The Complex Relationship Between Cyberbullying and Trust

Abstract

Theoretically, there are strong arguments for a relationship between cyberbullying and trust. On the one hand, trust is built on experiences; thus, experiences of malevolence such as cyberbullying might contribute to low trust. On the other hand, high trust may lead to risky online behavior such as self-disclosures that could increase the risk of cyberbullying. As first empirical evidence, we explored this relationship in two cross-sectional studies. Explorative Study 1 (N = 224) showed that negative experiences of family problems and cyber-perpetration predicted low generalized trust. Exploratory Study 2 (N = 196) showed no significant direct relationship, but trust was related to low online privacy concerns and the willingness to self-disclose online was positively related to cyber-victimization and cyber-perpetration. Thus, these studies show mixed evidence and demonstrate that the relationship between cyberbullying and trust might be more complex than assumed. Future longitudinal designs might be illuminating.

Keywords

Cyberbullying trust family problems self-disclosure

Trust is a complex phenomenon that has been addressed by researchers from different disciplines reaching from philosophy to marketing. Thus, there is no unitary theory or generally accepted definition of trust (for an overview see Blöbaum, 2016; Thielmann & Hilbig, 2015).

We consider trust a multidimensional construct with a combination of situation-specific and generalized facets and will shortly review the most relevant theories: One of the most recognized models of trust was proposed by Mayer, Davis, and Schoorman (1995) who defined trust as the “willingness of a party to be vulnerable” (p. 712). Consequently, trust enables people to act in a complex world characterized by multiple risks because people trust that these risks will not result in negative consequences. Trust is usually given by a trustor to a trustee which can be another person or an organization. Trust is expressed as risk-taking and is assumed to be based on multiple factors (Mayer et al., 1995): First, the trustor has to believe the trustee trustworthy; this trustworthiness is based on the trustee’s perceived ability, benevolence, and integrity. Second, the trustor’s generalized propensity to trust impacts the transition of these factors into trust. Third, the perceived level of risk also contributes to the decision to trust. Considering these diverse factors, trust is most likely based on particular situation-specific factors (e.g., the expected benevolence of a trustee in a specific situation) as well as on generalized factors (e.g., personal dispositions of the trustor that go beyond specific situations) (see also Blöbaum, 2016; Thielmann & Hilbig, 2015). Trusting others can have different outcomes: While positive outcomes are assumed to increase perceived trustworthiness of specific trustees and the trustor’s general propensity to trust, the reverse is assumed for negative outcomes. Thus, trust can be conceptualized as a continuous feedback loop.

While most models conceptualize trust in general, specific models for the digital context have also been suggested (e.g., Moll & Pieschl, 2016). On a general level, Bierhoff and Vornefeld (2004), for example, considered the following levels of trust relevant for the digital context: relational trust (in specific digital communication partners), generalized trust (in people in general; cf. propensity to trust), and system trust (in abstract digital systems such as the Internet or Facebook). While trust is a central variable in many relationships between people and organizations, it might be especially important in digital settings. Other and fewer trust cues might be utilized in digital communication for example because social cues such as facial expressions or gestures are missing (Kiesler, Siegel, & McGuire, 1984). Additionally, digital communication leaves persistent digital traces (Boyd, 2008) that offer new opportunities for abuse by other people for example by stalking, sexual harassment, and fraud via the Internet, or by institutions, for example by personalized advertisements. There is evidence for both directions of a relationship between trust and digital behavior. For example, trust predicted online privacy concerns and self-disclosures (Bansal, Zahedi, & Gefen, 2016; Joinson, Reips, Buchanan, & Schofield, 2010; Taddei & Contena, 2013) and playing violent video games reduced interpersonal trust (Rothmund, Gollwitzer, Bender, & Klimmt, 2015). However, trust does not necessarily translate into online behaviour as a focus group study showed where young people indicated explicit trust evaluations but based their decision to self-disclose rather on the benefits than on the potential risks (Bryce &Fraser, 2014).

One specific risk in digital communication is cyberbullying. It can be defined as “an aggressive, intentional act carried out by a group or individual, using electronic forms of contact, repeatedly and over time against a victim who cannot easily defend him or herself” (Smith et al., 2008, p. 376). Thus, the same defining criteria apply to cyberbullying as to bullying, namely intentionality, repetition, and imbalance of power. However, not all measurement instruments employ such strict definition (Berne et al., 2013) and occasional negative experiences seem a precursor to cyberbullying. For example, 5.4% of German students were classified as victims of cyberbullying because they had at least one experience per week while 14.1% of the same sample reportedexperiencing occasional negative cyber incidents (Riebel, Jäger, & Fischer, 2009). In other studies 6.0% or 8.0% of German students self-classified as victims of cyberbullying while 34.2% reported occasional negative cyber experiences (Behrens & Rathgeb, 2016; Porsch & Pieschl, 2014). Thus, reported prevalence rates depend on the utilized measure, construct definition, and sample. Interestingly for the relationship with trust, outing (i.e., betraying secrets online and thus also betraying trust) has been evaluated as one of the most distressing types of cyberbullying (Pieschl, Kuhlmann, & Porsch, 2015).

Based on trust models (e.g., Bierhoff & Vornefeld, 2004; Mayer et al., 1995) and on research on cyberbullying (e.g., Pieschl et al., 2015; Smith et al., 2008) a reciprocal relationship between (different facets of) trust and experiences of negative cyber incidents or cyberbullying (cyber-victimization and cyber-perpetration) can be assumed. On the one hand, cyberbullying can be interpreted as a negative outcome of trust that signals the malevolence of specific communication partners and the risk involved in using specific means of digital communication. Thus, cyberbullying might predict low relational and system trust. Furthermore, cumulative experiences with cyberbullying and other experiences of violence and malevolence might also predict low generalized trust. Consequently, a negative relationship between cyberbullying and trust can be predicted (Hypothesis 1a). On the other hand, high relational, system, and generalized trust might result in risky online behavior such as frequent self-disclosures. Self-disclosures and other risky online behavior seem to increase the likelihood of cyber-victimization and cyber-perpetration (Kowalski et al., 2014; Peluchette et al., 2015). Consequently, a positive relationship between trust and cyberbullying can be predicted (Hypothesis 1b). As there is not (yet) any empirical evidence regarding the relationship between cyberbullying and trust, we explored these contrasting hypotheses in two exploratorystudies.

Exploratory Study 1: Malevolent Experiences as Predictors of Trust?

This cross-sectional study explores if previous negative experiences of interpersonal problems, violence, and malevolence (family problems, verbal and physical violence, cyber-victimization and cyber-perpetration) are negatively related to different facets of trust (online relational trust, online system trust, generalized trust).

Method

Procedure

All participants were recruited at an open day of a local German university. They answered multiple questionnaires administered electronically via Unipark (© Globalpark). The main questionnaires focused on (1) family problems, (2) violence (verbal and physical), (3) trust (online relational trust, online system trust, and generalized trust), and (4) negative cyber experiences (cyber-victimization and cyber-perpetration).

Sample

The sample consists of 224 high school students most of whom attended grades 11 (n = 17), 12 (n = 123), and 13 (n = 81) of grammar (n = 177), vocational (n = 31), or comprehensive (n = 16) schools; three participants had already finished school. Participants were predominantly female (n = 167; male: n = 57), on average M = 18.02 years old (SD = 1.35), and mostly born in Germany (n = 211). All participants had Internet access at home, often with their own computer (n = 203). On average, they used the Internet for M = 2.35 hours (SD = 1.86) on a weekday. On a scale from 1 (= never) to 6 (= multiple times per day) they indicated frequent use of Social Networking Sites (SNS, M = 4.88, SD = 1.30).

Measures

Family Problems

This scale was adapted from 15 items of the sub-scale “home” of the German Problem Questionnaire for Young People (Problemfragebogen für Jugendliche; Süllwold & Berg, 1967). Adaptations concerned, among others, replacing outdated formulations and using a 5-point answer format (from 1 = definitely true to 5 = definitely not true). The utilized items cover a range of family problems, both explicitly between the participant and her/his parents (e.g., “I cannot talk about personal issues with my parents”) or between the participant and her/his siblings (e.g., “I cannot get along with my siblings”) as well as about the family as a whole (e.g., “I feel as if I do not really belong to this family”). We computed a mean Family Problems (FP) score with high values indicating highly problematic family relationships (Cronbach’s α= 0.89).

Violence

Verbal Violence (VV) within the family was captured by two adapted items from the Revised Conflict Tactics Scale (Straus et al., 1996); adaptations concerned using “someone” instead of “my partner” in the formulations (e.g., “Someone screamed at me”). Physical Violence (PV) within the family was captured by six adapted items from a questionnaire by Baier et al. (2009) which consists of items about light violence (e.g., “Someone slapped me in the face”) and serious violence (e.g., “Someone gave me a beating”). All items were answered on a 5-point scale (from 1 = never to 5 = always). We computed mean VV (Cronbach’s α= 0.79) and PV (Cronbach’s α= 0.77) scores; high values indicate frequent experiences of violence.

Trust

To measure online-specific trust we used 15 items from the sub-scales “reputation”, “risk”, and “trust” from Corritore et al. (2005) that had to be answered on 5-point scales (from 1 = definitely true to 5 = definitely not true). These items had to be answered once regarding participants’ favorite SNS (cf. System Trust, ST) and once regarding a (fictitious) person acquainted via this SNS (cf. Relational Trust, RT). The scale “trust” could not be replicated in this sample and was not considered further. The scale “risk” (RI) consisted of six items (e.g., “When I use this SNS/communicate with this person I take a risk”, Cronbach’s α= 0.81/0.86) and the scale “reputation” (RE) consisted of 4 items (e.g., “This SNS/this person is well respected”, Cronbach’s α= 0.65/0.84). As a proxy for Generalized Trust (GT) we used the scale Agreeableness of the German NEO-PI-R (Ostendorf & Angleitner, 2004). People with high values on this scale believe that others are honest and have good intentions while people with low values are more skeptical. All eight items of this scale (e.g., “My first reaction is to trust people”) were rated on 5-point scales (from 1 = definitely true to 5 = definitely not true; Cronbach’s α= 0.79). We computed mean ST-RI, ST-RE, RT-RI, RT-RE, and GT scores; high values indicate high risk, reputation, and trust.

Negative Cyber Experiences

This questionnaire (Pieschl et al., 2015) measures the frequency of cyber-perpetration (CP) and cyber-victimization (CV) in the last two months, each with five items (regarding harassment, denigration, impersonation, outing, and exclusion) on 5-point scales (from 0 = never to 4 = multiple times per week). Due to the diverse nature of these behavioral items, internal consistency is low (cyber-perpetration Cronbach’s α= 0.48; cyber-victimization Cronbach’s α= 0.52). Nonetheless, we computed CP and CV scores by adding values across the five scales (scores 0 – 20; high scores indicate frequent negative experiences) and classified participants as cyber-perpetrators or cyber-victims if they gave at least one answer other than “never”. We do not refer to these experiences as cyberbullying because they do not necessarily meet the defining criteria of intentionality, repetition, and imbalance of power.

Results

Descriptive Results

In general, the sample means (see Table 1) show that students have high generalized trust (GT), consider it fairly risky to communicate digitally with people (RT-RI) and to use their favorite SNS (ST-RI) but nonetheless ascribe moderate reputation to their favorite SNS (ST-RE) and potential communication partners (RT-RE). Risk and reputation are significantly inversely related regarding communication partners (RT) and favorite SNS (ST); the higher students consider the risk, the lower they consider the corresponding reputation. Furthermore, judged reputations and judged risks are each significantly and positively related across contexts (RT and ST); for example, if students consider communicating with others risky (RT-RI), they usually also consider using their favorite SNS risky (ST-RI). However, generalized trust (GT) is not significantly related to the online-specific trust facets with one exception: generalized trust is positively related to the judged reputation of students’ favorite SNS (ST-RE).

Table 1

Descriptive Data and Correlations between Variables Explorative Study 1

Variable	Correlation with Variable^a
	1	2	3	4	5	6	7	8	9	10
1 FP	–	0.408^***	0.317^***	0.137^*	0.038	–0.317^***	0.003	–0.123	0.110	–0.127
2 VV		–	0.556^***	0.264^***	0.182^**	–0.224^**	–0.056	0.040	0.009	–0.083
3 PV			–	0.261^***	0.217^**	–0.155^*	0.118	–0.041	0.038	–0.102
4 CV				–	0.439^***	–0.062	0.009	–0.114	–0.089	0.065
5 CP					–	–0.148^*	–0.005	–0.021	–0.032	–0.037
6 GT						–	0.026	–0.084	0.141^*	–0.052
7 RT-RE							–	–0.304^***	0.240^***	–0.111
8 RT-RI								–	–0.088	0.223^**
9 ST-RE									–	–0.279^***
10 ST-RI										–
M^b	2.02	2.44	1.24	0.65^c	0.53^c	3.42	2.57	3.63	3.42	3.14
SD	0.74	0.90	0.35	1.08	0.99	0.72	0.69	0.85	0.69	0.79
Cronbach’s α	0.89	0.79	0.77	0.52	0.48	0.79	0.84	0.86	0.65	0.81

Note. FP = Family Problems, VV = Verbal Violence, PV = Physical Violence, CV = Cyber-Victimization, CP = Cyber-Perpetration; GT = Generalized Trust, RT = Relational Trust, ST = System Trust, RE = Reputation, and RI = Risk. ^aSpearman’s rank-order correlations rho. ^b5-point scales; high values represent high family problems, violence, trust and reputation, or risk. ^cMean sum scores from 0 – 20; high values represent frequent negative cyber experiences. ^***p < 0.001, ^**p < 0.01, ^*p < 0.05.

In general, the sample means (see Table 1) show that students report few negative experiences: They report low family problems and infrequent experiences of physical and verbal violence in their families. Furthermore, even though 38% (n = 84) of participants were categorized as cyber-victims and 31% (n = 69) as cyber-perpetrators, the frequencies of negative cyber incidents are very low. All of these variables are significantly and positively interrelated with one exception: Family problems are not significantly related to cyber-perpetration. Notably, there is a highly significant positive correlation, and thus an overlap, between cyber-victimization and cyber-perpetration. Sex differences of all sample means were tested with the Mann-Whitney-U test for independent samples. We only found significant sex differences for RT-RI (p < 0.00). Females considered talking with people online riskier(RT-RI: M = 3.79, SD = 0.80) than males (M = 3.16,SD = 0.84).

Hypothesis Testing

In order to explore the relationship between negative experiences of interpersonal problems, malevolence, and violence (e.g., cyberbullying) and facts of trust, we looked at bivariate correlations and computed explorative stepwise regressions for each of the trust facets with the corresponding significantly correlated variables. The correlations (see Table 1) show that none of the negative life experiences (FP, VV, PV, CV, and CP) correlate significantly with any of the online-specific trust facets (RT-RE, RT-RI, ST-RE, and ST-RI). However, almost all the negative life experiences correlate significantly and negatively with generalized trust; reporting many family problems (FP), frequent experiences of verbal (VV) and physical violence (PV), and frequent cyber-perpetration (CP) is associated with low generalized trust (GT). A stepwise regression with generalized trust (GT) as criterion and all negative experiences (FP, VV, PV, CV, and CP) as well as age and sex as predictors indicates that two predictors explain 17% of the variance (Model 1: R² = 0.14, F_change (1,222) = 35.59, p < 0.001; Model 2: R² = 0.17, F_change (1,221) = 7.15, p = 0.008); family problems (β= –0.37, p < 0.001) and cyber-perpetration (β= –0.16, p = 0.008) both negatively predict generalized trust.

Discussion

Most of the negative experiences of interpersonal problems, malevolence, and violence were not significantly related to any trust facet. Only one relationship between cyberbullying and trust was significant: Frequently instigating negative cyber incidents as a cyber-perpetrator significantly predicted low generalized trust. The process of social projection may explain this result (Thielmann & Hilbig, 2015): Perpetrators of negative cyber incidents might have projected their own malevolence to others and therefore assumed that they cannot trust other people in general. Thus, this study shows weak support for Hypothesis 1a, predicting a negative relationship between cyberbullying andtrust.

However, these results have to be interpreted with the specific limitations of this study in mind: We recruited a homogeneous convenience sample ofpredominantly female high school students who used Social Media frequently, reported few negative life events, and were interested in studying psychology at university. While this sample is representative of typical students of psychology, we do not know if the results can be generalized to other populations. Furthermore, we only collected data regarding different kinds of negative experiences. However, in order to find a positive relationship between trust and cyberbullying it might be necessary to also collect more information about other relevant factors such as (risky) online behavior.

Exploratory Study 2: Trust as Predictor of Risky Online Behavior and Cyberbullying?

This cross-sectional study explores if the general propensity to trust (generalized trust) is positively related to riskier online behavior and cyberbullying (less privacy concerns, more self-disclosures, more cyber-victimization, and more cyber-perpetration).

Method

Procedure

Participants were recruited online via a database of research volunteers and via Facebook groups and volunteered to answer a short online survey administered via Unipark (© Globalpark). The main questionnaires focused on (1) online self-disclosures, (2) online privacy concerns, (3) (generalized) trust, and (4) negative cyber experiences (cyber-victimization and cyber-perpetration). A total of 250 participants started to answer the survey. Fifty-four participants (22%) were excluded from analyses because they stated at the end of the survey that they did not answer truthfully or took less than 120 seconds to complete all questionnaires.

Sample

The final sample consists of 196 participants aged 16 to 25 years (M = 22.84, SD = 2.00). Participants were predominantly female (n = 145; male: n = 51) and predominantly university students (n = 168; n = 6 high school students; n = 13 working people; n = 9 others). They used digital technologies such as smartphones (92%) or laptops (66%) daily, among others for frequent communication via WhatsApp (91%), Facebook (84%), or Email (64%).

Measures

Online Self-Disclosure

We adapted the scale by Norberg, Horne, and Horne (2007). For twelve pieces of information (e.g., first name, telephone number, or sexual orientation) participants had to indicate their willingness to disclose truthful information towards a new online acquaintance on a 5-point scale (from 1 = definitely not to 5 = definitely, Cronbach’s α= 0.86). We computed a mean Online Self-Disclosure (OSD) score with high values indicating high willingness to self-disclose.

Online Privacy Concerns

We used the Online Privacy Concern Scale by Buchanan et al. (2007). On 15 items (e.g., “Are you concerned about people online not being who they say they are?”) participants had to indicate their privacy concerns on a 5-point scale (from 1 = not concerned at all to 5 = very concerned, Cronbach’s α= 0.89). We computed a mean Online Privacy Concern (OPC) score with high values indicating high privacyconcerns.

Trust

We used the propensity to trust scale by Frazier, Johnson, and Fainshmidt (2013) that combined and condensed previous similar instruments. Participants had to agree or disagree to four items (e.g., “I usually trust people until they give me a reason not to trust them”) on a 5-point scale (from 1 = absolute disagreement to 5 = absolute agreement, Cronbach’s α= 0.89). We computed a mean Generalized Trust (GT) score with high values indicating hightrust.

Negative Cyber Experiences

This questionnaire (Pieschl et al., 2015) was adapted for this age group. For the time in their life with most negative cyber incidents participants retrospectively indicated the frequency of cyber-perpetration (CP) and cyber-victimization (CV), each with five items (regarding harassment, denigration, impersonation, outing, and exclusion) on 5-point scales (from 0 = never to 4 = multiple times per week). Due to the diverse nature of these behavioral items, internal consistency is low to moderate (cyber-perpetration Cronbach’s α= 0.65; cyber-victimization Cronbach’s α= 0.63). Nonetheless, we computed CP and CV scores by adding values across the five scales (scores 0–20; high scores indicate frequent negative experiences) and classified participants as cyber-perpetrators or cyber-victims if they gave at least one answer other than “never”. We do not refer to these experiences as cyberbullying because they do not necessarily meet the defining criteria of intentionality, repetition, and imbalance of power.

Results

Descriptive Results

In general, the sample means (see Table 2) show that participants have high generalized trust (GT) and simultaneously report moderate levels of online privacy concerns (OPC) and online self-disclosure (OSD). Trust is significantly negatively related to online privacy concerns; participants who indicate high trust also score lower on online privacy concerns. Online privacy concerns and online self-disclosure are also significantly negatively related; the more privacy concerned participants report less self-disclosures. Furthermore, even though 65% (n = 128) of participants were categorized as cyber-victims and 56% (n = 110) as cyber-perpetrators based on their retrospective reports, the overall frequencies of negative cyber incidents are very low. Notably, there is a significant positive correlation, and thus an overlap, between cyber-victimization and cyber-perpetration. Furthermore, online self-disclosure (OSD) is significantly positively related to both cyber-victimization and cyber-perpetration; participants who report more willingness to disclose information online also report more negative incidents as cyber-victim and cyber-perpetrator. Sex differences of all sample means were tested with the Mann-Whitney-U test for independent samples. We only found significant sex differences for OSD (p < 0.00). Females’ answers showed that they were significantly less likely to self-disclose personal information online (OSD: M = 2.71, SD = 0.60) than males (M = 3.19, SD = 0.65).

Table 2

Descriptive Data and Correlations between Variables Explorative Study 2

Variable	Correlation with Variable^a
	1	2	3	4	5
1 GT	–	0.137	–0.178^*	0.045	0.045
2 OSD		–	–0.150^*	0.160^*	0.229^**
3 OPC			–	0.097	0.020
4 CV				–	0.572^***
5 CP					–
M^b	3.33	2.83	2.92	1.38^c	1.15^c
SD	0.97	0.65	0.74	1.57	1.56
Cronbach’s α	0.89	0.86	0.89	0.63	0.65

Note. GT = Generalized Trust, OSD = Online Self-Disclosure, OPC = Online Privacy Concerns, CV = Cyber-Victimization, and CP = Cyber-Perpetration. ^aSpearman’s rank-order correlations rho. ^b5-point scales; high values represent high trust, self-disclosure, and privacy concerns. ^cMean sum scores from 0 –20; high values represent frequent negative cyber experiences. ^***p < 0.001, ^**p < 0.01, ^*p < 0.05.

Hypothesis Testing

In order to explore the relationship between trust and cyberbullying, we first looked at bivariate correlations. Generalized trust (GT) was not significantly related to either cyber-victimization (CV) or cyber-perpetration (CP) (see Table 2). Because of these non-significant results, we did not further analyze the data.

Discussion

We did not find any direct relationship between trust and cyberbullying in this study. Thus, the data neither seems to support Hypothesis 1a of a negative nor Hypothesis 1b of a positive relationship. In partial support of our overall argumentation, however, trust was positively related to low privacy concerns and risky online behavior (high willingness to self-disclose) was positively related to cyber-victimization and cyber-perpetration.

These results have to be interpreted with the specific limitations of this study in mind: We recruited an online convenience sample and therefore we cannot guarantee that all participants answered truthfully and we cannot rule out self-selection biases or selective drop-outs (Alessi & Martin, 2010). For example, especially distrustful persons in terms of privacy and online safety aspects may be underrepresented in the sample. Additionally, the sample of mostly female university students is not representative of the population; we do not know if our findings can be generalized to other populations. Furthermore, due to the (mostly adult) age of participants we asked them retrospectively about their experiences with cyber-victimization and cyber-perpetration; thus, we cannot rule out memory biases.

General Discussion Regarding Cyberbullying and Trust

These studies provide the first empirical pieces of evidence with regard to the relationship between cyberbullying and trust. Even though the results show mostly non-significant correlations between cyber-victimization or cyber-perpetration and trust (Exploratory Studies 1 and 2) and only weak evidence for a negative relationship (Exploratory Study 1), these studies are informative regarding which methodologies might be fruitful in future research.

The additional variables that we considered were significantly related to cyberbullying and/or trust. Therefore, it can be argued that future studies about the relationship between cyberbullying and trust need to control for other negative experiences of interpersonal problems, malevolence and violence (e.g. family problems; Exploratory Study 1) as well as for (risky) online behavior (e.g., online self-disclosure and online privacy-concerns; Exploratory Study 2). Additionally, general negative cyber experiences might not be specific enough to predict more fine-grained indicators of online-specific trust (Exploratory Study 1).

Last but not least a potential theoretical explanation for the overall pattern of findings also implies potentially fruitful future research designs: Trust is theoretically modelled as a continuous feedback loop (Mayer et al., 1995). In cross-sectional designs, the theoretically predicted positive and negative relationships between cyberbullying and trust might counteract each other, which might have contributed to our overall lack of significant findings. A solution would be to implement a fine-grained longitudinal design that measures cyberbullying and trust multiple times (e.g., every month) over a longer period. It could be that on such a micro level experiences with cyberbullying would immediately result in lower trust (e.g., online-specific trust in specific communication partners or SNSs); the other way round a trust-inspired high-risk behavior such as frequent and in-depth self-disclosures on Social Media sites might soon after result in negative cyber experiences (e.g., mean comments). By measuring these constructs repeatedly, memory biases could be reduced and the positive and negative paths between cyberbullying and trust could be tested separately. Of cause, another alternative would be to employ experimental designs where, for example, participants could be confronted with negative cyber experiences and their situation-specific trust could be measured as dependent variable.However, as such experiences can be very stressful for the victimized participants ethical factors need to be considered carefully. We hope that these suggestions will inspire further research into the complex relationship between cyberbullying andtrust.

Footnotes

Acknowledgments

Stefanie Nordenbrock and Frank Schmischke collected parts of the presented data as parts of their Bachelor of Science thesis requirements. Furthermore, we would like to thank Prof. Dr. Rainer Bromme and Dr. Ricarda Moll for their contributions to the design of these studies.

References

Alessi

E. J.

, & Martin

J. I.

(2010). Conducting an internet-based survey: Benefits, pitfalls, and lessons learned. Social Work Research, 34, 122–128. doi: 10.1093/swr/34.2.122

Baier

, Pfeiffer

, Simonson

, & Rabold

(2009). Jugendliche in Deutschland als Opfer und Täter von Gewalt. Erster Forschungsbericht zum gemeinsamen Forschungsprojekt des Bundesministeriums des Innern und des KFN[Young people in Germany as victims and perpetrators of violence. First research report on the joint research project of the Federal Ministry of the Interior and KFN]. (KFN Forschungsbericht No. 107)). Hannover: KFN.

Bansal

, Zahedi

F. M.

, & Gefen

(2016). Do context and personality matter? Trust and privacy concerns in disclosing private information online. Information & Management, 53, 1–21. doi: 10.1016/j.im.2015.08.001

Behrens

, & Rathgeb

(2016). JIM 2016. Jugend, Information, (Multi-) Media. Basisstudie zum Medienumgang 12- bis 19-Jähriger in Deutschland[Survey about the media use of 12-19 year olds in Germany]. Stuttgart: Medienpädagogischer Forschungsverbund Südwest (mpfs).

Berne

, Frisén

, Schultze-Krumbholz

, Scheithauer

, Naruskov

, Luik

, & … Zukauskiene

(2013). Cyberbullying assessment instruments: A systematic review. Aggression and Violent Behavior, 18, 320–334. doi: 10.1016/j.avb.2012.11.022

Blöbaum

(2016). Key factors in the process of trust. On the analysis of trust under digital conditions. In Blöbaum

(Ed.), Trust and communication in a digitized world. Models and concepts of trust research (pp. 3–25). Berlin: Springer.

Bryce

, & Fraser

(2014). The role of disclosure of personal information in the evaluation of risk and trust in young peoples’ online interactions. Computers in Human Behavior, 30, 299–306. doi: 10.1016/j.chb.2013.09.012

Bierhoff

H.-W.

, & Vornefeld

(2004). The social psychology of trust with applications in the Internet. Analyse & Kritik, Zeitschrift für Sozialtheorie, 26, 48–62.

Boyd

(2008). Taken out of context: American teen sociality in networked publics Doctoral Dissertation, University of California-Berkeley, School of Information). Retrieved from http://www.danah.org/papers/TakenOutOfContext.pdf

10.

Buchanan

, Paine

, Joinson

A. N.

, & Reips

U. D.

(2007). Development of measures of online privacy concern and protection for use on the Internet. Journal of the American Society for Information Science and Technology, 58, 157–165. doi: 10.1002/asi.20459

11.

Corritore

C. L.

, Marble

R. P.

, Wiedenbeck

, Kracher

, & Chandran

(2005). Measuring online trust of websites: Credibility, perceived ease of use, and risk. Americas Conference on Information Systems (AMCIS) Proceedings, 370, 2419–2427.

12.

Frazier

M. L.

, Johnson

P. D.

, & Fainshmidt

(2013). Development and validation of a propensity to trust scale. Journal of Trust Research, 3, 76–97. doi: 10.1080/21515581.2013.820026

13.

Joinson

A. N.

, Reips

, Buchanan

, & Schofield

C. P.

(2010). Privacy, trust, and self-disclosure online. Human-Computer Interaction, 25, 1–24. doi: 10.1080/07370020903586662

14.

Kiesler

, Siegel

, & McGuire

T. W.

(1984). Social psychological aspects of computer-mediated communication. American Psychologist, 39, 1123–1134. doi: 10.1037/0003-066X.39.10.1123

15.

Kowalski

R. M.

, Giumetti

G. W.

, Schroeder

A. N.

, & Lattanner

M. R.

(2014). Bullying in the digital age: A critical review and meta-analysis of cyberbullying research among youth. Psychological Bulletin, 140, 1073–1137. doi: 10.1037/a0035618

16.

Moll

, & Pieschl

(2016). Expecting collectiveprivacy - a new perspective on trust in online communication. In Blöbaum

(Ed.), Trust and communication in a digitalized world. Models and concepts of trust research, (pp. 239–251). Berlin: Springer.

17.

Mayer

R. C.

, Davis

J. H.

, & Schoorman

F. D.

(1995). An integrative model of organizational trust. The Academy Of Management Review, 20, 709–734. doi: 10.2307/25879

18.

Norberg

P. A.

, Horne

D. R.

, & Horne

D. A.

(2007). The privacy paradox: Personal information disclosure intentions versus behaviors. Journal of Consumer Affairs, 41, 100–126. doi: 10.1111/j.1745-6606.2006.00070.x

19.

Ostendorf

, & Angleitner

(2004). Manual zum NEO-Persönlichkeitsinventar nach Costa und McCrae (revidierte Fassung). [Manual for the NEO personality inventory according to Costa and McCrae (revised version).]Göttingen: Hogrefe.

20.

Peluchette

J. V.

, Karl

, Wood

, & Williams

(2015). Cyberbullying victimization: Do victims’ personality and risky social network behaviors contribute to the problem?Computers in Human Behor, 52, 424–435. doi: 10.1016/j.chb.2015.06.028

21.

Pieschl

, Kuhlmann

, & Porsch

(2015). Beware of publicity! Perceived distress of negative cyber incidents and implications for defining cyberbullying. Journal of School Violence, 14, 111–132. doi: 10.1080/15388220.2014.971363

22.

Porsch

, & Pieschl

(2014). Cybermobbing unter deutschen Schülerinnen und Schülern: Eine repräsentative Studie zu Prävalenz, Folgen und Risikofaktoren [Cyberbullying among German pupils: A representative study on prevalence, consequences and risk factors]. Diskurs Kindheits- und Jugendforschung, 9, 7–22.

23.

Riebel

J. R. S. J.

, Jäger

R. S.

, & Fischer

U. C.

(2009). Cyberbullying in Germany–an exploration of prevalence, overlapping with real life bullying and coping strategies. Psychology Science Quarterly, 51, 298–314.

24.

Rothmund

, Gollwitzer

, Bender

, & Klimmt

(2015). Short- and long-term effects of video game violence on interpersonal trust. Media Psychology, 18, 106–133. doi: 10.1080/15213269.2013.841526

25.

Smith

P. K.

, Mahdavi

, Carvalho

, Fisher

, Russell

, & Tippett

(2008). Cyberbullying: Its nature and impact in secondary school pupils. Journal of Child Psychology and Psychiatry, 49, 376–385. doi: 10.1111/j.1469-7610.2007.01846.x

26.

Straus

M. A.

, Hamby

S. L.

, Boney-McCoy

, & Sugarman

D. B.

(1996). The Revised Conflict Tactics Scales (CTS2). Development and preliminary psychometric data. Journal of Family Issues, 17, 283–316.

27.

Süllwold

, & Berg

(1967). Problemfragebogen für Jugendliche[Problem Questionnaire for Young People]. Göttingen: Hogrefe.

28.

Taddei

, & Contena

(2013). Privacy, trust and control: Which relationships with online self-disclosure?Computers in Human Behavior, 29, 821–826. doi: 10.1016/j.chb.2012.11.022

29.

Thielmann

, & Hilbig

B. E.

(2015). Trust: An integrative review from a person–situation perspective. Review of General Psychology, 19, 249–277. doi: 10.1037/gpr0000046