Peer Victimization and Social-Emotional Functioning

Abstract

The present longitudinal study used a social-ecological framework to explore the extent to which peer victimization and aggression were associated with changes in concentration problems and emotion regulation among elementary students in general versus special education, while accounting for student demographics and school contextual factors. Data come from a multilevel, longitudinal dataset including 7,314 students (kindergarten through Grade 2) across 37 elementary schools. Multilevel analyses indicated that on average, students in special education had worse trajectories, including higher concentration problem scores and lower emotion regulation skills over time than students in general education. Children who were victimized and in special education displayed significantly more concentration problems and worse emotion regulation skills at baseline as compared with other students. The findings highlight the risks associated with prolonged victimization among children in special education. Implications for prevention programming targeting social-emotional functioning and special education populations are discussed.

Keywords

special education peer victimization school environment

Research has shown that peer victimization negatively influences not only students’ current functioning, but also their long-term ability to succeed in the classroom (Bogart et al., 2014; Bradshaw, Waasdorp, O’Brennan, & Gulemetova, 2013; Ttofi, Farrington, & Lösel, 2012). There is growing evidence that students who receive special education services are particularly at risk of peer victimization and involvement in bullying (Rose, Monda-Amaya, & Espelage, 2011; Swearer, Wang, Maag, Siebecker, & Frerichs, 2012). However, much of the extant research has been cross-sectional, and thus, additional longitudinal research is needed. The present study aimed to examine peer victimization and aggression and their association with concentration problems and emotion regulation among elementary students in general versus special education. In addition, we examined whether changes in children’s concentration problems and emotion regulation skills over time vary by their role as a victim or aggressor, special education status, student demographics, and school contextual factors.

Peer Victimization in General and Special Education

In comparison with general education students, national data indicate that students with emotional, developmental, or behavioral problems (i.e., students receiving treatment and/or counseling services) are twice as likely to be victimized, three times as likely to bully other children, and three times as likely to be classified as a bully/victim (Van Cleave & Davis, 2006). Longitudinal studies on students with disabilities (e.g., learning disabilities, speech-language impairments, autism spectrum disorders) show that these youth experience more victimization as they move through school (Blake, Lund, Zhou, Kwok, & Benz, 2012). For example, Son, Parish, and Peterson (2012) used longitudinal data on peer victimization from 3-, 4-, and 5-year-olds in special education and found that parents’ reports of their child’s peer victimization increased from 21% to 30% over the 3-year period. However, less is known about these longitudinal trends in school-aged children with disabilities.

Although some research has documented differences between general and special education students, not all studies have found significant differences in peer victimization between these student populations. Rose, Espelage, Aragon, and Elliott (2011) found that among middle school youth, students with disabilities reported higher rates of victimization and significantly more physical altercations with peers than students without disabilities (i.e., attention deficit hyperactivity disorder [ADHD], emotional/behavioral disorders, learning disabilities); however, there were not significant differences between groups in terms of perpetration of bullying. Conversely, Kokkinos and Antoniadou (2013) found that elementary school students with specific learning disabilities were more likely than their peers to experience direct and indirect peer victimization and engage in bullying. Given these inconsistent findings, additional research is needed regarding elementary school students to determine patterns of students’ involvement in peer victimization across general and special education student groups.

Effects of Peer Victimization

Approximately one quarter of students in general education are frequently victimized by their peers (O’Brennan, Bradshaw, & Sawyer, 2009). Youth who are victimized by bullying tend to have poorer psychosocial functioning than their uninvolved peers (O’Brennan et al., 2009), and those problems appear to persist over time (Bogart et al., 2014). Specific to the present study, research has shown an association between peer victimization and subsequent concentration problems and emotion regulation; thus, we considered the literature related to these particular outcomes in greater detail.

Concentration Problems

Research shows a child’s ability to stay focused, on task, and successfully complete assignments is linked to his or her academic competence and displays of problem behavior in the classroom (Barkley, 2003; Thomas, Bierman, Thompson, Powers, & Conduct Problems Prevention Research Group, 2008). Students with cognitive impairments (e.g., learning disability, intellectual disability) often have difficulty completing academic tasks at the same pace as their peers, which may increase their likelihood for frustration at school. Attentional skills also help children process social information, such as non-verbal cues, when interacting with their peers (Barkley, 2003). Fine, Semrud-Clikeman, Butcher, and Walkowiak (2008) asked youth with autism spectrum disorders or ADHD to review social vignettes, and found that impaired social perceptions were related to the child’s attentional skills. Consequently, youth in special education may be more inclined to respond to a situation inappropriately (e.g., laughing at the wrong time) due to attentional difficulties, which can increase negative responses from peers. Thus, it was predicted that students in special education and those who are frequently victimized would show more problems with concentration compared with their peers in general education at baseline and over time.

Emotion Regulation

Children in special education may have particular difficulties with emotion regulation, especially when faced with challenging social situations. Students with emotional and behavioral disorders are more likely to engage in aggressive behavior and make verbal threats as compared with their peers (Rose, Forber-Pratt, Espelage, & Aragon, 2013). Although children in special education, particularly those with emotional and behavioral disorders, may be more likely to engage in aggressive behavior, it is unclear whether these problems with emotion regulation and concentration persist over time and whether they are exacerbated by involvement in peer victimization and aggression. In fact, research indicates that students who are able to regulate their emotions at an early age are more academically successful, both concurrently and in later years (Graziano, Reavis, Keane, & Calkins, 2007). Moreover, having the skills to attend to relevant stimuli, inhibit inappropriate behavior, and effectively socialize with adults and peers affects concurrent and later academic achievement (Spinrad & Eisenberg, 2009). Youth with positive social-emotional skills are also less likely to be involved in peer victimization (Swearer et al., 2012).

School Contextual Effects

Espelage and Swearer’s (2004) social-ecological model of bullying and peer victimization highlights the importance of contextual influences on interpersonal relationships across a child’s ecological domains (i.e., classroom, school, home). For instance, Thomas and colleagues (2008) found that individual inattentiveness and placement in a disorganized classroom exacerbated the risk for elementary school students to exhibit aggressive-disruptive behavior at school. Likewise, a multilevel study by O’Brennan, Bradshaw, and Furlong (2014) found that positive teacher perceptions of school climate (e.g., principal leadership, pride in school) and lower concentration of students living in poverty were associated with fewer disruptive behavior problems among elementary school youth. Moreover, research suggests that school-level indicators of disorder, such as a high student–teacher ratio, concentration of student poverty, suspension rate, and student mobility, were associated with an increased risk of aggressive–retaliatory attitudes (Bradshaw, Sawyer, & O’Brennan, 2009). As a result, we explored the extent to which aspects of the school, such as student–teacher ratio, student body composition, faculty and student mobility, and school climate were associated with the two student outcomes of social-emotional functioning and concentration problems.

Overview of the Current Study

The present study applied the social-ecological framework to examine the association between peer victimization/aggression and concentration problems and emotion regulation among elementary students in general versus special education. Furthermore, the study uses longitudinal data to determine whether changes in children’s concentration problems and emotion regulation skills over time vary by their role in peer victimization and special education status. Based on prior research, we hypothesized that students frequently involved in peer victimization would demonstrate worse concentration problems and emotion regulation skills compared with their peers who were not involved, both at baseline (i.e., intercept) and over time (i.e., slope). Likewise, youth in special education who are also frequently engaged in peer victimization were expected to show significantly worse concentration and emotion regulation outcomes as compared with general education students. Last, we aim to identify student demographics and school contextual factors that help explain these relationships. Given that students’ peer victimization and educational experiences are shaped by individual factors (e.g., special education status, demographics) as well as the school environment (e.g., student–teacher ratio, perceptions of school climate; Waasdorp, Pas, O’Brennan, & Bradshaw, 2011), individual and school factors were hypothesized to significantly predict student concentration and emotion regulation outcomes.

Method

Data for this study come from a large, population-based study of 37 elementary schools across the state of Maryland. The sample included 7,314 students who were in kindergarten, first grade, and second grade when the study was initiated. A total of 5 data points (fall and spring of Year 1, and spring of Years 2, 3, and 4) were collected over the course of 4 school years. These data were collected as part of a larger study of a school reform model.

Measures

Concentration problems and emotion regulation

The outcomes of interest were assessed through the Teacher Observation of Classroom Adaptation–Checklist (TOCA-C; Koth, Bradshaw, & Leaf, 2009; Werthamer-Larsson, Kellam, & Wheeler, 1991), a research-based measure of student behavior problems that has been well-validated. Specifically, the TOCA-C was used to assess concentration problems (“pays attention,” “stays on task”; seven items, α = .96), and emotion regulation (“stops and calms down when angry or upset”; four items, α = .89). Teachers responded on a 5-point Likert-type scale: 1 = never to 6 = almost always. Responses on all scale items were averaged, with higher scores on concentration problems indicating greater problems, but higher scores on emotion regulation reflecting better adjustment. These scales exhibit strong internal consistency, have a consistent factor structure over time (Koth et al., 2009), and relate to external criteria (Bradshaw, Waasdorp, & Leaf, in press; Stormshak, Bierman, Bruschi, Dodge, & Coie, 1999).

Peer victimization and aggression

Teacher reports of peer victimization and peer aggression were also assessed using items on the TOCA-C (Koth et al., 2009; Werthamer-Larsson et al., 1991). Teacher reports of victimization were assessed through three TOCA-C items (“is rejected,” “does not have many friends,” and “is not liked by classmates”; α = .84). The peer aggression subscale included four items (“teases classmates,” “yells at others,” “harms others,” and “fights”; α = .87), to which teachers responded on a Likert-type scale (1 = never to 6 = almost always). Confirmatory factor analysis indicated support for the Victimization Comparative Fit Index [CFI] = .997, Tucker–Lewis Index [TLI] = .995, Root Mean Square Error of Approximation [RMSEA] = .03) and Peer Aggression Scales (CFI = .995, TLI = .986, RMSEA = .07; two-factor model also had strong model fit, CFI = .995, TLI = .986, RMSEA = .05; also see Waasdorp, Bradshaw, & Leaf, 2012). Responses on both scales were averaged with higher mean scores indicating more victimization and peer aggression, respectively. To create peer victimization subtypes, the victimization and aggression subscales were each dummy coded. Students who had mean scores of 1 to 2 (reflecting an average response of “never” to “rarely”) were coded 0 (“low involvement”), and students who had mean scores above 2 (i.e., “sometimes” to “almost always”) were coded 1 (“frequent involvement”). If a student fell in the “frequent involvement” group on both the Victimization and Aggression subscales, he or she was characterized as an “aggressor-victim.” Separate variables were created to indicate when a student was receiving special education and was a victim, aggressor, or aggressor-victim based on the above-described dummy coding (see Table 1 for peer victimization subtypes).

Table 1.

Student and School Demographics.

Student characteristics (N = 7,314)	n (%)
Gender
Female	3,437 (47)
Male	3,876 (53)
Race/ethnicity
American Indian/Alaskan Native	44 (>1)
Asian/Pacific Islander	293 (4)
African American	2,926 (40)
White	3,803 (52)
Hispanic	219 (3)
Grade cohort
Kindergarten (aged up to third grade)	2,487 (34)
First (aged up to fourth grade)	2,414 (33)
Second (aged up to fifth grade)	2,414 (33)
Received free or reduced priced meals	3,438 (47)
Role in peer victimization
Aggressor	293 (4)
Victim	1,682 (23)
Aggressor-victim	1,024 (14)
Special education and aggressor	73 (10)^a
Special education and victim	292 (40)^a
Special education and aggressor-victim	219 (30)^a
School characteristics (N = 37 schools)	M (SD)
Student mobility	23.6 (8.2)
Student–teacher ratio	11.3 (3.3)
Faculty turnover rate	16.0 (7.6)
FARMS	40.0 (19.9)
School climate	2.96 (0.27)

Note. Percentages do not add up to 100 due to missing data. FARMS = free or reduced meals status.

Percentage of the total number of those in special education (N = 713).

Student demographics

Student demographic variables (grade, gender, race/ethnicity, free or reduced meals status [FARMS]) were obtained from the school districts or the Maryland State Department of Education (MSDE). Because the sample largely consisted of White (52%) and Black (40%) students, race/ethnicity was dummy coded with White being the reference group (1 = White students, 0 = all other racial/ethnic groups). Student records data indicated special education status (n = 713; 10.26%), which was dummy coded (1 = ever in special education or 0 = never in special education) at each of the five time points (see Table 1 for sample demographics).

School characteristics

Data on student–teacher ratio, faculty turnover (percentage of faculty new to the school that year), student mobility (sum of the percentage of students moving in and percentage of students moving out of the school in that school year), and percentage of students receiving free or reduced-cost meals (i.e., FARMS, an indicator of concentrated poverty) were obtained from the MSDE for the baseline school year. School staff members’ perceptions of school climate were assessed through the Organizational Health Inventory–Elementary School Version (OHI; Hoy & Woolfolk, 1993), which consisted of 37 items assessing organizational health (e.g., principal leadership, staff affiliation). All items were measured on a 4-point scale (1 = rarely occurs to 4 = very frequently occurs) and were scored such that higher values indicated a more favorable climate. An overall OHI score was calculated by averaging all the items for each school. Table 1 provides means and standard deviations for school-level variables.

Procedure

Data were collected from the teachers via an individually addressed survey packet. Teachers completed a brief student demographic form and behavior-rating checklist for each student, and completed the self-report OHI measure on their perceptions of school climate. To ensure confidentiality, teachers completed the study materials on their own time and returned the materials directly to the researchers through the U.S. mail. The staff response rate for the self-report materials (i.e., OHI) was 78%, whereas the TOCA-C completion rate was 91%. Staff provided written active consent, whereas there was a waiver of active parental consent for the teachers’ completion of TOCA-C ratings of the students. The institutional review board at the investigators’ institution approved this study.

Overview of Analyses

The primary aim of the present study was to examine the relationship between general and special education students’ pattern of peer victimization (as either a victim, aggressor, or both) and their subsequent concentration problems and emotion regulation skills, while taking into account student demographic and school contextual variables. Longitudinal, three-level hierarchical linear modeling (HLM) analyses were conducted using the HLM 7.1 software (Raudenbush et al., 2011). This approach accounts for the shared variability between the outcomes collected at multiple time points, as well as shared variance between students within the same school. Specifically, the repeated measure outcomes were modeled at Level 1, where special education status was modeled as a covariate. Because the assignment of special education did not necessarily precede the outcome (e.g., a student could have been placed into special education in the third or fourth year of the study but was victimized in the first year), the receipt of special education services was modeled at Level 1 to account for its time-varying nature, whereby a dummy-coded special education status indicator (i.e., 1 = receives special education, 0 = does not receive special education services) was included for each of the five time points (Hox & Stoel, 2005).

The remaining student-level predictors were dichotomized and modeled at Level 2. Specifically, this included which “cohort” (i.e., kindergarten, Grade 1, or Grade 2) the student was in when the data collection began and student demographics (i.e., FARMS, gender, and White [1 = White and 0 = other race/ethnicity]). Three dummy-coded variables were included to assess victimization status: victim (1 = frequent victim, 0 = no/low victimization), aggressor (1 = frequent aggressor, 0 = no/low aggression), and aggressor-victim (1 = frequent victim and aggressor, 0 = no/low victimization and aggression). The interaction between the receipt of special education and victimization was examined via three dummy-coded variables: Victim × Special education (1 = victim and in special education, 0 = all others), Aggressor × Special education (1 = aggressor and in special education, 0 = all other students), and Aggressor-victim × Special education (1 = aggressor-victim and in special education, 0 = all others). These interactions were tested to isolate the specific effects of being a student receiving special education and the role in peer victimization. All student variables were modeled to predict the intercept and slope (i.e., change over time) for the two outcomes of interest (concentration problems and emotion regulation). All continuous predictor variables were grand mean centered (Enders & Tofighi, 2007).

At Level 3, school-level characteristics (i.e., mobility, student-to-teacher ratio, faculty turnover, FARMS rate, and baseline OHI scores) were modeled on the intercept and slope. All predictor variables at this level were grand mean centered (Enders & Tofighi, 2007). Prior to analyzing data in HLM, the covariates used were examined in SPSS to ensure that collinearity was not a concern (Tabachnick & Fiddell, 2001). Once in HLM, the variables were added one at a time to ensure that changes in the direction of variable effects did not occur, which is another means for detecting collinearity (Raudenbush & Bryk, 2002).

Although the participation rate was consistently high, there were some missing data. We examined the patterns of missing data and found no evidence that the level of missingness was problematic (Schlomer, Bauman, & Card, 2010). For example, baseline scores on concentration problems were unrelated to subsequent missingness on this measure (adjusted odds ratio [AOR] = 1.00, 95% confidence interval [CI] = [0.96, 1.04], p > .05). Similarly, gender was unrelated to subsequent missingness on teacher ratings of students’ behavior. Therefore, our analyses assumed data were missing at random (MAR), which means that the reason for missingness is not related to the missing value itself, or is deemed random after controlling for the variables that are observed (Arbuckle, 1996; Little, 1995). Specifically, HLM adjusts parameter estimates for attrition using full-information maximum-likelihood (FIML) estimation, a widely recognized and appropriate means of handling missing data under the assumption that data are MAR (Raudenbush et al., 2011; Rumberger & Palardy, 2004). Furthermore, individuals can have missing data across any of the time points and still be included in the analyses; therefore, HLM is robust to missing data within repeated measures.

Below, we present the results for two separate models, first for concentration problems and then emotion regulation. For both models, special education status, victimization subtype, student demographics, and the school-level contextual variables were added into the models one at a time to create the final models presented in Table 2.

Table 2.

Three-Level HLM Results for Peer Victimization, Special Education, and Student and School Characteristics on Concentration Problems and Emotion Regulation.

	Concentration problems			Emotion regulation
	Coefficient	SE	t ratio	Coefficient	SE	t ratio
Intercept
Intercept	2.623	0.019	136.71***	4.566	0.023	201.42***
Student level
Generation	0.022	0.020	1.06	0.005	0.020	0.27
FARMS	0.382	0.036	10.58***	−0.260	0.027	−9.67***
Gender	0.447	0.029	15.15***	−0.393	0.026	−15.07***
White	−0.082	0.029	−2.79**	0.159	0.029	5.51***
Victim	0.461	0.034	13.69***	−0.391	0.028	−14.10***
Aggressor	0.438	0.079	5.62***	−0.720	0.060	−11.95***
Aggressor-victim	0.872	0.040	21.57***	−1.206	0.042	−28.87***
Victim/SpEd	0.474	0.044	10.88***	−0.176	0.052	−3.37***
Aggressive/SpEd	0.190	0.213	0.89	0.109	0.173	0.63
Aggressor-victim/SpEd	0.299	0.080	3.73***	−0.136	0.071	−1.91
School level
Student mobility	0.005	0.003	1.89	−0.006	0.003	−2.21*
Student–teacher ratio	0.008	0.004	1.96	−0.012	0.004	−2.31*
Faculty turnover	0.001	0.003	0.41	−0.001	0.003	−0.20
FARMS rate	−0.004	0.001	−3.46**	0.003	0.001	2.07*
School climate	−0.271	0.090	−3.20**	0.473	0.089	5.32***
Slope (growth)
Time intercept	0.045	0.011	4.12***	−0.014	0.012	−1.10
Student level
Generation	−0.048	0.013	−3.61***	0.041	0.014	2.89**
FARMS	0.041	0.011	3.65***	−0.019	0.011	−1.70
Gender	0.050	0.010	4.77***	−0.048	0.001	−5.02***
White	−0.003	0.014	−0.21	0.022	0.010	2.24*
Victim	−0.116	0.019	−6.18***	0.160	0.021	7.60***
Aggressor	0.002	0.035	0.05	0.117	0.022	8.73***
Aggressor-victim	−0.113	0.019	−5.95***	0.196	0.022	8.73***
Victim/SpEd	0.021	0.027	0.77	−0.044	0.028	−1.53
Aggressor/SpEd	−0.155	0.067	−2.32*	0.021	0.036	0.58
Aggressor-victim/SpEd	−0.061	0.029	−2.06*	0.021	0.036	0.58
School level
Student mobility	0.001	0.001	0.44	−0.003	0.002	−1.20
Student–teacher ratio	0.002	0.002	0.91	−0.002	0.003	−0.79
Faculty turnover	0.003	0.002	1.90	−0.002	0.002	−1.21
FARMS rate	0.002	0.001	2.25*	−0.001	0.001	−0.84
School climate	0.057	0.059	0.96	−0.052	0.054	−0.96
SpeEd intercept	0.227	0.030	7.50***	−0.079	0.027	−2.93**

Note. N = 7,314. Demographics include FARMS (1 = received FARMS, 0 = not received), gender (1 = male, 0 = female), and White (1 = White and 0 = other race/ethnicity). SpEd indicates a student was receiving special education services. Victim, aggressor, and aggressor-victim were coded 1 (student engaged frequently in behavior) and 0 (low/no involvement). The main effect for special education status was entered as time-varying covariate; the effect is listed as “SpeEd Intercept.” HLM = hierarchical linear modeling; SE = standard error; FARMS = free or reduced meals status.

p < .05. **p < .01. ***p < .001.

Results

Concentration Problems

Special education status

As noted above, special education status was modeled as a time-varying covariate because students were enrolled in special education at varying time points of the study (e.g., Year 1 vs. Year 3). Students in special education were rated as having greater concentration problems (standardized coefficient β = .227, p < .01) than students never enrolled in special education during the study time frame (see Table 2). At the student level, we examined the effect of being in special education and frequently involved in peer victimization. Students in special education who were frequently victimized (β = .474, p < .001) or characterized as an aggressor-victim (β = .299, p < .001) had higher concentration problems at baseline as compared with all other students. Peer aggressors who were also in special education did not significantly differ from their peers at baseline. An examination of the slope coefficients showed that victims in special education had score trajectories that were generally consistent with the general student population. Surprisingly, aggressor-victims in special education showed a significant decrease in their concentration problems scores, indicating minor improvements in attention skills (β = −.061, p < .05). Similarly, aggressors in special education also decreased in concentration problems (β = −.155, p < .05) to the point that their concentration problem scores were similar to all other students at the end of the 4 years.

Peer victimization subtypes

Compared with students (general and special education) not involved in peer victimization, victims (β = .461, p < .001), aggressors (β = .438, p < .001), and aggressor-victims (β = .872, p < .001) had significantly higher concentration problems scores at baseline (see Table 2) than non-victims, non-aggressors, and non-aggressor-victims. Examination of the slope coefficients suggests that the rate of change in concentration problems varied by involvement in peer victimization. As compared with youth not frequently involved in peer victimization, victims (β = −.116, p < .001) and aggressor-victims (β = −.113, p < .001) had a lower rate of change in concentration problems over time; however, their concentration problem scores still remained higher than their non-involved peers at the end of the 4 years (i.e., there was a narrowing of the gap between these groups of students). However, the slope term was not significant for aggressors.

Student demographics

Across the 4 years, all students in the study were rated as displaying increasing levels of concentration problems (i.e., β for time = .045, p < .001). However, at baseline, males were rated significantly higher (i.e., worse) on concentration problems as compared with females (β = .447, p < .001) and continued to receive worse scores (i.e., scores of concentration problems, which increased at a faster rate) than their female counterparts over time (β = .050, p < .001). White youth received lower scores on concentration problems at baseline (β = −.082, p < .01), yet the slopes for White and minority youth were not significantly different from each other. Last, students receiving FARMS were rated as displaying more concentration problems at baseline (β = .382, p < .001) and had greater increases in their concentration problems over the 4 years (β = .041, p < .001) as compared with students not receiving FARMS.

School-level factors

Positive school climate was associated with a lower student concentration problem intercept (β = −.271, p < .01) but not with the changes in student scores over time. Interestingly, students in schools with a higher FARMS rate had lower intercept scores on concentration problems (β = −.004, p < .01); however, over time, the ratings of student concentration problems at these schools tended to increase more rapidly compared with schools with fewer FARMS students (β = .002, p < .05). Student mobility, student–teacher ratio, and faculty turnover were not significantly related to the intercept of concentration problems or their growth over time (p > .05).

Emotion Regulation

Special education status

Inspection of the special education status time-varying covariate parameters indicated that students in special education were rated as having worse emotion regulation (β = −.08, p < .001) than students never enrolled in special education during the study time frame (Table 2). The inclusion of the Special education × Peer victimization coefficients (i.e., interaction terms) showed that youth who were victims and in special education had significantly lower emotion regulation scores than all other youth at baseline (β = −.176, p < .001). Youth in special education and who engaged in peer aggression did not significantly differ from other youth on their baseline emotion regulation scores. An examination of the slope coefficients revealed that changes in students’ emotion regulation scores did not differ by special education and victimization subtype (Table 2).

Peer victimization subtypes

Similar to the concentration problems model, victims (β = −.391, p < .001), aggressors (β = −.720, p < .001), and aggressor-victims (β = −1.206, p < .001) had lower baseline emotion regulation scores. The average student growth trajectory in emotion regulation was relatively flat (i.e., β for time = −.014, p > .10); however, those students frequently involved in peer victimization showed increases in their ratings of emotion regulation. Despite these gains for aggressors, victims, and aggressor-victims, their average mean ratings on emotion regulation were still lower than their peers at the end of the study.

Student demographics

At baseline, minority youth, males, and youth receiving FARMS received lower emotion regulation scores (p < .001; see Table 2). This pattern persisted over time for males (β = −.048, p < .001) who continued to decline in emotion regulation scores over the 4 years, whereas females’ scores remained relatively stable. Similarly, White students were rated as having greater increases in their emotion regulation over time as compared with minority students (β = .022, p < .05). Both students who received and did not receive FARMS had a similar rate of change in emotion regulation over time (i.e., non-significant slope finding).

School-level factors

Students in schools with higher student mobility (β = −.006, p < .05) and higher student–teacher ratios (β = −.012, p < .05) had significantly lower intercept scores on emotion regulation (p < .05). Students in schools with more positive school climates (β = .473, p < .001) and more students qualifying for FARMS (β = .003, p < .05) tended to have higher intercept scores on emotion regulation. Faculty turnover was not significantly related to any intercept scores. None of the school-level slope coefficients significantly predicted variation in emotion regulation scores.

Variance Accounted for by Models

To determine the amount of variability in concentration problems and emotion regulation scores over time and between schools, intraclass correlations (ICCs) were calculated using the fully unconditional model, which only includes time as a predictor variable at Level 1 (i.e., to ensure that the model is recognized as repeated measures) and no predictors at Level 2 or Level 3 (Raudenbush & Bryk, 2002). The final model explained 9.22% of between-student variance and 66.49% of between-school variance in concentration problem scores. For emotion regulation, the final model explained 19.91% of between-student variance and 69.87% of between-school variance. The Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) indices demonstrated better fit for the final conditional model than the unconditional model.

Discussion

Although cross-sectional research has well documented the various forms of peer victimization and its associations with other social, emotional, and behavioral problems (Card, Stucky, Sawalani, & Little, 2008; Espelage & Swearer, 2004; Hawker & Boulton, 2000), there remains a need for longitudinal research on at-risk populations to determine the short- and long-term effects of victimization. The aim of the present study was threefold; to: (a) examine the association between peer victimization/aggression and concentration problems and emotion regulation among elementary students in general versus special education, (b) determine whether changes in children’s concentration problems and emotion regulation skills vary over time by their role in peer victimization and special education status, and (c) identify student demographics and school contextual factors that help explain these relationships.

Peer Victimization Among Students in General and Special Education

The findings from the present study indicated that children frequently involved in peer victimization were rated by their teachers as having significantly worse concentration problems than their peers, no matter their special education status. Children’s role in peer victimization also explained variations in their ability to concentrate over time; however, the slope coefficients were only significant for victims and aggressor-victims. In other words, although all youth involved in peer victimization had significantly worse baseline scores on concentration problems, it was only victims’ and aggressor-victims’ concentration problems that continued to worsen over time. This may be related to potential anxiety and hyper-awareness that victims of peer aggression may experience, which further inhibits their abilities to attend to assignments in the classroom (Juvonen, Wang, & Espinoza, 2011). Children’s emotion regulation skills were also predicted by their exposure to peer victimization. At baseline, all three groups had lower emotion regulation scores (i.e., worse) as compared with those who were not frequently involved. This highlights the need for additional targeted interventions for these youth, to help them learn skills to appropriately deal with social stressors, such as being excluded from peer activities. Having children identify their emotions and then learn prosocial ways to cope with their feelings of anger or sadness may help reduce future victimization episodes and consequently increase their academic success (Takahashi, Koseki, & Shimada, 2009); in fact, this is a common target for social-emotional learning curricula, which have also demonstrated positive impacts on peer aggression (Bradshaw, 2013).

The inclusion of the special education time-varying covariate suggested that, on average, students in special education had worse trajectories, including higher concentration problem scores and lower emotion regulation skills over time than students in general education. For concentration problems, children who were victimized and in special education displayed significantly more concentration problems at baseline as compared with other students. This suggests that holding constant the effects of being victimized, students in special education have particular difficulties staying on task, completing assignments, and learning up to their abilities. This finding is consistent with research by Taylor, Saylor, Twyman, and Macias (2010), which revealed that students with ADHD and who are victimized have significantly more externalizing (aggression, social difficulties) and internalizing (withdrawn, thought disturbances) symptoms.

With regard to emotion regulation, victimized children in special education had significantly lower emotion regulation scores at baseline. Likewise, results showed that aggressor-victims and victimized children remained worse over time as compared with uninvolved students, despite slight increases in social emotional ratings for aggressor-victims in special education. Studies have found that youth with emotional difficulties/disturbances are at risk of repeated peer victimization more so than any other disability group (e.g., learning disability, autism, hearing impaired) and as compared with youth in general education (Blake et al., 2012). Thus, it would be beneficial for school-based programs to target emotion regulation skills as a preventive measure for youth who are victimized, particularly those who are in special education.

Interestingly, youth who are in special education and are aggressive toward their peers differed from the other subtypes with regard to their baseline scores and behavioral trajectories. Despite having higher concentration problem scores at baseline, aggressors in special education ended up resembling their uninvolved peers at the end of the study. Even more surprising, aggressive youth were rated as having higher (more positive) emotion regulation scores than their peers at baseline and throughout elementary school. These findings may suggest that aggressors in special education do not differ from aggressive children in general education. Although research has shown that youth with behavioral disorders, such as concentration difficulties and poor emotion regulation skills, are more likely to engage in aggressive acts (Swearer et al., 2012), special education status does not appear to increase this risk.

Student Demographics and School Contextual Effects

Individual- and school-level characteristics are known to be influential in both peer victimization and special education research (Bradshaw et al., 2009; Rose, Monda-Amaya, & Espelage, 2011). Being male, an ethnic/racial minority, and coming from a family of lower socioeconomic status (SES) were each associated with higher concentration problems and lower emotion regulation scores. Although these may not be malleable factors researchers can target for prevention, they may signal increased risk of social-emotional problems.

One of the unique aspects of this study was the inclusion of several school-level variables. Of the variables included, the school-level FARMS rate (e.g., percentage of the student body which was of low SES) and teachers’ perception of the school climate both were significantly related to student concentration problems and emotion regulation scores. Surprisingly, schools with higher FARMS rates tended to have less concentration problems and better emotion regulation skills at baseline; yet, over time, these scores tended to worsen. Although it is unclear what may be underlying this finding, perhaps teacher ratings of student behaviors tend to become less positive in these more impoverished schools. Similarly, teacher perceptions of the school’s climate (e.g., feelings of collegiality, leadership support) were related to more positive student outcomes. School climate research has consistently shown that schools with supportive and caring environments tend to cultivate positive student and staff behavior (O’Brennan, Bradshaw, & Furlong, 2014) and increase school staff members’ likelihood for intervening in bullying situations (O’Brennan, Waasdorp, & Bradshaw, 2014). Taken together, our findings are consistent with the social-ecological model, suggesting that future studies examining bullying among special education populations should include student- and school-level variables to better understand these contextual influences.

Limitations

We must consider some limitations when interpreting these findings. For example, the measures used to assess peer victimization and peer aggression were relatively brief due to the large scale of the study; future studies should utilize more comprehensive measures of bullying and victimization. Similarly, there were no indicators of the specific disability or level of service in the present study. Research has shown that there are differences in involvement in peer victimization depending on whether the child’s disability was observable (e.g., hearing impairments, mild mentally handicapped) or non-observable (e.g., learning disabilities; Swearer et al., 2012). Additional research should explore variation within special education populations (see Rose, Monda-Amaya, & Espelage, 2011). Moreover, although the aim of the present study was to compare general education students with special education students and peer aggression, there may be some confounding factors due to educational status when examining educational trajectories; therefore, future studies could compare special education students involved in peer victimization and aggression with youth in special education who are not involved in peer victimization. It would be important to have a larger sample of special education youth who are also involved in peer victimization to replicate the present findings. Finally, teacher reports of student behavior were the primary mode of measurement in this study. Although teacher surveys have been shown to be cost-effective and efficient for schools and educational researchers (Racz, King, Wu, Witkiewitz, & McMahon, 2013), and more accurate in predicting children’s later behavior problems than parent reports (Dwyer, Nicholson, & Battistutta, 2006), it would be helpful to use multiple informants (e.g., self-reports, peer reports, outside observations of bullying behaviors) to capture a more comprehensive assessment of peer victimization.

Conclusion and Future Directions

Research on school readiness suggests that a child’s level of cognitive, behavioral, and social-emotional functioning on entry into kindergarten is predictive of his or her later academic achievement (Duncan et al., 2007). Accordingly, it is important for educators to prevent children’s involvement in bullying and to help mitigate the effects of bullying when it does occur. Our findings indicate that children’s involvement in bullying, as either a victim, aggressor, or aggressor-victim, increases their likelihood for concentration problems and poor emotion regulation, both concurrently and prospectively. Frequent peer victimization in combination with academic difficulties may lead to increased distractibility and difficulties maintaining emotional composure during challenging situations. To offset the long-term adjustment problems associated with victimization, educators are encouraged to address bullying directly and subsequently get victimized students involved in bullying prevention programs available in the school (e.g., group counseling, individual skill building). In turn, general and special educators should teach and model prosocial replacement behaviors and encourage the adoption of school policies regarding the reporting of bullying situations, particularly in situations involving youth with disabilities.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Support for this project is funded by the National Institute of Mental Health (R01 MH67948-1A1, Principal Investigator [PI]: P. J. Leaf) and the National Institute of Mental Health Children’s Mental Health Services Training Program (T32 MH019545-21; PI: P. J. Leaf).

References

Arbuckle

J. L.

(1996). Full information estimation in the presence of incomplete data. In Marcoulides

G. A.

Schumacker

R. E.

(Eds.), Advanced structural equation modeling: Issues and techniques (pp. 243–277). Mahwah, NJ: Lawrence Erlbaum.

Barkley

R. A.

(2003). Issues in the diagnosis of attention-deficit/hyperactivity disorder in children. Brain & Development, 25, 77–83. doi:10.1016/S0387-7604(02)00152-3

Blake

J. J.

Lund

E. M.

Zhou

Kwok

O.-M.

Benz

M. R.

(2012). National prevalence rates of bully victimization among students with disabilities in the United States. School Psychology Quarterly, 27, 210–222. doi:10.1037/spq0000008

Bogart

L. M.

Elliott

M. N.

Klein

D. J.

Tortolero

S. R.

Mrug

Peskin

M. F.

. . .Schuster

M. A.

(2014). Peer victimization in fifth grade and health in tenth grade. Pediatrics, 133, 440–447. doi:10.1542/peds.2013-3510

Bradshaw

C. P.

(2013). Preventing bullying through Positive Behavioral Interventions and Supports (PBIS): A multi-tiered approach to prevention and integration. Theory Into Practice, 52, 288–295. doi:10.1080/00405841.2013.829732

Bradshaw

C. P.

Sawyer

A. L.

O’Brennan

L. M.

(2009). A social disorganization perspective on bullying-related attitudes and behaviors: The influence of school context. American Journal of Community Psychology, 43, 204–220. doi:10.1007/s10464-009-9240-1

Bradshaw

C. P.

Waasdorp

T. E.

Leaf

P. J.

(in press). Examining variation in the impact of School-Wide Positive Behavioral Interventions and Supports: Findings from a randomized controlled effectiveness trial. Journal of Educational Psychology. doi:10.1037/a0037630

Bradshaw

C. P.

Waasdorp

T. E.

O’Brennan

Gulemetova

(2013). Teachers’ and education support professionals’ perspectives on bullying and prevention: Findings from a National Education Association (NEA) Survey. School Psychology Review, 42, 280–297.

Card

N. A.

Stucky

B. D.

Sawalani

G. M.

Little

T. D.

(2008). Direct and indirect aggression during childhood and adolescence: A meta-analytic review of gender differences, intercorrelations, and relations to maladjustment. Child Development, 79, 1185–1229. doi:10.1111/j.1467-8624.2008.01184.x

10.

Duncan

G. J.

Dowsett

C. J.

Claessens

Magnuson

Huston

A. C.

Klebanov

. . .Japel

(2007). School readiness and later achievement. Developmental Psychology, 43, 1428–1446. doi:10.1037/0012-1649.43.6.1428

11.

Dwyer

S. B.

Nicholson

J. M.

Battistutta

(2006). Parent and teacher identification of children at risk of developing internalizing or externalizing mental health problems: A comparison of screening methods. Prevention Science, 7, 343–357. doi:10.1007/s11121-006-0026-5

12.

Enders

C. K.

Tofighi

(2007). Centering predictor variables in cross-sectional multilevel models: A new look at an old issue. Psychological Methods, 12, 121–138.

13.

Espelage

D. L.

Swearer

S. M.

(2004). Bullying in American schools: A social–ecological perspective on prevention and intervention. Mahwah, NJ: Lawrence Erlbaum.

14.

Fine

J. G.

Semrud-Clikeman

Butcher

Walkowiak

(2008). Brief report: Attention effect on a measure of social perception. Journal of Autism and Developmental Disorders, 38, 1797–1802.

15.

Graziano

P. A.

Reavis

R. D.

Keane

S. P.

Calkins

S. D.

(2007). The role of emotion regulation in children’s early academic success. Journal of School Psychology, 45, 3–19. doi:10.1016/j.jsp.2006.09.002

16.

Hawker

D. S. J.

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, 441–455. doi:10.1111/1469-7610.00629

17.

Hox

Stoel

R. D.

(2005). Multilevel and SEM approaches to growth curve modeling. In Everitt

B. S.

Howell

D. C.

(Eds.), Encyclopedia of statistics in behavioral science (pp. 1296–1305). Chichester, UK: John Wiley.

18.

Hoy

W. K.

Woolfolk

A. E.

(1993). Teachers’ sense of efficacy and the organizational health of schools. Elementary School Journal, 93, 355–372. doi:10.1086/461729

19.

Juvonen

Wang

Espinoza

(2011). Bullying experiences and compromised academic performance across middle school grades. The Journal of Early Adolescence, 31, 152–173. doi:10.1177/0272431610379415

20.

Kokkinos

C. M.

Antoniadou

(2013). Bullying and victimization experiences in elementary school students nominated by their teachers for specific learning disabilities. School Psychology International, 34, 674–690. doi:10.1177/0143034313479712

21.

Koth

C. W.

Bradshaw

C. P.

Leaf

P. J.

(2009). Teacher Observation of Classroom Adaptation—Checklist: Development and factor structure. Measurement and Evaluation in Counseling and Development, 42, 15–30. doi:10.1177/0748175609333560

22.

Little

R. J.

(1995). Modeling the dropout mechanism in repeated-measures studies. Journal of the American Statistical Association, 90, 1112–1121.

23.

O’Brennan

L. M.

Bradshaw

C. P.

Furlong

M. J.

(2014). Influence of classroom and school climate on teacher perceptions of student problem behavior. School Mental Health, 6, 125–136. doi:10.1007/s12310-014-9118-8

24.

O’Brennan

L. M.

Bradshaw

C. P.

Sawyer

(2009). Social–emotional problems among frequent bullies, victims, and bully/victims: Implications for prevention and intervention through schools. Psychology in the Schools, 46, 100–115. doi:10.1002/pits.20357

25.

O’Brennan

L. M.

Waasdorp

T. E.

Bradshaw

C. P.

(2014). Strengthening bullying prevention through school staff connectedness. Journal of Educational Psychology, 106, 870–880.

26.

Racz

S. J.

King

K. M.

Witkiewitz

McMahon

R. J.

(2013). The predictive utility of a brief kindergarten screening measure of child behavior problems. Journal of Consulting and Clinical Psychology, 81, 588–599. doi:10.1037/a0032366

27.

Raudenbush

S. W.

Bryk

A. S.

(2002). Hierarchical linear models: Applications and data analysis methods (2nd ed.). Thousand Oaks, CA: SAGE.

28.

Raudenbush

S. W.

Bryk

A. S.

Cheong

Y. K.

Fai

Y. F.

Congdon

R. T.

Jr. duToit

(2011). HLM 7: Hierarchical linear and nonlinear modeling. Lincolnwood, IL: Scientific Software International.

29.

Rose

C. A.

Espelage

D. L.

Aragon

S. R.

Elliott

(2011). Bullying and victimization among students in special education and general education curricula. Exceptionality Education International, 21, 2–14.

30.

Rose

C. A.

Forber-Pratt

A. J.

Espelage

D. L.

Aragon

S. R.

(2013). The influence of psychosocial factors on bullying involvement of students with disabilities. Theory Into Practice, 52, 272–279. doi:10.1080/00405841.2013.829730

31.

Rose

C. A.

Monda-Amaya

L. E.

Espelage

D. L.

(2011). Bullying perpetration and victimization in special education: A review of the literature. Remedial and Special Education, 32, 114–130. doi:10.1177/0741932510361247

32.

Rumberger

R. W.

Palardy

G. J.

(2004). Multilevel models for school effectiveness research. In Kaplan

(Ed.), Handbook of quantitative methodology for the social sciences (pp. 235–258). Thousand Oaks, CA: SAGE.

33.

Schlomer

G. L.

Bauman

Card

N. A.

(2010). Best practices for missing data management in counseling psychology. Journal of Counseling Psychology, 57, 1–10. doi:10.1037/a0018082

34.

Son

Parish

S. L.

Peterson

N. A.

(2012). National prevalence of peer victimization among young children with disabilities in the United States. Children and Youth Services Review, 34, 1540–1545. doi:10.1016/j.childyouth.2012.04.014

35.

Spinrad

T. L.

Eisenberg

(2009). Emotion regulation: Implications for positive youth development. In Gilman

Huebner

E. S.

Furlong

M. J.

(Eds.), Handbook of positive psychology in schools (pp. 119–130). New York, NY: Routledge.

36.

Stormshak

E. A.

Bierman

K. L.

Bruschi

Dodge

K. A.

Coie

J. D.

(1999). The relation between behavior problems and peer preference in different classroom contexts. Child Development, 70, 169–182.

37.

Swearer

S. M.

Wang

Maag

J. W.

Siebecker

A. B.

Frerichs

L. J.

(2012). Understanding the bullying dynamic among students in special and general education. Journal of School Psychology, 50, 503–520. doi:10.1016/j.jsp.2012.04.001

38.

Tabachnick

B. G.

Fiddell

L. S.

(2001). Using multivariate statistics (4th ed.). Needham Heights, MA: Allyn & Bacon.

39.

Takahashi

Koseki

Shimada

(2009). Developmental trends in children’s aggression and social problem-solving. Journal of Applied Developmental Psychology, 30, 265–272. doi:10.1016/j.appdev.2008.12.007

40.

Taylor

L. A.

Saylor

Twyman

Macias

(2010). Adding insult to injury: Bullying experiences of youth with attention deficit hyperactivity disorder. Children’s Health Care, 39, 59–72. doi:10.1080/02739610903455152

41.

Thomas

D. E.

Bierman

K. L.

Thompson

Powers

C. J.

Conduct Problems Prevention Research Group. (2008). Double jeopardy: Child and school characteristics that predict aggressive–disruptive behavior in first grade. School Psychology Review, 37, 516–532.

42.

Ttofi

M. M.

Farrington

D. P.

Lösel

(2012). School bullying as a predictor of violence later in life: A systematic review and meta-analysis of prospective longitudinal studies. Aggression and Violent Behavior, 17, 405–418. doi:10.1016/j.avb.2012.05.002

43.

Van Cleave

Davis

M. M.

(2006). Bullying and peer victimization among children with special health care needs. Pediatrics, 118, e1212–e1219. doi:10.1542/peds.2005-3034

44.

Waasdorp

T. E.

Bradshaw

C. P.

Leaf

P. J.

(2012). The impact of School-Wide Positive Behavioral Interventions and Supports (SWPBIS) on bullying and peer rejection: A randomized controlled effectiveness trial. Archives of Pediatrics & Adolescent Medicine, 116, 149–156. doi:10.1001/archpediatrics.2011.755

45.

Waasdorp

T. E.

Pas

E. T.

O’Brennan

L. M.

Bradshaw

C. P.

(2011). A multilevel perspective on the climate of bullying: Discrepancies among students, school staff, and parents. Journal of School Violence, 10, 115–132. doi:10.1080/15388220.2010.539164

46.

Werthamer-Larsson

Kellam

Wheeler

(1991). Effect of first grade classroom environment on shy behavior, aggressive behavior, and concentration problems. American Journal of Community Psychology, 19, 585–602.