DISCUSSION POST

288

School Psychology Review 2017, Volume 46, No. 3, pp. 288–303 DOI: 10.17105/SPR-2016-0004.V46-3

Cyber Victimization in High School: Measurement, Overlap With Face-to- Face Victimization, and Associations With Social–Emotional Outcomes

Christina Flynn Brown Community High School District 155 in Crystal Lake, IL

Michelle Kilpatrick Demaray Jaclyn E. Tennant

Northern Illinois University

Lyndsay N. Jenkins Eastern Illinois University

Abstract. Cyber victimization is a contemporary problem facing youth and adolescents (Diamanduros, Downs, & Jenkins, 2008; Kowalski & Limber, 2007). It is imperative for researchers and school personnel to understand the associations between cyber victimization and student social–emotional outcomes. This article explores (a) gender differences in rates of cyber victimization, (b) overlap between traditional and cyber victimization, (c) differences in social–emotional outcomes across victimization classes, and (d) associations among cyber victimization and social–emotional risk, internalizing problems, and externalizing problems while controlling for traditional victim- ization among 1,152 high school students. Boys reported significantly higher rates of cyber victimization than did girls. Ten percent of students reported experiencing low levels of both cyber and traditional victimization (low dual), 3% of students reported experiencing moderate levels of both cyber and traditional victimization (moderate dual), and 1% of students reported high levels of both types of victimization (high dual). Three percent of students reported experiencing traditional victimization but not cyber victimization (traditional). There were significant differences in social and emotional problems among youth involved in victimization in various groups (i.e., unin- volved, traditional, low dual, moderate dual, and high dual). Lastly, cyber victimization significantly predicted variance in social–emotional risk and internalizing problems above and beyond that predicted by traditional victimization.

Traditional victimization (i.e., receipt of physical, ver- bal, or relational aggression) has been widely researched in the United States and abroad. Alhough prevalence estimates vary, researchers agree that victimization is a pervasive prob- lem. In the United States, 30–35% of students are a bully or victim (Wang, Iannotti, & Nansel, 2009); however, students contend not only with traditional victimization, but also cyber victimization. As cell phones, smartphones, and personal computers become more accessible and relied on for commu- nication, cyber victimization will likely be more common. Researchers estimate that as many as 20–35% of children and adolescents report cyber victimization (Diamanduros et al., 2008; Kowalski & Limber, 2007), although these rates may

differ for girls and boys depending on the type of victimiza- tion examined. Both traditional and cyber victimization have been associated with social, emotional, and academic prob- lems (e.g., Hinduja & Patchin, 2008; Luk, Wang, & Simons- Morton, 2010; Machmutow, Perren, Sticca, & Alasker, 2012; Shariff, 2005; Sticca & Perren, 2013), but less information exists regarding how these two types of victimization may relate differently to outcomes.

The goals of the current study were to examine (a) gen- der differences in prevalence rates of cyber victimization among high school students, (b) the overlap between tradi- tional and cyber victimization, (c) differences in average levels of social–emotional outcomes across different victimization

Correspondence regarding this article should be addressed to Michelle Demaray, Department of Psychology, Northern Illinois University, DeKalb, IL; e-mail: [email protected]

Copyright 2017 by the National Association of School Psychologists. ISSN 0279-6015, eISSN 2372-966x

289

Cyber Victimization in High School

classes, (d) the association of cyber victimization with social– emotional outcomes, controlling for traditional victimization, and (e) the moderating effect of gender on the association between victimization and outcomes.

Traditional and Cyber Victimization Defined

Olweus (1997) stated, “A student is being bullied or victimized when he or she is exposed, repeatedly and over time, to negative actions on the part of one or more other students” (p. 496). According to this definition, for bullying to occur the behavior must be intentional and repeated over time. Bullying also involves a perceived or actual imbalance of power between the bully and the victim. Traditional vic- timization and cyber victimization share the same intent, but the mode of perpetration differs. With cyber victimization, technology is used to harass or intimidate the victim. Currently there is not a standard definition of cyber victimiza- tion, but all the definitions generally contain elements of intentional and repeated harm inflicted through the use of technology. For example, Mason (2008) defined cyber victim- ization as “an individual or a group willfully using informa- tion and communication involving electronic technologies to facilitate deliberate and repeated harassment or threat to another individual or group by sending or posting cruel text and/or graphics using technological means” (p. 323). Definitions by other researchers are similar (e.g., Erdur- Baker, 2010; Hinduja & Patchin, 2008).

Although both are rooted in aggression, there are sev- eral key differences between traditional and cyber victimiza- tion. First, if youth have access to computers and cell phones at home, cyberbullies can reach their victims almost all the time. Unlike traditional victimization, the cyber victimization audience can easily and quickly grow exponentially with a few key strokes on a cell phone, computer, or tablet. Third, cyber victimization is not a face-to-face interaction. Cyberbullies can maintain anonymity by hiding behind tech- nology. Finally, traditional bullies use social or physical power against victims, but in cyberbullying, power can take on different forms. The power differential can come from ano- nymity, the use of technology to rapidly spread embarrassing or false images or information, or threats of using physical or social power against the victim.

Gender Differences in Prevalence Rates

Research has revealed that boys and girls have different experiences with traditional victimization, so it is reasonable to expect gender differences in cyber victimization as well. For direct, physical victimization, boys are more likely to be victimized than girls (Nansel et al., 2001; Solberg & Olweus, 2003), but gender differences for indirect or relational aggres- sion are less clear. Some researchers have found that more girls than boys experience relational victimization (Crick & Grotpeter, 1995), while others (Prinstein, Boergers, & Vernberg, 2001) found no gender differences.

Gender differences in the frequency of cyber victimiza- tion are also inconsistent in the literature. For example, many studies have found no gender differences in the frequency of cyber victimization (Hinduja & Patchin, 2008; Kowalski & Limber, 2007; Patchin & Hinduja, 2006; Varjas, Henrich, & Meyers, 2009; Ybarra, Diener-West, & Leaf, 2007). However, there have been studies that have found girls are more likely to be victimized online than boys (Dehue, Bolman, & Völlink, 2008; Ybarra et al., 2007). More research is needed to find consensus on this issue.

Overlap in Traditional and Cyber Victimization

Unfortunately, many victims of traditional bullying are likely to be cyber victims as well (Fredstrom, Adams, & Gilman, 2011; Hinduja & Patchin, 2008; Mitchell, Jones, Turner, Shattuck, & Wolak, 2016). One study found that a large majority of students (∼90%) who experience cyber vic- timization also report experiencing face-to-face victimization (George & Odgers, 2015). However, a larger percentage of students report traditional victimization than cyber victimiza- tion. According to Kann et al. (2014), 19.6% of high school students reported being victimized while at school, and 14.8% reported being cyber victimized; however, not all students have access to forms of technology that enable cyber victim- ization. Therefore, some students who are traditionally vic- timized may not be cyber victimized because youth who cyberbully have no opportunity to target them.

Although most studies find that students who are vic- timized in person are also victimized online, this is not the case for all youth. Understanding outcomes for youth who only experience cyber victimization is still important, even if they are in the minority (i.e., only experience cyber victimiza- tion). For instance, Mitchell, Ybarra, and Finkelhor (2007) found that approximately 25% of youth who reported online victimization were not victimized offline. Although 75% of the participants were victimized online and in person, there is a portion of students who only experience cyber victimiza- tion. Another study (Ybarra et al., 2007) found that the major- ity of surveyed youth ages 10–15 who were harassed online were not bullied in the traditional sense. Research should examine characteristics of and outcomes for youth who only experience cyber victimization, even if this is not a typical experience. Although overlap in traditional and cyber victim- ization has been explored in other studies, the current study used a different approach, latent class analysis, to not just examine the correlation between these two types of victim- ization, but to derive groups of youth that experience different rates of one, both, or neither type of victimization.

Outcomes Related to Cyber Victimization

Due to the volume of research on bullying and victim- ization, it is well established that victims of traditional bully- ing are much more likely to have social, emotional, and academic difficulties (Hawker & Boulton, 2000; Nansel et al.,

School Psychology Review, 2017, Volume 46, No. 3

290

DOI: 10.17105/SPR-2016-0004.V46-3

2001; Parker & Asher, 1987; Salmon, James, & Smith, 1998). Initial research has demonstrated that being a victim of cyber- bullying can negatively affect an individual’s physical, social, emotional, and cognitive functioning, development, and gen- eral well-being (Beran & Li, 2005; Hinduja & Patchin, 2008; Luk et al., 2010; Mitchell et al., 2007; Sticca & Perren, 2013). In addition, behavioral and psychosocial problems have been found to increase as the intensity of online harassment increases (Ybarra & Mitchell, 2007).

Traditional and cyber victimization are associated with a multitude of negative social, emotional, and academic out- comes; however, less research has focused on the cumulative effect of both types of victimization. In 2016, Mitchell et al. found that youth who experienced both cyber and traditional victimization reported that the incidents had more of an emo- tional impact than students who experienced either type of vic- timization alone. Recently, Bonanno and Hymel (2013) found that in a sample of 399 students in eighth to 10th grades, involvement in cyberbullying as a bully or a victim uniquely contributed to the prediction of depressive symptoms and sui- cidal ideation over and above involvement in traditional forms of victimization. Hinduja and Patchin (2010) surveyed 1,963 middle school students and found that traditional and online victimization were significantly associated with increases in suicidal ideation. Luk et al. (2010) found that victimization was significantly associated with substance use and depression in boys and girls in 10th grade; the more frequently an individual was victimized, no matter the type of victimization, the higher the level of depression. Gradinger, Strohmeier, and Spiel (2009) found that students who were involved in victimization both online and offline had the most negative outcomes.

A number of studies have demonstrated that victimiza- tion across environments has a greater impact than victimiza- tion in a single domain. For example, Ybarra et al. (2007) found that victims of both online (i.e., cyber) and offline (i.e., traditional) bullying reported significantly more distress than those who were only bullied online. Raskauskas (2009) sug- gested the multiple victimization framework, which predicts that cumulative effects can result from adding text or online victimization to traditional victimization experiences. She found that students bullied physically and via text messages reported more depressive symptoms than those not bullied at all or those bullied in a single environment. Studies by Gradinger et al. (2009) and Fredstrom et al. (2011) support that theory. In their work, youth victimized both online and in person were at the highest risk for internalizing problems.

Gender-Specific Traditional and Cyber Victimization Outcomes

Prevalence rates, outcomes, and gender differences in outcomes have been studied in the traditional victimization literature, but much less has been done in the cyber victimiza- tion domain. Carbone-Lopez, Esbensen, and Brick (2010) found that negative outcomes associated with being bullied were related to gender, as well as frequency and type of

victimization. Repeated indirect victimization was associated with the most negative outcomes. Traditional victimization was related to delinquency, gang membership, and drug use for both girls and boys. Self-esteem, however, decreased fol- lowing victimization for girls only, suggesting that victimiza- tion is related to behavioral outcomes across gender but may have more negative psychological consequences for girls.

The cyber victimization literature provides very little information regarding outcomes by gender. This is not sur- prising, given that research has focused mainly on gender differences in terms of prevalence and data are inconclusive at best. There are some studies that looked at outcomes by gender in the context of cyber victimization. Brown, Demaray, and Secord (2014) found that cyber victimization did not pre- dict any outcomes for boys; however, cyber victimization was a significant, unique predictor of depression and social stress for girls. Results from this study suggest that girls are at greater risk of poor outcomes due to what happens online. Further research is needed, which may lead to a better under- standing of how boys and girls are impacted by cyber victim- ization involvement. Overall, it is not clear what outcomes are associated with cyber victimization above and beyond social and emotional problems associated with traditional victimiza- tion. In addition, there may be unique gender differences in the associations among traditional and cyber victimization and social and emotional outcomes.

The Current Study

Five research questions were addressed in the current study. First, are there gender differences in prevalence rates of cyber victimization? Previous research has not been consistent in terms of gender differences. Some studies found no gender differences at all (Beran & Li, 2005; Hinduja & Patchin, 2008) while others found that middle school girls, in particular, were more likely to be cyber victims (Dehue et al., 2008; Dempsey, Sulkowski, Nichols, & Storch, 2009; Kowalski & Limber, 2007). Understanding potential gender differences associated with cyber and traditional victimization rates will help practi- tioners more effectively target a specific group of individuals (such as male students). This is an important factor to consider because refining and narrowing the focus of interventions saves personnel time and financial resources.

Second, are there groups of students that experience dif- ferent rates of one, both, or neither type of victimization? Other studies have shown that students who experience cyber victim- ization are more likely to also experience traditional victimiza- tion (e.g., Hinduja & Patchin, 2008; Li, 2006; Mitchell et al., 2007; Raskauskas & Stoltz, 2007); however, no studies have examined the degree to which students experience different types of victimization using latent class analysis. Third, are there differences in social–emotional outcomes across victim- ization classes? We expect that students experiencing greater levels of both types of victimization will have more negative social–emotional outcomes than those with lower levels or no victimization (Mitchell et al., 2016).

291

Cyber Victimization in High School

Fourth, is cyber victimization associated with social– emotional outcomes (i.e., social–emotional risk, internalizing problems, externalizing problems) when controlling for tradi- tional victimization experiences? It was predicted that cyber victimization would be significantly and positively related to social–emotional risk, internalizing problems, and externaliz- ing problems for boys and girls above and beyond the role of traditional victimization (Beran & Li, 2005; Hinduja & Patchin, 2008; Shariff, 2005; Ybarra et al., 2007). Fifth, does the asso- ciation between cyber victimization and social–emotional out- comes differ based on gender? The literature to date has provided little research in terms of gender-specific outcomes; therefore, gender predictions were not made in the current study and were approached in an exploratory fashion.

METHOD

The research questions were addressed by collecting data with students in a high school. The participants, proce- dures, and measures are described below.

Participants

Participants for this study (N = 1,152) attended a large high school in northern Illinois. There were 1,482 students enrolled in the school at the time. Thus, approximately 78% of the student population participated in the study. There were 322 ninth graders, 304 tenth graders, 275 eleventh graders, and 251 twelfth graders. In addition, 578 of the participants were boys and 574 were girls. The majority of the students were White (75%) followed by Hispanic or Latino students (12%). Thirty-three percent of students qualified for free or reduced-price lunch, and 14% of participating students qual- ified for special education services. Data were originally col- lected on 1,200 students; however, all demographic data (e.g., gender, grade) were missing for 48 students. Given that gen- der was included in all analyses, the 48 cases with no demo- graphic data were deleted. At the time of data collection, the school’s student body was 66.7% Caucasian, 2.4% African American, 26.9% Hispanic American, 1.1% Asian American, 0.2% American Indian, and 1.6% multiracial, and 37.8% of students qualified as low income (based on free or reduced- price lunch status).

Measures

Four self-report rating scales and records data were uti- lized in the study. Records data included gender, grade level, and ethnicity. Cyber victimization was assessed via the Cyber Victimization Survey (CVS; Brown et al., 2014). Traditional victimization was assessed via the Revised Olweus Bully/ Victim Questionnaire (OBVQ; Olweus, 1996). Two measures were utilized that assessed problem behavior: the Strength and Difficulties Questionnaire (SDQ; R. Goodman, 2001) and the Behavior Assessment System for Children–Second Edition Behavioral and Emotional Screening System Student Self- Report form (BASC-2 BESS; Kamphaus & Reynolds, 2007).

Both problem behavior measures were utilized because the BESS provides a total score with overall level of risk and the SDQ provides subscale scores to investigate different areas of problem behaviors (i.e., internalizing problems, externalizing problems).

Cyber Victimization Survey The CVS is a 15-item measure that was developed to

assess cyber victimization (Brown et al., 2014). Participants were asked how often certain online incidents happened to them in the last 2–3 months. All items were answered on a 5-point scale (1 = it hasn’t happened at all in the past couple of months, 2 = only 1 or 2 times in the past couple of months, 3 = 2 or 3 times a month, 4 = about once a week, and 5 = sev- eral times a week).

Preliminary information regarding evidence of reliabil- ity and validity of the CVS was published by Brown et al. (2014) using a sample (N = 106) of middle school students. A factor analysis indicated that a single factor explained 52% of the variance. Item loadings on the single factor ranged from .62 to .81, and the coefficient alpha was .92. To provide evi- dence of validity, correlations were conducted between the victimization scale of the CVS and Hinduja and Patchin’s (2008) Cyberbullying and Online Aggression Survey Instrument as well as Kowalski and Limber’s (2007) measure, which were .59 and .52 (p < .01), respectively. For the current sample, internal consistency was .94. In the current study, the CVS was positively correlated (r = .71, p < .001) with the Cyberbullying and Online Aggression Survey Instrument (Hinduja & Patchin, 2007), which provides additional evi- dence of construct validity. The average of the 15 items (range 1–5) was utilized as the cyber victimization score in this study. See Appendix A for a copy of the measure.

Hinduja and Patchin’s Cyberbullying and Online Aggression Survey Instrument (Hinduja & Patchin, 2007) consists of nine victimization items that assess how often in the past 30 days youth have experienced different types of cyber victimization (e.g., “In the last 30 days have you received an e-mail from someone you know that made you really mad?” and “In the last 30 days has anyone posted any- thing about you online that you didn’t want others to see?”). Items are rated on a 5-point scale ranging from never to every day. Cronbach’s alpha coefficient for the victimization scale was .74 in the Hinduja and Patchin (2007) study; in the cur- rent study, the alpha coefficient was .89.

Revised Olweus Bully/Victim Questionnaire The OBVQ is a 40-item self-report questionnaire that

measures traditional bullying and victimization (Olweus, 1996); only the 11 victimization items were used in the cur- rent study. The measure begins with a detailed definition of bullying, and students are asked to respond to items based on a 2- to 3-month reference period. Response options are on a 5-point Likert type scale (1 = I haven’t been bullied or I hav- en’t bullied other students at school in the past couple of months, 2 = only once or twice, 3 = 2 or 3 times a month,

School Psychology Review, 2017, Volume 46, No. 3

292

DOI: 10.17105/SPR-2016-0004.V46-3

4 = about once a week, and 5 = several times a week). The measure is comprised of two global questions. The first asks, “How often have you been bullied at school in the past couple of months?” and the second asks, “How often have you taken part in bullying another student(s) at school in the past couple of months?” These items are each followed by more specific items related to victimization and bullying (Solberg & Olweus, 2003).

Internal consistency and test–retest reliability of the questionnaire is considered satisfactory (Genta, Menesini, Fonzi, Costabile, & Smith, 1996; Olweus, 1997). Victimization items have demonstrated good internal consistency, with coef- ficient alphas greater than .80 (Kyriakides, Kaloyirou, & Lindsay, 2006). Kyriakides et al. (2006) found acceptable con- struct validity and reliability for the measure using a Rasch model to test bullying and victimization scales. Other studies have reported additional evidence of reliability and validity that support the psychometric properties of the scale (Bendixen & Olweus, 1999; Kyriakides et al., 2006; Olweus, 1994). An average of the 11 victimization items (range 1–5) was utilized as the traditional victimization score in this study. For the cur- rent sample, internal consistency was .89.

Strength and Difficulties Questionnaire The SDQ is a 25-item screening questionnaire for

children ages 4–17 and is designed to assess emotional, behavioral, or concentration problems (R. Goodman, 2001). The SDQ was used to assess functioning in specific problem domains. The self-rated version for 11- to 17-year- olds was used in the current study. Ratings are given in 3-point Likert scales (0 = not true, 1 = sometimes true, and 2 = certainly true).

A total score provides a measure of social–emotional difficulties. Additional scores are provided across five domains: emotional problems, conduct problems, hyperactiv- ity/inattention problems, peer relationship problems, and pro- social behavior. However, for the current study, peer relationship problems and emotional problems were com- bined to create an internalizing problems composite. Cronbach’s alpha for the internalizing problems composite was .74. Additionally, the conduct problems subscale and hyperactivity/inattention problems subscale were summed to create an externalizing problem composite. Cronbach’s alpha for the externalizing problems composite was .72. Support for the use of these composite scores has been established by the creators of the SDQ (A. Goodman, Lamping, & Ploubidis, 2010).

The Behavior Assessment System for Children– Second Edition Behavioral and Emotional Screening System, Student Self-Report Form

The BESS is a brief inventory that assesses a wide array of behaviors representing both behavioral problems and strengths, including internalizing problems, externalizing problems, school problems, and adaptive skills (Kamphaus &

Reynolds, 2007). The BESS was used as a broad measure of overall social–emotional well-being, in contrast to the narrow fields assessed by the SDQ. Items are summed to get one total score that measures the level of social and emotional risk. The student form (Grades 3–12) was used in the current study. Respondents rate questions on a frequency-based 4-point Likert type scale ranging from never to almost always. The standard student form contains 30 items. According to the manual, the test–retest coefficient was r = .80. Internal con- sistency for the BESS is strong, with Cronbach’s alpha equal to .91 for the student form. Validity of the BESS is evidenced by correlations between the BESS and the BASC-2 Self- Report of Personality child/adolescent form (Reynolds & Kamphaus, 2004) ranging from.69–.86. The total score based on the combined sex norms (M = 50; SD = 10) was utilized in the current study.

Procedures

Participants for this study attended a large high school in the Midwest. The school was involved in collecting school- wide social–emotional data, including information on tradi- tional and cyberbullying and victimization, in Fall 2013. Consistent with school procedures, a passive consent approach was used to inform parents and students of the study and to allow parents and students the option not to participate. The letter announcing the study was followed by an e-mail/ voicemail blast informing parents that their children received an important letter at school that day. On the day of data col- lection, students were given the opportunity to opt out of the study. Teachers read scripts before students completed the paper and pencil surveys. The scripts included a description of the study, assurances of confidentiality, and an explanation that data would be used solely for research purposes unless students indicated that they were going to harm themselves or others. Participants were told that participation in the study was voluntary, and participation could be withdrawn at any time without penalty or prejudice.

Surveys were counterbalanced. Each survey had a Post-it® Note with the student’s name affixed to it, which students were asked to remove before completing. Participant identification numbers were on the first page of the survey so that school data such as ethnicity and free or reduced-price lunch status could be accessed. Data used by the researcher were deidentified, and institutional review board approval was granted for post hoc analysis of deidentified data. After all data were obtained, the primary investigator and a team of research assistants entered and verified all data.

Data Analytic Plan

The first goal of the study, to explore the gender differ- ences in prevalence rates of cyber and traditional victimiza- tion, was accomplished by running two regression analyses in Mplus (Muthén & Muthén, 2007) and comparing the mean ratings of traditional and cyber victimization for girls and

293

Cyber Victimization in High School

boys. The second goal of the study, to determine the overlap between cyber and traditional victimization, was achieved by conducting a latent class analysis (LCA) in Mplus to group students by their experiences with each type of victimization and determine the prevalence of each group in the sample. In addition, the traditional and cyber victimization scores were correlated. To answer the third research question, a multivar- iate analysis of variance (MANOVA) was conducted in SPSS (IBM Corp., 2013) to understand differences in social–emo- tional outcomes across the LCA groups. Finally, to answer the fourth and fifth research questions, three regression analyses were conducted in Mplus to understand the associations among cyber and traditional victimization and social–emo- tional outcomes (regression analyses) and the moderating effect of gender on these associations.

RESULTS

Means and standard deviations of all main variables for the total sample and by gender are presented in Table 1. Correlations between all variables are also presented in Table 1. Robust maximum likelihood (MLR) estimation was used for all regression analyses because it is robust against violations of normality and heteroscedasticity. All other assumptions of regression were met. The dependent variables and all inde- pendent variables, aside from gender, were continuous. Variance inflation factors and tolerance values were accept- able, and no independent variables were highly correlated (r > .90). Mean scores for CVS cyber victimization and OBVQ traditional victimization, BESS social–emotional risk scores, and sum scores for the SDQ composites (internalizing and externalizing problems) were calculated and used in all analyses. Missing data were accounted for using MLR in Mplus. Less than 4% of data were missing from any one variable.

Preliminary Analyses

Gender differences in the social–emotional outcome variables were investigated via three linear regressions with gender coded as a dummy variable (0 = girls and 1 = boys) and the social–emotional outcome variables entered as depen- dent variables. No significant difference was found between the boys and girls in externalizing problems. Significant gen- der differences were found for mean ratings of social–emo- tional risk scores (β = .12, p = .03) and internalizing problems (β = .29, p < .001). Specifically, girls reported lower rates of social–emotional risk scores and internalizing problems than did boys. Gender accounted for only a small proportion of the variance in social–emotional risk scores (R2 = .02) and inter- nalizing problems scores (R2 = .08).

Research Question 1: Gender Differences

Gender differences in the experience of traditional and cyber victimization were investigated via two linear regres- sions with gender coded as a dummy variable (0 = girls and 1 = boys) and the two types of victimization entered as depen- dent variables. No significant differences were found between the boys and girls in traditional victimization (β = .09, p = .15) with R2 = .01. However, significant gender differences were found for mean ratings of cyber victimization, (β = .14, p = .02). Specifically, girls reported lower rates of cyber vic- timization than boys. Gender accounted for only a small pro- portion of the variance in cyber victimization (R2 = .02).

Research Question 2: Victimization Groups

An LCA was performed on the two victimization scores (cyber victimization and traditional victimization) to deter- mine any overlap among youth in experiences of these types of victimization. The LCA was conducted utilizing maximum

Table 1. Descriptive Statistics by Total Sample and Gender and Correlations Among Study Variables

1 2 3 4 5 Total Girls Boys

M SD M SD M SD

1. Social–emotional risk† 1 50.89 11.20 49.49 11.22 52.27 11.07

2. Internalizing problems† .63** 1 15.59 3.55 14.53 3.24 16.60 3.52

3. Externalizing problems .62** .32** 1 16.19 3.24 16.22 3.33 16.14 3.16

4. Traditional victimization .48** .47** .28** 1 1.43 0.55 1.38 0.54 1.48 0.56

5. Cyber victimization .45** .39** .22** .72** 1 1.24 0.46 1.17 0.40 1.30 0.49

6. Gender .12** .29** −.01 .08** .14**

Note. Gender was dummy coded (0 = girls, 1 = boys). † denotes significant gender difference. Total N = 1,152 (574 girls, 578 boys). Social– emotional scores on the Behavior Assessment System for Children–Second Edition Behavioral and Emotional Screening System Student Self-Report form have a mean of 50 and a standard deviation of 10. Internalizing and externalizing problems composite scores on the Strengths and Difficulties Questionnaire composites range from 0–20. Traditional victimization scores on the Olweus Bully/Victim Questionnaire range from 1–5, and cyber victimization scores on the Cyber Victimization Survey range from 1–5. * p < .05. ** p < .01.

School Psychology Review, 2017, Volume 46, No. 3

294

DOI: 10.17105/SPR-2016-0004.V46-3

likelihood estimation with robust standard errors. The follow- ing fit indices were utilized to determine the number of latent classes: Bayesian information criterion (BIC), Vuong-Lo- Mendell-Rubin likelihood ratio test (VLMR LRT), bootstrap likelihood ratio test (bootstrap LRT), and entropy value. Models were considered to have converged if the maximum log-likelihood was replicated at least five times. Better model fit is determined by a lower BIC and higher entropy value (near 1.0). Both the VLMR LRT and the bootstrap LRT test whether the current model class size (K) being analyzed is significantly better than one less class size (K–1; e.g., Are four classes (K) significantly better than three (K–1)?).

Table 2 presents the fit results and subgroup prevalence for models with one to five latent classes. The five-class model

was retained as the best fit. Specifically, the five-class model had a lower BIC than the four-class model, and although the VLMR LRT was not significant, the bootstrap LRT was, demonstrating that the five-class model fits significantly better than the four-class model. In addition, the groups of the five- class model made theoretical sense and were provided labels to describe the groups (as follows). The six-class model did not converge. Results of the LCA are presented in Table 2 and the plot is presented in Figure 1. Note that the subgroup prev- alence in Table 2 is based upon estimated class probability. The final count of individuals in classes based on most likely latent class membership is described as follows.

In the final model of five latent classes, the classes in order of largest to smallest could be described as: (a) Class

Table 2. Results of Latent Profile Analysis (N = 1,152)

Solution Class

Log- Likelihood

BIC VLMR LRT p

Bootstrap LRT p

Entropy Value

Subgroup Prevalence (%)

1 2 3 4 5 6

1 −1,714.56 3,457.48 – – – 100

2 −1,038.48 2,126.57 .15 <.001 .99 95 5

3 −716.53 1,503.94 <.01 <.001 .98 91 8 1

4 −557.76 1,207.67 .06 <.001 .96 85 11 3 1

5 −416.21 945.84 .42 <.001 .97 83 10 3 3 1

Note. BIC = Bayesian information criterion; VLMR LRT = Vuong-Lo-Mendell-Rubin likelihood ratio test; Bootstrap LRT = bootstrap likelihood ratio test.

Figure 1. Results of Latent Class Analysis for Traditional and Cyber Victim Classes

Note. N = 1,152.

295

Cyber Victimization in High School

1 = uninvolved, a group not involved in either type of victim- ization (83%, n = 956); (b) Class 2 = low dual, a group with low levels of both types of victimization (10%, n = 117); (c) Class 3 = moderate dual, a group with moderate levels of both types of victimization (3%, n = 37); (d) Class 4 = traditional only, a group not cyber victimized but traditionally victimized (3%, n = 29); and (e) Class 5 = high dual, a group with high levels of both types of victimization (1%, n = 11).

Based on model results, most youth who were victim- ized at low, moderate, and high rates experienced both types of victimization (i.e., traditional and cyber) at these respective levels. Only one small group of youth (3%, n = 29) experi- enced traditional victimization and no cyber victimization. The LCA did not identify a group of youth that experienced cyber victimization and no traditional victimization. See Table 3 for a summary of the demographic characteristics (i.e., gender, grade) of the five classes and means and standard deviations of all variables by class.

The correlation between traditional and cyber victim- ization was also examined to determine the amount of asso- ciation between the two experiences. Traditional and cyber victimization were strongly and positively correlated (r = .72, p < .01) among the students in this sample.

To ensure that the groups identified in the LCA reported different levels of traditional and cyber victimization, a MANOVA was conducted on the victimization scores (tradi- tional and cyber) by LCA class. There was a statistically sig- nificant difference in levels of victimization based on the five classes: F (2, 8) = 836.73, p < .001, Wilk’s Λ = .066, for both OBVQ traditional victimization, F (1, 4) = 766.97, p < .001, and cyber victimization scores, F (1,4) = 2160.29, p < .001. Post hoc Scheffé comparisons were analyzed to examine dif- ferences in traditional and cyber victimization scores for each

class. All classes had significantly different traditional victim- ization scores and cyber victimization scores in the expected direction (e.g., the high dual group had higher scores than the moderate dual, low dual, and uninvolved groups; the moderate dual group had higher scores than the low dual and uninvolved groups; the low dual group had higher scores than the unin- volved group) with the exception of one comparison. The moderate dual and traditional only groups did not have sig- nificantly different traditional victimization scores. In each comparison, the uninvolved group had the lowest scores.

Research Question 3: Victimization Class Outcomes

To explore differences in social and emotional out- comes among the five groups identified in the LCA, a MANOVA was conducted on the social–emotional outcomes (social–emotional risk score, internalizing problems, exter- nalizing problems) by LCA class (see Table 4). There was a statistically significant difference in social–emotional out- comes based on the five classes, F (3, 12) = 30.82, p < .001, Wilk’s Λ = .731. Follow-up ANOVAs were significant for the social–emotional risk scores, F (1, 4) = 84.85, p < .001; internalizing problems scores, F (1, 4) = 61.69, p < .001; and externalizing problems scores, F (1,4) = 18.91, p < .001. Scheffé’s post hoc comparisons were analyzed to examine differences in social–emotional risk, internalizing problems, and externalizing problems for each class. Table 4 provides the full results of all comparisons. As expected, uninvolved students had significantly lower levels of all three outcomes compared to all other classes. Students in the low dual group reported worse scores than the uninvolved group, which supports the idea that students who experience even low levels of cyber and traditional victimization are at

Table 3. Characteristics of Victimization Classes (i.e., Gender, Grade) and Means and Standard Deviations of Variables by Class

Class Uninvolved Low Dual Moderate Dual Traditional Only High Dual

N 956 117 37 29 11

Percent female 52.30% 35.00% 30.60% 62.10% 33.30%

Percent 9th graders 27.30% 25.80% 25.00% 48.30% 58.30%

Percent 10th graders 26.70% 25.00% 33.30% 17.20% 25.00%

Percent 11th graders 22.70% 32.50% 33.30% 20.70% 16.70%

Percent 12th graders 23.40% 16.70% 8.30% 13.80% 0.00%

Mean social–emotional risk score 48.57 (9.84) 59.41 (10.17) 66.61 (10.23) 62.28 (7.95) 74.58 (13.31)

Mean internalizing problems score 14.95 (3.11) 17.69 (3.64) 19.85 (3.89) 19.56 (3.41) 22.67 (4.23)

Mean externalizing problems score 15.83 (3.06) 17.49 (3.43) 17.98 (3.35) 19.18 (3.78) 19.00 (4.07)

Mean traditional victimization score 1.25 (0.22) 1.80 (0.42) 2.80 (0.58) 2.82 (0.65) 4.10 (0.85)

Mean cyber victimization score 1.09 (0.13) 1.72 (0.21) 2.52 (0.35) 1.18 (0.19) 4.17 (0.38)

Note. Standard deviations follow means in parentheses. N = 1,152 (two cases were excluded due to missing data).

School Psychology Review, 2017, Volume 46, No. 3

296

DOI: 10.17105/SPR-2016-0004.V46-3

greater risk for social and emotional problems than students who do not experience victimization at all. Students in the traditional only group did not report significantly different levels of social and emotional difficulties compared to the low dual and moderate dual groups. See Table 4 for detailed results.

Research Questions 4 and 5: Social–Emotional Outcomes

All three of the regressions involving the social–emo- tional variables as dependent variables were significant. See Table 5 for regression results. Traditional victimization was positively related to each problematic social–emotional vari- able (social–emotional risk, internalizing problems, and exter- nalizing problems) for both genders (β = .26, p < .001; β = .28, p < .001; and β = .09, p < .001, respectively). Cyber victimization was significantly and positively related to social–emotional risk scores (β = .29, p < .001) and internal- izing problems (β = .11, p = .04) above and beyond the inclu- sion of traditional victimization.

Three significant gender differences were found among the associations between the types of victimization and social–emotional outcomes. The Gender × Traditional Victimization interaction was significantly and negatively related to Externalizing Problems (β = −.37, p < .01). An additional regression analysis was performed to identify the simple slope of the association between Traditional Victimization and Externalizing Problems (β = .12, p = .07) for boys by dummy coding gender as 0 = boys and 1 = girls. Although both slopes were positive, the simple slope between Traditional Victimization and Externalizing Problems (β = .36, p < .001) for girls was stronger than that for boys. See Figure 2 for a visual display of the interaction.

In contrast to Traditional Victimization, the Gender × Cyber Victimization interaction was significantly and posi- tively related to Externalizing Problems (β = .27, p < .05). Again, an additional regression analysis was performed to identify the simple slope of the association between Cyber Victimization and Externalizing Problems (β = .14, p = .02) for boys by dummy coding gender as 0 = boys and 1 = girls. The association between Externalizing Problems and Cyber Victimization was positive and significant for boys but not significant for girls. See Figure 3 for a visual display of the interaction.

The third and final significant gender interaction was among the Gender × Cyber Victimization interaction term and Social–Emotional Risk scores (β = −.27, p = .03). An addi- tional regression analysis was performed to identify the sim- ple slope of the association between Cyber Victimization and Social–Emotional Risk scores (β = .12, p = .02) for boys by dummy coding gender as 0 = boys and 1 = girls. Although both slopes were positive, the simple slope between Cyber Victimization and Social–Emotional Risk scores (β = .29, p < .001) for girls was stronger than that for boys. See Figure 4 for a visual display of the interaction.

The full collection of independent variables explained a small proportion of the variance in externalizing problems (R2 = .09) and a large proportion of the variance in social– emotional risk scores (R2 = .26) and internalizing problems (R2 = .28).

DISCUSSION

The current study sought to answer the following questions: (1) Are there gender differences in prevalence rates of cyber victimization? (2) Are there groups of students that experience different rates of one, both, or neither type

Table 4. Post Hoc Comparisons of Outcomes by Class

Class A Class B Social–Emotional Risk Internalizing Problems Externalizing Problems

Mean Difference p Mean Difference p Mean Difference p

Low dual Moderate dual −6.76 .013 −1.99 .034 −.13 1.000

Traditional only −3.68 .565 −1.81 .145 −1.77 .150

Uninvolved 11.04 .000 2.78 .000 1.68 .000

High dual −15.03 .000 −4.95 .000 −1.50 .647

Moderate dual Traditional only 3.08 .835 .18 1.000 −1.64 .396

Uninvolved 17.80 .000 4.77 .000 1.81 .024

High dual −8.27 .181 −2.95 .109 −1.37 .790

Traditional only

Uninvolved 14.72 .000 4.59 .000 3.45 .000

High dual −11.35 .029 −3.13 .099 .27 1.000

Uninvolved High dual −26.07 .000 −7.72 .000 −3.18 .016

Note. Values in the mean difference columns correspond to the difference between the Class A column and Class B column. N = 1,152. Bolded numbers are significant.

297

Cyber Victimization in High School

Table 5. Regressions With Traditional and Cyber Victimization in Relation to Social–Emotional Outcomes

β SE β R2 Sig.

Social–emotional risk .26*** <.01

Traditional victimization 0.29*** 0.05 <.01

Cyber victimization 0.29*** 0.05 <.01

Gender 0.17* 0.08 .03

Gender × Traditional Victimization 0.14 0.11 .21

Gender × Cyber Victimization −0.27* 0.01 .03

Internalizing problems .28*** <.01

Traditional victimization 0.37*** 0.05 <.01

Cyber victimization 0.11* 0.05 .04

Gender 0.27*** 0.08 <.01

Gender × Traditional Victimization 0.09 0.11 .44

Gender × Cyber Victimization −0.12 0.12 .32

Externalizing problems .09*** <.01

Traditional victimization 0.36*** 0.05 <.01

Cyber victimization −0.03 0.06 .63

Gender 0.05 0.09 .55

Gender × Traditional Victimization −0.37** 0.13 <.01

Gender × Cyber Victimization 0.27* 0.14 <.05

Note. N = 1,152; gender was dummy coded (0 = girls, 1 = boys). * p < .05. ** p < .01. *** p < .001.

Figure 2. The Effect of Gender on the Association Between Traditional Victimization and Externalizing Problems

Note. N = 1,152.

School Psychology Review, 2017, Volume 46, No. 3

298

DOI: 10.17105/SPR-2016-0004.V46-3

of victimization? (3) Are there differences in social– emotional outcomes across victimization classes? (4) Is cyber victimization associated with social–emotional out- comes (i.e., social–emotional risk, internalizing problems, externalizing problems) when controlling for traditional victimization experiences? and (5) Does the association between cyber victimization and social–emotional outcomes differ by gender?

Research Questions

Discussion of results are organized by research question.

Gender Differences The literature has reported mixed findings regarding gen-

der differences in cyber victimization experiences. While some studies have found no gender differences for cyber

Figure 3. The Effect of Gender on the Association Between Cyber Victimization and Externalizing Problems

Note. N = 1,152.

Figure 4. The Effect of Gender on the Association Between Cyber Victimization and Social– Emotional Risk Scores

Note. N = 1,152.

299

Cyber Victimization in High School

victimization (Beran & Li, 2005; Hinduja & Patchin, 2008; Kowalski & Limber, 2007; Li, 2006; Patchin & Hinduja, 2006; Raskauskas, 2009; Varjas et al., 2009; Williams & Guerra, 2007; Ybarra, 2004; Ybarra et al., 2007), the current study found high school girls reported significantly less cyber victim- ization than high school boys. Other studies have also found that girls are more likely to be victimized online than boys (Dempsey et al., 2009; Ybarra & Mitchell, 2007; Ybarra et al., 2007); however, Erdur-Baker (2010) found that boys were more likely than girls to be both cyber victims and cyberbullies.

The inconclusive gender results may be caused by dif- ferences in school populations, whether or not other internet usage factors like time spent online or engagement in risky online behavior were included as covariates, or differences in how victimization was specified and measured (Brown et al., 2014; Carbone-Lopez et al., 2010; Erdur-Baker, 2010). For example, school violence researchers have found misleading gender differences in victimization prevalence rates in the past because girls and boys engage in different types of victimiza- tion at different rates and measures may be better designed to capture one type (e.g., direct and overt victimization) over another (Carbone-Lopez et al., 2010).

Victimization Groups The current study also investigated the overlap between

traditional and cyber victimization experiences among high school youth by examining groups of students whom experi- ence different rates of one, both, or neither type of victimiza- tion. Although prior research has found an overlap between these two types of victimization, the exact nature of the over- lap was not clear (Fredstrom et al., 2011; Hinduja & Patchin, 2008; Li, 2006; Mitchell et al., 2007; Raskauskas, 2009; Raskauskas & Stoltz, 2007; Ybarra & Mitchell, 2004). Results indicate that most students who experience one type of vic- timization at a low, moderate, or high level also experience the other type of victimization at a similar level. Only one very small latent class (3%) was composed of students who experienced traditional victimization without cyber victim- ization. Furthermore, no latent class was identified that con- sisted of students experiencing cyber victimization without traditional victimization. Thus, for the 17% of students who experience victimization, most experience both types of vic- timization. For schools, this is important information. If the vast majority of students who experience one type of victim- ization are also victimized in other ways, schools and parents should work together to identify and address victimization on and off school grounds. General demographic characteristics of latent classes (i.e., gender, grade) were presented in Table 3; future research should be conducted to understand more about the characteristics of youth who make up these victim- ization groups.

Victimization Groups and Social–Emotional Outcomes There were no significant differences in the mean levels

of global risk of social–emotional difficulties, internalizing problems, and externalizing problems between each class.

However, important differences did emerge between some classes that are informative for practitioners. Uninvolved stu- dents reported the lowest level of social, emotional, and behavior problems for the outcomes in the current study. Additionally, students in the low dual group had higher levels of social and emotional problems compared to students who were not victimized; thus, youth who experience even low levels of cyber and traditional victimization are at greater risk for social and emotional problems than students who do not experience victimization at all. Therefore, schools and prac- titioners should not discount potential mental health difficul- ties among students who are victimized, even if it does not seem to occur frequently. Moreover, students in the low dual and moderate dual groups reported similar levels of social and emotional difficulties as the traditional only group. Although the mode of perpetration may differ, the negative impact on social and emotional functioning is similar. Being the victim of bullying, either traditional only or with low to moderate levels of both cyber and traditional, puts students at risk for difficulties. Practitioners should be keenly aware of peer vic- timization experiences and be prepared to address related social and emotional functioning difficulties.

Cyber Victimization and Social–Emotional Outcomes Given the association of cyber and traditional victim-

ization, it is important to include both constructs in correla- tional research designs (Erdur-Baker, 2010). If the current study only investigated the association of cyber victimization to the social–emotional outcomes, it may overemphasize the impact of cyber victimization because traditional victimiza- tion often co-occurs with cyber victimization and also affects students’ social–emotional outcomes. Furthermore, tradi- tional and cyber victimization have been found to be differ- entially related to social and emotional variables like depression and number of friends (Wang et al., 2009; Wang, Nansel, & Iannotti, 2011). The current study found that tradi- tional victimization was a unique significant predictor for all of the social–emotional outcomes, and cyber victimization was a unique significant predictor for two of the three included outcomes (a global indicator of risk of social–emotional dif- ficulties as well as an indicator of internalizing problems) when controlling for the impact of traditional victimization.

Cyber Victimization and Social–Emotional Outcomes and Gender

Gender differences were investigated in the association between victimization and the social–emotional outcomes in the study. Three gender differences were found across the two types of victimization and three social–emotional outcomes. For externalizing problems, there was a significant effect of the interaction between traditional victimization and gender. Although traditional victimization was positively associated with externalizing problems for girls and boys, the association was stronger for girls. This finding may be the result of dif- ferent social norms among girls and boys. For example, girls who demonstrate externalizing behaviors such as conduct

School Psychology Review, 2017, Volume 46, No. 3

300

DOI: 10.17105/SPR-2016-0004.V46-3

problems, hyperactivity, or other acting out behaviors follow- ing victimization may be more likely to be further victimized than boys who exhibit externalizing problems because such behaviors are more atypical for girls and prompt more social exclusion and bullying (Tiet, Wasserman, Loeber, McReynolds, & Miller, 2001). Alternatively, because conduct problems are less common among girls, girls who do demon- strate high levels of externalizing problems have more severe social–emotional problems overall and thus are likely to expe- rience victimization (Nock, Kazdin, Hiripi, & Kessler, 2006).

The interaction of cyber victimization and gender also significantly predicted externalizing problems. In this case, the association between cyber victimization and externalizing problems was positive and significant for boys but not signif- icant for girls. One of the venues in which cyber victimization can happen is through online gaming. Typically, boys are more likely to spend time engaged in this activity than are girls (Lenhart et al., 2008). Boys who have externalizing prob- lems and act out (e.g., have angry tantrums) may prompt online victimization. Girls with externalizing problems, how- ever, do not appear to be at greater risk for cyber victimization than girls without externalizing problems based on this result.

Finally, the interaction of cyber victimization and gen- der negatively and significantly predicted social–emotional risk scores. Although the association between cyber victim- ization and social–emotional risk scores was positive for both genders, it was stronger for girls than for boys. This pattern of results indicates that cyber victimization is associated with greater global social–emotional risk for both boys and girls, but even more so for girls. It is possible that cyber victimiza- tion is more strongly associated with social–emotional risk for girls than boys because it has a more global impact on their social lives. For example, most girls engage in many forms of social media online, whereas boys are less likely to be involved in social media and more likely to be involved in discrete activities like gaming. Girls may be targeted for cyber victimization on multiple social media sites for a variety of reasons, which may have a wide effect on their overall well-being. Potential factors underlying the gender differ- ences identified in this study should be hypothesized and examined for validity. If specific factors are identified, they can be used to guide interventions.

Limitations and Future Directions

Several limitations of the current study should be noted. While the sample size is large, the majority of the sample is Caucasian (75%). Thus, the sample was not ethnically diverse, with the next largest group of students being Latino (12%). Given that the sample is limited to high school, developmental differences were not examined. A sample of middle and high school students may provide a better understanding of devel- opmental differences. All measures used were self-reported; therefore, some students may have under- or overreported their experiences with victimization. Some may consider self-reported data to be a limitation, but cyber victimization

research to date has mainly relied on self-reported data. The anonymity and sometimes covert nature inherent in some cyber behaviors may mean that self-reported data is necessary to examine and understand this construct; however, self-re- porting was used for all measures in the study, which is a limitation because actual instances of bullying were not observed. Consequently, objective prevalence rates of tradi- tional and cyber victimization were not available to the authors. The CVS that was used to assess participant experi- ences with cyber victimization was used in a previously pub- lished study (Brown et al., 2014), but the sample in this validation study was quite small (N = 106) and consisted of middle school students only. The implications of this study are also limited because no data measuring students’ amount of internet/technology use were collected; thus, we could not control for time spent online. Students must use technology to provide cyberbullies access to them; those students who use technology most frequently are arguably the most accessible victims and may experience relatively higher levels of victim- ization as a result.

Another limitation to the study was that the fifth class of students identified via the LCA was quite small. This class was the high dual group and comprised only 1% of the sample (n = 11 students). Although this group was identified via sta- tistical techniques, more research needs to be conducted to learn about the prevalence rates of youth who experience high rates of both face-to-face and online victimization.

Research should delve into the conflicting gender results in terms of prevalence, levels, and outcomes associated with cyber victimization. A better understanding of why some students experience more negative outcomes due to the addi- tive effects of traditional and online victimization would help schools and practitioners more effectively identify at-risk stu- dents, intervene early, and implement prevention programs.

Summary and Implications for Practitioners

Bullying, both face-to-face and online, is a problem for our youth. While gender differences in levels of cyber victim- ization remain inconclusive in the literature, the current study found that levels of cyber victimization differed by gender, with boys reporting higher levels of online victimization than girls. Although most students in the current study were not involved in cyber or traditional victimization (83%), a fair portion of students were involved to some degree. Specifically, 10% had low involvement in both types, 3% had moderate involvement in both types, 1% had high involvement in both types, and 3% experienced traditional victimization but not cyber victimization. Although only a small percentage of stu- dents were moderately or highly victimized via traditional and online means (4.5% total), there are 52 youth from this sample that experience both cyber and traditional victimization on a regular basis. Given the negative outcomes associated with victimization and the overlap in experiencing both types of victimization, practitioners need to be aware of research- based prevention and intervention strategies.

301

Cyber Victimization in High School

Diamanduros et al. (2008) provided guidelines for how school psychologists can be involved in preventing and inter- vening in cyberbullying. School psychologists should pro- mote awareness of cyberbullying and its negative impact on youth, be involved in assessing the prevalence of cyberbully- ing, and develop polices and prevention programs for address- ing cyberbullying in schools. More specifically, Diamanduros et al. suggested that prevention and intervention programs include ways to educate all school stakeholders about the defi- nition and prevalence of cyberbullying and the importance of bystanders in intervening in or reporting cyberbullying, as well as ways to teach youth about internet safety, deciding when to share private information online, how cyberbullying can be traced, and legal actions that can occur.

Based on results of past research and the current study, students who are bullied both traditionally and electronically are a small but at-risk population. Teachers, parents, and prac- titioners who suspect a student is involved in one form of victimization should ask if that student is involved in other forms of victimization. School professionals and practitioners should be aware that the risk of negative social–emotional outcomes for these students is greater than the risk to children suffering only one form of victimization. These findings should allow for better identification of at-risk students, as well as better development of prevention and intervention strategies and programs.

APPENDIX A

Cyber Victimization Survey

REFERENCES

Bendixen, M., & Olweus, D. (1999). Measurement of antisocial behaviour in early adolescence and adolescence: Psychometric properties and sub- stantive findings. Criminal Behaviour and Mental Health, 9(4), 323–354. doi:10.1002/cbm.330

Beran, T., & Li, Q. (2005). Cyber-harassment: A study of a new method for an old behavior. Educational Computing Research, 32(3), 265–277. doi:10.2190/8YQM-B04H-PG4D-BLLH

Bonanno, R. A., & Hymel, S. (2013). Cyber bullying and internalizing difficul- ties: Above and beyond the impact of traditional forms of bullying. Journal of Youth and Adolescence, 42, 685–697. doi:10.1007/s10964-013-9937-1

1 = It hasn’t happened at all in the past couple of months 2 = Only 1 or 2 times in the past couple of months 3 = 2 or 3 times a month 4 = About once a week 5 = Several times a week

For the following questions, think about things that have happened to you. In the last 2–3 months …

Has someone lied about you online? 1 2 3 4 5

Have you been physically threatened online? 1 2 3 4 5

Has something posted online made others laugh at you? 1 2 3 4 5

Have you been called names online? 1 2 3 4 5

Has someone pretended to be you online in order to tease or hurt you? 1 2 3 4 5

Has someone intentionally shared a private message that you sent to a friend in order to tease or hurt you?

1 2 3 4 5

Have you seen conversations or pictures online that made you feel excluded? 1 2 3 4 5

Have you felt excluded while involved in an online activity? 1 2 3 4 5

Has someone posted pictures of you online in order to tease or hurt you? 1 2 3 4 5

Has someone intentionally shared an embarrassing picture or video of you in order to tease or hurt you?

1 2 3 4 5

Have you been made fun of online? 1 2 3 4 5

Have you been teased online? 1 2 3 4 5

Have rumors been spread about you online? 1 2 3 4 5

Has something posted online made you upset? 1 2 3 4 5

Has someone pretended to be someone else online in order to tease or hurt you? 1 2 3 4 5

School Psychology Review, 2017, Volume 46, No. 3

302

DOI: 10.17105/SPR-2016-0004.V46-3

Brown, C. F., Demaray, M. K., & Secord, S. M. (2014). Cyber victimization in middle school and relations to social emotional outcomes. Computers in Human Behavior, 35, 12–21. doi:10.1016/j.chb.2014.02.014

Carbone-Lopez, K., Esbensen, F.-A., & Brick, B. T. (2010). Correlates and consequences of peer victimization: Gender differences in direct and indi- rect forms of bullying. Youth Violence and Juvenile Justice, 8(4), 332–350. doi:10.1177/1541204010362954

Crick, N. R., & Grotpeter, J. K. (1995). Relational aggression, gender, and social-psychological adjustment. Child Development, 66, 710–722. doi:10.2307/1131945

Dehue, F., Bolman, C., & Völlink, T. (2008). Cyberbullying: Youngsters’ experiences and parental perception. CyberPsychology & Behavior, 11(2), 217–223. doi:10.1089/cpb.2007.0008

Dempsey, A. G., Sulkowski, M. L., Nichols, R., & Storch, E. A. (2009). Differences between peer victimization in cyber and physical settings and associated psychosocial adjustment in early adolescence. Psychology in the Schools, 46, 962–972. doi:10.1002/pits.20437

Diamanduros, T., Downs, E., & Jenkins, S. J. (2008). The role of school psychologists in the assessment, prevention, and intervention of cyberbul- lying. Psychology in the Schools, 45, 693–704. doi:10.1002/pits.20335

Erdur-Baker, Ö. (2010). Cyberbullying and its correlation to traditional bul- lying, gender and frequent and risky usage of internet-mediated commu- nication tools. New Media & Society, 12(1), 109–125. doi:10.1177/1461444809341260

Fredstrom, B. K., Adams, R. E., & Gilman, R. (2011). Electronic and school- based victimization: Unique contexts for adjustment difficulties during adolescence. Journal of Youth and Adolescence, 40, 405–415. doi:10.1007/ s10964-010-9569-7

Genta, M. L., Menesini, E., Fonzi, A., Costabile, A., & Smith, P. K. (1996). Bullies and victims in schools in central and southern Italy. European Journal of Psychology of Education, 11(1), 97–110. doi:10.1007/BF 03172938

George, M. J., & Odgers, C. L. (2015). Seven fears and the science of how mobile technologies may be influencing adolescents in the digital age. Perspectives on Psychological Science, 10(6), 832–851. doi:10.1177/ 1745691615596788

Goodman, A., Lamping, D. L., & Ploubidis, G. B. (2010). When to use broader internalising and externalising subscales instead of the hypothe- sised five subscales on the Strengths and Difficulties Questionnaire (SDQ): Data from British parents, teachers and children. Journal of Abnormal Child Psychology, 38(8), 1179–1191. doi:10.1007/s10802 -010-9434-x

Goodman, R. (2001). Psychometric properties of the Strengths and Difficulties Questionnaire (SDQ). Journal of the American Academy of Child and Adolescent Psychiatry, 40(11), 1337–1345. doi:10.1097/ 00004583-200111000-00015

Gradinger, P., Strohmeier, D., & Spiel, C. (2009). Traditional bullying and cyberbullying: Identification of risk groups for adjustment problems. Journal of Psychology, 217(4), 205–213. doi:10.1027/0044-3409 .217.4.205

Hawker, D. S., & Boulton, M. J. (2000). Twenty years’ research on peer victimization and psychosocial maladjustment: A meta-analytic review of cross-sectional studies. Journal of Child Psychology and Psychiatry, 41(4), 441–455. doi:10.1111/1469-7610.00629

Hinduja, S., & Patchin, J. (2007). Cyberbullying and online aggression survey instrument 2007 version. Cyberbullying Research Center.

Hinduja, S., & Patchin, J. W. (2008). Cyberbullying: An exploratory analysis of factors related to offending and victimization. Deviant Behavior, 29(2), 129–156. doi:10.1080/01639620701457816

Hinduja, S., & Patchin, J. W. (2010). Bullying, cyberbullying, and suicide. Archives of Suicide Research, 14(3), 206–221. doi:10.1080/13811118.20 10.494133

IBM Corp. (2013). IBM SPSS Statistics for Windows (Version 22.0) [Computer software]. Armonk, NY: IBM Corp.

Kamphaus, R. W., & Reynolds, C. R. (2007). BASC-2 Behavioral and Emotional Screening System manual. Circle Pines, MN: Pearson.

Kann, L., Kinchen, S., Shanklin, S. L., Flint, K. H., Hawkins, J., Harris, W. A., … Zaza, S. (2014). Youth Risk Behavior Surveillance—United States, 2013. Morbidity and Mortality Weekly Report, 63(SS04), 1–168.

Kowalski, R. M., & Limber, S. P. (2007). Electronic bullying among middle school students. Journal of Adolescent Health, 41, S22–S30. doi:10.1016/ j.jadohealth.2007.08.017

Kyriakides, L., Kaloyirou, C., & Lindsay, J. (2006). An analysis of the revised Olweus Bully/Victim Questionnaire using the Rasch measurement model. British Journal of Educational Psychology, 76(4), 781–801. doi:10.1348/ 000709905X53499

Lenhart, A., Kahne, J., Middaugh, E., Macgill, A. R., Evans, C., & Vitak, J. (2008). Teens, video games, and civics: Teens’ gaming experiences are diverse and include significant social interaction and civic engagement. (Pew Internet & American Life Project Report). Retrieved from http:// www.pewinternet.org/2008/09/16/teens-video-games-and-civics/

Li, Q. (2006). Cyberbullying in schools: A research of gender differences. School Psychology International, 27, 157–170. doi:10.1177/01430 34306064547

Luk, J. W., Wang, J., & Simons-Morton, B. G. (2010). Bullying victimization and substance use among U.S. adolescents: Mediation by depression. Prevention Science, 11(4), 355–359. doi:10.1007/s11121-010-0179-0

Machmutow, K., Perren, S., Sticca, F., & Alsaker, F. D. (2012). Peer victim- isation and depressive symptoms: Can specific coping strategies buffer the negative impact of cybervictimisation? Emotional and Behavioural Difficulties, 17(3-4), 403–420. doi:10.1080/13632752.2012.704310

Mason, K. L. (2008). Cyberbullying: A preliminary assessment for school personnel. Psychology in the Schools, 45, 323–348. doi:10.1002/pits.20301

Mitchell, K. J., Jones, L. M., Turner, H. A., Shattuck, A., & Wolak, J. (2016). The role of technology in peer harassment: Does it amplify harm for youth? Psychology of Violence, 6(2), 193–204. doi:10.1037./a0039317

Mitchell, K. J., Ybarra, M., & Finkelhor, D. (2007). The relative importance of online victimization in understanding depression, delinquency, and substance use. Child Maltreatment, 12(4), 312–324. doi:10.1177/1077 559507305996

Muthén, L. K., & Muthén, B. O. (2007). Mplus user’s guide (6th ed.). Los Angeles, CA: Muthén & Muthén.

Nansel, T. R., Overpeck, M., Pilla, R. S., Ruan, W. J., Simons-Morton, B., & Scheidt, P. (2001). Bullying behaviors among US youth: Prevalence and association with psychosocial adjustment. Journal of the American Medical Association, 285(16), 2094–2100. doi:10.1001/jama.285.16.2094

Nock, M. K., Kazdin, A. E., Hiripi, E., & Kessler, R. C. (2006). Prevalence, subtypes, and correlates of DSM-IV conduct disorder in the National Comorbidity Survey Replication. Psychological Medicine, 36(5), 699– 710. doi:10.1017/S0033291706007082

Olweus, D. (1994). Annotation: Bullying at school: Basic facts and effects of a school based intervention program. Journal of Child Psychology and Psychiatry, 35(7), 1171–1190. doi:10.1111/j.1469-7610.1994. tb01229.x

Olweus, D. (1996). The Revised Olweus Bully/Victim Questionnaire, Mimeo. Bergen, Norway: Research Center for Health Promotion (HIMIL), University of Bergen.

Olweus, D. (1997). Bully/victim problems in school: Knowledge base and an effective intervention program. The Irish Journal of Psychology, 18(2), 170–190. doi:10.1080/03033910.1997.10558138

Parker, J. G., & Asher, S. R. (1987). Peer relations and later personal adjust- ment: Are low-accepted children at risk? Psychological Bulletin, 102(3), 357–389. doi:10.1037/0033-2909.102.3.357

Patchin, J. W., & Hinduja, S. (2006). Bullies move beyond the schoolyard: A preliminary look at cyberbullying. Youth Violence and Juvenile Justice, 4(2), 148–169. doi:10.1177/1541204006286288

Prinstein, M. J., Boergers, J., & Vernberg, E. M. (2001). Overt and relational aggression in adolescents: Social-psychological adjustment of aggressors and victims. Journal of Clinical Child & Adolescent Psychology, 30(4), 479–491. doi:10.1207/S15374424JCCP3004_05

Raskauskas, J. (2009). Text-bullying: Associations with traditional bullying and depression among New Zealand adolescents. Journal of School Violence, 9(1), 74–97. doi:10.1080/15388220903185605

Raskauskas, J., & Stoltz, A. D. (2007). Involvement in traditional and elec- tronic bullying among adolescents. Developmental Psychology, 43, 564– 575. doi:10.1037/0012-1649.43.3.564

303

Cyber Victimization in High School

Reynolds, C. R., & Kamphaus, R. W. (2004). Behavior Assessment System for Children–Second Edition. Circle Pines, MN: AGS Publishing.

Salmon, G., James, A., & Smith, D. M. (1998). Bullying in school: Self reported anxiety, depression, and self esteem in secondary children. British Medical Journal, 317, 924–925. doi:10.1136/bmj.317. 7163.924

Shariff, S. (2005). Cyber-dilemmas in the new millennium: School obliga- tions to provide student safety in a virtual school environment. McGill Journal of Education, 40(3), 467–487.

Solberg, M., & Olweus, D. (2003). Prevalence estimation of school bullying with the Olweus Bully/Victim Questionnaire. Aggressive Behavior, 29(3), 239–268. doi:10.1002/ab.10047

Sticca, F., & Perren, S. (2013). Is cyberbullying worse than traditional bully- ing? Examining the differential roles of medium, publicity, and anonymity for the perceived severity of bullying. Journal of Youth and Adolescence, 42, 739–750. doi:10.1007/s10964-012-9867-3

Tiet, Q. Q., Wasserman, G. A., Loeber, R., McReynolds, L. S., & Miller, L. S. (2001). Developmental and sex differences in types of conduct prob- lems. Journal of Child and Family Studies, 10, 181–197. doi:10.1023/A:1016637702525

Varjas, K., Henrich, C. C., & Meyers, J. (2009). Urban middle school stu- dents’ perceptions of bullying, cyberbullying, and school safety. Journal of School Violence, 8(2), 159–176. doi:10.1080/15388220802074165

Wang, J., Iannotti, R. J., & Nansel, T. R. (2009). School bullying among adolescents in the United States: Physical, verbal, relational, and cyber. Journal of Adolescent Health, 45(4), 368–375. doi:10.1016/ j.jadohealth.2009.03.021

Wang, J., Nansel, T. R., & Iannotti, R. J. (2011). Cyber and traditional bully- ing: Differential association with depression. Journal of Adolescent Health, 48(4), 415–417. doi:10.1016/j.jadohealth.2010.07.012

Williams, K. R., & Guerra, N. G. (2007). Prevalence and predictors of Internet bullying. Journal of Adolescent Health, 41(6), S14–S21. doi:10.1016/j.jadohealth.2007.08.018

Ybarra, M. L. (2004). Linkages between depressive symptomatology and Internet harassment among young regular Internet users. CyberPsychology & Behavior, 7(2), 247–257. doi:10.1089/109493104323024500

Ybarra, M. L., Diener-West, M., & Leaf, P. J. (2007). Examining the overlap in Internet harassment and school bullying: Implications for school inter- vention. Journal of Adolescent Health, 41(6), S42–S50. doi:10.1016/j. jadohealth.2007.09.004

Ybarra, M. L., & Mitchell, K. J. (2004). Youth engaging in online harassment: Associations with caregiver–child relationships, Internet use, and personal characteristics. Journal of Adolescence, 27, 319–336. doi:10.1016/ j.adolescence.2004.03.007

Ybarra, M. L., & Mitchell, K. J. (2007). Prevalence and frequency of Internet harassment instigation: Implications for adolescent health. Journal of Adolescent Health, 41(2), 189–195. doi:10.1016/j.jadohealth.2007. 03.005

Date Received: September 16, 2015 Date Accepted: October 12, 2016

Associate Editor: Erin Dowdy

AUTHOR BIOGRAPHICAL STATEMENTS

Christina Brown is a school psychologist at Community High School District 155 in Crystal Lake, IL. She received her doctoral degree in School Psychology from Northern Illinois University in May 2014. She is inter- ested in how increased technology in the classroom impacts social–emotional and academic outcomes as well as supporting students with anxiety and depression to increase resiliency to better manage academic and social demands.

Michelle K. Demaray is a professor in the School Psychology program at Northern Illinois University and editor of the Journal of School Psychology. Her research interests include the role of social support in the lives of youth. In addition, she studies bullying and victimization in schools, including cyber victimization and the role of bystanders in bullying.

Jaclyn Tennant is a School Psychology doctoral student at Northern Illinois University and 2016 recipient of a Society for the Study of School Psychology (SSSP) Dissertation Grant Award. Her research interests include peer victimization, emotion regulation, social support, and social–emotional well-being. She is interested in social, emotional, and cognitive factors that differentiate active from passive bystanders as well as factors that promote resiliency for students involved in bullying and other types of adversity.

Lyndsay N. Jenkins is an assistant professor in the School Psychology program and Department of Psychology at Eastern Illinois University. Her research interests focus on bullying and victimization, defending behaviors in youth and young adults, and social and emotional barriers to academic achievement.

Copyright of School Psychology Review is the property of National Association of School Psychologists and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.