Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Feb 16.
Published in final edited form as: J Prev Health Promot. 2022 May 5;3(2):147–165. doi: 10.1177/26320770221094032

The Longitudinal Relationship Between Peer Relations and Empathy and their Joint Contribution to Reducing Bullying in Middle School: Findings from a Randomized Trial of Cooperative Learning

Mark J Van Ryzin 1,*, Cary J Roseth 2
PMCID: PMC9934011  NIHMSID: NIHMS1864733  PMID: 36818659

Abstract

The positive peer relations arising from cooperative learning can contribute to the development of affective empathy, which in turn can reduce bullying (Van Ryzin & Roseth, 2019). However, from a theoretical perspective, the direction of effects between peer relations and empathy could be in the opposite direction, or bi-directional. In the current paper, we employed a process-oriented approach (i.e., cross-lag difference score modeling; McArdle, 2009) to investigate the longitudinal relationship between positive peer relations and affective empathy, as well as their joint effect on bullying. Using four waves of data from a cluster randomized trial including 15 middle schools (7 intervention and 8 control schools; N=1,890 students, 47.1% female, 75.2% White), we found a bi-directional or reciprocal relationship between peer relations and affective empathy, and change in both constructs predicted lower levels of bullying. Cooperative learning predicted positive change in peer relations and affective empathy, as well as lower levels of bullying. These results suggest that the structured social interactions that occur during cooperative learning can enhance student interpersonal relations, and simultaneously the experiential skill building of cooperative learning can contribute to a more profound understanding of the emotional states of others. These effects amplify one another and, in turn, significantly reduce bullying in middle school. Given that cooperative learning has already been demonstrated to enhance academic motivation and achievement (Roseth et al., 2008), we argue that cooperative learning offers an effective, attractive alternative to traditional curriculum-based bullying prevention programs (National Library of Medicine [NLM], NCT03119415).

Keywords: bullying, empathy, peer relations, cooperative learning, middle school

Despite prevention efforts, aggressive and deliberate victimization of others (i.e., bullying) remains a significant problem in U.S. schools. At least one-fifth, and perhaps up to one-third, of all students are bullied by peers at some point during their school years (Craig et al., 2009; Musu et al., 2019; World Health Organization, 2012). This behavior increases significantly following the transition to middle school (Pellegrini & Long, 2002; Robers, Kemp, & Truman, 2013), and early adolescents are particularly vulnerable to these types of social stressors due to a developmental lag in self-regulatory capability (Casey et al., 2011). Thus, the highest levels of bullying in schools generally occur when the victims are least prepared to cope with it. In addition, experiencing victimization by peers has been linked to a variety of negative behavioral and emotional outcomes for students at this age, including increased levels of anxiety, depression, drug use, and delinquency, and lower levels of self-esteem, school attendance, and academic achievement (Barchia & Bussey, 2010; Juvonen et al., 2000; Nishina et al., 2005; Rueger et al., 2011; Storch et al., 2005; Sullivan et al., 2006; Thijs & Verkuyten, 2008).

Although bullying prevention programs have had a degree of success, generating small to moderate effects (Gaffney et al., 2019; Kennedy, 2020), there is also significant heterogeneity in effects, potentially reflecting a degree of uncertainty in the theoretical foundation for these programs. As we noted previously (Van Ryzin & Roseth, 2019), one area of significant uncertainty is the role of empathy in bullying prevention. Although empathy training is a core component of many of these programs (Espelage & Swearer, 2003), there has been confusion around the precise roles of cognitive and affective empathy, as well as a lack of evidence that existing programs can actually have significant effects on empathy.

Previously, we proposed an innovative approach to building empathy as a means to prevent bullying using a particular form of small-group instruction called cooperative learning. We found that the positive peer relations that arise from cooperative learning (Roseth et al., 2008) contributed to the development of both cognitive and affective empathy, but only affective empathy was linked to reductions in bullying (Van Ryzin & Roseth, 2019). These results clarified the role of empathy in the prevention of bullying and highlighted cooperative learning as a means of prevention that did not require the implementation of a stand-alone curriculum.

Cooperative Learning

In contrast to traditional instruction where the teacher is the focal point and students work individually, cooperative learning places students into small learning groups. These groups are not intended for brief, informal conversations; rather, cooperative learning specifies extended collaboration among students, and thus requires the careful and purposeful implementation of several key design features to ensure that the collaboration is successful. First, a cooperative learning lesson must create conditions of positive interdependence, in which individual goal attainment also promotes the goal attainment of others (in contrast to more common educational situations, in which individual goal attention either has no impact, or has a negative impact, on the goal attainment of others). In a cooperative learning lesson, many different types of positive interdependence may exist (Johnson et al., 2013). For example, teachers may require a single finished product from a group (goal interdependence), or may offer an incentive to the group if everyone achieves above a certain threshold on assessment (reward interdependence). The lesson may specify that each member of the group be issued different materials that they must share with others in their group to complete the lesson (resource interdependence), or that each member of the group do something specific for a lesson to be completed successfully, such as fulfill a unique role (role interdependence, e.g., tracking the group status, or taking notes on group discussions) or complete a unique task (task interdependence, e.g., each student has a different component of a project or presentation).

In less effective implementations of cooperative learning, goal interdependence is often used exclusively, but multiple forms of positive interdependence can (and should) be added to a single lesson, increasing the incentive for students to collaborate. For example, in addition to goal interdependence, teachers can also implement either task or role interdependence during the lesson, and combine this with reward interdependence linked to later assessments.

Second, in addition to positive interdependence, cooperative learning activities must also provide individual accountability to ensure that students have a strong incentive to contribute to the success of the group (Johnson et al., 2013). Individual accountability can include an end-of-unit assessment to be taken individually (with the potential for group rewards as discussed above), or something as simple as a random oral quiz by the teacher as he or she supervises the group work during class time. When students know that they are going to be held accountable, they are more likely to engage and fulfil their role in their learning group (Johnson & Johnson, 1989). Peer expectations can also contribute to accountability, particularly if reward interdependence has been implemented, because a student can benefit from the success of the others in the group.

Third, high-quality cooperative learning lessons must also include the explicit coaching of students in collaborative social skills (e.g., encouraging participation, checking for understanding, sharing ideas, asking for clarification), which includes explicit scaffolding of a specific skill, setting expectations and goals for group behavior, and monitoring by the teacher to identify and reward examples of such behavior. Finally, cooperative learning requires guided processing of group performance after the lesson is completed. This involves groups discussing what they did well, setting targets for improvement in the future, and providing positive reinforcement to one another for behavior during the lesson that contributed to group success (Johnson et al., 2013).

When all of these features are present in a well-designed cooperative learning lesson, students are incentivized to promote the success of others through instrumental and emotional support. This promotive social interaction, and the successful attainment of group goals, together create a positive shared emotional experience among group members (Deutsch, 2011). In turn, these positive experiences stimulate more positive peer relations, reduce outgroup prejudices and biases, and reduce bullying, with moderate to large effect sizes (Choi et al., 2011; Johnson & Johnson, 1989; Johnson et al., 1983; Roseth et al., 2008; Van Ryzin & Roseth, 2018a).

The Present Study

In previous work, we established that peer relations contributed to the development of empathy, which in turn reduced bullying (Van Ryzin & Roseth, 2019). At that time, we did not test a competing hypothesis, whereby the development of affective empathy can contribute to more positive peer relations, even though there is empirical support for such a hypothesis (De Weid et al., 2007; Gleason et al., 2009; Portt et al., 2020). Thus, in this study, we evaluated a more complex relationship between affective empathy and peer relations and how they jointly contribute to reductions in bullying.

We employed a process-oriented approach using cross-lagged latent difference score modeling, which decomposes the overall growth trajectory into a series of segments representing the amount of change from one measurement wave to the next. Difference scores are then used in an autoregressive cross-lag framework to assess the degree in which one variable influences change in the other over time (McArdle, 2009). These difference scores are latent constructs that represent the amount of change between adjacent waves, which enables us to obtain accurate assessments of the influence of one variable (X) on the net change in another variable (Y) while controlling for the influence of baseline levels of Y. Typical cross-lag models cannot provide this capability (McArdle, 2009). We provide an example model in Figure 1 where we estimate change between waves (e.g., wave 1 to wave 2) in affective empathy and peer relatedness and evaluate how each influences subsequent change in the other. We also evaluate the impact of the two constructs on change in bullying. We controlled for intervention condition in the model. Given that some research has found sex differences in the link between empathy and bullying (Caravita et al., 2009), we evaluated sex as a moderator of model paths.

Figure 1.

Figure 1.

Open in a new tab

Cross-lag difference score model.

Method

We received approval for all aspects of this study from the Institutional Review Board (IRB) at the Oregon Research Institute. The study was registered as trial NCT03119415 in ClinicalTrials.gov.

Sample

The sample was derived from a small-scale randomized trial of cooperative learning in 15 rural middle schools in the Pacific Northwest. Schools were matched based upon size and demographics (e.g., free/reduced-price lunch percentage) and randomized to condition (i.e., intervention vs. waitlist control) with a random number generator. We were concerned about the likelihood of losing schools assigned as controls, so the authors randomized an extra school to this condition (i.e., 8 waitlist-control vs. 7 intervention schools).

Our analytic sample included N = 1,890 students who enrolled in the project during the 2016–2017 or 2017–2018 school years; all students in participating schools who were in 7th grade during the first year of the project and in 8th grade during the second year were eligible to participate, and any student who participated in at least one wave of data collection was included in the sample. We achieved greater than 80% student participation at each data collection point by using a passive consent procedure and providing research staff to oversee the data collection. We also offered compensation to the schools for participating in the project, and enrolled participating students in a prize raffle. We provide student demographics by school in Table 1, and enrollment data by wave and condition in Table 2. Overall, the sample was 47.1% female (N = 890) and 75.2% White (N = 1,421). Other racial/ethnic groups included Hispanic/Latino (13.2%, N = 249), multi-racial (5.3%, N = 100), and American Indian/Alaska Native (3.1%, N = 58); our sample included less than 1% Asian, African-American, and Native Hawaiian/Pacific Islander. Overall, 13.9% (N = 262) were reported as having Special Ed status, 78.6% (N = 1486) did not have Special Ed status, and 7.5% (N = 142) were missing this designation. Free/reduced-price lunch status was not made available by the schools, although we added school-level figures (obtained from state records) to Table 1.

Table 1.

Intervention condition, sample size (number of students), sex, race/ethnicity, special education, and free/reduced-price lunch data by school

School Intervention N % female % White % Special Ed % FRPLa
1 Yes 282 47.9 73.0 11.7 53
2 Yes 121 47.1 90.1 19.8 71
3 Yes 112 50.0 83.0 15.2 72
4 Yes 110 40.0 60.9 n/a 62
5 Yes 105 46.7 78.1 10.5 57
6 Yes 84 33.3 72.6 4.8 95
7 Yes 61 52.5 75.4 16.4 66
8 No 239 51.0 48.5 13.0 84
9 No 197 49.2 90.4 11.7 66
10 No 183 44.8 65.0 17.5 61
11 No 114 47.4 93.0 24.6 65
12 No 108 51.9 80.6 15.7 46
13 No 71 45.1 81.7 19.7 45
14 No 53 41.5 92.5 18.9 33
15 No 50 48.0 88.0 16.0 39
Open in a new tab
a

State records.

Note. One school did not provide Special Ed status. FRPL = free/reduced-price lunch.

Table 2.

Enrollment data (number of students) by wave and intervention condition

Wave Enrolled Lost to Follow-Upa
Intervention Control Intervention Control
1 668 792 24 48
2 104 106 22 30
3 97 112 14 18
4 6 5 - -
Total 875 1015 60 96
Open in a new tab
a

Students do not appear in any subsequent waves.

Note. Data analysis (Maximum Likelihood) included all students. Data collection was conducted in September/October and March/April of the 2016–2017 and 2017–2018 school years (4 waves in total, about 6 months apart).

Procedure

Training for intervention school staff began in the fall of 2016 and continued through the 2016–2017 school year, consisting of 3 half-day in-person sessions, periodic check-ins via videoconference, and access to resources (e.g., newsletters). The three in-person training sessions per school were conducted in (1) late September and early October, (2) late October through early December, and (3) late January through late March. Training sessions were conducted by D. W. and R. T. Johnson, supported by the authors, and utilized Cooperation in the Classroom, 9th Edition by Johnson et al. (2013); each staff member received a copy of the book. Due to the geographic dispersal of the schools, each school received training individually according to their own schedule for professional development. Finally, we conducted a one-day administrator training during the summer of 2017, and a half-day follow-up training in the second year.

Under the Johnson’s approach, cooperative learning can include reciprocal teaching (e.g., Jigsaw), peer tutoring, collaborative reading, and other methods in which peers help each other learn in small groups under conditions of positive interdependence. The Johnsons’ approach also emphasizes individual accountability, explicit coaching in collaborative social skills, a high degree of face-to-face interaction, and guided post-lesson processing of group performance. Cooperative learning is viewed as a conceptual framework within which teachers can apply the basic concepts to design their own group-based activities using existing curricula.

Measures

Student data collection was conducted in September/October and March/April of the 2016–2017 and 2017–2018 school years (4 waves in total, about 6 months apart) using on-line surveys (i.e., Qualtrics; https://www.qualtrics.com/). To assess fidelity of implementation, we also conducted teacher observations. A Certificate of Confidentiality was obtained for these data from NIAAA (#CC-AA-17–011). To shrink the overall number of items and reduce participant burden, existing data from other studies were used to select the highest-loading items from each scale below (additional information available from the first author).

Positive Peer Relations

We used 4 items from the Peer Relatedness Scale, which has been used in previous research as a predictor of positive school adjustment in adolescents (Furrer & Skinner, 2003). Items included “When I’m with my classmates, I feel accepted” and “When I’m with my classmates, I feel unimportant” (reverse scored). Students responded on a 4-point scale from 1 (Not at all true) to 4 (Very true). Items were averaged to arrive at the scale score. Cronbach’s alpha (reliability) was between .71 and .84.

Affective Empathy

We assessed empathy using a subset of items from the Basic Empathy Scale (Jolliffe & Farrington, 2006). We used 3 items, including “After being with a friend who is sad about something, I usually feel sad” and “I can often understand how people are feeling even before they tell me”. Students responded on a 5-point scale from 1 (Strongly disagree) to 5 (Strongly agree) and items were averaged to arrive at the final scores. Cronbach’s alpha (reliability) was between .69 and .78.

Bullying

We assessed bullying using a subscale from the University of Illinois Bully Scale (Espelage & Holt, 2001). We used 5 items which ask about behavior over the previous 30 days, including “I teased other students while we were in a group” and “I spread rumors about other students”. Students responded on a 5-point scale from 0 (Never) to 4 (7 or more times) and items were averaged to arrive at the final scores. Cronbach’s alpha (reliability) was between .74 and .83.

Demographics

Sex was collected from school records and coded Male (0), Female (1).

Intervention Fidelity

Research staff blind to intervention assignment observed teaching practices in intervention and control schools. Using an established observation protocol for key aspects of cooperative learning (e.g., positive interdependence; Krol et al., 2008; Veenman et al., 2002), the second author trained observers to 100% reliability using simulated data before they were permitted to conduct observations in actual classrooms. Observations were conducted once in the late fall/early winter and again in the spring. Observers remained in a classroom for an entire class period.

Analysis Plan

We used Structural Equation Modeling (SEM) to fit our cross-lag difference score model, which provides a number of advantages. For example, we can constrain all cross-lag coefficients in the model to be identical, which provides empirical tests of whether the degree of influence of X on Y is identical to the influence of Y on X (i.e., reciprocity). In addition, we conducted tests of moderation through a multiple-group comparison with a deviance test to explore whether effects hold for subgroups (e.g., boys vs. girls).

We fit our cross-lag difference score model using Mplus 7.4 (Muthén & Muthén, 1998–2012) and Maximum Likelihood (ML) estimation with robust standard errors, which can provide unbiased estimates in the presence of missing data and/or non-normal distributions (Enders & Bandalos, 2001). Mplus also enabled us to account for the nesting in the data and calculate appropriate standard errors; however, sample size limitations prevented us from including random effects in the model, so all effects were fixed.

We provide standard measures of fit, including the chi-square (χ2), comparative fit index (CFI), nonnormed or Tucker-Lewis index (TLI), and root mean square error of approximation (RMSEA). CFI values greater than .95, TLI values greater than .90, and RMSEA values less than .05 indicate good fit (Bentler, 1990; Hu & Bentler, 1999).

Results

We provide descriptive data and correlations in Table 3. Female students reported lower levels of peer relations (r = −.05 to −.17) and bullying (r = −.06 to −.08), with some correlations being non-significant, and higher levels of affective empathy (r = .24 to .29). Multi-level models indicated that students in intervention and control schools did not differ in terms of baseline levels of bullying (B = −.03, SE = .04, ns), peer relatedness (B = −.01, SE = .05, ns), or empathy (B = −.10, SE = .09, ns). With regards to fidelity observations, analysis indicated significantly higher levels of observed positive interdependence in intervention schools as compared to control schools (B = .04, SE = .02, p < .05, R2 = .10).

Table 3.

Correlations, means, and standard deviations for all study measures

Variable 1 2 3 4 5 6 7 8 9 10 11
1. Peer Relatedness (W1) —
2. Peer Relatedness (W2) .50*** —
3. Peer Relatedness (W3) .42*** .45*** —
4. Peer Relatedness (W4) .35*** .38*** .53*** —
5. Affective Empathy (W1) −.06* −.08* −.06* −.05 —
6. Affective Empathy (W2) .02 .11*** .02 .02 .47*** —
7. Affective Empathy (W3) .04 .01 .20*** .11*** .38*** .47*** —
8. Affective Empathy (W4) .01 .04 .12*** .28*** .30*** .38*** .50*** —
9. Bullying (W1) −.17*** −.07* −.13*** −.12*** −.07* −.09** −.13*** −.11*** —
10. Bullying (W4) −.04 −.02 −.18*** −.30*** −.04 −.08** −.19*** −.26*** −.32*** —
11. Sex −.05* −.11*** −.17*** −.10*** .24*** .29*** .26*** .25*** −.06* −.08** —
N 1447 1513 1562 1481 1449 1531 1566 1470 1453 1476 1856
M 3.07 2.97 2.94 2.84 3.38 3.40 3.37 3.39 .26 1 .39 .48
SD .68 .76 .79 .84 .93 .96 .97 1.01 .50 .65 .50
Open in a new tab
*

p < .05.

**

p < .01.

***

p < .001.

Note. Sex was coded Male (0), Female (1), so positive correlations indicate that females were higher. Peer relatedness scores ranged from 1 to 4. Affective empathy scores ranged from 1 to 5. Bullying scores ranged from 0 to 4.

We fit our cross-lag difference score model to the data and model fit was adequate, χ2(43) = 114.65, p < .001; CFI = .98; TLI = .97; RMSEA = .030 (90% C.I.: .023-.036). The results (presented in Figure 2) indicated a bi-directional or reciprocal relationship between peer relations and affective empathy across time. Specifically, peer relations predicted positive change in affective empathy (β = .34-.45, p < .001, 95% C.I.: .20 to .28|.49 to .62) and affective empathy predicted positive change in peer relations (β = .21-.22, p < .01, 95% C.I.: .06|.37 to .39). When these reciprocal effects were constrained to be equal, model fit was significantly impacted [χ2(1) = 9.14, p < .001] suggesting that the effect of peer relations on affective empathy was stronger than the effect of empathy on peer relations.

Figure 2.

Figure 2.

Open in a new tab

Fitted model (with standardized betas and 95% confidence intervals).

Change in both constructs over time predicted lower levels of bullying at wave 4 (peer relations: β = −.48, p < .001, 95% C.I.: −.65|−.32; affective empathy: β = −.31, p < .001, 95% C.I.: −.48|−.14), controlling for bullying at Wave 1. When these effects were constrained to be equal, model fit was significantly impacted [χ2(1) = 4.30, p < .05], suggesting that the effect of peer relations on bullying was stronger than that of affective empathy.

The autoregressive pathways (i.e., the prediction from baseline to the change score) were non-significant for peer relations (β = −.02, ns, 95% C.I.: −.24 to −.21|.17 to .20) but significant for affective empathy (β = −.24 to −.28, p < .001, 95% C.I.: −.44 to −.38|−.10 to −.12), suggesting a degree of stability or ceiling effects (i.e., affective empathy increased less for those who were already scoring higher). The predictive path from bullying at Wave 1 to Wave 4 was also significant, again suggesting a degree of stability in this behavior (β = .34, p < .001, 95% C.I.: .26|.42).

With regards to the intervention, cooperative learning predicted positive change in peer relations (β = .40, p < .001, 95% C.I.: .26 to .27|.52 to .54) and affective empathy (β = .25-.28, p < .001, 95% C.I.: .14 to .17|.36 to .40), as well as lower levels of bullying at Wave 4, controlling for Wave 1 levels (β = −.19, p < .01, 95% C.I.: −.25|−.13); this latter effect is independent of the effects of peer relations and affective empathy (these results are not presented in Figure 2 to enhance clarity). Overall, model paths did not differ by sex, χ2(10) = 13.69, ns.

Discussion

These results suggest that the positive effects of cooperative learning build upon one another over time. The social nature of cooperative learning enhances student perceptions of peer relations, and at the same time the experiential skill building in small learning groups can contribute to a more profound understanding of the emotional states of others (i.e., affective empathy). These effects strengthen one another in a positive feedback loop; as peer relations improve, they contribute to the development of affective empathy, and affective empathy contributes to the development of more positive peer relations.

In addition, both positive peer relations and affective empathy contributed independently to reductions in bullying over time. Researchers have long argued that those with high affective empathy will be better able to experience the emotional reactions of others and will thus be less inclined to bully their peers (Miller & Eisenberg, 1988), and these results add to that literature. Interestingly, peer relations also exhibited a significant suppressive effect on bullying that was even stronger than affective empathy; this may be due to the social nature of cooperative learning, which provides a way to attain a degree of acceptance or status with peers without resorting to bullying. In other words, cooperative learning may alter social norms such that peer acceptance or status becomes a function of prosocial behavior (Van Ryzin, Roseth, & Biglan, 2020) rather than antisocial acts, such as bullying. This hypothesis deserves additional research. From a practical perspective, our results suggest that bullying prevention should focus as much (or more) on improving peer relations as opposed to developing student empathy as a means to reduce bullying.

The lack of autoregressive effects in peer relations across time indicates that students with higher levels of perceived peer relations did not necessarily experience the most positive change, suggesting that cooperative learning can build peer relations even among students who are less widely accepted by peers. The autoregressive effects for affective empathy suggest that this skill may accumulate over time, requiring a significant time investment to achieve positive change. This may be why existing empathy enhancement programs designed for adolescents have not demonstrated significant effects (Teding van Berkhout & Malouff, 2016).

We found no sex differences in these links, i.e., sex did not moderate these relationships. Although previous research has found sex differences in empathy (De Wied et al., 2007) and peer relations (Rose & Rudolph, 2006), research has not found sex differences in links between empathy and peer relations (De Wied et al., 2007; Gleason et al., 2009), suggesting that these reciprocal processes operate similarly for boys and girls.

This study is limited in several ways. First, it is based upon a relatively homogeneous sample of rural students that was about three-quarters White, which limits the external validity (generalizability) of the results. Second, all student measures were self-report, which limits internal validity. Future research should consider additional data sources, such as teachers and/or parents, and more diverse populations. And third, the small number of schools in our sample (i.e., 15) limited the complexity of the models that we were able to fit to the data and may have prevented us from finding significant effects in some cases.

Conclusion

This study provides further evidence that both positive peer relations and affective empathy are enhanced through the use of cooperative learning. Unlike more informal approaches to small-group work, cooperative learning establishes specific structures and contingencies (i.e., positive interdependence, individual accountability, etc.) that support more extensive, and more positive, social interactions among students. With cooperative learning, students have clear roles and tasks, specific incentives to cooperate, and accountability mechanisms that make a positive social and academic experience much more likely as compared to less structured approaches. Without these structures and contingencies, small-group instruction can lead to confusion, disagreement, or an unequal distribution of work, which can result in negative social and academic experiences for students.

This study also found that the effects of peer relatedness and affective empathy are mutually reinforcing over time, whereby empathy contributes to greater success in establishing and maintaining positive peer relations, and positive peer relations serve as an experiential context to support further growth in empathy. Peer relations contributed more strongly to empathy than the reciprocal pathway, possibly indicating the degree to which peer relations are a key context where empathy can be developed and strengthened; in contract, empathy is just one of many skills that are used in the maintenance of positive peer relations. Both constructs have significant suppressive effects on bullying (likely via different mechanisms) and thus both deserve attention in any effort to reduce bullying in middle school – particularly peer relations, which demonstrated stronger suppressive effects than empathy.

In previous work, we reported that cooperative learning has significant positive effects on student behavior, including reductions in alcohol and tobacco use and higher levels of prosocial behavior (Van Ryzin & Roseth, 2018b, c), as well as positive effects on student mental health (Van Ryzin & Roseth, 2021), and the design of our study (i.e., cluster randomization, longitudinal data) adds strength to these results. Previous work found that cooperative learning can promote greater academic motivation and achievement (see meta-analyses by Johnson et al., 2014; Roseth et al., 2008). Thus, cooperative learning is a way for schools to reduce bullying and while enhancing academic, behavioral, and social-emotional outcomes for students simultaneously. Given that cooperative learning can be used in any subject at any grade level from kindergarten to graduate school, we argue for an increased emphasis on cooperative learning as a core aspect of instruction in K-12 education.

Finally, we present one caveat: cooperative learning is a complex pedagogy, and is far more effective when implemented with fidelity (Roseth et al., 2008). Thus, we suggest that teachers and schools obtain specialized training or resources (e.g., Johnson et al., 2013) or enlist specialized technology support (e.g., https://www.PeerLearning.net/) before attempting to implement cooperative learning in their classroom. In particular, technology can reduce the learning curve and support teachers in implementing cooperative learning quickly and easily, and with a higher degree of fidelity.

Acknowledgements

The National Institute on Alcohol Abuse and Alcoholism (NIAAA) provided financial support this project (R34 AA024275; PI: M. J. Van Ryzin). The content of this manuscript is solely the responsibility of the authors and does not necessarily represent the official views of NIAAA or the National Institutes of Health.

Footnotes

The authors declare that they have no conflicts of interest.

References

  1. Barchia K, & Bussey K (2010). The psychological impact of peer victimization: Exploring social-cognitive mediators of depression. Journal of Adolescence, 33(5), 615–623. 10.1016/j.adolescence.2009.12.002 [DOI] [PubMed] [Google Scholar]
  2. Bentler PM (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107(2), 238–246. 10.1037/0033-2909.107.2.238 [DOI] [PubMed] [Google Scholar]
  3. Caravita SC, Di Blasio P, & Salmivalli C (2009). Unique and interactive effects of empathy and social status on involvement in bullying. Social Development, 18(1), 140–163. 10.1111/j.1467-9507.2008.00465.x [DOI] [Google Scholar]
  4. Casey BJ, Jones RM, & Somerville LH (2011). Braking and accelerating of the adolescent brain. Journal of Research on Adolescence, 21(1), 21–33. 10.1111/j.1532-7795.2010.00712.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Choi J, Johnson DW, & Johnson R (2011). Relationships among cooperative learning experiences, social interdependence, children’s aggression, victimization, and prosocial behaviors. Journal of Applied Social Psychology, 41(4), 976–1003. 10.1111/j.1559-1816.2011.00744.x [DOI] [Google Scholar]
  6. Craig W, Harel-Fisch Y, Fogel-Grinvald H, Dostaler S, Hetland J, Simons-Morton B, ... & Pickett W (2009). A cross-national profile of bullying and victimization among adolescents in 40 countries. International Journal of Public Health, 54, 216–224. 10.1007/s00038-009-5413-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. De Wied M, Branje SJ, & Meeus WH (2007). Empathy and conflict resolution in friendship relations among adolescents. Aggressive Behavior, 33(1), 48–55. 10.1002/ab.20166 [DOI] [PubMed] [Google Scholar]
  8. Deutsch M (2011) Cooperation and Competition. In Coleman P (Ed.), Conflict, Interdependence, and Justice. Peace Psychology Book Series, vol 11. Springer. 10.1007/978-1-4419-9994-8_2 [DOI] [Google Scholar]
  9. Enders CK, & Bandalos DL (2001). The relative performance of full information maximum likelihood estimation for missing data in structural equation models. Structural Equation Modeling, 8(3), 430–457. 10.1207/S15328007SEM0803_5 [DOI] [Google Scholar]
  10. Espelage DL, & Holt M (2001). Bullying and victimization during early adolescence: Peer influences and psychosocial correlates. Journal of Emotional Abuse, 2(2–3), 123–142. 10.1300/J135v02n02_08 [DOI] [Google Scholar]
  11. Espelage DL, & Swearer SM (2003). Research on school bullying and victimization: What have we learned and where do we go from here?. School Psychology Review, 32(3), 365–383. 10.1080/02796015.2003.12086206 [DOI] [Google Scholar]
  12. Furrer C, & Skinner E (2003). Sense of relatedness as a factor in children’s academic engagement and performance. Journal of Educational Psychology, 95(1), 148–162. 10.1037/0022-0663.95.1.148 [DOI] [Google Scholar]
  13. Gaffney H, Farrington DP, & Ttofi MM (2019). Examining the effectiveness of school-bullying intervention programs globally: A meta-analysis. International Journal of Bullying Prevention, 1(1), 14–31. 10.1007/s42380-019-0007-4 [DOI] [Google Scholar]
  14. Gleason KA, Jensen-Campbell LA, & Ickes W (2009). The role of empathic accuracy in adolescents’ peer relations and adjustment. Personality and Social Psychology Bulletin, 35(8), 997–1011. 10.1177/0146167209336605 [DOI] [PubMed] [Google Scholar]
  15. Hu L, & Bentler PM (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1–55. 10.1080/10705519909540118 [DOI] [Google Scholar]
  16. Johnson DW, & Johnson R (1989). Cooperation and competition: Theory and research Interaction Book Company. [Google Scholar]
  17. Johnson DW, Johnson R, & Holubec E (2013). Cooperation in the classroom (9th ed.) Interaction Book Company. [Google Scholar]
  18. Johnson DW, Johnson RT, & Maruyama G (1983). Interdependence and interpersonal attraction among heterogeneous and homogeneous individuals: A theoretical formulation and a meta-analysis of the research. Review of Educational Research, 53(1), 5–54. 10.3102/00346543053001005 [DOI] [Google Scholar]
  19. Johnson DW, Johnson RT, Roseth CJ, & Shin T-S. (2014). The relationship between motivation and achievement in interdependent situations. Journal of Applied Social Psychology, 44(9), 622–633. 10.1111/jasp.12280 [DOI] [Google Scholar]
  20. Jolliffe D, & Farrington DP (2006). Development and validation of the Basic Empathy Scale. Journal of Adolescence, 29(4), 589–611. 10.1016/j.adolescence.2005.08.010 [DOI] [PubMed] [Google Scholar]
  21. Juvonen J, Nishina A, & Graham S (2000). Peer harassment, psychological adjustment, and school functioning in early adolescence. Journal of Educational Psychology, 92(2), 349–359. 10.1037/0022-0663.92.2.349 [DOI] [Google Scholar]
  22. Kennedy RS (2020). A meta-analysis of the outcomes of bullying prevention programs on subtypes of traditional bullying victimization: Verbal, relational, and physical. Aggression and Violent Behavior, 101485. 10.1016/j.avb.2020.101485 [DOI] [Google Scholar]
  23. Krol K, Sleegers P, Veenman S, & Voeten M (2008). Creating cooperative classrooms: Effects of a two‐year staff development program. Educational Studies, 34(4), 343–360. 10.1080/03055690802257101 [DOI] [Google Scholar]
  24. McArdle JJ (2009). Latent variable modeling of differences and changes with longitudinal data. Annual Review of Psychology, 60, 577–605. 10.1146/annurev.psych.60.110707.163612 [DOI] [PubMed] [Google Scholar]
  25. Miller PA, & Eisenberg N (1988). The relation of empathy to aggressive and externalizing/antisocial behavior. Psychological Bulletin, 103(3), 324–344. 10.1037/0033-2909.103.3.324 [DOI] [PubMed] [Google Scholar]
  26. Musu L, Zhang A, Wang K, Zhang J, and Oudekerk BA (2019). Indicators of School Crime and Safety: 2018 (NCES 2019–047/NCJ 252571). National Center for Education Statistics, U.S. Department of Education, and Bureau of Justice Statistics, Office of Justice Programs, U.S. Department of Justice Available at: https://nces.ed.gov/pubs2019/2019047.pdf.
  27. Muthén LK, & Muthén BO (1998–2012). Mplus User’s Guide (7th Edition). Muthén & Muthén. [Google Scholar]
  28. Nishina A, Juvonen J, & Witkow MR (2005). Sticks and stones may break my bones, but names will make me feel sick: The psychosocial, somatic, and scholastic consequences of peer harassment. Journal of Clinical Child and Adolescent Psychology, 34(1), 37–48. 10.1207/s15374424jccp3401_4 [DOI] [PubMed] [Google Scholar]
  29. Pellegrini AD, & Long JD (2002). A longitudinal study of bullying, dominance, and victimization during the transition from primary school through secondary school. British Journal of Developmental Psychology, 20(2), 259–280. 10.1348/026151002166442 [DOI] [Google Scholar]
  30. Portt E, Person S, Person B, Rawana E, & Brownlee K (2020). Empathy and positive aspects of adolescent peer relationships: A scoping review. Journal of Child and Family Studies, 29, 2416–2433. 10.1007/s10826-020-01753-x [DOI] [Google Scholar]
  31. Robers S, Kemp J, & Truman J (2013). Indicators of School Crime and Safety: 2012 (NCES 2013–036/NCJ 241446). National Center for Education Statistics, U.S. Department of Education, and Bureau of Justice Statistics, Office of Justice Programs, U.S. Department of Justice
  32. Rose AJ, & Rudolph KD (2006). A review of sex differences in peer relationship processes: potential trade-offs for the emotional and behavioral development of girls and boys. Psychological Bulletin, 132(1), 98–131. 10.1037/0033-2909.132.1.98 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Roseth CJ, Johnson DW, & Johnson RT (2008). Promoting early adolescents’ achievement and peer relationships: The effects of cooperative, competitive, and individualistic goal structures. Psychological Bulletin, 134(2), 223–246. 10.1037/0033-2909.134.2.223 [DOI] [PubMed] [Google Scholar]
  34. Rueger SY, Malecki CK, & Demaray MK (2011). Stability of peer victimization in early adolescence: Effects of timing and duration. Journal of School Psychology, 49(4), 443–464. 10.1016/j.jsp.2011.04.005 [DOI] [PubMed] [Google Scholar]
  35. Storch EA, Masia‐Warner C, Crisp H, & Klein RG (2005). Peer victimization and social anxiety in adolescence: A prospective study. Aggressive Behavior, 31(5), 437–452. 10.1002/ab.20093 [DOI] [Google Scholar]
  36. Sullivan TN, Farrell AD, & Kliewer W (2006). Peer victimization in early adolescence: Association between physical and relational victimization and drug use, aggression, and delinquent behaviors among urban middle school students. Development and Psychopathology, 18(1), 119–137. 10.1017/S095457940606007X [DOI] [PubMed] [Google Scholar]
  37. Teding van Berkhout E, & Malouff JM (2016). The efficacy of empathy training: A meta-analysis of randomized controlled trials. Journal of Counseling Psychology, 63(1), 32. 10.1037/cou0000093 [DOI] [PubMed] [Google Scholar]
  38. Thijs J, & Verkuyten M (2008). Peer victimization and academic achievement in a multiethnic sample: The role of perceived academic self-efficacy. Journal of Educational Psychology, 100(4), 754–764. 10.1037/a0013155 [DOI] [Google Scholar]
  39. Van Ryzin MJ, & Roseth CJ (2016–2018). Enlisting peer cooperation and prosociality in the service of substance use prevention in middle school. Identifier: NCT03119415. https://clinicaltrials.gov/ct2/show/NCT03119415 [DOI] [PMC free article] [PubMed]
  40. Van Ryzin MJ, & Roseth CJ (2018a). Cooperative learning in middle school: A means to improve peer relations and reduce victimization, bullying, and related outcomes. Journal of Educational Psychology, 110(8), 1192–1201. 10.1037/edu0000265 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Van Ryzin MJ, & Roseth CJ (2018b). Enlisting peer cooperation in the service of alcohol use prevention in middle school. Child Development, 89(6), e459–e467. 10.1111/cdev.12981 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Van Ryzin MJ, & Roseth CJ (2018c). Peer influence processes as mediators of effects of a middle school substance use prevention program. Addictive Behaviors, 85, 180–185. 10.1016/j.addbeh.2018.06.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Van Ryzin MJ, & Roseth CJ (2019). Effects of cooperative learning on peer relations, empathy, and bullying in middle school. Aggressive Behavior, 45(6), 643–651. 10.1002/ab.21858 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Van Ryzin MJ, & Roseth CJ (2021). The cascading effects of reducing student stress: Cooperative learning as a means to reduce emotional problems and promote academic engagement. The Journal of Early Adolescence, 41(5), 700–724. 10.1177/0272431620950474 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Van Ryzin MJ, Roseth CJ, & Biglan A (2020). Mediators of effects of cooperative learning on prosocial behavior in middle school. International Journal of Applied Positive Psychology, 5(1), 37–52. 10.1007/s41042-020-00026-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Veenman S, van Benthum N, Bootsma D, van Dieren J, & van der Kemp N (2002). Cooperative learning and teacher education. Teaching and Teacher Education, 18(1), 87–103. 10.1016/S0742-051X(01)00052-X [DOI] [Google Scholar]
  47. World Health Organization (2012). Risk behaviours: Being bullied and bullying others. In Currie C et al. (Eds.), Social determinants of health and well-being among young people. Health behaviour in school-aged children (HBSC) study: International report from the 2009/2010 survey (pp. 191–200). WHO Regional Office for Europe. [Google Scholar]

RESOURCES