© 2014 American Psychological Association 1045-3830/14/$12.00 http://dx.doi.org/10.1037/spq0000070
School Psychology Quarterly 2014, Vol. 29, No. 3, 272-286
Decreasing Bullying Behaviors in Middle School: Expect Respect Rhonda N. T. Nese and Robert H. Horner
Celeste Rossetto Dickey
University of Oregon
Placer County Office of Education, Auburn, California
Brianna Stiller and Anne Tomlanovich Eugene School District 4J, Eugene, Oregon A nonconcurrent multiple baseline across 3 middle schools was used to assess the impact that teaching all students to follow the Bullying and Harassment Prevention in Positive Behavior Support: Expect Respect intervention had on bullying behaviors. The 3 schools were using School-Wide Positive Behavioral Interventions and Supports and as part of this effort all students in each school had been taught to discriminate “respectful” versus “nonrespectful” behavior. The Expect Respect intervention in cluded, 3 1-hr lessons over a 6-month period to learn (a) how to signal “stop” when encountering nonrespectful behavior, (b) how to follow a “stopping routine” when asked to stop, (c) how to utilize the “bystander routine” when you are a witness to disrespectful behavior that does not stop even after the perpetrator has been asked to, and (d) how to recruit adult support if bullying behaviors endured. Before intervention implementation, 8 students from each school were engaged in focus groups to define the perceived need for bully prevention, and the bully prevention routines that best fit the social culture of their school. Data assessing the fidelity of intervention implemen tation indicate that the program was used with high fidelity and that in each of the 3 schools a reduction of verbal or physical aggression in the cafeteria was documented via direct observation. No consistent patterns were found with respect to the conditional probabilities that bystanders or recipients of bullying would use the bully prevention routines. No consistent changes were reported in student pre-post rating of school climate. Keywords: bullying prevention, middle school intervention, PB1S, universal-tier
Bullying remains a major concern in schools (Swearer & Espelage, 2011). Behaviors such as physical aggression, taunting, teasing, name-
This article was published Online First May 12, 2014. Rhonda N. T. Nese and Robert H. Homer, Educational and Community Supports, University of Oregon; Celeste Rossetto Dickey, Placer County Office of Education, Au burn, California; Brianna Stiller and Anne Tomlanovich, Eugene School District 4J, Eugene, Oregon. This research was supported by the Office of Special Education Program s US D epartm ent of Education (H326S9S0003). Opinions expressed herein are those of the authors and do not necessarily reflect the position of the US Department of Education, and such endorsements should not be inferred. Correspondence concerning this article should be ad dressed to Rhonda N. T. Nese, Educational and Community Supports, University of Oregon, 1235 University of Oregon, Eugene, OR 97403-1235. E-mail: [email protected]
calling, threatening, social exclusion, and ha rassment have negative effects both socially and academically for students engaging in the be haviors and those targeted (Arseneault et al., 2006; Gini & Pozzoli, 2009). Prevalence esti mates suggest that bullying behavior is not lim ited to an isolated few students, but occurs across all strata and subgroups in most schools (Bradshaw & Waasdorp, 2011). A quarter to a third of students in elementary and middle schools report that they engage in or experience the effects of bullying behaviors (Cook, Wil liams, Guerra, & Kim, 2010). Boys are more likely than girls to perform bullying behaviors (Cook et al., 2010), and bullying is most prev alent in middle school (Nansel et al., 2001; Pellegrini et al., 2010), especially at points of transition from elementary to middle school, and middle school to high school. 272
MIDDLE SCHOOL BULLY PREVENTION
Many programs have been proposed to pre vent or reduce bullying behavior in schools (Vernberg & Biggs, 2010; Waasdorp, Brad shaw, & Leaf, 2012). Although evidence sup ports reduction in student-reported bullying be haviors for some programs (Olweus, 1995, 1997), the support for formal strategies that reduce observed patterns of bullying behavior is limited (Merrell, Gueldner, Ross, & Isava, 2008; Vernberg & Biggs, 2010). In fact there is growing concern that programs centered around teaching students about bullying, and then en couraging them to reduce those behaviors, may inadvertently result in increased incidences of bullying (Dishion, McCord, & Poulin, 1999). Ross and Horner (2009) report one approach to bully prevention that follows guidelines en couraged by Bradshaw and Waasdorp (2011) to (a) build a positive, school-wide social climate, (b) teach students constructive responses to minimize the social rewards of bullying behav ior, and (c) provide consistent staff responses to instances of bullying. Ross and Homer (2009) examined bullying behaviors across three ele mentary schools within a multiple baseline re search design in which two children from each school who were identified as engaging in bul lying behaviors were observed on the play ground over 60 school days. Data were col lected on the incidence of verbal or physical aggression toward peers, and the responses from recipients and bystanders following an instance of verbal or physical aggression. The intervention, Bully Prevention in Positive Be havior Supports (BP-PBS: Ross, Horner, & Stiller, 2008), involved teaching all students first to identify “respectful” behavior. The term “bullying,” however, was never part of the stu dent instruction. Building from the conceptual logic that a high proportion of bullying behav iors endure because they result in social atten tion from recipients or bystanders, the interven tion then focused on teaching all students in the school three routines: (a) a “stop” signal that they could use either as a recipient or bystander when they encountered verbal or physical ag gression, (b) a “stopping routine” that should be followed if someone asks you to stop, and (c) a protocol for obtaining adult support if someone does not stop when asked (Ross et al., 2008). Results from the study demonstrated a 72% reduction in verbal and/or physical aggression on the playground. In addition, the descriptive
273
data indicated an increased use of the “stop” routine and decreased conditional probability that verbal or physical aggression would be followed by peer social rewards. The present study builds from these efforts to develop a social-ecological approach to bully prevention based on behavioral theory (Brad shaw & Waasdorp, 2011; Waasdorp et al., 2012). The specific goal of the present analysis was to assess the effectiveness of the Ross et al. (2008) approach when it was adapted to fit the social context of middle schools. The Bullying and Harassment Prevention in Positive Behav ior Support: Expect Respect program (Expect Respect; Stiller, Nese, Tomlonovich, Horner & Ross, 2013) incorporates the same basic instruc tional goals used by Ross et al. (2008) but with three major adaptations: First, a student focus group component was built into Expect Respect to obtain student input about (a) the student perception about the significance of aggression, harassment, and intimidation in the school; (b) a socially acceptable way to deliver a “stop” mes sage that could be used by all students; (c) a socially acceptable “stopping routine” that could be taught to all students; (d) a “bystander” routine to be used by witnesses of disrespectful behaviors; and (e) the parameters of adult re sponses when instances of bullying are reported by students. The purpose of the student focus group is to adapt the messages of the program to the local school context, as research has shown that information gathered from focus groups help to identify existing cultural norms and practices needed for improving service delivery (Baker-Henningham, 2011; Bronheim, 2013). The second major adaptation was to build three 1-hour lesson plans that emphasized op portunities for students to shape how the content of the four routines (stop, stopping, bystander, and seeking support) fit with the language, cus toms, and culture of their school. Lesson one, led by school personnel in student groups sim ilar to class sizes, presents students with a basic review of the importance of respectful behavior, how to deliver the “stop” signal if the disre spectful behavior is encountered, how to use the “stopping routine” if someone asks you to stop, how to use the “bystander routine” if you wit ness someone using the stop signal and the perpetrator does not stop, and the process and protocol for seeking supports from adults if disrespectful behaviors endure. Lessons two
274
NESE, HORNER, DICKEY, STILLER, AND TOMLANOVICH
and three are designed to be more studentguided and frame problem-solving situations where the basic “stop,” “stopping,” “by stander,” and “seeking support” routines are being applied with and without success. The third adaptation focused on adult behav ior. Training and coaching for faculty and staff was built into the implementation of Expect Respect. Initial field-testing identified that a ma jor concern of middle school students was the consistency of adult responses to instances of bullying behavior. The goal of this portion of the program was to both maximize consistency and ensure that students were informed about what they could expect when reporting an in stance of disrespectful behavior. The primary research question guiding the study was whether Expect Respect, when imple mented with fidelity, was functionally related to an overall reduction in the level of verbal or physical aggression observed in uncontrolled settings like the cafeteria. Secondary, descrip tive questions focused on the extent to which the conditional probability of bystander and re cipient responses to verbal or physical aggres sion became less rewarding after introduction of the program, and if student perceptions of the overall safety of the school, and the quality of social interactions within the school changed after introduction of the program.
Todd et al., 2003). Adequate implementation of SWPBIS includes the following: establishing and teaching behavior expectations, establish ing and teaching a school-wide strategy for re warding appropriate behavior, and establishing a continuum of responses to problem behaviors (Homer, Sugai, Todd, & Lewis-Palmer, 2005). All sixth-, seventh-, and eighth-grade stu dents in the three participating middle schools received training on Expect Respect, following passive parental consent. Parents were notified that direct observations would be taking place during lunchtime and that no individual student would be targeted for observation. Parents were also informed that they could choose to remove their student from participation in the Expect Respect intervention or from lunchtime direct observations at any time; however, no parent declined consent for their student’s participa tion. To protect student confidentiality, no iden tifying student information was recorded on the daily direct observation sheets other than stu dent sex. Additionally, students provided writ ten assent before completing any surveys. The parental consent, student assent, and all materi als and procedures were approved by our Insti tutional Review Board (IRB) as well as the participating school districts.
Method
Data were collected between November and May of the 2011-2012 school year. The pri mary data collected focused on direct observa tion of bullying behaviors, as well as the behav iors of both recipients and bystanders immediately following an instance of bullying. In addition, survey data were collected to assess students’ perception of “school climate,” and self-assessment checklists completed by school staff were collected to assess the fidelity with which Expect Respect was implemented. Bullying behaviors. The primary depen dent measure was the frequency of physical or verbal aggression during 20-min direct observa tions in the cafeteria during lunchtime. An ad ministrator at each school identified a section of the cafeteria in which bullying behaviors were most likely to be reported. This was often an area with low adult supervision and high student foot traffic. During baseline and intervention, trained University data collectors positioned themselves near the identified area so that they
Setting and Participants Three middle schools in two Pacific North west school districts participated in this study during the 2011-2012 school year. Participating schools were attended by 508, 511, and 691 students, and eligibility for free or reducedprice lunch ranged from 43% to 65% of stu dents. Additionally, the student population at all three schools was primarily White, with stu dents of color ranging from 26% to 32%. To be eligible for participation, selected schools were made up of Grades 6 through 8, had not re ceived training on or utilization of BP-PBS or Expect Respect before this study, and had im plemented School-Wide Positive Behavioral In terventions and Supports (SWPBIS: Horner, Sugai, & Anderson, 2010) for at least three years with adequate fidelity (e.g., ^80% Total Score on the School-Wide Evaluation Tool:
Measurement
MIDDLE SCHOOL BULLY PREVENTION
could see and hear all interactions that took place between students. All observations were conducted in this identified area at the same time of day throughout the study. No student was singled-out for observation in this study; therefore the behaviors of any student that in volved aggression in the observed area were recorded. For the entire 20-min observation pe riod, observers recorded whether a bullying event took place and whether the event involved physical aggression, verbal aggression, or both. Physical aggression was operationally defined as “hitting, biting, kicking, chocking, stealing, throwing objects, or restricting freedom of movement” (Ross & Homer, 2009, p. 751). Verbal aggression was operationally defined as “the direction of negative communication either verbal or gestural, toward one or more focus students including teasing, taunting, threaten ing, negative body language, or negative ges tures” (Ross & Horner, 2009, p. 751). Two to four direct observation sessions per week took place at each participating school, with no more than five days between observation sessions. This five-day span without data collection only took place when schools were out-of-session for Spring Break. Recipient and bystander responses to bul lying behaviors. The second and third depen dent measures were recipient and bystander re sponses to bullying behaviors. During each observation session, observers recorded how re cipients and bystanders of bullying behaviors responded to the perpetrator. A recipient was operationally defined as a student receiving the bullying behaviors. A bystander was operation ally defined as a student located within three minutes of the bullying event that is witnessing the bullying behaviors but does not take part. Once an instance of bullying occurred, observ ers would document whether the recipient and any bystander responded by telling the perpe trator to “Stop,” as taught in the Expect Respect curriculum, or by either ignoring the perpetra tor, or providing social attention (e.g., laughing, yelling back) to the perpetrator within five sec onds of each instance of a bullying behavior. Observers also recorded when recipients and bystanders provided no response. Each recipient and bystander response was mutually exclusive, and all responses to an incident were recorded. For example, in a case where a bystander said
275
“stop” but then started to yell at the perpetrator, both “stop” and “attention” were recorded. Observer training and interobserver agree ment. Five graduate students from a Pacific Northwest university served as trained observ ers for this study. All observers were blind to the goals of the present study and when the intervention was started at each school. Before beginning data collection, observers were trained to an 85% interobserver agreement cri terion on each target behavior. Observers par ticipated in three training sessions with the first author that involved reviewing the measures, procedures, and data collection tool. During the first session, observers became familiar with the operational definitions of each behavior that were coded, and began practicing coding sam ple videotapes as a group. The second and third training sessions involved reviewing the opera tional definitions, having specific questions an swered, and coding more sample videotapes. Total agreement was at or greater than 85% for all target behaviors before the observers began coding baseline data for the present study. Ses sions coded within the study were assessed for interobserver agreement on 40% of observa tions during baseline and intervention phases for each of the three schools. During these ses sions, a second observer independently coded during the same lunchtime observation. Total agreement was calculated by dividing the num ber of events that both observers agreed a bul lying behavior occurred by the total number of events observed, and multiplying that number by 100%. Table 1 displays total interobserver agreement across all three schools, which ex ceeded 85% in both baseline and Expect Re spect phases. Fidelity of implementation. Fidelity of Expect Respect implementation was assessed Table 1 Average (Range) Interobserver Agreement
School 1 2 3
Bullying behaviors
Baseline total agreement
Intervention total agreement
Physical Verbal Physical Verbal Physical Verbal
98% 99% 99% 99% 95% 92%
96% 95% 96% 97% 97% 98%
276
NESE, HORNER, DICKEY, STILLER, AND TOMLANOVICH
using a 5-item fidelity checklist that measured Procedures staff adherence to the program components. Experimental design. A nonconcurrent School staff members completed fidelity check multiple baseline design (Kazdin, 2011) across lists twice: between weeks 1 and 3 and weeks 6 three middle schools was used to examine the and 8 after launching the intervention. Items on effectiveness of Expect Respect to reduce bul the fidelity checklist include the extent to which lying behaviors and improve bystander and re teacher and school staff (a) interrupt bullying cipient responses to bullying behaviors. The behaviors they observed in or around the design involved two phases: baseline and Ex school, (b) reflectively listen when students re pect Respect implementation. The total duration port incidents of bullying, (c) praise students for of the study was 6 months. School 1 partici reporting such incidents, (d) encourage students to use an assertive word or phrase to interrupt pated in the study between November and May, bullying incidents or walk away, and (e) check School 2 participated between February and back in with students following a report to de May, and School 3 participated between Janu termine if the issue had been resolved. Scores ary and May. For each school, a baseline period on the checklist were recorded on a Likert scale was followed by a staff training on the curric from 1 to 4, with higher scores indicating better ulum, a meeting with a focus group of students to define the routines to be taught, and a wholeadherence to Expect Respect procedures. School climate survey. Students’ percep school presentation of Lesson 1, with Lessons 2 tions of their school climate were assessed dur and 3 taught one to seven weeks later after ing the first week of implementation of the scheduling availability defined for each school. Baseline. In baseline, students were ob Expect Respect intervention and again at the end served during unstructured lunchtime in the of the study. The researchers developed a school cafeteria. No teaching procedures to 9-item survey that assessed students’ percep modify any bullying behaviors were in effect. A tions of their school safety, respectfulness among students and teachers toward one an minimum of five data points were collected for other, school members’ interest in creating a each school during the baseline phase, and in safer school environment, students’ use of strat cidents of bullying behaviors and conditional egies to interrupt disrespectful behavior, and probabilities of recipient and bystander re the school’s dedication to bullying preven sponses were recorded. Each data point repre tion. The items on the survey were developed sents a single 20-min observation session. Expect Respect. Implementation of Expect from the targets of the Expect Respect manual, Respect involved three components. In the first and therefore focus on the extent to which stu dents perceive each other as being treated with component, the first or second author and a respect, and if they have a clear idea about what school coordinator provided training to the en to do when they encounter disrespectful behav tire school staff on the Expect Respect curricu ior. Five of the nine items were rated on a lum. Training consisted of a 1-hr workshop on 4-point scale (1 = strongly disagree, 2 = dis the intervention components, including the agree, 3 = agree, 4 = strongly agree), with classroom lessons, methods of supervising and higher scores indicating a more favorable im interrupting inappropriate behaviors in unstruc pression of the school climate (see Figure 3), tured environments, appropriate adult responses and the other four items could either be an to students’ reports of bullying and harassment, swered Yes or No (see Figure 4). On average, and staff responsibilities for supporting students the survey took students approximately five through the reporting and responding process. minutes to complete. All students at the three Staff members were given the opportunity to participating schools were asked to complete practice delivering the classroom lessons during the survey online, and before beginning the sur staff training. Once the entire school staff had vey, they were prompted on the computer screen been trained on the Expect Respect curriculum, to elect to complete it or not complete. Each implementation moved on to the second com school chose when and how they wanted the stu ponent. The second component involved the organi dents to complete the survey, with most classes circulating through the computer lab during an zation of a student focus group, which com prised approximately 8 to 12 students who eiintervention or flex block period.
MIDDLE SCHOOL BULLY PREVENTION
ther volunteered or were nominated by their teachers. The student focus group, along with the staff members overseeing implementation of Expect Respect, held a 1-hr meeting where the students selected the school-wide STOP sig nal, discussed and practiced the “stop,” “stop ping,” “bystander,” and “seeking support” strat egies, and discussed a plan for gathering student buy-in to the program. The student focus group is a fundamental component of the Expect Re spect intervention, as students have noted that for a program to be successful at the middle school level, middle school students need to be actively involved in planning and administering the lessons, and organizing events related to improving their school climate. In the third component, which occurred in the week following staff training, teachers, school staff, and student focus group members used the three Expect Respect lesson plans to provide training for all students in Grades 6 through 8 during 1-hr intervention block class periods. The administrators at each school were able to decide how far apart they wanted to space the three lessons out, based on their building sched ules and holidays. However, all schools admin istered lesson one within one week of Expect Respect implementation. During lesson one, students were led through a basic review of the importance of respectful behavior and the school-wide expectations around being respect ful. After this review was complete, students were then instructed on the four strategies for interrupting disrespectful behavior: how to de liver the “stop” signal if the disrespectful be havior is encountered, how to use the “stopping routine” if someone asks you to stop, how to use the “bystander routine” if you witness someone using the stop signal and the perpetrator does not stop, and the process and protocol for seek ing supports from adults if disrespectful behav iors endure. Lesson two provided an opportu nity for students to practice the four routines in a simulated activity on a school bus, and to problem solve around situations where the rou tines were and were not successful. Lesson three had two components. The first component focused on identifying ways of interrupting and removing social attention given to disrespectful behaviors identified in online videos. During the second component, students were guided through completing a “seeking support” work sheet for their day planners, in which they iden
277
tified individuals within the school and outside the school they could go to for support. They also wrote out three strategies to use when they are either the recipient or the witness of disre spectful behaviors. After all three 1-hr lessons were imple mented, school staff utilized the “Repeat and Repair” activity as a 15-min refresher for stu dents and staff on the trained routines. This lesson was either delivered a few times through out the school year or at the end of implemen tation. Results For each of the three participating schools, the effects of the Expect Respect intervention are reported on (a) fidelity of implementation, (b) the incidents of physical and verbal aggres sion, (c) recipient and bystander response prob abilities, and (d) school climate. Note that there were no points of missing data in this study. Fidelity of Implementation Fidelity of Expect Respect implementation was assessed by evaluating a convenience sam ple of staff self-reported adherence to program components. A total of 82 staff members across all three schools completed two fidelity check lists during implementation of Expect Respect. All staff members who attended their school’s monthly staff meeting when fidelity data were being collected were given the fidelity checklist to complete (no descriptive data were collected on the teachers). The overall median was cal culated for each item by determining the per centage of staff who endorsed a 3 (Mostly) or 4 (Always), and then determining the median value as well as the range of values across the averages. Results indicated that staff members reported high adherence to the proper proce dures when responding to witnessed or reported incidents of bullying, and that their adherence increased over time, reporting an overall me dian of 83% at Time 1 (range = 44% to 96%) and 91% at Time 2 (range = 82% to 100%). Specifically, a median of 82% of staff members across schools indicated that they “interrupted bullying behaviors they observed” (range = 67% to 94%), 91% stated that they “reflectively listened when students report incidents of bul lying” (range = 81% to 100%), 91% reported
278
NESE, HORNER, DICKEY, STILLER, AND TOMLANOVICH
that they “praised students for reporting such incidents” (range = 75% to 96%), 88% indi cated that they “encouraged students to use an assertive word or phrase to interrupt bullying incidents or walk away” (range = 69% to 91%), and a median of 85% of staff members stated that they “checked back in with students fol lowing a reported incident” (range = 44% to
88%). Incidents of Physical and Verbal Aggression The incidents of physical and verbal aggres sion across all three schools are presented in Figure 1. Students in each middle school en gaged in physical, and/or verbal aggression dur ing Baseline. For School 1, baseline observa tions scored with aggression averaged 4 incidents per 20-min session (range = 2 to 7), with high variability and a decreasing trend. For School 2, baseline aggression averaged 2.44 events per 20-min observation (range 1-4) but with a moderate level of variability, and stable trend. School 3 students were observed to en gage in aggression an average of 2.37 times per observation (range 0-5) with higher variability and a slight increasing trend. Each school demonstrated reduction in rates of physical and verbal aggression after introduc tion of the Expect Respect intervention. Schools 1, 2, and 3 averaged .89, .88, and .64 incidents of aggression respectively per 20-min observa tion across the Expect Respect phase. Note, however, an unexplained increase in the level of aggression in School 1 between Sessions 15-19 in conjunction with the second Expect Respect lesson. Levels of aggression later in the inter vention phase for each school (especially after Spring Break) were at or near zero. Taken to gether the data from the Expect Respect phases represent a 78%, 64%, and 73% reduction in level of aggression, respectively, for Schools 1, 2, and 3, with reduction in variability also noted for each school. Recipient and Bystander Response Probabilities Information on recipient and bystander re sponses to bullying behaviors was recorded at the time of each aggressive incident. Figure 2 presents the conditional probabilities of these
responses prior to and following implementa tion of the Expect Respect curriculum. During baseline at School 1, when an inci dent of bullying occurred, recipients said “Stop” 5% of the time, rewarded (reinforced) bullying behaviors 45% of the time, and ignored bullying behaviors 35% of the time. After implementa tion of Expect Respect, recipients increased their delivery of the “Stop” phrase to 19% of the time, increased rewarding of bullying behaviors at 58% of the time, and decreased ignoring of bullying behaviors to 29% of the time. Note that sessions where no bullying behaviors were ob served did not contribute to the conditional probabilities. Bystanders at School 1 said “Stop” during baseline 5% of the time, rewarded bullying be haviors 35% of the time, and ignored bullying behaviors 35% of the time. During the interven tion phase, bystanders decreased their delivery of the “Stop” phrase to 3% of the time. Delivery of a rewarding response after bullying behaviors increased to 55% of the time, and ignoring maintained at 35%. During baseline at School 2, recipients said “Stop” 32% of the time, rewarded bullying be haviors 55% of the time, and ignored bullying behaviors 36% of the time. After implementa tion of Expect Respect, recipients decreased their delivery of the “Stop” phrase to 14% of the time, increased their level of reward for bully ing behaviors to 71% of the time, and decreased ignoring of bullying behaviors to 29% of the time. Bystanders at School 2 said “Stop” during baseline 0% of the time, as they were not ob served demonstrating this behavior during any observation session. Bystanders rewarded bul lying behaviors 45% of the time, and ignored bullying behaviors 45% of the time. During the intervention phase, bystanders increased their delivery of the “Stop” phrase to 7% of the time. Delivery of a rewarding response following bul lying behaviors increased to 50% of the time, and ignoring decreased to 43% of the time. During baseline at School 3, recipients said “Stop” 29% of the time, rewarded bullying be haviors 44% of the time, and ignored bullying behaviors 33% of the time. After implementa tion of Expect Respect, recipients increased their delivery of the “Stop” phrase to 33% of the time, maintained the same level of reward for
MIDDLE SCHOOL BULLY PREVENTION
Baseline
279
Expect Respect Intervention
Sessions Figure 1. Incidents of physical and verbal aggression during baseline and PBPBS-ER implementation for all three participating schools during 20-min observations. Phase change lines indicate the date of Lesson 1, and asterisks indicate the date of Lessons 2 and 3 per school. The double hash marks indicate when each school was out-of-session for Spring Break.
bullying behaviors at 44%, and decreased ignor ing of bullying behaviors to 22% of the time. Bystanders at School 3 said “Stop” during baseline 4% of the time, rewarded bullying be haviors 56% of the time, and ignored bullying
behaviors 42% of the time. During the interven tion phase, bystanders’ delivery of the “Stop” phrase decreased to 0% of the time, as they were not observed demonstrating this behavior during any observation session. Delivery of a
280
NESE, HORNER, DICKEY, STILLER, AND TOMLANOVICH
Recipient Response Probabilities
STOP
Reward
Bystander Response Probabilities
Ignore
STOP
Reward
Ignore
Responses Figure 2. The conditional probabilities of recipient and bystander responses to bullying behaviors during 20-min observations during lunchtime in the cafeteria, [graybox] Preinter vention; [blackbox] Postintervention.
rewarding response after bullying behaviors maintained at 56% of the time, and ignoring decreased to 33%. School Climate Students’ perceptions of their school climate were assessed pre- and postintervention via a brief online survey. Table 2 displays the demo graphic information for students who completed
the survey, as well as the sample sizes and the number of students who declined to participate. Pre-post samples varied by size and demo graphic data as a result of logistics and sched uling constraints related to circulating students through the computer lab to complete the survey (see Table 2). It should also be noted that de mographic data were not available for School 1 during the preintervention survey. The pro-
MIDDLE SCHOOL BULLY PREVENTION
281
Table 2 Frequency (%) o f Demographic Characteristics fo r Students Who Were Administered the School Climate Survey School 1
School 3
School 2
Characteristic
Pre
Post
Pre
Post
Pre
Post
Completed Declined Sex Female Male Not indicated Race White Non-White Not indicated
309 (96) 12(4)
353 (93) 26(7)
53 (98) 1(2)
70 (93) 5(7)
234 (98) 6(2)
81 (91) 8(9)
166 (47) 176 (50) 11 (3)
29 (55) 24 (45) 0(0)
40 (57) 30 (43) 0(0)
122 (52) 112 (48) 0(0)
29 (36) 51 (63) 1 (1)
240 (68) 89 (25) 24(7)
42 (79) 9(17) 2(4)
45(64) 20 (29) 5(7)
167(71) 59 (25) 8(3)
47 (58) 33 (41) 1(1)
— — —
— —
—
Note. Preintervention demographic data were not available for School 1.
portion of students who indicated Strongly Agree or Agree to the first five items of the school climate survey can be found in Figure 3, and the percentage of students who indi cated Yes to the last four items of the survey can be found in Figure 4. On average, students across all schools rated their school climate as high before and after completion of the Expect Respect program. Be fore intervention implementation, the mean pro portion of students across all three schools who indicated Strongly Agree or Agree to the fol lowing questions were as follows: Do you feel safe? was .93, Do other students treat you re spectfully? was .78, Do you treat other students respectfully? was .95, Do adults treat you re spectfully? was .90, and Do you treat adults in your school respectfully? was .95. After inter vention implementation, the mean proportion of students across all three schools who indicated Strongly Agree or Agree to the following ques tions were as follows: Do you feel safe? was .91, Do other students treat you respectfully? was .80, Do you treat other students respect fully? was .95, Do adults treat you respectfully? was .86, and Do you treat adults in your school respectfully? was .93. Figure 3 depicts the re sults for student responses at each school. Al though these data indicate a small increase in the proportion of students who reported that other students treated them respectfully, a twoproportion z test revealed no statistically signif icant pre-post differences at the p < .05 level (see Table 3).
Preintervention mean percentages across all 596 students who indicated Yes to the following questions were as follows: Has anyone treated you disrespectfully? was 53%, Have you asked someone to STOP? was 59%, Has anyone asked you to STOP? was 20%, and Have you seen someone else treated disrespectfully? was 65%. Postintervention mean percentages across all 504 students who indicated Yes to the following questions were as follows: Has anyone treated you disrespectfully? was 56%, Have you asked someone to STOP? was 59%, Has anyone asked you to STOP? was 22%, and Have you seen someone else treated disrespectfully? was 62%. Figure 4 depicts the results for student re sponses at each school. Although these data indicate a small decrease in the percentage of students who reported seeing others being treated disrespectfully and a small increase in other students using a school-wide STOP phrase, a two-proportion z test revealed no sta tistically significant pre-post differences at the p < .05 level (see Table 3). Finally, 53% of students across all schools rated Expect Respect as being either very helpful or helpful for their schools, and 60% of students felt that Expect Respect should be taught again next year. Discussion The present study focused on a program de signed to build the social-ecological conditions recommended by Swearer, Espelage, Vaillancourt, and Hymel (2010) and Waasdorp et al.
282
NESE, HORNER, DICKEY, STILLER, AND TOMLANOVICH
D o y o u fe e l s a fe ? D£
•a
a a H
J=
School 2
-
■a
s
ce
■8 & i e t—
School Psychology Quarterly 2014, Vol. 29, No. 3, 272-286
Decreasing Bullying Behaviors in Middle School: Expect Respect Rhonda N. T. Nese and Robert H. Horner
Celeste Rossetto Dickey
University of Oregon
Placer County Office of Education, Auburn, California
Brianna Stiller and Anne Tomlanovich Eugene School District 4J, Eugene, Oregon A nonconcurrent multiple baseline across 3 middle schools was used to assess the impact that teaching all students to follow the Bullying and Harassment Prevention in Positive Behavior Support: Expect Respect intervention had on bullying behaviors. The 3 schools were using School-Wide Positive Behavioral Interventions and Supports and as part of this effort all students in each school had been taught to discriminate “respectful” versus “nonrespectful” behavior. The Expect Respect intervention in cluded, 3 1-hr lessons over a 6-month period to learn (a) how to signal “stop” when encountering nonrespectful behavior, (b) how to follow a “stopping routine” when asked to stop, (c) how to utilize the “bystander routine” when you are a witness to disrespectful behavior that does not stop even after the perpetrator has been asked to, and (d) how to recruit adult support if bullying behaviors endured. Before intervention implementation, 8 students from each school were engaged in focus groups to define the perceived need for bully prevention, and the bully prevention routines that best fit the social culture of their school. Data assessing the fidelity of intervention implemen tation indicate that the program was used with high fidelity and that in each of the 3 schools a reduction of verbal or physical aggression in the cafeteria was documented via direct observation. No consistent patterns were found with respect to the conditional probabilities that bystanders or recipients of bullying would use the bully prevention routines. No consistent changes were reported in student pre-post rating of school climate. Keywords: bullying prevention, middle school intervention, PB1S, universal-tier
Bullying remains a major concern in schools (Swearer & Espelage, 2011). Behaviors such as physical aggression, taunting, teasing, name-
This article was published Online First May 12, 2014. Rhonda N. T. Nese and Robert H. Homer, Educational and Community Supports, University of Oregon; Celeste Rossetto Dickey, Placer County Office of Education, Au burn, California; Brianna Stiller and Anne Tomlanovich, Eugene School District 4J, Eugene, Oregon. This research was supported by the Office of Special Education Program s US D epartm ent of Education (H326S9S0003). Opinions expressed herein are those of the authors and do not necessarily reflect the position of the US Department of Education, and such endorsements should not be inferred. Correspondence concerning this article should be ad dressed to Rhonda N. T. Nese, Educational and Community Supports, University of Oregon, 1235 University of Oregon, Eugene, OR 97403-1235. E-mail: [email protected]
calling, threatening, social exclusion, and ha rassment have negative effects both socially and academically for students engaging in the be haviors and those targeted (Arseneault et al., 2006; Gini & Pozzoli, 2009). Prevalence esti mates suggest that bullying behavior is not lim ited to an isolated few students, but occurs across all strata and subgroups in most schools (Bradshaw & Waasdorp, 2011). A quarter to a third of students in elementary and middle schools report that they engage in or experience the effects of bullying behaviors (Cook, Wil liams, Guerra, & Kim, 2010). Boys are more likely than girls to perform bullying behaviors (Cook et al., 2010), and bullying is most prev alent in middle school (Nansel et al., 2001; Pellegrini et al., 2010), especially at points of transition from elementary to middle school, and middle school to high school. 272
MIDDLE SCHOOL BULLY PREVENTION
Many programs have been proposed to pre vent or reduce bullying behavior in schools (Vernberg & Biggs, 2010; Waasdorp, Brad shaw, & Leaf, 2012). Although evidence sup ports reduction in student-reported bullying be haviors for some programs (Olweus, 1995, 1997), the support for formal strategies that reduce observed patterns of bullying behavior is limited (Merrell, Gueldner, Ross, & Isava, 2008; Vernberg & Biggs, 2010). In fact there is growing concern that programs centered around teaching students about bullying, and then en couraging them to reduce those behaviors, may inadvertently result in increased incidences of bullying (Dishion, McCord, & Poulin, 1999). Ross and Horner (2009) report one approach to bully prevention that follows guidelines en couraged by Bradshaw and Waasdorp (2011) to (a) build a positive, school-wide social climate, (b) teach students constructive responses to minimize the social rewards of bullying behav ior, and (c) provide consistent staff responses to instances of bullying. Ross and Homer (2009) examined bullying behaviors across three ele mentary schools within a multiple baseline re search design in which two children from each school who were identified as engaging in bul lying behaviors were observed on the play ground over 60 school days. Data were col lected on the incidence of verbal or physical aggression toward peers, and the responses from recipients and bystanders following an instance of verbal or physical aggression. The intervention, Bully Prevention in Positive Be havior Supports (BP-PBS: Ross, Horner, & Stiller, 2008), involved teaching all students first to identify “respectful” behavior. The term “bullying,” however, was never part of the stu dent instruction. Building from the conceptual logic that a high proportion of bullying behav iors endure because they result in social atten tion from recipients or bystanders, the interven tion then focused on teaching all students in the school three routines: (a) a “stop” signal that they could use either as a recipient or bystander when they encountered verbal or physical ag gression, (b) a “stopping routine” that should be followed if someone asks you to stop, and (c) a protocol for obtaining adult support if someone does not stop when asked (Ross et al., 2008). Results from the study demonstrated a 72% reduction in verbal and/or physical aggression on the playground. In addition, the descriptive
273
data indicated an increased use of the “stop” routine and decreased conditional probability that verbal or physical aggression would be followed by peer social rewards. The present study builds from these efforts to develop a social-ecological approach to bully prevention based on behavioral theory (Brad shaw & Waasdorp, 2011; Waasdorp et al., 2012). The specific goal of the present analysis was to assess the effectiveness of the Ross et al. (2008) approach when it was adapted to fit the social context of middle schools. The Bullying and Harassment Prevention in Positive Behav ior Support: Expect Respect program (Expect Respect; Stiller, Nese, Tomlonovich, Horner & Ross, 2013) incorporates the same basic instruc tional goals used by Ross et al. (2008) but with three major adaptations: First, a student focus group component was built into Expect Respect to obtain student input about (a) the student perception about the significance of aggression, harassment, and intimidation in the school; (b) a socially acceptable way to deliver a “stop” mes sage that could be used by all students; (c) a socially acceptable “stopping routine” that could be taught to all students; (d) a “bystander” routine to be used by witnesses of disrespectful behaviors; and (e) the parameters of adult re sponses when instances of bullying are reported by students. The purpose of the student focus group is to adapt the messages of the program to the local school context, as research has shown that information gathered from focus groups help to identify existing cultural norms and practices needed for improving service delivery (Baker-Henningham, 2011; Bronheim, 2013). The second major adaptation was to build three 1-hour lesson plans that emphasized op portunities for students to shape how the content of the four routines (stop, stopping, bystander, and seeking support) fit with the language, cus toms, and culture of their school. Lesson one, led by school personnel in student groups sim ilar to class sizes, presents students with a basic review of the importance of respectful behavior, how to deliver the “stop” signal if the disre spectful behavior is encountered, how to use the “stopping routine” if someone asks you to stop, how to use the “bystander routine” if you wit ness someone using the stop signal and the perpetrator does not stop, and the process and protocol for seeking supports from adults if disrespectful behaviors endure. Lessons two
274
NESE, HORNER, DICKEY, STILLER, AND TOMLANOVICH
and three are designed to be more studentguided and frame problem-solving situations where the basic “stop,” “stopping,” “by stander,” and “seeking support” routines are being applied with and without success. The third adaptation focused on adult behav ior. Training and coaching for faculty and staff was built into the implementation of Expect Respect. Initial field-testing identified that a ma jor concern of middle school students was the consistency of adult responses to instances of bullying behavior. The goal of this portion of the program was to both maximize consistency and ensure that students were informed about what they could expect when reporting an in stance of disrespectful behavior. The primary research question guiding the study was whether Expect Respect, when imple mented with fidelity, was functionally related to an overall reduction in the level of verbal or physical aggression observed in uncontrolled settings like the cafeteria. Secondary, descrip tive questions focused on the extent to which the conditional probability of bystander and re cipient responses to verbal or physical aggres sion became less rewarding after introduction of the program, and if student perceptions of the overall safety of the school, and the quality of social interactions within the school changed after introduction of the program.
Todd et al., 2003). Adequate implementation of SWPBIS includes the following: establishing and teaching behavior expectations, establish ing and teaching a school-wide strategy for re warding appropriate behavior, and establishing a continuum of responses to problem behaviors (Homer, Sugai, Todd, & Lewis-Palmer, 2005). All sixth-, seventh-, and eighth-grade stu dents in the three participating middle schools received training on Expect Respect, following passive parental consent. Parents were notified that direct observations would be taking place during lunchtime and that no individual student would be targeted for observation. Parents were also informed that they could choose to remove their student from participation in the Expect Respect intervention or from lunchtime direct observations at any time; however, no parent declined consent for their student’s participa tion. To protect student confidentiality, no iden tifying student information was recorded on the daily direct observation sheets other than stu dent sex. Additionally, students provided writ ten assent before completing any surveys. The parental consent, student assent, and all materi als and procedures were approved by our Insti tutional Review Board (IRB) as well as the participating school districts.
Method
Data were collected between November and May of the 2011-2012 school year. The pri mary data collected focused on direct observa tion of bullying behaviors, as well as the behav iors of both recipients and bystanders immediately following an instance of bullying. In addition, survey data were collected to assess students’ perception of “school climate,” and self-assessment checklists completed by school staff were collected to assess the fidelity with which Expect Respect was implemented. Bullying behaviors. The primary depen dent measure was the frequency of physical or verbal aggression during 20-min direct observa tions in the cafeteria during lunchtime. An ad ministrator at each school identified a section of the cafeteria in which bullying behaviors were most likely to be reported. This was often an area with low adult supervision and high student foot traffic. During baseline and intervention, trained University data collectors positioned themselves near the identified area so that they
Setting and Participants Three middle schools in two Pacific North west school districts participated in this study during the 2011-2012 school year. Participating schools were attended by 508, 511, and 691 students, and eligibility for free or reducedprice lunch ranged from 43% to 65% of stu dents. Additionally, the student population at all three schools was primarily White, with stu dents of color ranging from 26% to 32%. To be eligible for participation, selected schools were made up of Grades 6 through 8, had not re ceived training on or utilization of BP-PBS or Expect Respect before this study, and had im plemented School-Wide Positive Behavioral In terventions and Supports (SWPBIS: Horner, Sugai, & Anderson, 2010) for at least three years with adequate fidelity (e.g., ^80% Total Score on the School-Wide Evaluation Tool:
Measurement
MIDDLE SCHOOL BULLY PREVENTION
could see and hear all interactions that took place between students. All observations were conducted in this identified area at the same time of day throughout the study. No student was singled-out for observation in this study; therefore the behaviors of any student that in volved aggression in the observed area were recorded. For the entire 20-min observation pe riod, observers recorded whether a bullying event took place and whether the event involved physical aggression, verbal aggression, or both. Physical aggression was operationally defined as “hitting, biting, kicking, chocking, stealing, throwing objects, or restricting freedom of movement” (Ross & Homer, 2009, p. 751). Verbal aggression was operationally defined as “the direction of negative communication either verbal or gestural, toward one or more focus students including teasing, taunting, threaten ing, negative body language, or negative ges tures” (Ross & Horner, 2009, p. 751). Two to four direct observation sessions per week took place at each participating school, with no more than five days between observation sessions. This five-day span without data collection only took place when schools were out-of-session for Spring Break. Recipient and bystander responses to bul lying behaviors. The second and third depen dent measures were recipient and bystander re sponses to bullying behaviors. During each observation session, observers recorded how re cipients and bystanders of bullying behaviors responded to the perpetrator. A recipient was operationally defined as a student receiving the bullying behaviors. A bystander was operation ally defined as a student located within three minutes of the bullying event that is witnessing the bullying behaviors but does not take part. Once an instance of bullying occurred, observ ers would document whether the recipient and any bystander responded by telling the perpe trator to “Stop,” as taught in the Expect Respect curriculum, or by either ignoring the perpetra tor, or providing social attention (e.g., laughing, yelling back) to the perpetrator within five sec onds of each instance of a bullying behavior. Observers also recorded when recipients and bystanders provided no response. Each recipient and bystander response was mutually exclusive, and all responses to an incident were recorded. For example, in a case where a bystander said
275
“stop” but then started to yell at the perpetrator, both “stop” and “attention” were recorded. Observer training and interobserver agree ment. Five graduate students from a Pacific Northwest university served as trained observ ers for this study. All observers were blind to the goals of the present study and when the intervention was started at each school. Before beginning data collection, observers were trained to an 85% interobserver agreement cri terion on each target behavior. Observers par ticipated in three training sessions with the first author that involved reviewing the measures, procedures, and data collection tool. During the first session, observers became familiar with the operational definitions of each behavior that were coded, and began practicing coding sam ple videotapes as a group. The second and third training sessions involved reviewing the opera tional definitions, having specific questions an swered, and coding more sample videotapes. Total agreement was at or greater than 85% for all target behaviors before the observers began coding baseline data for the present study. Ses sions coded within the study were assessed for interobserver agreement on 40% of observa tions during baseline and intervention phases for each of the three schools. During these ses sions, a second observer independently coded during the same lunchtime observation. Total agreement was calculated by dividing the num ber of events that both observers agreed a bul lying behavior occurred by the total number of events observed, and multiplying that number by 100%. Table 1 displays total interobserver agreement across all three schools, which ex ceeded 85% in both baseline and Expect Re spect phases. Fidelity of implementation. Fidelity of Expect Respect implementation was assessed Table 1 Average (Range) Interobserver Agreement
School 1 2 3
Bullying behaviors
Baseline total agreement
Intervention total agreement
Physical Verbal Physical Verbal Physical Verbal
98% 99% 99% 99% 95% 92%
96% 95% 96% 97% 97% 98%
276
NESE, HORNER, DICKEY, STILLER, AND TOMLANOVICH
using a 5-item fidelity checklist that measured Procedures staff adherence to the program components. Experimental design. A nonconcurrent School staff members completed fidelity check multiple baseline design (Kazdin, 2011) across lists twice: between weeks 1 and 3 and weeks 6 three middle schools was used to examine the and 8 after launching the intervention. Items on effectiveness of Expect Respect to reduce bul the fidelity checklist include the extent to which lying behaviors and improve bystander and re teacher and school staff (a) interrupt bullying cipient responses to bullying behaviors. The behaviors they observed in or around the design involved two phases: baseline and Ex school, (b) reflectively listen when students re pect Respect implementation. The total duration port incidents of bullying, (c) praise students for of the study was 6 months. School 1 partici reporting such incidents, (d) encourage students to use an assertive word or phrase to interrupt pated in the study between November and May, bullying incidents or walk away, and (e) check School 2 participated between February and back in with students following a report to de May, and School 3 participated between Janu termine if the issue had been resolved. Scores ary and May. For each school, a baseline period on the checklist were recorded on a Likert scale was followed by a staff training on the curric from 1 to 4, with higher scores indicating better ulum, a meeting with a focus group of students to define the routines to be taught, and a wholeadherence to Expect Respect procedures. School climate survey. Students’ percep school presentation of Lesson 1, with Lessons 2 tions of their school climate were assessed dur and 3 taught one to seven weeks later after ing the first week of implementation of the scheduling availability defined for each school. Baseline. In baseline, students were ob Expect Respect intervention and again at the end served during unstructured lunchtime in the of the study. The researchers developed a school cafeteria. No teaching procedures to 9-item survey that assessed students’ percep modify any bullying behaviors were in effect. A tions of their school safety, respectfulness among students and teachers toward one an minimum of five data points were collected for other, school members’ interest in creating a each school during the baseline phase, and in safer school environment, students’ use of strat cidents of bullying behaviors and conditional egies to interrupt disrespectful behavior, and probabilities of recipient and bystander re the school’s dedication to bullying preven sponses were recorded. Each data point repre tion. The items on the survey were developed sents a single 20-min observation session. Expect Respect. Implementation of Expect from the targets of the Expect Respect manual, Respect involved three components. In the first and therefore focus on the extent to which stu dents perceive each other as being treated with component, the first or second author and a respect, and if they have a clear idea about what school coordinator provided training to the en to do when they encounter disrespectful behav tire school staff on the Expect Respect curricu ior. Five of the nine items were rated on a lum. Training consisted of a 1-hr workshop on 4-point scale (1 = strongly disagree, 2 = dis the intervention components, including the agree, 3 = agree, 4 = strongly agree), with classroom lessons, methods of supervising and higher scores indicating a more favorable im interrupting inappropriate behaviors in unstruc pression of the school climate (see Figure 3), tured environments, appropriate adult responses and the other four items could either be an to students’ reports of bullying and harassment, swered Yes or No (see Figure 4). On average, and staff responsibilities for supporting students the survey took students approximately five through the reporting and responding process. minutes to complete. All students at the three Staff members were given the opportunity to participating schools were asked to complete practice delivering the classroom lessons during the survey online, and before beginning the sur staff training. Once the entire school staff had vey, they were prompted on the computer screen been trained on the Expect Respect curriculum, to elect to complete it or not complete. Each implementation moved on to the second com school chose when and how they wanted the stu ponent. The second component involved the organi dents to complete the survey, with most classes circulating through the computer lab during an zation of a student focus group, which com prised approximately 8 to 12 students who eiintervention or flex block period.
MIDDLE SCHOOL BULLY PREVENTION
ther volunteered or were nominated by their teachers. The student focus group, along with the staff members overseeing implementation of Expect Respect, held a 1-hr meeting where the students selected the school-wide STOP sig nal, discussed and practiced the “stop,” “stop ping,” “bystander,” and “seeking support” strat egies, and discussed a plan for gathering student buy-in to the program. The student focus group is a fundamental component of the Expect Re spect intervention, as students have noted that for a program to be successful at the middle school level, middle school students need to be actively involved in planning and administering the lessons, and organizing events related to improving their school climate. In the third component, which occurred in the week following staff training, teachers, school staff, and student focus group members used the three Expect Respect lesson plans to provide training for all students in Grades 6 through 8 during 1-hr intervention block class periods. The administrators at each school were able to decide how far apart they wanted to space the three lessons out, based on their building sched ules and holidays. However, all schools admin istered lesson one within one week of Expect Respect implementation. During lesson one, students were led through a basic review of the importance of respectful behavior and the school-wide expectations around being respect ful. After this review was complete, students were then instructed on the four strategies for interrupting disrespectful behavior: how to de liver the “stop” signal if the disrespectful be havior is encountered, how to use the “stopping routine” if someone asks you to stop, how to use the “bystander routine” if you witness someone using the stop signal and the perpetrator does not stop, and the process and protocol for seek ing supports from adults if disrespectful behav iors endure. Lesson two provided an opportu nity for students to practice the four routines in a simulated activity on a school bus, and to problem solve around situations where the rou tines were and were not successful. Lesson three had two components. The first component focused on identifying ways of interrupting and removing social attention given to disrespectful behaviors identified in online videos. During the second component, students were guided through completing a “seeking support” work sheet for their day planners, in which they iden
277
tified individuals within the school and outside the school they could go to for support. They also wrote out three strategies to use when they are either the recipient or the witness of disre spectful behaviors. After all three 1-hr lessons were imple mented, school staff utilized the “Repeat and Repair” activity as a 15-min refresher for stu dents and staff on the trained routines. This lesson was either delivered a few times through out the school year or at the end of implemen tation. Results For each of the three participating schools, the effects of the Expect Respect intervention are reported on (a) fidelity of implementation, (b) the incidents of physical and verbal aggres sion, (c) recipient and bystander response prob abilities, and (d) school climate. Note that there were no points of missing data in this study. Fidelity of Implementation Fidelity of Expect Respect implementation was assessed by evaluating a convenience sam ple of staff self-reported adherence to program components. A total of 82 staff members across all three schools completed two fidelity check lists during implementation of Expect Respect. All staff members who attended their school’s monthly staff meeting when fidelity data were being collected were given the fidelity checklist to complete (no descriptive data were collected on the teachers). The overall median was cal culated for each item by determining the per centage of staff who endorsed a 3 (Mostly) or 4 (Always), and then determining the median value as well as the range of values across the averages. Results indicated that staff members reported high adherence to the proper proce dures when responding to witnessed or reported incidents of bullying, and that their adherence increased over time, reporting an overall me dian of 83% at Time 1 (range = 44% to 96%) and 91% at Time 2 (range = 82% to 100%). Specifically, a median of 82% of staff members across schools indicated that they “interrupted bullying behaviors they observed” (range = 67% to 94%), 91% stated that they “reflectively listened when students report incidents of bul lying” (range = 81% to 100%), 91% reported
278
NESE, HORNER, DICKEY, STILLER, AND TOMLANOVICH
that they “praised students for reporting such incidents” (range = 75% to 96%), 88% indi cated that they “encouraged students to use an assertive word or phrase to interrupt bullying incidents or walk away” (range = 69% to 91%), and a median of 85% of staff members stated that they “checked back in with students fol lowing a reported incident” (range = 44% to
88%). Incidents of Physical and Verbal Aggression The incidents of physical and verbal aggres sion across all three schools are presented in Figure 1. Students in each middle school en gaged in physical, and/or verbal aggression dur ing Baseline. For School 1, baseline observa tions scored with aggression averaged 4 incidents per 20-min session (range = 2 to 7), with high variability and a decreasing trend. For School 2, baseline aggression averaged 2.44 events per 20-min observation (range 1-4) but with a moderate level of variability, and stable trend. School 3 students were observed to en gage in aggression an average of 2.37 times per observation (range 0-5) with higher variability and a slight increasing trend. Each school demonstrated reduction in rates of physical and verbal aggression after introduc tion of the Expect Respect intervention. Schools 1, 2, and 3 averaged .89, .88, and .64 incidents of aggression respectively per 20-min observa tion across the Expect Respect phase. Note, however, an unexplained increase in the level of aggression in School 1 between Sessions 15-19 in conjunction with the second Expect Respect lesson. Levels of aggression later in the inter vention phase for each school (especially after Spring Break) were at or near zero. Taken to gether the data from the Expect Respect phases represent a 78%, 64%, and 73% reduction in level of aggression, respectively, for Schools 1, 2, and 3, with reduction in variability also noted for each school. Recipient and Bystander Response Probabilities Information on recipient and bystander re sponses to bullying behaviors was recorded at the time of each aggressive incident. Figure 2 presents the conditional probabilities of these
responses prior to and following implementa tion of the Expect Respect curriculum. During baseline at School 1, when an inci dent of bullying occurred, recipients said “Stop” 5% of the time, rewarded (reinforced) bullying behaviors 45% of the time, and ignored bullying behaviors 35% of the time. After implementa tion of Expect Respect, recipients increased their delivery of the “Stop” phrase to 19% of the time, increased rewarding of bullying behaviors at 58% of the time, and decreased ignoring of bullying behaviors to 29% of the time. Note that sessions where no bullying behaviors were ob served did not contribute to the conditional probabilities. Bystanders at School 1 said “Stop” during baseline 5% of the time, rewarded bullying be haviors 35% of the time, and ignored bullying behaviors 35% of the time. During the interven tion phase, bystanders decreased their delivery of the “Stop” phrase to 3% of the time. Delivery of a rewarding response after bullying behaviors increased to 55% of the time, and ignoring maintained at 35%. During baseline at School 2, recipients said “Stop” 32% of the time, rewarded bullying be haviors 55% of the time, and ignored bullying behaviors 36% of the time. After implementa tion of Expect Respect, recipients decreased their delivery of the “Stop” phrase to 14% of the time, increased their level of reward for bully ing behaviors to 71% of the time, and decreased ignoring of bullying behaviors to 29% of the time. Bystanders at School 2 said “Stop” during baseline 0% of the time, as they were not ob served demonstrating this behavior during any observation session. Bystanders rewarded bul lying behaviors 45% of the time, and ignored bullying behaviors 45% of the time. During the intervention phase, bystanders increased their delivery of the “Stop” phrase to 7% of the time. Delivery of a rewarding response following bul lying behaviors increased to 50% of the time, and ignoring decreased to 43% of the time. During baseline at School 3, recipients said “Stop” 29% of the time, rewarded bullying be haviors 44% of the time, and ignored bullying behaviors 33% of the time. After implementa tion of Expect Respect, recipients increased their delivery of the “Stop” phrase to 33% of the time, maintained the same level of reward for
MIDDLE SCHOOL BULLY PREVENTION
Baseline
279
Expect Respect Intervention
Sessions Figure 1. Incidents of physical and verbal aggression during baseline and PBPBS-ER implementation for all three participating schools during 20-min observations. Phase change lines indicate the date of Lesson 1, and asterisks indicate the date of Lessons 2 and 3 per school. The double hash marks indicate when each school was out-of-session for Spring Break.
bullying behaviors at 44%, and decreased ignor ing of bullying behaviors to 22% of the time. Bystanders at School 3 said “Stop” during baseline 4% of the time, rewarded bullying be haviors 56% of the time, and ignored bullying
behaviors 42% of the time. During the interven tion phase, bystanders’ delivery of the “Stop” phrase decreased to 0% of the time, as they were not observed demonstrating this behavior during any observation session. Delivery of a
280
NESE, HORNER, DICKEY, STILLER, AND TOMLANOVICH
Recipient Response Probabilities
STOP
Reward
Bystander Response Probabilities
Ignore
STOP
Reward
Ignore
Responses Figure 2. The conditional probabilities of recipient and bystander responses to bullying behaviors during 20-min observations during lunchtime in the cafeteria, [graybox] Preinter vention; [blackbox] Postintervention.
rewarding response after bullying behaviors maintained at 56% of the time, and ignoring decreased to 33%. School Climate Students’ perceptions of their school climate were assessed pre- and postintervention via a brief online survey. Table 2 displays the demo graphic information for students who completed
the survey, as well as the sample sizes and the number of students who declined to participate. Pre-post samples varied by size and demo graphic data as a result of logistics and sched uling constraints related to circulating students through the computer lab to complete the survey (see Table 2). It should also be noted that de mographic data were not available for School 1 during the preintervention survey. The pro-
MIDDLE SCHOOL BULLY PREVENTION
281
Table 2 Frequency (%) o f Demographic Characteristics fo r Students Who Were Administered the School Climate Survey School 1
School 3
School 2
Characteristic
Pre
Post
Pre
Post
Pre
Post
Completed Declined Sex Female Male Not indicated Race White Non-White Not indicated
309 (96) 12(4)
353 (93) 26(7)
53 (98) 1(2)
70 (93) 5(7)
234 (98) 6(2)
81 (91) 8(9)
166 (47) 176 (50) 11 (3)
29 (55) 24 (45) 0(0)
40 (57) 30 (43) 0(0)
122 (52) 112 (48) 0(0)
29 (36) 51 (63) 1 (1)
240 (68) 89 (25) 24(7)
42 (79) 9(17) 2(4)
45(64) 20 (29) 5(7)
167(71) 59 (25) 8(3)
47 (58) 33 (41) 1(1)
— — —
— —
—
Note. Preintervention demographic data were not available for School 1.
portion of students who indicated Strongly Agree or Agree to the first five items of the school climate survey can be found in Figure 3, and the percentage of students who indi cated Yes to the last four items of the survey can be found in Figure 4. On average, students across all schools rated their school climate as high before and after completion of the Expect Respect program. Be fore intervention implementation, the mean pro portion of students across all three schools who indicated Strongly Agree or Agree to the fol lowing questions were as follows: Do you feel safe? was .93, Do other students treat you re spectfully? was .78, Do you treat other students respectfully? was .95, Do adults treat you re spectfully? was .90, and Do you treat adults in your school respectfully? was .95. After inter vention implementation, the mean proportion of students across all three schools who indicated Strongly Agree or Agree to the following ques tions were as follows: Do you feel safe? was .91, Do other students treat you respectfully? was .80, Do you treat other students respect fully? was .95, Do adults treat you respectfully? was .86, and Do you treat adults in your school respectfully? was .93. Figure 3 depicts the re sults for student responses at each school. Al though these data indicate a small increase in the proportion of students who reported that other students treated them respectfully, a twoproportion z test revealed no statistically signif icant pre-post differences at the p < .05 level (see Table 3).
Preintervention mean percentages across all 596 students who indicated Yes to the following questions were as follows: Has anyone treated you disrespectfully? was 53%, Have you asked someone to STOP? was 59%, Has anyone asked you to STOP? was 20%, and Have you seen someone else treated disrespectfully? was 65%. Postintervention mean percentages across all 504 students who indicated Yes to the following questions were as follows: Has anyone treated you disrespectfully? was 56%, Have you asked someone to STOP? was 59%, Has anyone asked you to STOP? was 22%, and Have you seen someone else treated disrespectfully? was 62%. Figure 4 depicts the results for student re sponses at each school. Although these data indicate a small decrease in the percentage of students who reported seeing others being treated disrespectfully and a small increase in other students using a school-wide STOP phrase, a two-proportion z test revealed no sta tistically significant pre-post differences at the p < .05 level (see Table 3). Finally, 53% of students across all schools rated Expect Respect as being either very helpful or helpful for their schools, and 60% of students felt that Expect Respect should be taught again next year. Discussion The present study focused on a program de signed to build the social-ecological conditions recommended by Swearer, Espelage, Vaillancourt, and Hymel (2010) and Waasdorp et al.
282
NESE, HORNER, DICKEY, STILLER, AND TOMLANOVICH
D o y o u fe e l s a fe ? D£
•a
a a H
J=
School 2
-
■a
s
ce
■8 & i e t—
