Prevalence and personality correlates of Facebook bullying among university undergraduates

Highlights

• Self-reported FB bullying was examined among 226 Greek university students.

• One third of the participants had bullied through FB at least once.

• None of the studied variables predicted FB bullying for females.

• Low Agreeableness and time spent on FB predicted males' FB bullying.

Abstract

The purpose of the present study was to examine the prevalence of cyber-bullying through Facebook in a sample of 226 Greek university undergraduates, and to explore whether big five personality characteristics, narcissism, as well as attitudes toward Facebook, technological knowledge and skills were predictive of such behavior. Participants completed a self-report questionnaire measuring the above constructs. Results indicated that almost one third of the sample reported Facebook bullying engagement at least once during the past month, with male students reporting more frequent involvement than females. Bullying through Facebook was predicted by low Agreeableness and more time spent on Facebook only for males, whereas for females none of the studied variables predicted engagement in Facebook bullying. Findings are discussed in terms of prevention and intervention strategies.

Introduction

Technology-mediated communication, such as social network sites (SNS), has significantly influenced the nature of everyday social interactions. Facebook (FB) is a member-based Internet community that allows users to post personal information and to communicate with others in innovative ways such as sending public or private online messages or sharing photos online. However, while the use of social media can have positive benefits associated with community engagement, education, social connectedness, and identity development, it can also lead to risks linked to social rejection, depression as well as other negative effects for well-being such as cyber-bullying (CB¹) (O'Keefee, Clarke-Pearson, & Council on Communication and Media, 2011). Hinduja and Patchin (2009) defined CB as an intentional act carried out by a group or an individual using electronic forms of contact, repeatedly and over time, against a victim who cannot easily defend him/herself. Most of the definitions include similar characteristics such as durability, repeatability, harassment, disrespect, anonymity, publicity, the intention of the perpetrator and the situation that the victim is defenseless (e.g., Hinduja and Patchin, 2008, Slonje and Smith, 2008, Smith et al., 2008, Tokunaga, 2010). Nevertheless, some researchers suggest that the aspect of repeatability is not essential for CB to occur since when something is uploaded online, it could be seen by thousands of users, especially on FB which bears public friendly features and could be particularly challenging for cyber-bullies (Slonje, Smith, & Frisén, 2013). In all, CB definition is still unclear, especially in the field of SNS, where the criteria are numerous and complicated (e.g., Dredge, Gleeson, & de la Piedad Garcia, 2014).

Cyber-bullying is carried out through the use of mobile devices or personal computers (Smith et al., 2008) including behaviors such as flaming, harassment, denigration, impersonation, outing, trickery, exclusion and cyber stalking (Willard, 2007). Nocentini et al. (2010) proposed four categories of CB through SNS: written or verbal behaviors which could be occurred with instant messages through FB, voice messages, comments and chats, visual behaviors by uploading or posting material such as pictures or videos on FB, segregation by intentionally excluding someone from a FB group and impersonation by imitation, stealing passwords and invading into someone's FB profile account. Other examples of FB bullying include offending or ridiculous comments, invading photos, liking a humiliating photo or reposting it, sending abusive inbox messages, posting false information about others, hacking someone's profile, underrating someone's reputation, uploading publicly nasty and embarrassing photos of someone, posting public humiliating status in someone's profile page, sending cruel or threatening private messages to someone (Dredge et al., 2014).

Although there seems to be a consensus on the behaviors that constitute CB, the measurement of the construct is typically guided by the aim and purpose of each study, with self-report assessments most commonly being used (both paper and pencil, as well as online surveys) since they are easier to implement, better at gaining the individual's perspective and therefore are more likely to reflect intention and power imbalance (Furlong, Sharkey, Felix, Tanigawa, & Green, 2010). However, the majority of cyber-bullying instruments lack the minimum psychometric standards of scale development as a recent review of 44 cyber-bullying instruments concluded, since only 12 of them had been derived using exploratory or confirmatory factor analysis (Thomas, Connor, & Scott, 2014). Additionally, almost half of the instruments included in the review did not use the explanation of the concept of cyber-bullying in the instructions (Berne et al., 2013). In terms of measuring cyber-bullying through SNS, there have been limited efforts worldwide to investigate the phenomenon, but research regarding CB in FB exclusively is extremely limited. For example, Kwan and Skoric (2013) only investigated FB bullying behaviors in secondary school students using a questionnaire based on the scales developed by Cassidy, Jackson, and Brown (2009) and Patchin and Hinduja (2010).

Facebook reported that the number of active monthly users reached 1.06 billion globally in December 2014 (Facebook, 2015), with 890 million daily active users, and 745 million mobile daily active users, with approximately 82.4% of the daily active users to be outside the US and Canada. Tam (2013) reported that of the 193 million U.S. and Canada users, the 25–34 (24.4%) and 18–34 (23.7%) age groups appear to be the two largest groups of the North America users. What is more, U.S. college students reported using FB an average of 10–30 min daily (Ellison, Steinfield, & Lampe, 2007). Greece has a total population of almost 11 millions, and average Internet penetration (56%; approx. 6 million users; European average 68%; global average 34%) (We Are Social, 2014).

Information and Communication Technologies (ICT) offer college students many opportunities to communicate with their peers. However, in their social interactions via FB, students can be confronted with undesirable phenomena such as cyber-bullying. Thus, while many FB related studies have explored the reasons behind FB use, there has been a lack of systematic investigation examining factors that might explain users' engagement in risky FB use, such as CB. Previous researchers have looked at the association between personality traits and Internet use, in general (Devaraj, Easley, & Crant, 2008) and social media, such as FB, in particular (e.g., Amichai-Hamburger & Vinitzky, 2010), suggesting that certain personality features are better predictors of FB use than others (Moore & McElroy, 2012). However, it is not yet clear if the same personality dimensions are also responsible for online aggression through FB. Therefore, the present study set out to investigate the role of big five personality factors (Neuroticism, Extraversion, Openness-to-Experience, Agreeableness and Conscientiousness), and Narcissism in relation to FB bullying. Moreover, while personality traits may potentially influence the way individuals use FB, it is not clear how attitudes toward FB may be associated with CB involvement. While there has been plenty of research looking into CB across different platforms, a limited number of studies, particularly among adolescents, have specifically examined bullying over FB. Since university students are among the heavy FB users, it is not yet clear how prevalent this type of aggression is among this group of users. Therefore, this study will also provide evidence regarding this issue.

Given the dearth of research on CB through FB among college students, an overview of CB prevalence will be presented. Research with adolescents (13–17 yrs old) shows that more than half of FB users have experienced at least one form of FB bullying in the past year (Kwan & Skoric, 2013). Nevertheless, although CB has been theorized to peak in early adolescence and then to significantly decrease after high school (Tokunaga, 2010), CB among college students ranges from 8% (Slonje & Smith, 2008) to 9% (MacDonald & Roberts-Pittman, 2010). Cyber-bullying research among university undergraduates in Greece showed rates of perpetrators varying between 16% and 14% (Sygkollitou et al., 2010, Kokkinos et al., 2014). Cyber bullies/victims in the latter study represented 33% of the participants. In the US, Gibb and Devereux (2014) found that 14.3% of their participating college students were cyber bullies. Likewise, Whittaker and Kowalski (2015) found that almost 12% of their undergraduate participants committed CB, whereas CB and cyber-victimization were positively related.

Nevertheless, rates of victimization have been higher – between 22% (MacDonald & Roberts-Pittman, 2010) to over 50% (Gibb & Devereux, 2014), or from 9% to 34% (Baldasare, Bauman, Goldman, & Robie, 2012). Arıcak (2009) and Dilmac (2009) found that over half (54.4% and 55.3%, respectively) of Turkish college students had been cyber-bullied in their student life, and approximately one-fifth (19.7% and 22.5%, respectively) had cyber-bullied others. Although prevalence rates among college students vary widely, the results from all the studies suggest that a substantial portion of college students are victims and/or perpetrators of CB.

Thus, although the quantification of CB prevalence rates among college students has been attempted by a number of researchers, the findings vary from study to study due to the use of different CB definitions, time frames (from lifetime prevalence to the last two months), item wording and number, response options, as well as the behaviors studied. Nevertheless, the existing evidence shows that CB is not unknown among college students.

Greece is a country with average Internet and social media penetration compared to the rest of the European countries, whereas in terms of the global average it is well above. Therefore, one would expect that increased Internet use would be linked to online aggression. Recent evidence with a Greek sample of university undergraduates showed that the frequency of Internet use was positively associated with CB, especially in the case of cyber-bully/victims, who also used chat-rooms more frequently, as well as IM programs, and SNS compared to pure bullies and pure victims (Kokkinos et al., 2014). Findings from similar research suggest that time spent on SNS, the most common medium of CB among college and university students, predicts involvement in CB (Lindsay and Krysik, 2012, Walker et al., 2011). Thus, there appears to be systematic evidence linking time spent online and CB prevalence, confirming that Internet use (e.g., via mobile devices) is a strong predictor of CB and aggressive online behavior (e.g., Juvonen and Gross, 2008, Rosen et al., 2013, Suler, 2004). Juvonen and Gross (2008) reported that ICT are not the reason nor the motive of CB but the means used to bully others, suggesting that the more electronic communication tools someone uses, the more the odds for him/her to use them antisocially. Livingstone and Haddon, 2009, De Haan and Livingstone, 2009 and Livingstone, Haddon, Görzig, and Ólafsson (2011) linked Internet connection with online opportunities and risks, meaning the more opportunities someone has on the Internet, the more dangerous it could be. On the other hand, O'Neill and Dinh (2015) indicated that ICT penetration is simply one of the factors leading to CB. Finally, Li (2007) suggested that ICT penetration, and generally technology access, tend to increase the possibilities of online aggression. She also found that the frequency of computer use was a significant predictor of possible CB incidents.

Engagement in CB was found to be positively related, to some degree, to technological skills (Hinduja and Patchin, 2008, Li, 2007). Moreover, the intensity of FB use, as well as the advanced Internet use were also positively related to FB bullying among adolescents (Kwan & Skoric, 2013). A recent survey by Ellison, Steinfield, and Lampe (2011) revealed that FB use may even be on the rise among college students, with their sample averaging more than 80 min of FB use a day, although Kowalski, Giumetti, Schroeder, and Reese (2012) found that time spent online was not associated with CB.

As for the Internet use, cyber-bullies were found to make more Internet use in general (Smith et al., 2008, Ybarra, 2004) and use of IM programs (e.g., FB chat) in particular (Ybarra & Mitchell, 2004b). Aricak et al. (2008) found a positive association between frequency of Internet use and certain CB forms (i.e. sending infected emails and saying things online that would not be said face to face), while Telkok, Cakir, and Tural (2015) reported that risky Internet users spent more time online per day than non risky ones, a finding which resembles other research evidence (e.g., Erdur-Baker, 2010). Likewise, Kwan and Skoric (2013) indicated that the intensity of FB use does not play an important role on FB bullying incidents, but the risky usage is the leading factor of engaging in such behaviors.

In Ybarra and Mitchell's (2004a) study, Internet aggressors rated themselves as having almost expert or expert knowledge of the Internet. Rigby (2007) claimed that the power in the online world is associated with superior technological knowledge. Indeed, recent technological advances are vast pointing to a digital divide (instead of generational gap) which underlines the differences among the types, uses, and knowledge of technology between the generations, meaning that current generations waste more time using, and are more familiar with, ICT (Singh, Veron-Jackson, & Cullinane, 2008). Whittaker and Kowalski (2015) reported that most college students knew how to respond to a CB incident on social media, including blocking the person who did the action and reporting him. Nevertheless, taking a picture using a mobile phone camera and posting it online or creating a fake SNS profile require only basic skills and information, whereas other forms of CB (i.e. modifying pictures) entail more advanced skills, although these CB forms are relatively less common (Smith et al., 2008).

Personality has been associated with the way individuals interact with and maintain their social relationships. In 1974, Rosengren argued that individual differences, such as age, gender, and personality, influence the use of mass media, a theory that has been successfully applied in research relating to preferences for movies, music, and television shows, as well as books and cultural activities (Ryan & Xenos, 2011). Due to its relevance to social behavior, the Big Five factors have recently been employed to investigate the use of certain forms of online social media, such as SNS (Amichai-Hamburger and Vinitzky, 2010, Ross et al., 2009). In addition to predicting general online behaviors, big five personality factors have also been found to correlate with patterns of FB use.

The Five-Factor Model of personality (McCrae & Costa, 1997) represents the dominant conceptualization of personality structure in the current literature. This model posits that the Big Five personality factors of Neuroticism (N), Extraversion (E), Openness to Experience (O), Agreeableness (A), and Conscientiousness (C) reside at the highest level of the personality hierarchy, and are considered to include the entire domain of more narrow personality traits that fall at lower-levels of the hierarchy.

Narcissism is conceptualized as a normative personality trait, including both adaptive and maladaptive behaviors (Miller & Campbell, 2010), particularly grandiose ones. Narcissism is predictive of aggression proneness, authoritarian interpersonal problems, resistance to negative feedback, and manipulativeness (e.g., Locke, 2009), but has also been positively associated with psychological health, and well-being (e.g., Brown, Budzek, & Tamborski, 2009). Individuals with narcissistic traits are attracted to the Internet and particularly to SNS, since they provide them with the ability for instant display to a large crowd. However, when these individuals perceive social neglect through FB, are more likely to be angered about not getting attention paid to their FB status updates, suggesting that the grandiose exhibitionist aspect of narcissism appears to be its more strongly anti-social aspect with relation to FB use (Carpenter, 2012). Taken from a socio-cultural perspective, narcissism is viewed as a characteristic of societal change towards individualism due to the rapid growth of technology, a finding that has been confirmed among U.S. college students after the year 2000 (Twenge, Konrath, Foster, Campbell, & Bushman, 2008). Therefore, the characteristics of SNS foster a generation with extreme digital narcissism which anticipates constant feedback and validation (Keen, 2007).

Research findings on FB bullying and personality are limited. The existing studies focusing on CB engagement and personality suggest that cyber bullies are mainly characterized by low levels of A, C, O and high levels of N as well as narcissism (e.g., Karl, Peluchette, & Schlaegel, 2010). The evidence about Extraversion are conflicting, in that some researchers argue that cyber bullies tend to report higher levels of E (e.g., Baldasare et al., 2012) whereas others, low (e.g., Mishna, Cook, Gadalla, Daciuk, & Solomon, 2010).

Research has shown that people with low levels of A, due to low empathy, lack of effective management of interpersonal hostility and disagreements (McCullough, Bellah, Kilpatrick, & Johnson, 2001), are more likely to be rude, argumentative, vengeful, inconsiderate, and uncooperative, as well as inclined toward such interpersonal deviant behavior. They are more likely to express inappropriate and antisocial behaviors on FB such as making fun of others or saying something hurtful (Karl et al., 2010), and may seek for revenge through Internet because in that way they feel more powerful (e.g., Baldasare et al., 2012).

Cyber bullies with low levels of E tend to be lonely, introverted, and pessimistic, with low self-esteem. They feel weak in the real world and therefore may choose the Internet as a space where they can behave aggressively to compensate for their weakness (Mishna et al., 2010). Instead, bullies with high levels of E, could be characterized by determination, selfishness, impulsiveness, tendency to dominate, as well as leadership and narcissistic behavior (e.g., Baldasare et al., 2012).

Conscientiousness is associated with possessing an interpersonal orientation, an adherence to social norms, and engaging in positive, pro-social behaviors (Roberts, Jackson, Fayard, Edmonds, & Meints, 2009). Therefore, those high in C are less likely to exhibit inappropriate and antisocial Internet behavior. Moore and McElroy (2012) found that C was associated with greater regret over inappropriate FB posts. Moreover, individuals who bully online and have low levels of C and A are more likely to score higher on psychoticism, an individual's tendency to behave violently and cruelly without empathy or compassion (Arıcak, 2009, Ybarra and Mitchell, 2007).

Individuals with neurotic personality characteristics tend to be anxious, emotionally unstable when socially exposed (McCrae & Costa, 1999). Moreover, high levels on this trait predict unsuitable and unfriendly behavior on FB (Karl et al., 2010). Cyber bullies with neurotic traits may be frustrated individuals who have low self-esteem, low self-control, poor social perception, elevated depressive symptoms, and sensitivity to develop negative thinking about themselves and their social life, while they might behave aggressively on the Internet to compensate their weakness for the real world (Mishna et al., 2010, Sontag et al., 2011).

Openness to Experience is most likely to be associated with using an SNS to seek out new and novel experiences (Butt & Phillips, 2008). Therefore, these individuals are more likely to be active FB users. High O persons have a greater tendency to be sociable through FB and disclose personal information (Amichai-Hamburger & Vinitzky, 2010). However, since FB has become a relatively mainstream communication tool for university students (Ellison et al., 2007), it may no longer constitute a “unique” experience for them. Although there appears to be no evidence regarding bullying through FB for open college students, studies with preadolescents have shown the negative association between O and CB (Kokkinos, Antoniadou, Dalara, Koufogazou, & Papatziki, 2013).

Individuals who score high on narcissism may show a broad, positive and unrealistic self-concept and use self-regulatory strategies for positive self-promotion maintenance (Campbell & Foster, 2007). As a result, these individuals are more likely to update their profile frequently and post more provocative photos (Carpenter, 2012). In case they discover negative comments about themselves or do not receive the social support and the attention they seek, they might exhibit hostile behavioral tendencies.

Research findings regarding the engagement of men and women in CB are contradictory. Some studies have found that men are more likely to report being involved in CB (Li, 2006, Slonje and Smith, 2008) with most of these behaviors being encountered in SNS (MacDonald & Roberts-Pittman, 2010), whereas other studies have found that women were more likely to participate in CB (Rivers & Noret, 2010). Still however others have failed to find significant gender differences in CB involvement (Kowalski et al., 2012, Smith et al., 2008), while Hinduja and Patchin (2010) concluded that the observed gender differences represent an aftereffect of different CB behaviors, with women spreading rumors (indirect form), whilst men responding aggressively and negatively in comments or posting humiliating photos and/or videos with mischievous behaviors (direct form).

The evidence regarding gender differences in FB use appear to be consistent among studies with college students. Women comprise over 56% of the overall FB population (Facebook, 2010), and are more active on FB in terms of actual time spent than men (Lenda & Aiello, 2010). Furthermore, 74% of women reported FB as their favorite website compared to 60% of men (Hoy & Milne, 2010). The Rapleaf study (Macmanus, 2008) showed that of approximately 2.5 million people (most of them in the U.S.), aged between 18 and 24 years, about 1.7 million women had used FB, compared to 977,753 men. Similarly, Thomson, Ches, and Lougheed (2012) found that female undergraduates in their study reported spending almost 62% of their Internet time on FB compared to 44% of males. They also reported significant differences between female and male “heavy” FB users (more than one hour per day) and in the minutes spent daily examining others' FB profiles (females – 24 min; males – 10 min). Therefore, women may be at a greater risk for getting involved in negative online behaviors compared to men.

In Greece, Facebook penetration is 41% (European average 40%; global average 26%), with 54.3% users being males and 45.7% females (Allin1social, 2015). In terms of users' age breakdown, the most frequent FB users belong in the 25–34 years age group (28%; 54% males). The age group of 18–24 years, where university students are more likely to belong, is the third more frequent category of FB users (22%; 53% males), and those aged between 35 and 44 years belong to the second category (23%; 54% males) (We Are Social, 2015). Meanwhile, an unpublished study by Kakogiannis (2012) showed an elevated percentage of 83.56% among Greek university FB users, 44.86% of whom were males and 55.14% females.

Research on the attitudes toward FB is limited. There is evidence to suggest that attitudes have some effect on how members or non members are behaving on FB (Acquisti & Gross, 2006), whereas Kayri and Cakir (2010) noted that individuals who frequently use FB tend to have more positive attitudes towards FB. Teppers, Luyckx, Klimstra, and Goossens (2014) found that women are likely to hold more positive attitudes toward FB, while Lampe, Ellison, and Steinfield (2008) in their longitudinal research reported that FB members tend to have more positive attitudes as they grew older.

Positive attitudes toward FB were found to be positively associated with high levels of E, O and N (Orr et al., 2009, Teppers et al., 2014), personality traits that are likely to characterize cyber bullies. Those who possess these traits and use FB may behave narcissistically, impulsively and aggressively, be resentful and ruthless, and are more likely to cyber bully on FB (e.g., Baldasare et al., 2012, Mishna et al., 2010).

In all, research findings seem to agree that attitudes toward FB are positively related with the time spent on FB, low levels of A and C, characteristics that interface CB (e.g., Baldasare et al., 2012, Carpenter, 2012, Smith et al., 2008, Ybarra and Mitchell, 2007), while a positive association between attitudes toward FB and FB bullying among adolescents was reported by Kwan and Skoric (2013).

Given the explosion in Facebook's popularity, the present study set to identify the personality characteristics of those university undergraduates, FB users, who are more likely to create an aggressive atmosphere on FB, and to predict those personality and attitudinal characteristics of FB cyber-bullying involvement.

The research discussed so far suggests that FB users' involvement in CB may differ as a function of their individual personality characteristics. In keeping with previous research, this study focused on the narrow traits of Big Five personality factors and narcissism. In addition, attitudes towards FB were examined, as positive attitudes toward FB seem to be associated with personality factors that are implicated in cyber bullying. It is also important to identify, what, if any, personality characteristics are predictive of CB engagement through FB, along with attitudes toward FB, time spent online and technological knowledge and skills.

Thus, the aims of the current study were threefold: to examine the prevalence of cyber bullying behaviors through FB among University undergraduates, to explore the possibility that students with certain personality characteristics were more likely to be engaged in FB cyber bullying, and finally to ascertain whether these characteristics, along with attitudes towards FB and technological knowledge and skills were predictive of this behavior. Therefore, it is postulated that big five personality factors, as well as narcissism would be predictive of FB cyber bullying. Specifically, based on the evidence linking these personality traits to aggressive behavior, four major hypotheses were formulated:

Hypothesis 1

Higher scores in A (H1a), C (H1b) and O (H1c) will be negatively associated with FB cyber-bullying.

Hypothesis 2

Higher scores in narcissism (H2a), as well as high N (H2b) will be positively associated with cyber bullying through FB.

Hypothesis 3

University students who use frequently the Internet (H3a), spent more time on FB (H3b), and possess more technological knowledge and skills (H3c) will be more likely engaged in cyber bullying through FB.

Hypothesis 4

Participants who hold positive attitudes toward FB will report higher levels of E (H4a), O (H4b) and N (H4c). With regards to the association between E and FB cyber-bullying, no specific hypotheses were formulated due to inconclusive evidence.