namely the antecedents of TTB. Guided by the social theories underpinning this research, this theme was refined, and four subthemes that correspond with Johnson's (2011:56) four interrelated systems, were identified.
To advance the trustworthiness of my study, I attempted to facilitate transferability by presenting thick descriptions that may "enable judgements about how well the research context fits other contexts" (cf. Li, 2004:305) and placed the findings within the existing body of knowledge on TTB (Shenton, 2004). Readers of this article can freely access the data on which this article is based (https://theeducatorsroom.com/the-bulliedteacher/). Readers will, therefore, be able to critique my findings. My decision to use comments from The Educator's Room (2016) (and not another website) as research data was influenced by ease of access, the fact that the site is publicly available and has been operational for some time, and the number and range of comments.
The Educator's Room (2016), a USA-based website launched in May 2012, is not password-or firewall-protected. The website is thus accessible to most Internet users worldwide. Comments by the victims of TTB have been placed in the public domain. The commenters were requested to post their comments anonymously. It is, however, difficult to determine whether or not all of the commenters' usernames are pseudonyms. Some comments were posted under what may be presumed to be the commenter's name or name and surname (e.g. "Deborah," "Marilyn Bullard," "Lisa" and "Jacque Tobin"); others used pseudonyms ("Anon in So Ca" and "Bullied by mean girls"). My decision to use the usernames of the commenters as these appear on the website is supported by the premise that "online pseudonyms already afford users a chosen degree of anonymity and their inclusion in research is appropriate given that the comments are in the public domain" (Raby & Raddon, 2015:170). In addition, Taylor and Pagliari (2018:3) note that "consent to the use of social media data in research is rarely obtained through informed choice but rather assumed on the basis that users have chosen to place it in the public domain." I thus failed to adhere to the traditional ethical principles of anonymity and informed consent. Connected to fears over the infringement of social media users' anonymity is the "risk of harm" that researchers place on the users of social media. Townsend and Wallace (n.d.:7) write that "risk of harm" is imminent for social media users when "more sensitive data" is revealed to "new audiences" such as members of academe, and the possibility that the exposure may result in the "embarrassment, reputational damage or prosecution" of the users. The use of the painful experiences of the bullied teachers purely for the sake of research, and not for the aim to gain greater understanding into the phenomenon, is not in keeping with qualitative research and the interpretative paradigm. It is thus important to state that I engaged with the commenters' postings in a respectful manner and did not trivialise their humiliating experiences and pain. I consulted with the terms and conditions of The Educator's Room's website. I sent an email to an address provided on the website. Nobody responded to my email in which I stated that I was planning to use the comments in response to Sorge's (2013) article as data for a research project. I was however added to The Educator's Room's electronic newsletter distribution list.

Findings and Discussion
Johnson's (2011) ecological model provides the structure for the discussion of the findings of the content analysis. TTB is a multidimensional problem. The discussion highlights that the antecedents of TTB should be sought in four interrelated systems, namely the microsystem (the bully and the victim), the mesosystem (the immediate workgroup), the exosystem (the organisation) and the macrosystem (society).

The Microsystem Level The victims of TTB
Previous studies that explored the antecedents of TTB emphasised the demographic characteristics of the victims and perpetrators of TTB (cf. literature study). The findings from the current study, however, move beyond statistics and highlight commenters' understanding of the personal and professional traits of both victims and perpetrators.
Of the 48 commenter victims of TTB whose gender could easily be identified, 39 were female persons. These commenters used typical female names, such as "Saundra," "Tiffany," "Kath," "Mary" and "Carol." The aforementioned five female commenters mentioned that they were the victims of cyber-bullying (Saundra and Tiffany), vandalism (Kath), "false allegations" and threats of bodily harm (Carol). Mary was bullied on a daily basis. Seven of the commenters could be identified as male persons as they identified themselves with names such as "Carlos," "Jacque," "Pablo" or "Mr O," or wrote "I'm a male teacher." While Jacque wrote that the students made his life a "complete misery," Pablo noted that the bullies "team up" to disrupt his classes. The gender of 11 of the commenters could not be identified from either their usernames or their narratives, for example "Bullied Teacher," "JAD," "Bonpet" and "Anon in So Ca." Superficially, it seems as if female teachers are more prone to being bullied than their male colleagues. Cognisance should, however, be taken of findings from research on gender differences of Internet users and the gender composition of teachers. The fact that more female than male teachers posted comments on The Educator's Room may be indicative of gender differences in social media use and the unbalanced male-female teacher ratio in schools. The Pew Research Center (2017) found that 73% of adult USA female persons and 65% of their male counterparts used at least one social media site during January 2018. During the period from 2015 to 2016, 75% of all teachers in the USA were female (Humanities Indicators, 2017). It would, thus, be unwise to claim that female teachers are more prone to TTB than their male colleagues. This assumption is supported by previous quantitative research. Özkılıç (2012) found, for example, that there is no statistically significant difference between male and female teachers regarding exposure to TTB. In addition, De Wet and Jacobs (2006) found that male teachers are more prone to TTB than their female colleagues. Theoretically, teacherswhether male or female personshave "power" over their learners on the grounds of "their position as teachers" (De Wet, 2019:98). The bullied teachers have, however, lost their power and are placed in a subservient position. One of the core characteristics of bullying is an imbalance of power (De Wet, 2010;Olweus, 1993). Male and female bullied teachers' forced abdication of power at the hands of their learners, and not their gender, makes them "typical" victims or targets of TTB.
A large number of the teachers who wrote about their bullying experiences emphasised the fact that they had been teaching for several years. They often referred to themselves as "veteran" teachers or cited the number of years they had been teaching (e.g. twenty or twenty-five years). Two of the teachers who wrote about their negative experiences noted that teaching was their second career. R wrote in this regard, "I honorably served twenty years in the military before starting my teaching career." These commenters' suggestions that teachers are being bullied despite their maturity or years of experience working with children is supported by previous research (De Wet & Jacobs, 2006). A few of the commenters, however, noted that inexperienced teachers are more often the targets of TTB than their more experienced colleagues. This finding is in line with previous research (De Wet, 2012, Emmerová & Kohútová, 2017Özkılıç, 2012;Pervin & Turner, 1998;Terry, 1998). The data used, namely comments written by victims of TTB out of their own free will on a website, make it impossible to ascertain whether or not age or years' working experience may have a definite influence on teachers' susceptibility to TTB.
Several of the commenters were proud of their accomplishments as teachers, students and artists, as well as their standing in the community. Nat Jones wrote, for example, that she held "two master's degrees and a BA" and was an "accomplished musician." Georgia also noted that she held a master's degree. In one of the narratives, a commenter noted that he or she had been "a respected member of this community and an excellent teacher for over fifteen years." On 1 April 2017, AJ Coco wrote,

I am a dance studio owner and have coached dancers and taught English in my community for over a decade. I've never gotten a bad evaluation and never gotten a complaint. I was always admired by my students and respected by my colleagues.
It seems from the comments that the position of substitute teachers is especially sensitive. Mary, a substitute teacher, wrote, "I am bullied on a daily basis by the students … When I tell the teachers this has happened, the answer is to cancel my future bookings and take me off the request lists." Carlos Miranda, who had been a substitute teacher in New York public schools for six years, warned the readers, "Remember, you are not their teacher, you are just A NANNY … Try and teach them, and it will be to your downfall by false accusations." The comments of the two substitute teachers, Mary and Carlos, recognise that bullying is about power (cf. De Wet, 2019;Olweus, 1993). Mary's plight (being bullied by her learners on a daily basis, as well as colleagues' unwillingness to support her) emphasises the lack of power she holds, and her sense of helplessness. Mary's power is reduced in two ways: in relation to the learners who bully her (microsystem level) and in relation to her colleagues (mesosystems level). Carlos also has no power and his sense of self is reduced to be "just A NANNY" rather than a teacher.
The findings from this study imply that, in accordance with the attribution theory, teachers are being bullied despite positive professional attributes, such as excellent qualifications and a high standing among colleagues and learners, as well as "circumstances" or "situations" outside the victims' control (Kauppi & Pӧrhӧlä, 2012a:1061, such as their position as substitute teachers, age and gender.

The perpetrators of TTB
The teachers who posted comments on the website were extremely critical of their bullies. Shannon Stoney characterised her bullies as "sophisticated, manipulating and intimidating individuals," who bully teachers into giving them better marks. Margaret's post is in line with Shannon's. According to Margaret, she became the target of bullies because the learners were not happy with her high standards -"They just want good grades and no homework." Marilyn Bullard typified her bullies as "selfish [and] entitled teenagers." Deborah's bully was described as "very outspoken and a bit of a ringleader amongst the class (and probably a bit of a bully to some of the other pupils / or they were all scared of her)." Katherine believed that "the privileged status of the kid as a star in his school team" should be regarded as the reason why this boy bullied her. The commenters, furthermore, described their bullies as individuals who "act either aggressively or passive-aggressively toward the teacher," are "full of anger and hatred," "mean," a "drama queen," "accountable to no one" and "a kid on probation." The commenters' description of their bullies as devious, power hungry persons who purposefully rob their teachers of their power, once again portray the bullied teachers as disempowered individuals. The teachers lost their status as professional persons (cf. discussion of the victims of TTB). Only one of the commenters, Jill Mehlinger, gave a medical explanation for her bully's aggressive behaviour, namely attention deficit hyperactivity disorder (ADHD). According to Jill, she "has no control" over this learner who is running havoc in her class.
For victims of TTB to harshly condemn their learners as devious, confrontational, power hungry, inept and lazy individuals is not exceptional in TTB research (cf. De Wet, 2012 for similar findings). Victims of TTB who completed Pervin and Turner's (1998:6) survey identified their bullies as learners with learning disabilities, who were "less able than their peers," disinterested in education and "came from a background where education was not highly valued, and therefore channelled their energy into undermining teachers rather than into the pursuit of knowledge." The suggestion by Deborah that perpetrators of TTB may also be guilty of learner-on-learner bullying is supported by Kauppi and Pӧrhӧlä's (2012b:410) finding that bullying can be seen as "a typical model of behaviour" for learners who bully their teachers.
Even though this article focuses on the bullying of teachers by their learners, it seems from the comments that the bullying of teachers by learners goes hand in hand with the bullying of teachers by their bullies' parents. A beginner teacher wrote that she received "intimidating phone calls/messages from parents." Marilyn Bullard similarly wrote that she was "screamed at and cursed out by aggressive, angry, foul-mouthed, misinformed, so-called 'parents.'" The reason for parents' aggressive behaviour towards teachers may be found in Margaret's commentary. She wrote that children often tell malicious lies about their teachers to cover their own misbehaviour or dreadful marks. This may result in parents either confronting the teachers or lodging complaints against them with management. De Wet (2012:241) similarly found that "parents' aggressive attitudes towards educators" may be seen as a reason for TTB. Pervin and Turner (1998:7) found that parents "did not appreciate the existence of TTB" and thus denied any wrongdoing on the part of their children.
Teachers' inability to stand up to learnerbullies and their "foul-mouthed, misinformed … parents" illustrate targeted teachers' powerlessness. Margaret's and Marilyn Bullard's comments are illustrative of how targeted teachers relinquish their power to parents who may argue that they have the right to encroach onto teachers' professional spheres. This encroachment is made possible by colleagues' (mesosystems level) and school management's (exosystems level) unwillingness to support victims of TTB.

Mesosystems Level: Colleagues
An analysis of the data revealed that antecedents of TTB on a mesosystems level could be found in colleagues' unprofessional conduct, ineptness and apathetic, and unsupportive attitudes towards victims of TTB. Disempowered victims of TTB try to rationalise their victimisation by surveying and critiquing their colleagues.
In a lengthy narrative, Marie blamed lazy and incompetent colleagues for the bullying of diligent teachers: The teachers are horrible and they don't teach, and the students love them for this reason. If a new teacher is hired and actually comes in with the intent of being a good teacher by actually teaching, the kids will bully you to tears and do whatever it takes to get rid of you.
Marie's negative portrayal of her colleagues and her assumption that their indolent, incompetent and unprofessional behaviour should be seen as an antecedent of TTB, is in agreement with research by De Wet (2012). She found that hardworking teachers, who may be seen as disciplinarians, blame their colleagues' lack of work ethics, and their inability to discipline unruly learners, for their plight.
Three of the commenters wrote that they received little support from their colleagues. Wendy Thorpe wrote, for example, that she heard colleagues saying that she was being bullied because of her "lack of emotional intelligence." According to this commenter, her colleagues tended to support the bullying learners "because they have issues." Teachers who are bullied by learners (microsystems level) are further disempowered by their colleagues (mesosystems level) who are reinforcing the message that they are not good enough. Colleagues are thus guilty of victim-blaming. This may have serious negative consequences for the victims' professional and private lives, as attested to by Kath ("I had severe panic attacks"), NateNC ("I'd be lying if I said this hasn't taken a toll on my mentally and physically"), Ellen ("… now I am like a dictator in there. It is too bad, but I have to be") and Jill Mehlinger ("I feel I am being forced into early retirement … and settle for a lower pension since I am not 65 years"). These negative consequences of TTB add to the victims' feelings of powerlessness. A finding of this study, namely that a lack of collegial support contributes to the disempowerment of victims of TTB, is in line with previous research (De Wet, 2012;Steffgen & Ewen, 2007).

The Exosystems Level
A common allegation by the commenters was the lack of support or the inappropriate actions taken by principals and administration when victims of TTB reported their victimisation. According to the Collins Dictionary (2018a:para. 3), "[t]he administration of a company or institution is the group of people who organize and supervise it." The school principal is "the person in charge of the school" (Collins Dictionary, 2018b:para. 2). The principal is also the head of the administration of a school. While some of the commenters used the term "administration" in the analysed comments, others referred specifically to the school principal. It was ambiguous from the reading of the comments whether commenters who critiqued the administration were referring to a specific person, for example the school principal, or, as per definition, to the "group of people," including the principal, who "organize and supervise" the school. In the ensuing account, I respect the preferred term used by the different commenters.
Many of the commenters highlighted the lackadaisical or even bullying attitude of the administration towards victims of TTB. The commenters noted that instead of getting support from the administration, their calls for help when being victimised were used against them. Jill Mehlinger, who was being bullied by a child who had ADHD, reported the learner's misbehaviour to administration. This did not have the required result. On the contrary: "Admin now says I have poor classroom management, and signed me up for behavioral management classes." Deborah noted that she received no "respect or support" from management when false accusations were brought against her by "one horrible lying school pupil." Karen H wrote, "no matter what I said or did, Admin ALWAYS believes the students." The commenters were thus suggesting that the unsympathetic, even antagonistic attitude of the administration resulted in their being bullied by both the learners and the administration.
The plight of the commenters who turned to their principals for help seemed to be a replication of those who turned to their administration: the victims' cries for help were either ignored or the situation deteriorated. Wendy wrote, for example, that she decided not to tell her principal about her victimisation, because she knew beforehand that the principal would "do nothing." Bonpet, who was bullied by five learners in his or her class, reported the bullying to the principal. Bonpet wrote the following about the "support" he or she had received from the principal: My principal promised that he would read them the riot act, and instead, when he came in, gave them the warm fuzzies of how much he loves them, and then left my room. … my admin is willing to throw me under the bus in order to keep a parent happy.
"Anon in So Ca" likewise wrote that the principal was "siding with the kids to advance his agenda. … I'm pretty much being forced out." "Bullied by mean girls" was put on "a performance improvement plan" by the new principal who was, according to this commenter, unaware of the teacher's excellent track record.
The administration and school principals' actions towards the commenters were power laden. They blamed the victims. While the administration told Jill Mehliner that she had poor management skills, Karen H's and Wendy's grievances were disregarded. Bonpet and "Anon in Sa Ca" lamented that their principals sided with the bullies. These power-laden actions sanctioned the negative behaviour of the learners. This sanctioning by people in power (exosystems level) reinforces the perception that the victimised teachers were wrong, that the learners' bullying behaviour was acceptable, and that they were not to be held accountable or responsible. Not being held accountable for their bullying by people in power has the ability to ensure the continuation of the bullying. Whereas the above-mentioned behaviour by the administration and school principals may have reinforced the power of the bullies, it left the victims with a sense of feeling betrayed by the people they turned to for help. While Deborah wrote that she received no "respect or support" from management, Bonpet noted that "my admin is willing to throw me under the bus in order to keep a parent happy." These teachers were disempowered at institutional level.
The current study highlights the negative attitude of school principals and management towards victims of TTB as an antecedent of TTB. In their study on the possible impact of "teacher working environment fit," that is, the interrelation between teachers and their professional community, Pyhältӧ et al. (2015:264) argue that the unwillingness of school principals to support victims of TTB may increase the victims' vulnerability. While the current study, as well as De Wet (2012), Pervin and Turner (1998) and Pyhältӧ et al. (2015) found that school leadership was apathetic towards victims of TTB, Terry (1998) found that the majority (66.3%) of senior staff were sympathetic and supportive towards victims of TTB.

Macrosystems Level
Although researchers concede that socio-cultural, political and economic factors may have an impact on bullying, there appears to be a paucity of research on the issue (De Wet, 2012). This silence is also apparent among the commenters. During my analysis of the data, I could identify only five comments that could be linked to the macrosystems level. A comment by Doreen ("the feelings towards teachers in our society are in a downward spiral") and Marie's statement that TTB started "when education went from a privilege to an entitlement," suggest that society's disregard for education could be seen as an antecedent of TTB. Saundra Delgado believes that she was being targeted by learners because she was female and a member of a minority group (Mexican American).
Two commenters placed the blame for what they perceived to be an escalation in TTB on changes in the United States' federal policy regarding the disciplining of unruly learners (cf. U.S. Department of Education, 2014). R's comments (April 2018) serves as an example: "My first years as a teacher went well. But after the presidential policy of 'leniency' was implemented … classroom behavior became increasingly difficult. I encountered severe teacher bullying by my students because there were NO ramifications for their actions." Commenters who implied that socio-political factors have impacted their relationships with learners, seem to have nowhere to turn for help. The victims were placed in a helpless position as a result of society's disregard for education, cultural biases and what was perceived to be the federal government's condoning of ill discipline. Disrespect for education (and teachers), the entitlement and empowerment of bullies ("when education went from a privilege to an entitlement"), racial or cultural biases, and an unwillingness of government to strengthen the position of teachers, may permeate and spiral down to the school administration, colleagues and individual teachers (victims) and learners (bullies). This may leave the victims of TTB completely at the mercy of their tormentors.
It seems from the data that a school district per se cannot be regarded as a reason for TTB. The commenters were bullied while teaching in upper middle class, affluent, inner-city and poor school districts. Khoury-Kassabri et al. (2009), on the other hand, found that TTB was more prevalent in Israeli schools located in poverty-stricken areas.

Conclusion
Guided by an ecological model and the attribution theory, this article focuses on the attributes of victims and perpetrators, colleagues' indifference and unprofessionalism, school management's lack of leadership and failure to address the problem, as well as socio-cultural and policy changes as antecedents of TTB. The multi-dimensional interplay between what is happening on the different ecological levels should be taken into consideration when trying to explain why learners bully their teachers, and why there seems to be an unwillingness or inability to address this scourge in schools. Bullied teachers have been placed in a subservient position as a result of parents', colleagues', school management's and government's unwillingness to support them. The explicit and implicit condoning of TTB on all four ecological system levels may leave victims with the message that they brought the bullying on themselves.
The aim of this article was not to blame the victims of TTB for their plight. The data analysis has shown that factors outside the realm of the victims may contribute to their victimisation. It should, however, be noted that the commenters blamed the learners, parents, administration, the school principal, colleagues, culture and the change in education policy for their bullying, and suggested that demographic variables (age, years teaching experience and position as, for example, a substitute teacher) may have an impact on victims' vulnerability. These insights can be understood in terms of the attribution theory. The theory suggests that people are "inclined to attribute negative behavior directed at them as being caused by other people or by the characteristics of the situation" (Kauppi & Pӧrhӧlä, 2012a:1066. It may seem as if some of the victims of TTB are unwilling or unable to critically reflect on their own (negative) character traits. The aim of this qualitative study was, however, to report on the commenters' understanding of the antecedents of TTB, and not to derive universal "truths," critique the commenters' "selfserving bias" (cf. Kauppi & Pӧrhӧlä, 2012a:1062, or blame the victims for their plight. As with all interpretative research, the aim of this study was not to generalise findings, but to gain insight into a specific group of commenters' understanding of the antecedents of TTB. There is, however, a need to "compare [my findings] with what goes on in different times and places" (Cohen et al., 2009:22). Very little research has been done on the topic in the South African context (De Wet, 2010Wet, , 2012Wet, , 2019De Wet & Jacobs, 2006;Woudstra et al., 2018). The use of Internet postings in which South African teachers share their TTB experiences as research data may expand the existing body of knowledge on TTB, while taking into consideration South Africa's unique socio-political, economic and cultural landscape. Such a study may give new insights into the ways that parents, colleagues and school administration encourage TTB by either condoning the bullying behaviour of learners, or by blaming the victims.
Being bullied by learners can be a humiliating experience for a teacher, and many such teachers suffer in silence for fear of being labelled "unsuccessful teachers." The advent of the Internet and social media has created immense opportunities for targets of TTB to break this silence and share their plight with colleagues and researchers worldwide. The current study utilised data posted on a USAbased website accessible to most global Internet users. The worldwide increase of access to the Internet to 67% in 2015, and the expansion of social media use by adults, especially those residing in emerging economies, make social media a feasible data source when investigating sensitive topics in developed and developing or emerging countries (Poushter, 2016). Internet access rates are still very low in some parts of Africa, for example Tanzania (21%), Uganda (11%) and Ethiopia (8%) (Poushter, 2016). Restricted and/or limited access to the Internet and social media may impede the use of social media as a data source in some countries. Even though less than half of South African adults (42%) have Internet access (Poushter, 2016), the possibility of using social media as a data source to investigate educational issues such as TTB and violence in schools, should be considered. Numerous videos of South African learners attacking their teachers have been posted on YouTube (e.g. Masuku, 2019). Although this makes social media a rich and easily assessable data source, researchers should be aware of ethical dilemmas when using data that publically shame victims of abuse. The juxtaposing of findings from the current study with findings from previous research on the topic highlights the usability of commenters' posts on the Internet as a rich data source. Users of social media as a data source should, however, always be aware of the fact that they are working with data that were not originally intended to be used in research, and that the commenters have not given informed consent for the use of their comments.
Moving into somewhat unchartered waters regarding the research topic and data collection method, this article addresses a subject often ignored or downplayed by parents, fellow teachers, education leaders and policy makers. Only when confronted with sound empirical research will there be an acknowledgment that teaching can be an extremely stressful profession due to teachers' disempowerment, continual humiliation and abuse at the hands of their learners. This article adds to the limited body of knowledge on TTB and I hope to create an awareness of the plight that some teachers have to face on a regular basis.

Notes
i. Poushter (2016:26) used the World Bank income classification, the per-capita Gross Domestic Product (GDP) based on purchasing power parity, the GDP in current prices and average GDP growth rate over the past ten years to create their economic classification of the 40 countries in their survey on Smartphone ownership and Internet use.

ii.
A commenter is "a person who expresses an opinion or engages in discussion of an issue or an event, especially online in response to an article or blog post" (