Please take our brief survey

Blueprints Programs = Positive Youth Development

Return to Search Results

Promising Program Seal

Steps to Respect

Blueprints Program Rating: Promising

A 12-14 week anti-bullying curriculum, plus a grade appropriate literature unit, delivered to third to sixth grade students to reduce bullying and destructive bystander behaviors, increase prosocial beliefs related to bullying, and increase social-emotional skills.

  • Bullying
  • Prosocial with Peers

    Program Type

    • Bullying Prevention
    • School - Environmental Strategies
    • School - Individual Strategies
    • Social Emotional Learning
    • Teacher Training

    Program Setting

    • School

    Continuum of Intervention

    • Universal Prevention (Entire Population)

    A 12-14 week anti-bullying curriculum, plus a grade appropriate literature unit, delivered to third to sixth grade students to reduce bullying and destructive bystander behaviors, increase prosocial beliefs related to bullying, and increase social-emotional skills.

      Population Demographics

      All third to sixth grade students are targeted for intervention.

      Age

      • Late Childhood (5-11) - K/Elementary

      Gender

      • Male and Female

      Race/Ethnicity

      • All Race/Ethnicity

      Race/Ethnicity/Gender Details

      The program is appropriate for all youth, and evaluations have reflected the race/ethnic makeup of the population in general. The strongest study included 52% White, 7% African American, 6% Asian American, 35% Other Mixed Race.

      • Individual
      • School
      Risk Factors
      • Individual: Bullies others*
      Protective Factors
      • Individual: Clear standards for behavior, Problem solving skills, Skills for social interaction*
      • School: Rewards for prosocial involvement in school

      *Risk/Protective Factor was significantly impacted by the program.

      See also: Steps to Respect Logic Model (PDF)

      Steps to Respect is designed to decrease school bullying problems by increasing staff awareness and responsiveness, fostering socially responsible beliefs, and teaching social-emotional skills to counter bullying and promote healthy relationships. All school staff receive an overview of program goals and key features of the program content. Teachers, counselors and administrators receive additional training in how to coach students involved in bullying. The student curriculum consists of skill and literature based lessons presented by third- through sixth-grade teachers over a 12-14 week period. Ten semi-scripted skill lessons focus on social-emotional skills for positive peer relations, emotion management, and recognizing, refusing and reporting bullying behavior. A 45-minute skill lesson and 15-minute follow-up booster are taught weekly. Upon completion of skill lessons, teachers implement a grade-appropriate literature unit, based on existing children's books, providing further opportunities to explore bullying-related themes. Parents are provided information about the program throughout the implementation of the classroom curriculum.

      Steps to Respect is designed to decrease school bullying problems by (a) increasing staff awareness and responsiveness, (b) fostering socially responsible beliefs, and (c) teaching social-emotional skills to counter bullying and promote healthy relationships. Thus, the program also aims to promote skills (e.g., group joining, conflict resolution) associated with general social competence. The program is comprised of a school-wide program guide, staff training, and classroom lessons for students in grades 3-6. All school staff receive an overview of program goals and key features of the program content. Teachers, counselors, and administrators receive additional training in how to coach students involved in bullying. The student curriculum consists of skill and literature based lessons presented by third- through sixth-grade teachers over a 12-14 week period. Ten semi-scripted skill lessons focus on social-emotional skills for positive peer relations, emotion management, and recognizing, refusing and reporting bullying behavior. A 45-minute skill lesson and 15-minute follow-up booster are taught weekly. Upon completion of skill lessons, teachers implement a grade-appropriate literature unit, based on existing children's books, providing further opportunities to explore bullying-related themes. Parents are provided information about the program throughout the implementation of the classroom curriculum.

      The social-ecological model of bullying is the platform on which the Steps to Respect program is based. This model views youth behavior as shaped by multiple factors within nested contextual systems. Steps to Respect addresses both the social context and individual characteristics that are likely to contribute to the processes underlying bullying and victimization. The underlying theory of the program is that peer attitudes, norms, and behaviors influence bullying behavior. Because bullying is a social process that is strongly influenced by the reactions and behaviors of peers, the program seeks to change attitudes about the acceptability of bullying through clearly labeling bullying behavior as unfair and wrong, increasing empathy for students who are bullied, and educating students about their responsibilities as bystanders to bullying.

      • Cognitive Behavioral
      • Normative Education
      • Skill Oriented

      Steps to Respect has been evaluated twice. In the first evaluation, six elementary schools in two school districts were matched and paired, within each district, and each school in the pair was randomly assigned to the bullying intervention or to a control condition. Children in grades 3 to 6 (n = 1126) completed pre- and posttest surveys of behaviors and teachers rated beliefs and attitudes. Observers coded playground behavior of a random sub-sample (n = 544). The second evaluation matched 33 schools in North Central California and then randomly assigned matched pairs to treatment or control conditions. From each of the 33 participating schools, 128 classrooms were randomly selected for data collection: grades 3 (n=52), 4 (n=62) and 5 (n=11). There were also 2 third- and fourth-grade split classrooms and 1 fourth- and fifth-grade split classroom. Implementation ran from December, 2008 and ended May, 2009, at which time posttest data was collected.

      In the original evaluation (Frey, et al., 2005), self-reports showed that students in intervention schools were significantly less accepting of bullying/aggression, felt more responsibility to intervene with friends who bullied, and reported greater adult responsiveness than those in control schools. Most intervention students in grades 5 and 6 perceived the difficulty of responding assertively to bullying to be lower than did their peers in control schools. There were no significant differences between groups on self-reported bullying/aggression or teacher ratings of interaction skills. Effects for self-reported bullying victimization were only marginally significant.

      Increases in observed bullying from pretest to posttest were significantly lower for intervention youth versus controls, but there were only marginally significant declines in bystander encouragement of bullying, and there were no effects for victimization by bullying or non-bullying aggression. Children in the intervention group did show significant decreases in argumentative behavior and increases in agreeable behavior.

      Program replication (Brown, et al., 2011) found significant student outcomes on student climate perceptions in intervention schools compared to control schools, teacher/staff bullying prevention behavior, student bullying intervention behavior, teacher/staff bullying intervention behavior, and positive bystander behavior. Measures of staff and teacher reports indicate significant program effects on school environment and student behavior, specifically social competency and prevalence of physical bullying perpetration.

      Significant outcomes for intervention students, compared to control students, included (Frey, et al., 2005):

      • Less acceptance of bullying/aggression, more responsibility to intervene with friends who bullied, and greater adult responsiveness, according to self-report,
      • Perceived difficulty of responding assertively to bullying lower in grades 5 and 6,
      • Increases in observed bullying were lower,
      • Decreases in observed argumentative behavior and increases in observed agreeable behavior.

      Significant replication outcomes for intervention schools compared to control schools included (Brown, et al., 2011):

      • Greater increases in school anti-bullying policies and strategies,
      • Improved student and staff climate reported by school staff,
      • Students more willing to intervene in bullying situations reported by school staff,
      • Less school bullying-related problems reported by school staff,
      • Lower levels of physical bullying perpetration reported by teachers,
      • Higher levels of student climate and positive bystander behavior reported by students,
      • Less decline in teacher/staff bullying prevention during the year and greater increases in students and teachers/staff willing to intervene in bullying, reported by students.

      Significant Program Effects on Risk and Protective Factors:

      • Higher levels of student social competency reported by teachers.

      No formal test of mediation.

      The adjusted odds ratio for Physical Bullying Perpetration, as assessed by teachers, was 0.61. Other effect sizes from student report measures ranged from 0.115 for Student Bullying Intervention to 0.187 for Student Climate. The average effect size across five social-ecological school context variables was 0.296 (range = 0.212 for staff climate to 0.382 for antibullying policies and strategies).

      The results from this study may be generalized to elementary school youth in grades 3 to 6. The effects are largely consistent across gender and grade.

      In the original study (Frey, et al., 2005), although schools were randomly assigned to treatment or control conditions, attrition rates were substantial. While all students in the intervention schools participated in the program, there is a 36% attrition rate of those eligible (across intervention and control schools), due to failure to obtain parental consent. Posttest sample sizes on individual scales used in analysis ranged from (due to incomplete data) 907 to 919. This represents 47.8% to 48.4% attrition for those who participated (in intervention schools) or who were eligible to participate. Results were mixed. While the observational subgroup was randomly selected and assigned by student, rather than by school as the entire program population, significant effects were present for only one of four measures of aggressive behavior. It is unclear whether observers were blind to condition. Significant effects on self-report items emerged only for attitudes and not for any behavioral measures and teacher-reports produced no effects.

      In the larger replication study (Brown et al., 2011), the following limitations are noted. Effect sizes were small. Also a limitation was the short-term duration of the study. More robust effects and sustainability are not available without longer-term follow-up of the students. Finally, bullying perpetration and victimization (major outcomes) were not significant according to student reports, although teacher reports supported the reduction of physical bullying perpetration among intervention youth, but not non-physical bullying.

      • Blueprints: Promising
      • Crime Solutions: Effective
      • OJJDP Model Programs: Effective

      Brown, E. C., Low, S., Smith, B. H., & Haggerty, K. P. (2011). Outcomes from a school-randomized control trial of Steps to Respect. School Psychology Review, 40(3), 423-443.

      Frey, K. S., Hirschstein, J. L., Snell, J. L., Van Schoiack Edstrom, L., MacKenzie, E. P., & Broderick, C. J. (2005). Reducing playground bullying and supporting beliefs: An experimental trial of the Steps to Respect program. Developmental Psychology, 41(3), 479-491.

      Sally Vilardi
      Director of Marketing & Outreach
      Committee for Children
      2815 Second Avenue, Suite 400
      Seattle, WA 98121
      800-634-4449
      Email: svilardi@cfchildren.org
      www.cfchildren.org

      Study 1

      Frey, K. S., Hirschstein, J. L., Snell, J. L., Van Schoiack Edstrom, L., MacKenzie, E. P., & Broderick, C. J. (2005). Reducing playground bullying and supporting beliefs: An experimental trial of the Steps to Respect program. Developmental Psychology, 41(3), 479-491.

      Study 2

      Brown, E. C., Low, S., Smith, B. H., & Haggerty, K. P. (2011). Outcomes from a school-randomized control trial of Steps to Respect. School Psychology Review, 40(3), 423-443.

      Frey, K. S., Hirschstein, J. L., Snell, J. L., Van Schoiack Edstrom, L., MacKenzie, E. P., & Broderick, C. J. (2005). Reducing playground bullying and supporting beliefs: An experimental trial of the Steps to Respect program. Developmental Psychology, 41 (3), 479-491.

      Evaluation Methodology

      Design: Six elementary schools in the Pacific Northwest were matched in pairs by size, ethnicity, and percent receiving free lunch, within each of two school districts. Schools in each pair were then randomly assigned to the bullying intervention or to a control condition. Conditions for inclusion were that (a) 80% of all staff voted to participate, (b) staff agreed to random assignment to intervention or wait-list control conditions, and (c) principals agreed to refrain from introducing similar interventions during the study. Measures were taken at baseline and after one year of implementation through student surveys, teacher ratings, and playground observations. Although all students in grades 3 to 6 received the program, parental consent was required for their participation in the evaluation. Sixty-four percent (n = 1126) of all students in grades 3 to 6 participated in the evaluation. 36 experimental and 36 control teachers (89.4% of whom were female) also participated in completing assessments.

      A subset of 620 students was randomly selected at pretest for playground observation. This represented 10 students in each control classroom and each grade five and six intervention classroom. In addition, 12 children in each grade 3 and 4 intervention classroom were selected (comprising a longitudinal sample). Playground observations were collected across 2.5 months in both pre- and posttest periods. Each child was observed for 5-minute sessions approximately once a week over the two 10-week periods. Only children meeting a minimum of 40 minutes of observation at pre- and posttest (n = 544) were included in the analyses.

      Sample Characteristics: The sample was equally divided by gender (49.4%) female. Sample sizes for grades 3-6 were comprised of, respectively, 278, 312, 277, and 259 students. Student ethnic background and English proficiency were reported by teachers. The student sample was 9% African American, 12.7% Asian American, 7.0% Hispanic American, 1.3% Native American, and 70.0% European American.

      Measures: Teachers rated students with the Peer-Preferred Social Behavior subscale. The subscale consists of 17 items measuring social behaviors valued by peers. The Student Experience Survey was used to measure students self-reports of their beliefs and behaviors. It is a 60-item questionnaire measuring behavior, attitudes about behaviors, and impressions of adults. Items were combined into the following scales: acceptance of bullying and aggression, bystander responsibility, difficulty of responding assertively, spectator interest, direct bullying/aggression, indirect bullying/aggression, victimization.

      Child observations were combined into five composite codes aggregated across ten sessions. The five composite codes were as follows: (1) bullying; (2) non-bullying aggression; (3) encouragement of bullying; (4) agreeable social behavior; (5) argumentative social behavior. Bullying included physical, verbal, and indirect aggression involving either a) a discernible power imbalance between aggressors and target and/or b) repeated aggression, during the same observation session toward a nonretaliating peer. Encouragement of bullying consisted of focal children laughing or cheering ongoing bullying events or passively watching a bullying event within 15 feet. Nonbullying aggression was coded for physical, verbal, or indirect aggression that did not involve a discernible power imbalance or repeated nonreciprocal aggression. Agreeable social behavior was coded when a child directed neutral or positive acts or statements toward another and argumentative social behavior was coded for nonaggressive, negative acts or statements directed towards another child.

      Analysis: Multilevel models examined student outcomes in terms of individual- and classroom-level variables while concurrently adjusting for shared error due to classroom nesting. The basic model tested for the fixed effects of group and three covariates: gender, grade level, and fall pretest scores.

      Outcomes

      Attrition and Group Equivalence: The researchers were unable to determine if the 36% of children for whom parental consent was not given differed from those who participated in the study. Otherwise, comparisons of pretest student surveys and teacher ratings revealed no significant group differences, nor group by retention status interactions. There were no significant differences between treatment and control conditions in ethnic makeup or English proficiency.

      There were posttest assessments for 1,023 of the students (8.7% attrition from point of consent, but 42% attrition of those who participated). Due to subjects' failure to respond to specific survey items and to the students' who were unavailable for surveying at either pre- or posttest, sample sizes on individual scales ranged from 907 to 919 (18.4 to 19.4% attrition of those who consented; 47.8 to 48.4% attrition of those who participated).

      There were also posttests for 502 of the observation sub-sample (92.4% from the 544 who were analyzed and 81% of the 620 who were randomly selected). Retention rates for the observed group did not differ by condition either. Pretest behaviors of the students in the final sub-sample were compared to the other students for whom they had at least 40 minutes of observation time at pretest. There was one significant interaction of group and retention status for encouragement of bullying. Follow-up tests revealed a significant group difference at pretest in the original sample, but not in the final sample of 544 students.

      Posttest: Self-reports showed students in intervention schools were less accepting of bullying/aggression, felt more responsibility to intervene with friends who bullied, and reported greater adult responsiveness than those in control schools. Most intervention students in grades 5 and 6 perceived the difficulty of responding assertively to bullying to be lower than did their peers in control schools. Intervention boys tended to regard assertive responses as marginally less difficult than did boys in the control group (p < .06). Students in the intervention group tended to report less victimization at posttest than those in the control group (p < .10). There were no significant differences between groups on self-reported bullying/aggression or teacher ratings of interaction skills.

      For observed bullying behaviors, intervention students showed a near-significant decline in bystander encouragement of bullying. Victimization by bullying and non-bullying aggression showed no significant group differences. Children in the intervention group displayed significant decreases in argumentative behavior and increases in agreeable behavior. The observed main effect for change in bullying was due primarily to reductions among those who bullied at pretest and as a result, the evaluators explain the need to discount the small mean difference.

      Brown, E. C., Low, S., Smith, B. H., & Haggerty, K. P. (2011). Outcomes from a school-randomized control trial of Steps to Respect. School Psychology Review, 40(3), 423-443.

      This replication study looks to expand the efficacy and overcome some of the limitations of the original study by using a larger study sample and conducting analysis at the school level.

      Evaluation Methodology

      Design: A randomized control trial was used for this study. Thirty-three schools in three geographic areas in north-central California participated in the study. Schools were chosen based on a set of criteria which included: 1) having broad socioeconomic and racial/ethnic diversity, 2) having an established relationship with the Committee for Children, 3) expressing a strong need or desire for school bullying prevention programming, and 4) not currently implementing a bullying prevention program. The schools were matched based on characteristics of the school environment (i.e., total school enrollment, change in student enrollment from 2006-07, number of teachers) and characteristics of the student population (i.e., percentage of students eligible for free/reduced price lunch, race/ethnicity, percentage of ESL students). Schools within each matched pair were randomly assigned to either the intervention or waitlist control condition who received the program one year after intervention schools. From each of the 33 participating schools, 128 classrooms were randomly selected for data collection (grades 3 (n=52), 4 (n=62) and 5 (n=11); there were also 2 third- and fourth-grade split classrooms and 1 fourth- and fifth-grade split classroom).

      Implementation ran from December, 2008 and ended May, 2009, at which time the posttest was given. Students received 11 semi-scripted lessons focusing on social-emotional skills for positive peer relations; emotion management; and recognizing, refusing, and reporting of bullying behavior. Each weekly lesson was an hour in length, taught over 2-3 days. Once skill lessons were taught, teachers implemented a grade-appropriate literature unit, based on existing children's books, to further explore bullying-related themes. Parents also received scripted information that covered the key concepts and skills and described activities to support their use at home. They also received program information as well as the school's bullying policy and procedures.

      Sample: Twenty-five of the schools were from rural areas, 10% were from small towns, 50% were suburban, and 15% were located in midsized cities. Across all schools, approximately 40% of students received free or reduced-price lunch, there was a mean of 479 students per school, and a mean of 24 teachers per school. All school staff (paid and volunteer) were asked to participate in the study. At pretest, 1,307 participants completed the survey (77% of the total population of school staff). At posttest, 1,296 participants (76%) completed the survey. The majority of staff respondents were teachers (58%) and female (90%). Eighty-five percent of staff identified their ethnicity as "Spanish, Latino, or Hispanic." The average age of staff was 46 years.

      A total of 3,119 students from the selected classrooms were targeted to participate in the study. Of these, 2,940 (94%) completed both pre- and post-test data. The mean age of intervention participants was 8.9 years, with 49% percent male and 52% White. Among the control condition participants, the mean age was 8.9 years, 52% were male, and 54% were White.

      Baseline analysis revealed no significant differences between intervention and control groups on outcome measures except that intervention school students reported higher pretest levels of student bullying intervention than did students in the control schools.

      Measures: In order to assess the social-ecological context of the school environment, school staff completed the School Environment Survey (SES), which asked staff about their perceptions of their school's climate regarding trust, willingness to help, and cooperation among students and school staff; willingness for students and staff to intervene in observed bullying; perceptions of bullying-related problems among students in their school; anti-bullying policies and strategies in their school. Teachers also completed the Teacher Assessment of Student Behavior (TASB), to assess students' classroom behavior, scholastic aptitudes, and student demographic information. Outcome measures included Social Competency, Academic Competency, Academic Achievement, and teacher observed Physical Bullying Perpetration and Nonphysical Bullying. On both staff and teacher measures, scaled scores for outcome measures were created as the mean of all nonmissing items on the scale and were considered missing if more than 67% of items in that scale were missing.

      Students completed the Bullying Prevention Initiative Student Survey, which measured perceptions of the School Climate, School Connectedness, Student Support, Attitudes Against Bullying, Attitudes Towards Bullying Intervention, Student Bullying Intervention and Teacher/Staff Bullying Intervention, Teacher/Staff Bullying Prevention, Positive Bystander Behavior, School Bullying-Related Problems, Bullying Perpetration, and Bullying Victimization. Again, scales were set to missing if they were missing 67% of scale items.

      Analysis: A mixed-model ANCOVA was used to analyze the data from the TASB and student surveys. Hierarchical linear modeling was used for continuous outcomes and hierarchical generalized linear modeling was used for binary, count, and ordered categorical outcomes. Outcome measures were modeled by level, with students nested in classrooms, and classrooms nested within schools. The SES data was analyzed using a two-level modeling strategy with school staff members (level 1) nested within schools (level 2), controlling for gender, race/ethnicity, age, duration of school employment, and position at school (administrative or nonacademic, relative to a teaching position) at level 1 and geographic area at level 2.

      Overall rates of missing data were low (less than 10%), and there were no differences between those who completed measures compared to those who had missing data, except in geographic area, where one area had lower rates of missing data than the other two geographic areas. Missing data were accounted for using multiple imputation analyses.

      Outcomes:

      Implementation Fidelity: Treatment fidelity was monitored through weekly self-report checklists completed by teachers. According to their responses, 83% of teachers in the intervention schools completed 80% of the lessons and 91% of teachers reported teaching 60% of the lessons. Teachers also reported teaching 99.2% of all classroom skills lessons. Approximately 75% of students were exposed to at least 95% of all lessons and 92% of teachers reported completing all objectives.

      Posttest:

      School Context: On the measures of school environment, all outcomes were significant except for staff bullying intervention. The average effect size for the 5 significant measures was .296. The five significant results indicated greater increases in school antibullying policies and strategies, student climate, staff climate, less decrease in student bullying intervention, and a larger decrease in school bullying related problems.

      Teacher Perception of Student Behavior: On the measures of teacher perception of student behavior, there were significant program effects for social competency and prevalence of physical bullying perpetration. Effect sizes were .131 for the standardized difference in Social Competency means and AOR = .609, indicating a reduction of 31% in the likelihood of Physical Bullying Perpetration in intervention schools relative to control schools.

      Student Perceptions: There were significant program effects on 5 of the 13 measured student outcomes. These outcomes included student climate, teacher/staff bullying prevention behavior, student bullying intervention behavior, teacher/staff bullying intervention behavior, and positive bystander behavior. Effect sizes associated with standardized differences in adjusted means between intervention and control schools for significant outcomes ranged from .115 for Student Bullying Intervention to .187 for Student Climate.

      Limitations

      Small effect sizes may be seen as a limitation on the significance of the research findings. Also a limitation was the short-term duration of the study. More robust effects and sustainability are not available without longer-term follow-up of the students. Finally, bullying perpetration and victimization (major outcomes) were not significant according to student report, although teacher report supported the reduction of physical bullying perpetration among intervention youth, but not non-physical bullying.

      Video

      http://www.cfchildren.org/steps-to-respect.aspx