This study examined the effectiveness of the Don’t Laugh at Me curricula for improving classroom climate, using a quasi-experimental design. Participants (N = 67) included elementary school students in 4th and 5th grade. Forty-six students received the intervention curriculum, and 21students served as controls. Treatment group participants were exposed to selected modules of the Don’t Laugh at Me (DLAM) curricula for 10 weeks. Outcome measures were administered as a pretest and posttest. Students in the experimental group exhibited significant gains in psychological sense of school membership, relational experiences, the desire to stop dissing and ridicule, and significant declines in bullying, compared to the control group. No significant gender differences were found. DLAM revealed 3 factors and showed acceptable evidence of construct validity. Teacher perceptions of the process and implementation issues with respect to DLAM are also reported. Study findings and their implications are discussed.
There is an emerging consensus that effective character education must include efforts to promote “communities of caring” within classrooms and schools (Battistich, Solomon, Watson, & Schaps, 1997). According to this view, the development of character is fundamentally linked to the quality of interpersonal relationships that children experience within their educational settings. Social processes that contribute to students’ perception of acceptance, caring, support and belonging are critical components of socio-moral development (Berkowitz & Bier, in press). A school climate that encourages social and emotion bonding and promotes positive interpersonal experiences is one that provides the minimum necessary grounding for the formation of character. Indeed, as Berkowitz (2002) put it, “Relationships are critical to character education, so character education must focus on the qualityof relationships at school” (p. 58-59). Specialized curricula, programs, and pedagogy are likely to have scant effect if the child’s daily relational experiences at school are characterized by victimization, bullying and humiliation.
There is mounting empirical evidence to support the putative linkage between children’s perception of the sense of community and sociomoral and prosocial development (Solomon, Watson, & Battistich, 2002). Student perception that one’s classroom or school is a caring community has been linked to greater concern for others, more acceptance of outgroups, better conflict resolution skills, intrinsic prosocial motivation and altruistic behavior (Battistich, Solomon, Kim, Watson & Schaps, 1995; Solomon, Battistich, Watson, Schaps, & Lewis, 2000). The sense of belonging and community has also been linked to prosocial behavior (Solomon, Watson, Delucchi, Schaps, & Battistich, 1988), academic motivation and effort (Goodenow, 1993), reduced drug use and delinquent behavior (Battistich & Hom, 1997), social problem-solving (Battistich, Solomon, Watson, Solomon, & Schaps, 1989) and civic virtue (Oakes, Quartz, Ryan, & Lipton, 2000).
Hence there is clear theoretical and empirical warrant for grounding character education on relational and social processes within classrooms and schools. In addition to its sociomoral implications, the quality of affective and relational experiences also influences a broader set of educational outcomes. For example, research has shown that when students perceive their school climate positively, they develop a strong sense of belonging and school bonding (Eccles & Midgley, 1989; Goodenow, 1993; Wentzel, 1998). Students who perceive their teachers and peers as socially and emotionally supportive report academic motivation and a positive self-concept (Blankenship, 1990; Dart et al., 1999; Fouts & Myers, 1992; Osterman, 2000). Support from teachers and peers, embedded in a positive classroom climate, appears to create a strong sense of school membership that encourages educational persistence (Finn, 1989; Fraser, 1994; Goh & Fraser, 1998; Goh & Tobin, 1999; Wehlage, 1989).
School climate and the sense of community, then, are crucial variables for understanding the academic, sociomoral and characterological aspects of education. In this context there has been understandable urgency among educators to combat the deleterious effects of peer bullying and victimization. Research suggests that the effects of bullying and teasing are not limited to the several students who are its target, but rather has pervasive effects on a school climate (Hoover & Hazler, 1991). Indeed, the culture of bullying in classrooms and schools subverts the formation of caring school communities and works against efforts to promote forms of social interaction that support sociomoral development. Moreover, bullying and peer victimization appear to be a sadly ubiquitous feature of American schools (Slee & Taki, 1999; Smith et al., 1999). For example, a recent survey of bullying in South Carolina middle schools showed that 25% of youngsters reported that they had been bullied several times in a 3-month period; 10% said they had been bullied at least once a week (Institute for Families in Society, 1997).
Victims of bullies can suffer both short-term and long-term effects (Limber, 1997). In the short-term students may show symptoms of depression, loss of interest in school work or even attending school. Moreover, the effects of bullying can persist into adulthood. A recent longitudinal study of male victims who were bullied during childhood found that in their 20s they were more depressed and had lower self-esteem than their counterparts who had not been bullied in childhood (Olweus, in press). Bullying can also indicate a serious problem for the bully as well as the victim. Olweus’ study (in press) found that about 60% of the boys who were identified as bullies in middle school had at least one criminal conviction (and about one third had three or more convictions) in their 20s, a far higher percentage than nonbullies. Consequently interventions that hold promise for altering the social dynamics that encourage or sustain the culture of bullying and victimization is attracting much interest.
The Don’t Laugh at Me (DLAM) program is one such intervention. This program was conceived by Operation Respect and Educators for Social Responsibility (2000). It attempts to sensitize elementary and middle-school youngsters to the painful effects of peer ridicule, rejection, ostracism, and bullying, and to inspire them to create classrooms that are “climates of respect” and “ridicule free zones.” By means of special classroom activities and curricula DLAM attempts to transform the culture of schools into places of caring, compassion, and cooperation. It attempts to teach children how to work together to recognize intolerance, increase respect, and enhance sensitivity to students’ emotional needs. It encourages children to recognize the contributions of each member of the classroom and school community to resolve conflicts creatively, and to celebrate diversity by acknowledging differences without judgment. The DLAM approach views school not only as a place where children acquire academic skills to become successful, but also be a place where they acquire the tools to help them grow up to be ethical, compassionate citizens of strong character, healthy self-esteem, and humane sensibilities.
Although DLAM is widely promoted and actively implemented in schools throughout the country, and although the standard DLAM program evaluation survey is routinely distributed with the program, we are unaware of published studies on whether aspects of the DLAM curriculum are effective in improving the climate of respect within elementary classrooms. In this article we report the first such evidence. The present study has a number of additional novelties. To our knowledge ours is the first DLAM efficacy study to employ a quasi-experimental methodology that involves a pretest and posttest of treatment and within-school control groups. In addition, we attempted to measure program effectiveness using a broader array of assessments than is currently the practice in DLAM program evaluation. Hence, in addition to the DLAM survey, we also measured students’ sense of school membership, their perception of classroom climate, and their perception of bullying and victimization. We also examined whether the program influences boys and girls similarly. Finally, we present the first evidence on the construct validity of the standard evaluation instrument that is distributed with DLAM curricula.
Method
Participants
Participants included 67 students from fourth- and fifth-grade classes of a public elementary school. Classrooms were randomly assigned to treatment and control conditions. There were two fourth grade classes (N = 27), one was a control (n = 9), one was treatment (n = 18). In the fourth grade, there were 14 females altogether, 5 in the control and 9 in the treatment group. Thirteen of the fourth grade subjects were males, 4 were the control and 9 were in the treatment group. There were 3 fifth-grade classes (N = 40), one was a control (n = 12) with an equal number of males and females in each group. Two fifth-grade classes formed the treatment group (n = 28), one class had 10 subjects, five males and five females. The other fifth grade class had 18 subjects, 8 males and 10 females. The average class size at the school is 18 students per class. Only students who returned signed permission slips participated in the study. This is the reason for the uneven small class sizes. The distribution of participants by grade and condition is summarized in Table 1. Overall, the student racial composition was 62% White, 25% Black, 9% Multiracial, 3% Hispanic and 1% Asian. Specifically, 25 of the subjects were White, 27 Black, 2 Hispanic, 3 Native American, 8 Multiracial, and 2 Other. Eighty-six percent of the students in the school qualify for free or reduced lunch.
Procedure
Teacher Training
Teachers whose classrooms were randomly assigned to the treatment condition attended a full-day workshop administered by a professional DLAM trainer. The complete DLAM program consists of four major thematic units, and a total of 11 modules across the four thematic units. As children move through the first unit (“Being You, Being Me, Being Us”) they progressively learn how to identify and monitor emotions, and they practice alternatives to acting-out when angry or upset. In the second unit (“I Care, You Care, We Care”) children identify and monitor aversive behavior, learn how to generate classroom commitment to a “ridicule free zone,” and identify specific caring behaviors. The major focus of the third unit (“Words That Hurt, Words That Heal”) is conflict resolution. It encourages children to practice perspective-taking, to see all sides of a conflict, and to understand the “you” and “I” messages that serve to escalate or de-escalate conflict. One module in this unit also teaches children how to respond assertively to bullying (“being a bully buster”), and the appropriate conditions to intervene. Finally, the fourth unit (“Together We Can”) encourages children to accept and celebrate differences, to be sensitive to the effects of teasing, prejudice and bullying, and to recommit to making their classroom and school a Ridicule Free Zone.
As can be seen, every unit ends with a “Take It To The Community” activity that encourages students to share what they have learned with the larger classroom and school community (and, for one unit, the community beyond the school), and to engage in some concrete activity that solidifies the lessons of the unit. Table 2 summarizes the four major thematic units of DLAM, as well as the instructional objectives of each module.
Implementation
A pretest was administered at the end of February before program implementation. The program was implemented over an 8-week period from early March through early May. The posttest was given 2 weeks after the program was completed. Participating classrooms implemented the program in two 30-minute segments per week. Teachers in the control group taught their regular curricula without incorporating DLAM concepts in their lessons. The ideal design would require teachers to implement systematically all of the modules under each of four units. This was not possible in the present school given other demands placed on teachers. Consequently teachers were given some flexibility about which units and modules to implement in their respective class. In our view whatever limitations this introduces are offset by the ecological validity of the implementation. The reality in most schools and classrooms is that character education and other sociomoral programs compete with direct instruction of the standard academic curriculum, especially in the present context that emphasizes academic accountability. Hence it is a matter of empirical interest to know what minimum number of program modules must be implemented in order to secure an intervention effect.
In this study all participating teachers completed at least four modules, one module from each thematic unit. The following modules were covered by all participating teachers: The Torn Heart; The Ridicule-Free Zone Constitution of Caring; Walk in Another’s Shoes; and Family Ties and Fabric Tales. The fourth-grade teacher covered two additional modules; The Caring Being and Being a Bullying Buster. One fifth-grade teacher covered one additional module; Being a Bullying Buster. Each module took approximately 10 minutes to prepare and 20 minutes for the activity. Student outcomes were assessed in a pretest and posttest. To control for possible differences due to differential reading proficiency, each question on a survey was read aloud. At the end of the program implementation, participating teachers completed the Teacher Survey included in the DLAM program, asking them to give their personal comments with respect to the effectiveness of the DLAM program.
Instruments
DLAM Program Scale
A 30-item survey is circulated with the DLAM curricula as the standard test of program effectiveness. Two items (“This is how I feel about my school” and “This is how much I like the children in this class”) are rated along a five-step Likert continuum (e.g., “I like it a lot” to “I don’t like it at all”). The remaining 28 items are rated along a three-step continuum that varies somewhat with the rating stem. Six items, for example, ask participants to rate the degree to which they are happy, enthusiastic, sad, bored, angry or lonely in “this class” along a continuum that includes these response options: “Never or almost never,” “Sometimes” and “A lot of the time.” Four items ask participants to rate the degree to which children in this class like and respect “me” and “each other” using these response options: “A lot,” “OK” and “Not very much.” One item asks if “the adults in this school care about me” and is rated along the same continuum. Five items ask how often various things happen to the participant, such as another child “says something nice to me,” or smiles, teases, say something mean or bullies. These items are rated as “Almost or almost never, ” “Sometimes,” and “A lot.” This scale is also used for the next five items that inquires how often the child smiles at someone, says something nice, says something mean, teases or bullies. The next three items ask the child how safe the child feels (“Very Safe,” “OK,” and “Not Safe”) in the classroom, on the playground and in the hallways. Finally, four items ask how important it would be (“Kind of” and “Not Important”) to stop name-calling, dissing, bullying, and making fun of someone (ridiculing).
The traditional use of this scale has been to look for program effects or group (treatment) differences on an item-by-item basis. This would require, of course, 30 different comparisons, one for each survey item. One clear drawback of this procedure is that it inflates the Type 1 error rate. One option is to reduce the number of comparisons to determine if scale items would coalesce into well-defined subscales. To this end we subjected the 30-item scale to an exploratory factor analysis (extracting principal components). This analysis revealed three factors in accordance with both scree and parallel analysis (see Lautenschlager, 1989) criteria, accounting for 41% of the variance. To clarify the factor structure a second principal components analysis was calculated with oblique rotation (using the direct oblimin method). The resulting rotated component matrix is reported in Table 3. Note that only those items with eigenvalue loadings < .40 are reported.
The first factor consists of nine items that possess strong internal consistency (α = .87). The dominant theme of this factor is the relational experience of the classroom (e.g., “Children in this class smile/respect me”; “Another child in this class smiles/bullies/says something nice/mean to me”), and therefore was labeled Relational Experience. Note that low scores on this scale indicate a more positive relational experience. The second factor consists of four items (α = .86) that reflect the desire to stop ridicule, bullying, dissing and teasing, and therefore was labeled “Stop Ridicule.” High scores on this scale indicate a greater commitment to stop ridicule. The third factor consists of nine items (α = .77) whose dominant theme is the personal affective experience of the classroom and school (e.g., “I feel angry/happy/safe in the class/hallway/school/ playground”), and therefore was labeled Per-sonal Affect. High scores indicate more positive personal affect.
Classroom Climate
In addition to the DLAM, students also responded to the Classroom Climate Survey (Fraser, 1994). The Classroom Climate Survey is designed to assess four underlying dimensions of classroom climate in late elementary and middle school classrooms. It assesses students’ perceptions of their class and school in terms of satisfaction, friction, teacher support, and cohesiveness. The satisfaction subscale (5-items, α = .72) assesses the extent to which students are involved in the environment and take ownership and interest in class activities (e.g., “The students enjoy their school work in my class”). The friction subscale (4-items, α = .70) identifies the ways in which students experience, respond and resolve conflicts in their environment (e.g., “Students are always fighting with each other in my class”). The teacher support subscale (4-items, α = .75) assesses the extent to which teachers support and help their students (e.g., “The teacher goes out of his way to help students”). The cohesiveness subscale (5-items, α = .63) assesses the nature and intensity of personal relationships within the environment, how students support and help each other (e.g., “All students in my class are friends”). The response format is dichotomous “Yes” = 3, or “No” = 1.
The Classroom Climate Survey is widely used in the literature (Fouts & Myers, 1992; Fraser, 1994; Fraser & Fisher, 1986; Goh & Fraser, 1998). Test-retest reliability ranges from "= .91 (Fraser, 1994) to "= .95 (Goh & Fraser, 1998). Reliability from previous studies has been ranging from "= .76 for friction and α =.90 for teacher support (Fraser, 1994, Goh & Fraser, 1998). These classroom climate scales have been shown to correlate significantly with student achievement (Levy, Brok, Wubbels, & Brekelmans, 2003), self efficacy, and attitude toward school (Dart et al., 1999; Roelofs, Visser, & Terwel, 2003).
Psychological Sense of School Membership
Students were also administered the Psychological Sense of School Membership scale developed by Goodenow (1993). This scale consists of 18 items that assesses students’ perceived belonging or psychological membership in the school environment (e.g., “I feel like a part of (name of school);” “Sometimes I feel as if I don’t belong here”) The response format is on a five-point scale, where 1 = Strongly disagree, and 5 = Strongly agree. All 18 items are summed up to yield each subject’s total score of a sense of belonging. The scale has demonstrated strong psychometric properties (Bosworth, Espelage, DuBay, Dahlberg, & Daytner, 1996; Craig, 1998; Goodenow, 1993).
Bullying
We also administered a bullying survey developed by Bosworth and colleagues, which consists of 23 items that examine the frequency of bullying and victimization incidences in the school (Bosworth, Espelage, & Simon, 1999). Students are asked to rate the number of times they have bullied other students or been victims of bullying (e.g., “I upset other students for the fun of it”; “I fought students I could easily beat”; “Other students picked on me/called me names”). The response format is on a five-point scale, 1 = Never, and 5 = 7 or more times. For analysis purposes items were grouped into two categories based on their thematic content; perpetrators/bully and victim. Perpetrator/bully consisted of 19 items (α = .91) that examined the number of times a student had engaged in bullying others. The victim scale consisted of four items (α = .91) and examined the number of times a student had been a victim of bullying. High scores on this scale indicate more bullying (as either perpetrator or victim). Bullying scale is used extensively in research, with strong internal consistency (Bosworth, Espelage, & Simon, 1999; Bosworth, et al., 1996; Craig, 1998; Craig & Pepler, 1997).
Results
A preliminary analysis was conducted to determine if there were differences among the dependent measures by the number of modules completed. No significant differences were found. A second preliminary analysis was conducted to determine if there were differences among the dependent variables by grade. No grade differences were evident. Consequently fourth and fifth-grade pupils were combined in subsequent analyses. Tests of means were conducted on all measures with treatment group (experimental-control) × gender (male-female) × test (pretest-posttest) MANOVA, with repeated measures on the last factor. All pair wise contrasts were adjusted by the Bonferroni method. Means and standard deviations for all significant effects are reported in Table 4.
Don’t Laugh at Me
The first set of analyses examined the three Don’t Laugh at Me measures (Relational Experience, Stop Ridicule, Personal Affect). The analysis of relational experiences revealed a significant treatment group x test interaction, Pillai Trace = .151, F(1, 56) = 9.94, p = .03, ή2 = .151. Post hoc comparisons showed that the treatment group reported significantly improved relational experiences from pretest to post test than did the control group (recall that lower scores on this scale indicate better relational experience). No other effects were statistically significant. Means and standard deviations for all comparisons are reported in Table 4.
The analysis of Stop Ridicule also revealed a significant Treatment x Test interaction, Pillai Trace = .106, F(1, 55) = 6.51, p = .014, ή2 = .106, indicating that the difference between experimental and control groups was most pronounced at the post test, with experimental students increasing their commitment to stop ridicule from pretest to post test while the importance of this diminished over time for students in the control group. No other effects were statistically significant.
Finally, a significant Test effect was evident for Personal Affect, Pillai Trace = .189, F(1, 55) = 12.82, p = .001, ή2 = .189. Students in both the experimental and control group showed significant improvement in personal affect from pretest to post test. It would appear that the gains were greatest in the experimental group, although the interaction was not statistically significant.
Bullying
Similar analyses were conducted on the bully-perpetrator and bully-victim scales. Significant effects on the bully-perpetrator scale were observed for the treatment × test interaction, Pillai Trace = .235, F(1, 36) = 11.05, p = .002, ή2 = .235 and for the gender × test interaction, Pillai Trace = .135, F(1, 36) = 5.59, p = .023, ή2 = .135. As one can see in Table 4 the incidence of bullying decreased significantly from pretest to post test for students in the experimental group, but bullying increased for students in the control group. Females also showed significant declines in reported incidences of bullying from pretest (M = 36.00) to posttest (M = 27.39) while boys reported an increase of scores from pretest (M = 26.05) to posttest (M = 28.03). The analysis of bullying-victim revealed no significant effects.
Classroom Climate
No significant effects were observed for the analysis of the climate variables satisfaction, cohesion or teacher support. A significant test effect emerged for friction, Pillar = .086, F(1, 48) = 4.52, p = .039, ή2 = .039, indicating that the perception of lack of friction increased significantly, but modestly, from a pretest (M = 6.56, SE = .41) to posttest (M = 7.28, SE = .447) across the experimental and control groups.
Psychological Sense of School Membership
A significant effect emerged for the test main effect, Pillar = .553, F(1, 52) = 64.24, p < .00, ή2 = .55) and for the Treatment x Test interaction, Pillar = .446, F(1, 52) = 41.82, p < .00, ή2 = .45). The main effect indicated that psychological sense of school membership increased from a pretest (M = 62.00, SE = 1.04) to posttest (M = 71.92. SE = .93) across groups. The interpretation of this main effect is qualified by the interaction. As can be seen in Table 4, there was a significant increase in the psychological sense of school membership from pretest to post test in the experimental group Δ = 17.98), but a quite modest increase evident in the control group Δ = 1.81).
Correlational Analysis
One purpose of this study is to contribute to the construct validity of the DLAM assessments. To this end we report in Table 5 the pattern of pretest correlations between the three DLAM scales and other measures used in this study. As one can see, the Personal Affect component is significantly correlated with all of the measures of school membership, school climate and bullying. The Stop Ridicule scale is significantly correlated with two indices of school climate (satisfaction and perceived teacher support), while the Relational Experience scale is significantly and positively correlated with psychological sense of school membership, and negatively correlated with bully-victimization. The correlations among the DLAM scales are modest or nonsignificant, indicating that the three scales are tapping somewhat different dimensions of the perceived classroom experience.
Discussion
The purpose of this study was to examine the effectiveness of the Don’t Laugh at Me curriculum for improving classroom climate, using a quasi-experimental design. To our knowledge this is the first efficacy study to include within-school experimental and control groups and a pretest-posttest methodology. Moreover, we attempted to measure program effectiveness using a broader array of assessments than is currently the practice in DLAM program evaluation. To this end we included assessments of school membership, perceptions of classroom climate and of bullying. We also examined whether the program influences boys and girls similarly. Finally, we present the first evidence on the structural properties and construct validity of the standard evaluation instrument that is distributed with DLAM curricula.
The results for the DLAM intervention are most encouraging. Students in the DLAM experimental group exhibited significant gains in psychological sense of school membership, in relational experiences, in the desire to stop dissing and ridicule, and significant declines in bullying, compared to the control group. Although significant differences were not apparent on the Classroom Climate Survey it is clear that DLAM holds considerable promise in advancing a nurturing and respectful classroom environment, as measured by the DLAM and bullying variables. One possible reason for the lack of significant difference on classroom climate could be the fact that the school is based on an open-concept system. The present results also vindicate the strategy of casting a wide nomological net in the assessment of outcome variables. The DLAM intervention appears to have had similar influences on both boys and girls except on bullying. Consistent with previous research, boys reported bullying their peers significantly more than girls (Craig, 1998; Bosworth et al., 1999).
The psychometric properties of the DLAM assessment instrument were also examined. The common practice heretofore was to analyze outcome data using this scale at the level of individual items. Our goal, then, was to determine if standard scale reduction techniques would reveal parsimonious factors that would facilitate more efficient uses of the scale in data analysis. The present data showed that the DLAM items coalesce into three distinct factors, each of which possesses adequate internal consistency. The construct validity of the scales was also demonstrated in the present study, although much more work will have to be done on this score in future research. The Personal Affect scale, in particular, is pervasively correlated with numerous dimensions of classroom climate.
One perspective that is not heard in these data is the voice of teachers who implemented DLAM. As one teacher reported,
I thought the program allowed students a way to communicate with each other about sensitive issues. I also thought the program was very useful in getting to know my students.
Some teachers felt less prepared to implement effectively at least some parts of the program. For example, some teachers had difficulty dealing with students who felt uncomfortable with some of the exercises. One teacher remarked
I think some children feel put on the spot by all of the “feelings sharing” and it makes them withdraw somewhat if they are told they don’t have to participate, then they feel just as awkward or singled out.
Another sentiment among teachers was that the program ideally should be implemented consistently from the earliest grades. One of the teachers had the following to say about the program,
I feel that this program can only be truly beneficial over time. It needed to be implemented in kindergarten and continued throughout elementary. If there is no consistency, there won’t be substantial progress.
Another teacher said
I think that it is difficult to measure the impact of a program such as this one unless you study the same group of kids over a 3 or 4-year period. At 5th grade, attitudes are already ingrained, and it was difficult to get the kids in the class to take things seriously.
Those that did take things seriously were the ones who did not need the program as much. I hope to have a classroom of 2nd grade next year and I plan on implementing the program with them because I think that it can be beneficial.”
Not everyone concurred with this idea though. Some teachers thought the program was a bit advanced for the lower grades. As one of the teachers reported,
I think that the project is worthwhile, however, I feel that it is not designed for first graders. Many of the lesson plans were modified so it is hard to get a good idea of how effective the program is.
An extended or in-depth training may guide some of these teachers in modifying their lesson plans but still retaining the intended objective of the lesson and the program.
One limitation of the present study is that we are unable to determine if there are certain DLAM modules that are more effective than others. Nor are we able to determine if there is some minimum number of modules that are necessary for program success. All of the treatment classrooms implemented at least four modules in the present study, although future research should attend to this variable more systematically. In addition, we had no control over the class sizes. Only students who returned signed consent forms participated in the study. This small, uneven class sizes calls for caution in interpreting the findings.
The present study also shares the limitations of all quasi-experimental designs. Students were not randomly assigned to treatment and control classrooms, and therefore selection effects cannot be ruled out definitively. Although the design features of the present study have many decided advantages that give one greater confidence in the efficacy of DLAM, future research will still need to aim for randomized trials.
In summary, the results of the present study suggest that implementing at least four modules of the DLAM program has a significant effect on the socio-moral effect climate of elementary school classrooms. Students who participated in the DLAM program reported improved relational experiences within the classroom, a stronger commitment to stop peer ridicule and improved personal affect. Students in DLAM classrooms also reported less bullying and a stronger psychological sense of school membership than did students in control group classrooms. The promise of the DLAM program for improving the climate of tolerance and mutual respect appears to be supported by the present data.
