Media Effects Research Lab - Research Archive

Cyberbullying and its Relationship with Long Term Psychological and Behavioral Effects

Student Researcher(s)

Richard Caneba (Ph.D Candidate); Qiguan Jiang (Masters Candidate)

Faculty Supervisor

Dr. S. Shyam Sundar

This paper was based on a project as part of the “COMM 506: Research Methods in Communications” course.

INTRODUCTION

In an information society, our ever increasing exposure to communication technology has given rise to a new form of negative behaviors: cyberbullying.  While there is already a significant body of literature and study concerning traditional bullying behaviors, cyberbullying as a relatively new phenomenon has a number of unique features that are still being studied and understood.  As a result of cyberbullying’s youth as a concept, many of the long-term effects of exposure are still being understood and examined.

Our research is aimed to find out the relationship between previous exposure to cyberbullying and psychological effects. We break down each variable into several sub concepts. Cyberbullying was broken down into six sub-concepts: 1) flaming/harassment, 2) denigration, 3) impersonation, 4), outing, 5) exclusion, 6) cyberstalking.  We studied the relationship between these sub-concepts insofar as they composed the construct of cyberbullying and public speaking confidence, self esteem and anxiety level.

RESEARCH QUESTIONS/HYPOTHESES  

For Americans and Canadians aged from 18-35,
Controlling for gender,
RQ 1: What is the relationship between exposure to cyberbullying in the past and public speaking confidence?
RQ 2: What is the relationship between exposure to cyberbullying in the past and self-esteem?
RQ 3: What is the relationship between exposure to cyberbullying in the past and anxiety level?

Hypotheses:
H1: There is a negative correlation between previous exposure to cyberbullying and public speaking confidence.
H2: There is a negative correlation between previous exposure to cyberbullying and self-esteem.
H3: There is a positive correlation between previous exposure to cyberbullying and anxiety level. 

METHOD

Following a small pilot study where we gathered feedback from our social networks in regards to to the composition of our survey, we deployed this survey questionnaire through Amazon Mechanical Turk to 400 respondents (n=272 after filtering for age and IP redundancies), done in 3 batches over 3 days.  The survey first asked respondents to accept a consent form, and then were asked for their age.  If respondents failed the age check, they were sent to the end of the survey.  Respondents between the ages of 18 and 35 were sent to the full survey.  After answering several demographics questions, they were exposed to the IV and DV Likert-style scales.  Respondents were first asked to report on their previous exposure to a number of cyberbullying incidents, and then responded to 3 well established scales purporting to measure their respective dependent variables: a subset of items from the PRCA-24 scale to measure public speaking confidence, the Rosenberg Self-Esteem scale to measure self-esteem, and the Beck Anxiety Inventory to measure anxiety levels.  Each variable was assigned an index that was generated by summing their respective items, after reverse coding the negatively worded items.  This data was then analyzed through JMP (v. 13.0.0), treating each index as a continuous, ratio level variable.

RESULTS
Through a simple linear regression analysis, we found that cyberbullying exposure was negatively correlated with self-esteem (H2):

Table 1

Table 1. Cyberbullying exposure and self-esteem regression

Cyberbullying exposure also was shown to be positively correlated with anxiety levels (H3):

Table 2

Table 2. Cyberbullying exposure and anxiety levels regression

Unintuitively, we found that cyberbullying exposure was positively correlated with public speaking confidence, confirming the opposite of H1:

Table 3

Table 3. Cyberbullying exposure and public speaking confidence regression

A refinement of these findings found that cyberstalking targeting was the only sub-concept of cyberbullying that had a significant effect on the variance of anxiety levels:

Table 4

Table 4. Effect Tests for Cyberbullying Sub-Concepts on Anxiety

Likewise, we discovered that flaming/harassment targeting exposure was the only sub-concept that had a significant effect on the variance of self-esteem levels:

Table 5

Table 5. Effect Tests for Cyberbullying Sub-Concepts on Self-Esteem

A moderator analysis revealed that gender was a significant moderator of the relationship between cyberstalking and anxiety:

Table 6

Table 6Effect tests table indicating interaction effect of gender on cyberstalking to anxiety relationship

A formal test of mediation revealed that anxiety levels were a significant mediator of the relationship between flaming/harassment targeting and self-esteem

CONCLUSIONS/DISCUSSION
While the confirmation of H2 and H3 was unsurprising (but noteworthy nonetheless), the unintuitive confirmation of the opposite of H1 was rather surprising.  There were a couple of possible theories that we entertained, and should consider for future study: the effect of visibility (which is a natural outcome of regularly speaking in public) on cyberbullying exposure, and aggression as a compensatory mechanism in response to cyberbullying.  Gender moderation of the relationship between cyberstalking and anxiety level we theorized as something indicative of the lower level of anxiety that the female group had at low or no exposure to cyberbullying, insofar as the same or comparable amount of exposure to cyberbullying will exhibit stronger effects on a population less accustomed to it, i.e. males in this case.  Mediation of the relationship between flaming/harassment exposure (a sub-concept of cyberbullying) and self-esteem by anxiety captures a fascinating notion: that anxiety is a strong predictor of self-esteem.

 

For more details regarding the study contact

Dr. S. Shyam Sundar by e-mail at sss12@psu.edu or by telephone at (814) 865-2173

More Articles From: