926610

research-article2020
RSEXXX10.1177/0741932520926610Remedial and Special EducationSinclair et al.

Special Series: Universal Social Emotional Intervention

Remedial and Special Education

Effects of a Universal Classroom

2021, Vol. 42(1) 18­–30
© Hammill Institute on Disabilities 2020
Article reuse guidelines:
Management Intervention on Middle sagepub.com/journals-permissions
DOI: 10.1177/0741932520926610
https://doi.org/10.1177/0741932520926610

School Students With or At Risk of rase.sagepub.com

Behavior Problems

James Sinclair, PhD1,2, Keith C. Herman, PhD1, Wendy M. Reinke, PhD1,

Nianbo Dong, PhD3, and Melissa Stormont, PhD1

Abstract
Universal interventions implemented in schools have the potential to impact large numbers of students on a multitude
of behavioral and academic outcomes. In the context of a large group randomized controlled trial, the current study
examined whether teacher-perceived student need for behavior support plans at baseline moderated the effects of a
middle school universal classroom management training program, CHAMPS, which stands for Conversation, Help, Activity,
Movement, Participation, and Success, on student outcomes. We hypothesized that students in CHAMPS classrooms who
were identified by teachers at baseline as needing behavior support plans would have greater end-of-year improvements
on behavior and academic outcomes relative to comparable youth in non-CHAMPS classrooms. Results indicated baseline
teacher-perceived student behavior support need moderated some intervention outcomes; in particular, youth at risk
in CHAMPS classrooms demonstrated improvements in concentration problems (b = –.19) and communication arts
outcomes (b = .13) compared with youth at risk in control group classrooms. Implications of these findings and future
research are discussed.

Keywords
universal prevention, youth at risk, classroom management

Disruptive behavior problems among students in class- Middle School Context

rooms present a significant challenge for teachers and
schools. In response, many schools have adopted and One area that many teachers report struggling with is imple-
implemented universal prevention interventions to help menting effective classroom behavior management strate-
reduce problem behaviors that disrupt student’s educa- gies (Buell et al., 1999; Pavri, 2004). In fact, teachers
indicated that they considered classroom behavior manage-
tional attainment (McIntosh et al., 2018). Universal pre-
ment to be the most challenging aspect of their job and one
vention methods are interventions that impact all students
in which they received the least amount of training (Reinke
in a classroom or a school building. They are an efficient,
et al., 2011). Furthermore, problematic behaviors in class-
low-cost attempt to reduce multiple risk factors and pro- rooms can disrupt learning and also take time away from
mote protective factors in the youth population (Greenberg instruction. Thus, teachers have also expressed concerns
& Abenavoli, 2017). about the impact that problem behaviors have on the aca-
Universal interventions have been developed to make a demic performance of their students (Lassen et al., 2006).
broad impact on those who are exposed to it. The expecta-
tion is that approximately 85% of students will respond suc-
1
cessfully to preventive universal strategies, whereas the University of Missouri, Columbia, USA
2
University of Oregon, Eugene, USA
other 15% of students will benefit from more targeted sup- 3
The University of North Carolina at Chapel Hill, USA
ports (Stormont et al., 2012; Sugai & Horner, 2006). What
Corresponding Author:
is less known is whether students at risk of behavior prob-
James Sinclair, Department of Special Education and Clinical Sciences,
lems, who are identified as in need of more intensive inter- Center on Human Development, University of Oregon, 355 Clinical
ventions, also benefit from the universal interventions they Services Building, 901 E. 18th Ave., Eugene, OR 97403, USA.
receive as part of the larger student body. Email: jamesin@uoregon.edu
Sinclair et al. 19

Middle schools prove to have their own specific chal- and high-quality instruction shifts to the classroom teacher
lenges that enhance the potential for increased problematic (Lassen et al., 2006).
behaviors. Students in middle schools are faced with envi- While research has suggested that teachers can play a
ronmental changes, more independence from teachers, and critical role in supporting the behavior of students, it has also
increased peer influence (Rusby et al., 2011). Furthermore, indicated that many teachers are unaware of evidence-based
students who express problem behaviors throughout middle practices that might ameliorate the disruptive problem
school and later are at increased risk of more punitive con- behaviors they struggle to manage and, conversely, increase
sequences, including often being suspended or expelled, positive outcomes for students in their classrooms (Stormont
which decreases academic learning time and opportunity et al., 2011). For example, Stormont et al. (2011) docu-
for classroom engagement (Tobin & Sugai, 1999). mented that only 44% of educators were confident that the
Students who continue to experience problem behaviors interventions they used had the desired impact on their stu-
or are at risk of developing problem behaviors are vulnera- dents. This lack of knowledge may be associated with a lack
ble to experiencing persistent long-term negative outcomes of sufficient experience, training, or support (Reinke et al.,
(Wagner & Newman, 2012). Of particular concern are the 2011). Thus, there is a need for interventions that focus on
lasting effects of problem behaviors through high school building teachers’ knowledge and skills to promote the use
where these behaviors are predictive of negative outcomes of effective universal classroom management strategies.
such as high school dropout (Darney et al., 2013; Sweeten,
2006) and poorer academic performance (McIntosh et al.,
2008). The changing dynamics between elementary and
Conversation, Help, Activity,
secondary education settings also require more specific Movement, Participation, and Success
skill competence and acquisition to be successful in the (CHAMPS)
classroom. Students in middle school must learn to self- CHAMPS is a universal classroom-wide intervention that
regulate, communicate clearly, stay on task for longer peri- can be implemented by middle school teachers. CHAMPS
ods of time, become critical thinkers, and be able to was developed for implementation in middle schools over a
problem-solve. Furthermore, students become more aware
decade ago, with the intention to help teachers develop
of environmental impacts to their education and develop-
strategies to reduce off-task behavior, promote prosocial
ment; for instance, students who experience poverty or
behavior, and increase academic opportunities to respond.
attend schools in lower socioeconomic areas may also
CHAMPS consists of a modular series of training and
experience poorer academic achievement and be more at
coaching materials designed to help classroom teachers
risk of lower attendance (Morrissey et al., 2014).
The unique challenges that middle schools face in develop an effective classroom management plan that is
addressing the needs of their students creates an opportu- proactive, positive, and instructional (Sprick et al., 1998).
nity for professionals to increase prevention and interven- Emphasis is placed on promoting positive teacher–student
tion efforts and intervene to curb problem behaviors and interactions and in helping teachers structure their class-
the potential long-lasting impacts it can cause. Providing rooms in ways that prompt responsible student behavior.
the opportunity for students to interact with effective class- The program prepares teachers to explicitly teach students
room management practices implemented at the universal how to behave responsibly, to give attention and energy
level lays the groundwork for learning and engagement, toward acknowledging student responsible behavior, and to
which in turn reduces problem behaviors (Conroy et al., being consistent, calm, and brief when correcting misbe-
2008). Promoting and implementing universal-level inter- havior. Teachers trained in CHAMPS receive ongoing
ventions offers a higher probability of capturing youth who coaching to help support the implementation of newly
may be considered at risk or are “under the radar”; these learned skills in the classroom. Coaches use evidence-based
students thus would not have the opportunity to benefit coaching strategies, including modeling and providing per-
from targeted or individualized interventions (Wilson & formance feedback.
Lipsey, 2007). Universal interventions can be beneficial CHAMPS is a fully developed and widely disseminated
for those under the radar because not all students with prevention program. Books, planning materials, and DVDs
behavioral challenges will have a diagnosable disorder and support the implementation of this program in precise and
receive special education services. In fact, it has been repeatable ways (Sprick, 2009). In a recent randomized
widely hypothesized that youth with emotional and behav- controlled trial of the CHAMPS intervention, teachers who
ioral disorders are under-identified (Walker et al., 2000), received training in and implemented the CHAMPS inter-
which could lead to students who need special education vention were found to use more proactive behavioral class-
services not receiving those services. Without proper sup- room strategies than teachers in the control condition
ports and services provided by the school, the sole respon- (Herman et al., under review). In addition, the effects found
sibility of simultaneous effective classroom management for the treatment teachers were maintained over time.
20 Remedial and Special Education 42(1)

Students in the treatment classrooms also showed better Method

academic and behavior outcomes (Herman et al., under
review); specifically, they had significantly higher scores Participants
on a state standardized communication arts achievement Participants were recruited as part of a randomized con-
measure and on a math problem-solving achievement test trolled trial of the CHAMPS behavior management inter-
as well as significantly lower levels of disruptive behaviors vention funded by the Institute of Education Sciences. The
at the end of the school year adjusting for baseline randomized controlled trial recruited participants from nine
covariates. middle schools from an urban school district in the Midwest.
Teachers and students were recruited from Grades 6 to 8
Current Study English language arts and mathematics classrooms.
Teachers and caregivers of participating students provided
Universal interventions have the capacity to impact the broad written informed consent and students provided verbal and
spectrum of students in a classroom. Typically, studies that written assent prior to participating in the study.
report the results of a universal intervention’s main effects do A total of 102 teachers and 1,405 students consented to
so across all students. Fewer studies disaggregate students participate in the study. Teachers identified primarily as
and report how these interventions may benefit students who female (79.1%). Teachers’ racial and ethnic identities included
have been identified as at risk. Yet, the realities are that youth White (70.9%), Black (25.6%), Asian (2.3%), and Other
at risk of problem behaviors are included in general educa- (1.2%). The mean age of teacher participants was 37.8 (SD =
tion settings, where universal prevention interventions are 8.8, range = 23–63 years) and the mean years of teaching was
implemented (Lane et al., 2014). Thus, to broaden our under- 10.4 (SD = 6.3, range = 1–23 years). Fifty-one teachers
standing of the impact that universal interventions can have, (50%) were randomly assigned to the intervention condition.
we must examine how they affect individual students who Students were identified, by teacher report, as either hav-
are at risk as well as classrooms as a whole. ing a need for individualized behavioral support (i.e., stu-
The purpose of this study was to evaluate whether stu- dents at risk) or no need. Approximately, 44% (n = 639) of
dents identified by teachers as needing behavioral supports the total student sample were students who were identified
in classrooms of teachers who were randomly assigned to as having need for individualized behavioral support. Of
receive the CHAMPS intervention demonstrated improve- those students, 32% were in Grade 6, 44% were in Grade 7,
ments in disruptive behavior, concentration problems, emo- and 24% were in Grade 8. Fifty-nine percent of students
tional regulation, prosocial behaviors, and academic were male and 41% female. Racial and ethnic categories for
outcomes in comparison with students identified by teach- students who were identified as at risk included 87% Black,
ers as needing behavioral supports in classrooms of teach- 11% White, 0.9% Latinx, 0.6% Asian, and 0.5% multira-
ers who did not receive the training. Specific research cial. Seventy-six percent received free/reduced-price lunch
questions included the following: (FRL). Furthermore, of students who were receiving special
education services at the time of the study (n = 138), 60%
Research Question 1: To what extent do students per- of these students were identified as at risk based on teacher
ceived by teachers to be in need of behavioral supports in report that the student would benefit from individualized
classrooms of teachers receiving CHAMPS have more behavioral supports. Disability category was not available
positive end-of-year behavioral outcomes than students in the data set. Of those students considered not at risk of
perceived to be in need of behavioral supports in the con- behavior problems, 38% were in Grade 6, 34% were in
trol classrooms? Grade 7, and 28% were in Grade 8. Forty-two percent were
Research Question 2: To what extent do students per- male and 58% were female. Racial and ethnic categories for
ceived by teachers to be in need of behavioral supports in students who were not identified as in need of individual-
classrooms of teachers receiving CHAMPS have more ized behavioral support plans included 72% Black, 23%
positive end-of-year academic outcomes than students White, 0.1% American Indian, 3% Latinx, 0.1% Pacific
perceived to be in need of behavioral supports in the con- Islander/Native Hawaiian, 1.4% Asian, and 0.9% multira-
trol classrooms? cial. Sixty-five percent of students qualified for FRL.

We hypothesized that students identified as at risk (i.e.,

teachers reported that a student would benefit from an indi-
Procedures
vidualized behavioral support plan) in CHAMPS class- The university’s institutional review board and the partici-
rooms would demonstrate improved outcomes, including pating school district approved the study protocol. Teachers
decreased disruptive behavior, improved prosocial behavior were recruited and students in their classrooms were eligi-
and emotional regulation, increased concentration, and ble to participate. Only students with written parental con-
improved academic outcomes. sent and student assent for participation were included in
Sinclair et al. 21

the study (75% of all eligible students). All pre-intervention providing teachers with explicit performance feedback fol-
assessments were completed at the start of the academic lowing coach observations of the implementation of the
year in the month of October prior to teachers being ran- CHAMPS intervention using a partnership framework
domized to condition. All post-intervention assessments (rather than coach as expert model; see Sprick et al., 2009).
were completed in the Spring term after all intervention In between each workshop session, the CHAMPS coach
activities were completed in April or May of the same aca- conducted observations of teacher practices and student
demic year. behaviors, followed by a meeting with each teacher individ-
ually for up to 1 hr on a weekly basis. A minimal dose of
Intervention condition. Teachers randomly assigned to coaching for each teacher was set at a minimum of four vis-
receive the CHAMPS training attended three full-day group its with the coach. The first coaching visit focused on estab-
training sessions in October and one additional session in lishing rapport and setting goals. The second coaching visit
late November/early December. All trainings were facili- focused on providing the teacher with explicit feedback
tated by a certified CHAMPS trainer supervised by the pro- based on the coach’s classroom observations and developing
gram developer. The CHAMPS trainer also provided a plan based on the teacher’s own goals. Subsequent coach-
ongoing coaching to support teacher implementation of ing visits were tailored to teacher needs and goals. During
newly learned skills following training sessions. coaching sessions, the coach provided performance feed-
CHAMPS is a comprehensive curriculum for improving back, reviewed workshop content, modeled effective prac-
teacher classroom management and relationship skills. The tices, supported goals setting and action planning, and
acronym STOIC highlights the key principles for an orga- processed barriers to implementation. The mean time spent
nized and effective classroom: Structure for success, with a teacher by the coach, outside of classroom observa-
Teacher expectations, Observe systematically, Interact posi- tions, was 147 min (range = 48–358 min).
tively, and Correct calmly. The training focuses on building
teacher competence in each of these five domains. Teachers Control condition. The teachers who were randomly assigned
receive training across seven modules: (a) developing a to the control condition (business as usual) were instructed
vision, (b) organizing and structuring the classroom, (c) to continue teaching without alteration and to participate in
teaching expectations, (d) using proactive teaching, (e) sup- any professional development opportunities that were pro-
porting student motivation, (f) using data to make decisions, vided during the study period. At the end of the school year,
and (g) providing effective corrections. CHAMPS includes teachers were asked the number of hours of professional
a host of well-developed and user-friendly materials to sup- development they had received. Control teachers reported
port teacher implementation of the practices, including receiving on average 19 hr of professional development,
companion books and DVD compendiums. whereas intervention teachers reported on average receiv-
Adherence and fidelity of implementation of the ing 28 hr of professional development. Due to the wait-list
CHAMPS workshops and teacher implementation skills design, control condition teachers were offered the
were monitored over the course of the year, including dose/ CHAMPS intervention after all of the follow-up data were
exposure to training and coaching, teacher ratings of work- collected and the study was concluded.
shop quality, and observation of classroom implementation.
Regarding dose/exposure to the training, teachers in the
Measures
intervention were all exposed to the training workshops;
nearly all teachers attended all three workshops (attendance Student demographics. The participating school district pro-
rate was 92%–100% for each workshop) and the few teach- vided student-level data on race, sex, FRL, and special edu-
ers who missed a workshop due to illness or any other rea- cation status. Race was coded as Black, White, or Other.
son met with the CHAMPS coach to review missed material. Student sex was coded as 1 for male and 0 for female. FRL
Teachers rated the workshops with high satisfaction and was coded as 1 if they received FRL and 0 if not. Finally,
likelihood of recommending the training to others (mean with regard to special education services, students were
ratings of 4.80 and 4.87 on a scale from 1 to 5, with high coded as 1 for receiving services and 0 if they were not
scores indicating greater satisfaction). In addition, teachers receiving services.
reported that they expected good results from receiving the
training (4.60), agreed with the approach to behavior change Teacher implementation fidelity to CHAMPS. Independent
(4.69), and were confident it would be helpful with current observers conducted direct observations of teacher imple-
(4.33) and future (4.38) behavior problems in their mentation fidelity using the STOIC Rating Form (Sprick,
classrooms. 2013) three times across the school year, including immedi-
ately after the CHAMPS training in October, after the addi-
CHAMPS coaching. In this study, the CHAMPS coach was a tional CHAMPS training session in December, and at the
doctoral-level special educator. The coaching model includes end of the school year. Data were collected for both
22 Remedial and Special Education 42(1)

intervention and control classrooms. STOIC provides coefficients. Specifically, reliability of the communication
global ratings of each of the five key domains of CHAMPS arts test was .87 for Grade 6, .90 for Grade 7, and .91 for
practices: Structure for success, Teacher expectations, Grade 8, and the mathematics test produced reliability coef-
Observe systematically, Interact positively, and Correct ficients of .88 for the Grade 6, .90 for the Grade 7, and .87
calmly. Independent observers rated each of these five for the Grade 8 versions of the test (Missouri Department of
domains on a 0 (no evidence) to 4 (full evidence) rating Elementary and Secondary Education, 2015).
scale, and we computed a summary score of these ratings as In addition, we administered subtests of the 10th edition
a measure of adherence. The intraclass correlation for of the Stanford Achievement Test (SAT-10; Harcourt, 2004)
STOIC summary scores for this sample ranged between .92 pre-intervention, post-intervention, and in the spring of the
and .97 at each measurement time point. following year. The Stanford 10 is a widely used group-
We only collected post-intervention STOIC ratings administered standardized measure of academic achieve-
across all cohorts, so we used the baseline Classroom ment developed around national and state curriculum
Assessment Scoring System–Secondary (CLASS-S; Pianta standards as well as those trends promoted by national pro-
et al., 2008) Climate subscale as a covariate to equate class- fessional educational groups (Harcourt, 2004). It is designed
rooms on baseline climate. The CLASS Climate subscale to estimate academic achievement in reading, math, lan-
has been shown to be highly reliable and to predict student guage arts, and science. Extensive research has documented
achievement and social outcomes in a number of studies of the reliability and construct validity of the SAT-10 (Harcourt,
studies (Allen et al., 2013). The interclass correlation for the 2004). Subtest coefficient alphas all exceeded .80. We used
Climate subscale across all time periods was .751. two subtests, the Comprehension subtest for students in
reading classes and the Problem-Solving subtest for stu-
Behavioral outcome variables. The Teacher Observation of dents in math classes. Assessment occurred post-interven-
Classroom Adaptation–Checklist (TOCA-C; Koth et al., tion in April of the same school year.
2009) was completed by teachers and used as the primary
behavioral outcome measure. The TOCA-C is a measure of Moderator variable. Students were designated as at risk of
child behavior and is completed on each student. When behavioral problems if their teachers answered the question
completing the TOCA-C, teachers were asked to reflect on “Do you feel this student would benefit from individualized
the student’s behaviors in the last 3 weeks. Teachers rate behavioral support in the classroom?” in the affirmative.
student behaviors on a scale of 1 (never) to 6 (almost Although a teacher perceived and indicated that additional
always) on four major subscales (disruptive behaviors, con- supports would be beneficial, no behavior support plans
centration problems, emotional dysregulation, and proso- were developed as a result. Thus, this variable was used
cial behavior). Internal consistency for the TOCA-C for this simply as a teacher-perceived indicator of risk of behavior
sample was .95 to .97 for the concentation problems sub- problems. This variable was scored as a “1” if teachers indi-
scale, .88 to .94 for the disruptive behavior subscale, .85 to cated yes, and “0” if teachers indicated no. Being at risk of
.89 for the prosocial subscale, and .87 to .91 for the emotion behavior problems was positively correlated with students
dysregulation subscale. receiving in-school suspensions (r = .28, p < .001) and out
of school suspension (r = .30, p < .001) during the same
Academic outcome variables. Two district-level assessments school year, indicating that teacher perception of behavioral
were used as academic outcome variables. The Missouri risk at the start of the year was associated with exclusionary
Assessment Program (MAP) is a standardized, statewide punishment during the school year. This variable was also
assessment administered to all students across Grades 3 to correlated with baseline levels of observed disruptive
8. This assessment was designed to measure achievement behavior of the students gathered by independent observers,
toward statewide standards. Students received scores in (r = .15, p < .001), indicating that students who were iden-
communication arts and mathematics. Data included in the tified at risk had higher rates of observed disruptive class-
current study were from the end-of-year Mathematics and room behaviors.
Communication Arts subtests of the MAP. Since 2014, the
MAP assessments have been online and administered by the
Design
district’s testing vendor, CTB-McGraw-Hill. Scale scores
produced for each student describes achievement on a con- This study utilized a block cluster random assignment
tinuum that spans Grades 3 to 8. MAP scores of adjacent design in which teachers were either randomly assigned to
grades can be compared within a content area. The content receive the CHAMPS intervention or assigned to a wait-list
for the MAP assessments were developed using grade-level (business as usual) control condition. Randomization
blueprints that map onto the Missouri Learning Standards. occurred within school with the limitation that each group
MAP scaled scores had acceptable Cronbach’s alpha (intervention or control condition) not outnumber the other
Sinclair et al. 23

by more than 1. Teachers were recruited for the study across Level 2 (class):
three cohorts.
α 0 jk = β00 k + β01k (Condition) jk + µ jk
, u jk ~ N (0, τ22 ),
α qjk = βq 0 k + βq1k (Condition) jk , q = 1,..., Q.
Data Analysis
Missing data. A total of 102 teachers and 1,450 students
Level 3 (school):
across nine schools participated in the study that represent
our original sample. Of the 102 teachers, only one was not β00 k = γ 000 + ξk
2
able to complete the study due to moving out of the district β01k = γ 001 , ξk ~ N (0, τ3 ) ,
during the school year. Preliminary review of data for miss-
βq 0 k = γ q 00 , q = 1,..., Q.
ingness revealed that most missing data were from outcome
measures. The missing rates for the pretests of four social where X qijk represents student-level covariates, which
and behavioral outcome measures was 0.5%, whereas the include pretest, age at pretest, gender, race, FRL, special
missing rates for the posttests of four social and behavioral education status, grade level, cohort year in the study, and
outcome measures was 14.2% in the overall sample. Mobil- being at risk of behavior problems. (Condition)jk is a binary
ity (e.g., students moving out of the school district during variable indicating treatment condition (Condition = 0 for
the year) was the most predominant reason for missing stu- control group and Condition = 1 for treatment group). The
dent data. The differential missing rates between the treat- parameter, βq1k , estimates the moderator effects of the qth
ment and control groups were 0.7% for the pretest and 2.7% covariate and was assumed as constant across schools
for the posttest. Students with missing posttests were ( βq1k = γ q10 ).
excluded from the final analytic sample. The final analysis
included nine schools, 101 teachers, and 1,244 students.
Data from all teachers and students were included in the Results
analyses of social and behavioral outcomes. Data from
47 math teachers (587 students for Problem-Solving and Teacher Implementation Adherence to CHAMPS
594 students for MAP Math) and 54 reading teachers (632 Repeated-measure analyses of covariance (ANCOVAs) of
students for Reading Comprehension and 646 students for STOIC ratings from post-intervention in October and
MAP Communication) were included in the analyses of December, and at the end of the school year, adjusting
academic achievement outcomes. Multiple imputation, for baseline climate scores, revealed a significant interven-
using a Markov Chain Monte Carlo (MCMC) method in tion effect on STOIC summary scores across three post-
SAS PROC MI was used to impute missing data on pretest intervention observations adjusting for pretest scores, F(1,
and other covariates. The data were imputed five times. 96) = 7.51, p = .007, η2p = .074. The CHAMPS teachers
had consistently higher ratings of STOIC, and the average
Moderation analysis. Moderation analyses were conducted differences across three posttest observations represented a
to examine whether the treatment effects on student out- small effect (d = .28).
comes differed by teacher ratings on student needs for indi-
vidualized behavioral support in the classroom. For each of
the five imputed data sets, three-level hierarchical linear Moderation Analyses
models (HLM), in which students (Level 1) are nested
See (Herman et al., under review) for the results of the over-
within teachers (Level 2) and teachers are nested within
all main effects of the CHAMPS intervention. Moderating
schools (Level 3), were conducted using SAS PROC
effects of students identified by their teachers as at risk of
MIXED to examine the moderation effects on student
behavior problems have not been examined from the
behavior and academic outcomes. Each student’s pretest
CHAMPS randomized controlled efficacy trial prior. Tables
and demographic information were included at Level 1, the
1 and 2 provide the moderation effects of students identified
treatment variable was at Level 2, and its coefficient was
by their teachers as at risk of behavior problems on behav-
assumed constant across Level 3. We included the treatment
ioral and academic outcomes. Students at risk in the treat-
condition to predict the coefficient of the moderator vari-
ment group were found to have improvement on
able (being at risk of behavior problems). SAS PROC
concentration problems in comparison with peers at risk in
MIANALYZE was used to combine the results from the
the control group (b = –.19, p < .05, g = −0.07), meaning
analyses of five data sets. The statistical model follows:
that students identified as needing behavior supports by
Level 1 (student): their teachers who received the intervention demonstrated a
Q decrease in concentration problems in comparison with
Yijk = α 0 jk + ∑α qjk X qijk + eijk ,
eijk ~ N (0, σ2 ),
similar peers in control classrooms. No significant moderat-
q =1 ing effects were found for disruptive behavior, emotional
24 Remedial and Special Education 42(1)

Table 1. Behavioral Outcome Variables for the CHAMPS Study.

Concentration problems Prosocial behavior Emotional regulation Disruptive behavior

Variable b SE p b SE p b SE p b SE p
Intercept 1.51 0.64 .05 4.91** 0.38 <.01 2.37** 0.42 <.01 1.60** 0.39 <.01
Age 0.10 0.05 .05 −0.01 0.04 .69 −0.01 0.03 .82 0.01 0.03 .71
Female −0.16** 0.03 <.01 0.08* 0.04 .03 0.00 0.06 .93 −0.03 0.04 .42
Lunch status 0.03 0.06 .54 −0.04 0.06 .55 0.00 0.03 .98 0.05 0.02 .06
Support Time 1 0.10 0.06 .11 −0.17** 0.05 <.01 0.10 0.06 .09 0.10* 0.04 .01
Black 0.23** 0.06 <.01 −0.10* 0.04 .01 0.09** 0.02 <.01 0.11** 0.04 <.01
Other race −0.04 0.05 .45 0.19** 0.05 <.01 −0.13* 0.05 .01 −0.01 0.04 .70
Year 2 −0.03 0.10 .74 0.08 0.07 .29 0.01 0.08 .88 −0.01 0.07 .90
Year 3 0.13 0.08 .09 0.07 0.10 .48 0.01 0.09 .94 0.02 0.10 .86
Year 4 0.27* 0.08 .01 0.08 0.12 .53 0.06 0.09 .53 0.07 0.08 .40
Grade 7 −0.21 0.15 .14 0.07 0.10 .47 −0.10 0.07 .19 −0.05 0.07 .53
Grade 8 −0.13 0.16 .41 −0.04 0.14 .75 0.16 0.11 .13 0.08 0.08 .29
Pretest 0.72 0.03 <.01 0.71** 0.02 <.01 0.72** 0.02 <.01 0.75** 0.03 <.01
Intervention 0.09 0.14 .51 −0.03 0.11 .78 −0.16 0.15 .30 −0.11 0.09 .20
Intervention × Support −0.19* 0.07 .01 0.08 0.09 .34 0.04 0.10 .64 0.02 0.05 .67

Note. Test used was the Teacher Observation of Classroom Adaptation–Checklist (TOCA-C; Koth et al., 2009). CHAMPS = Conversation, Help, Activity,
Movement, Participation, and Success.
*p < .05. **p < .01 or less.

Table 2. Academic Outcome Variables for the CHAMPS Study.

Reading comprehensiona Problem-solvinga Communication artsb Mathematicsb

Variable b SE p b SE p b SE p b SE p
Intercept 699.90** 28.67 <.01 642.04** 19.60 <.01 1.64* 0.73 .02 1.29 1.12 .24
Age −4.16 2.73 .12 0.52 1.92 .78 −0.15* 0.07 .02 −0.10 0.09 .26
Female 2.28 1.96 .24 0.86 1.45 .55 0.08** 0.03 <.01 −0.01 0.03 .80
Lunch status −4.35 2.44 .07 −4.14** 1.39 <.01 −0.08** 0.03 <.01 −0.08 0.04 .08
Support Time 1 −9.03* 3.91 .02 −7.27* 3.32 .02 −0.22** 0.04 .01 −0.23** 0.08 <.01
Black −0.94 4.90 .84 −7.89 4.72 .09 0.04 0.09 .64 −0.16 0.10 .11
Other race 11.25* 4.64 .01 7.09* 3.37 .03 0.24** 0.09 <.01 0.07 0.04 .10
Year 2 −2.43* 0.99 .01 1.88 4.38 .66 0.06 0.05 .19 −0.01 0.08 .89
Year 3 −8.43** 1.24 .01 0.11 4.19 .97 −0.10 0.11 .35 −0.03 0.08 .68
Year 4 −6.91 3.92 .07 0.58 4.36 .89 −0.26** 0.07 <.01 −0.34** 0.09 <.01
Grade 7 15.21** 4.79 <.01 13.51** 4.56 <.01 0.25** 0.09 <.01 0.20 0.10 .05
Grade 8 42.96** 8.11 <.01 22.12** 7.42 <.01 0.44* 0.18 .01 0.53** 0.12 <.01
Pretest 25.35** 1.87 <.01 17.77** 2.05 <.01 0.71** 0.03 <.01 0.59** 0.04 <.01
Intervention −3.06 7.16 .66 8.56 5.85 .14 −0.08 0.08 .30 0.14 0.17 .41
Intervention × Support 4.37 5.47 .42 −1.87 4.17 .65 0.13* 0.06 .03 0.02 0.09 .84

Note. CHAMPS = Conversation, Help, Activity, Movement, Participation, and Success.

a
Subtest of the Stanford Achievement Test–10th Edition (SAT-10; Harcourt, 2004). b Subtest of the Missouri Assessment Program (MAP).
*p < .05. **p < .01 or less.

dysregulation, or prosocial behavior. With regard to aca- the intervention improved academically in comparison with
demic outcomes, students at risk in the intervention were similar peers in the control classrooms. No other significant
found to have a significant increase in their scores on the moderating outcome findings in relation to academic out-
Communication Arts subtest of the MAP assessment (b = comes were found. Figures 1 and 2 present the differential
.13, p < .05, g = 0.06) in comparison with peers at risk in effect sizes and their 95% confidence intervals between stu-
the control group. Thus, students who were identified by dents identified as at risk of behavior problems or not for
their teachers as at risk of behavior problems and received concentration problems and communication arts academic
Sinclair et al. 25

Figure 1. Differential effect sizes and 95% confidence intervals on concentration problems varying by the student risk of behavior
problems.

Figure 2. Differential effect sizes and 95% confidence intervals on communication arts academic performance varying by the student
risk of behavior problems.

performance, respectively. The figures indicate that the Discussion

intervention favored students identified as at risk by their
Universal prevention-oriented interventions have the abil-
teachers in reducing concentration problems and improving ity to not only affect students receiving the intervention in
communication arts academic performance. Although the which the intervention supports classroom behavioral
effect sizes for the moderator subgroups (at risk or not) expectations and academic instruction, but also specifi-
were not significant, the differences on the effect sizes cally those students who may be at risk of future problem
between the moderator subgroups were significant. behaviors or academic concerns (Lane et al., 2014; Wilson
26 Remedial and Special Education 42(1)

& Lipsey, 2007). In this study, we hypothesized that stu- content during the academic year and in turn have higher
dents who were perceived and identified as at risk by their communication arts scores than students at risk in compari-
teachers for needing individualized behavioral supports son classrooms.
would demonstrate improved behavioral and academic It is also noteworthy that the items on the Concentration
outcomes in response to the CHAMPS intervention. Problems subscale (concentrates, pays attention, works
Results from the moderation analysis found a significant hard, stays on task, is easily distracted, completes assign-
interaction between behavior risk and intervention status, ments, learns up to ability) are similar to the educational
indicating that students identified by their teachers as at construct of student engagement. In the main effect analy-
risk of behavior problems had significant improvements ses, intervention effects were partially mediated by direct
on teacher-reported concentration problems and commu- observations of students’ time-on-task (Herman et al., under
nication arts achievement scores compared with compa- review); that is, one reason CHAMPS benefited the com-
rable peers in wait-listed classrooms. This is consistent munication arts skills of all students was because students in
with the concept that academic and behavioral problems CHAMPS classrooms spent more time on task. Engagement
are interconnected and not isolated concerns (McIntosh and concentration skills are often measured by time-on-task
et al., 2008; Reinke et al., 2008). Thus, if a universal class- and have been identified as skills for student success
room management intervention can improve both aca- (Chafouleas et al., 2013). Student engagement has been
demic and behavioral outcomes for youth with increased researched and continues to be an important indicator of stu-
risk of behavior problems, it is worth exploring further dent academic and long-term success (see Brophy, 2010). In
implementation of CHAMPS and its preventive properties particular, there have been three common subcategories of
with regard to reducing the need for more resource-heavy student engagement that include behavioral, emotional, and
interventions or programs. cognitive engagement (Fredricks et al., 2004). Specifically,
The finding that the CHAMPS intervention had particular the concentration problems subtest of the TOCA is aligned
benefit on the concentration skills of students identified by with the principles of both behavioral and cognitive engage-
teachers as being in need of behavioral supports is notewor- ment. Behavioral engagement has been conceptualized as
thy. Main effect analyses reported a positive, although non- students’ conduct while they attend school, such as involve-
significant, trend on concentration problems for the whole ment in learning and academic tasks (e.g., asking questions,
sample (Herman et al., under review, Herman et al., 2018, attending to class discussions; Fredricks et al., 2004).
2019). The present findings suggest that changes in class- Cognitive engagement has been referred to as the idea of a
rooms where the CHAMPS intervention was implemented student’s investment in learning (e.g., persistence with dif-
affected student outcomes such that they largely contributed ficult tasks, using self-regulation strategies to guide learn-
to the positive trends. Concentration problems assessed in the ing; Finn & Zimmer, 2012). These two constructs are useful
present study align with inattention and symptoms of atten- in conceptualizing the implications of our findings. The
tion-deficit/hyperactivity disorder (ADHD). Youth with inat- measures administered post-intervention give insights to the
tention and ADHD symptoms are at high risk of negative short-term outcomes of students at risk of behavior problems
social, emotional, and academic outcomes. Thus, the finding who received the intervention; the engagement literature
that a universal classroom management intervention helps provides additional evidence for long-term outcomes
reduce these problems for youth in need of individualized because student engagement with school is associated with
behavioral support suggests that classroom-wide interven- positive academic and behavioral outcomes.
tions such as CHAMPS may be an efficient way to alter The effect sizes for findings in this study were modest;
engagement in school. The small effect sizes indicate that it is however, small effect sizes are common in universal pre-
likely that most of these youth would continue to need addi- vention studies (Durlak et al., 2011; Flay et al., 2005).
tional behavior support in the context of CHAMPS class- Because the entire population receives the intervention,
rooms; however, improvements in concentration problems small effects are expected, given the varying degrees of risk
may help these students benefit more from subsequent (i.e., many individuals would not develop behavior or aca-
behavior and academic support interventions and serve as a demic problems even without the intervention); yet, very
catalyst for further social and emotional development. small effects on a population level can result in dramatic
The significant moderation effects on communication improvements in public health outcomes (National Research
arts is also consistent with the concentration benefits expe- Council and Institute of Medicine, 2009). The relatively
rienced by students identified as at risk of behavior prob- small effects found among a subsample within this univer-
lems by teachers in this study. Inattention may interfere sal prevention trial points to the manner in which universal
with academic success. Thus, selective improvements in prevention interventions reach youth who otherwise might
attention skills—due to changes in classroom management need additional supports.
and instruction due to CHAMPS—for students at risk It is important to note that not all hypotheses were sup-
likely made it easier for students to access academic ported. Teacher perceptions of students at risk of behavior
Sinclair et al. 27

problems at baseline did not moderate intervention effects indicators of risk, such as special education status. It is
on disruptive behaviors, prosocial behaviors, self-regula- notable that this method resulted in a high rate of students
tion skills, and either measure of math skills used in the deemed to be in need of support (44%). Teacher report of
study. The main effects reported on disruptive behaviors student need is a commonly used method of referral to
and problem-solving skills suggest that these are universal behavior support teams, thus our method was in line with
effects; that is, on average, all students, regardless of behav- common school practices. Teachers are the most common
ior risk, were more likely to experience improvements in source of students’ social behavior and special education
these domains relevant to the comparison conditions. On referrals (Zima et al., 2005), and have been shown to pre-
the contrary, the null main effects on prosocial behaviors dict social behavioral problems (Koth et al., 2009; Reinke
and self-regulation suggest that CHAMPS did not affect et al., 2008). In addition, because they interact with large
these particular student outcomes either as a universal or numbers of youth during their careers, teachers provide a
selective effect. It is likely that prosocial and self-regulation valuable normative perspective on youth behaviors.
skill development requires explicit instruction not currently Moreover, the method we used to identify student risk is
offered by the CHAMPS intervention; whereas disruptive generalizable, as any school could ask the single-item
behaviors can be reduced by providing teachers the question of teachers in their building; it is this population
CHAMPS skills focused on classroom management. of students identified by teachers as being in need of
behavior support that the findings reported here generalize
to. Furthermore, Pas et al. (2011) reported that teacher use
Study Strengths and Limitations of office discipline referrals was a moderately valid indi-
The study had several notable strengths. The findings were cator of student behavior problems, suggesting that teacher
from a rigorous group-randomized trial with longitudinal referral can be a reliable source for the identification of
data. Analyses were conducted with multiple covariates and students who may experience problem behavior. That
accounted for clustering of students within classrooms. The being said, future research may address this limitation
large student sample across groups included a high percent- with the use of a standardized measure as an additional
age of underrepresented groups (nearly 70% Black and gate for understanding student behavioral support needs in
FRL). The study also had relatively high rates of assent and addition to teacher report.
participation from students and relatively low levels of Finally, multiple hypotheses were not confirmed,
missing data. Finally, prior to the start of training, teachers including intervention effects on outcomes on disruptive
thought it would be helpful and expected positive results behaviors, prosocial behaviors, self-regulation, and math
from implementation of the intervention; although small skills. Additional research using the CHAMPS interven-
effect sizes were detected, the clinically meaningful result tion is needed, as positive outcomes have been found
of positive behavior change is a strength. through its main effects on students. Furthermore, because
At the same time, this study is not without some limita- the CHAMPS protocol is focused on changing teacher
tions. First, while the sample includes underrepresented classroom behaviors, we may need additional explicit
groups, it is important to acknowledge that it is unknown instructions on how to improve student outcomes such as
how the findings will generalize to students in different self-regulation. Students at risk who benefited with dem-
educational settings or sociodemographic characteristics onstrating improvement in concentration problems and
due to the fact that this study had a high proportion of stu- communication arts may continue to demonstrate improve-
dents receiving FRL. We acknowledge that there is a com- ment longitudinally. Thus, additional research looking at
plex relation between schooling and experience of poverty. the impact of universal interventions on students at risk
Our findings must be considered with this in mind. While into the future may also be warranted. While behavioral
attempting to address the impact that incidence, depth, and academic outcomes are associated with one another,
duration, and timing of poverty has on student outcomes additional academic and instructional supports could
(Ferguson et al., 2007; Hopson & Lee, 2011), which is out- potentially bolster the effects of non-statistically signifi-
side of the scope of this article, we do recognize that the cant outcomes.
experience of poverty has been found to be a risk factor for
poorer academic and behavioral outcomes (Hopson & Lee,
Implications for Research and Practice
2011). Other aspects of poverty (e.g., limited resources,
quality of resources) may also contribute to students’ expe- For youth in middle school, exposure to preventive prac-
riencing poorer outcomes (Lacour & Tissington, 2011), yet tices is important as they move into an academic and social
with structural changes such as classroom-wide interven- behavior context that can shape their future. Risk and
tions like CHAMPS, risk may be reduced. opportunity compound as students begin to move away
Second, the study relied on teacher report of student from adult influence in favor of peers. By middle school,
need for individualized behavior support rather than other the prevention clock is running out of time to deter
28 Remedial and Special Education 42(1)

life-altering consequences, including drug and alcohol use, and also need to have time and support for sustained use of
pregnancy, and school dropout (Eccles et al., 1993; Henry such practices. Teachers may also benefit from the use of
et al., 2012). Teacher time and student/teacher ratios are standardized behavioral assessments that can be interpreted
more restrictive and not as conducive to individualized and used to help identify evidence-based interventions for
interventions as youth transition to middle school. To this classroom behavioral support. The use of standardized mea-
end, the findings from this study help direct research and sures may help reduce bias in behavioral reporting when the
practice in understanding how universal prevention inter- identification for student behavioral support needs are con-
ventions can affect all students, including those who may ducted. The findings from this research indicate that the
benefit from additional behavioral supports. investment in the CHAMPS intervention had positive
As previously mentioned, the student engagement litera- effects for students identified by teachers as at risk of
ture reports that students who are engaged emotionally, behavioral problems.
behaviorally, and cognitively with school ultimately have
shown better in-school and post-school outcomes than stu- Conclusion
dents who are not similarly engaged (Rumberger &
Rotermund, 2012). One of the main concerns of low student Current research has typically taken a reactionary approach
engagement during middle school is what happens when to fixing a problem. Whether teachers come to researchers
these students enter high school, including the risk of school or researchers go to teachers, a problem is identified and
dropout (Eccles et al., 1993; Henry et al., 2012). Dropout interventions are put into place to “fix” the problem. Yet,
has been long been thought of a long-term process rather through preventive approaches and increasing the use of
than a short-term decision (Christenson et al., 2001). This evidence-based behavioral management practices such as
process has been thought to start in middle school (Balfanz CHAMPs, teachers can help youth at risk of behavior prob-
et al., 2007) and certain populations, including students lems who are in need of support with less intensive strate-
with disabilities or students at risk of academic or behav- gies. This can allow those youth who truly need those more
ioral problems, have been shown to be at increased risk of intensive individualized supports to rise to the top, allow-
dropout (McFarland et al., 2016). ing for limited resources to be allocated accordingly. The
Successful dropout prevention practices are aligned with moderation analysis conducted to answer our hypothesis
the literature on student engagement. Because CHAMPS is indicated that youth identified by teachers to be at risk ben-
a universal classroom management intervention, it has the efited from the intervention in comparison with their peers
in control conditions. Further investigation into the impact
ability to affect a larger number of students than some drop-
of preventive supports on students demonstrating greater
out prevention interventions, which tend to target individual
risk in school is needed. Long-term impacts to prevent
students. This study provides evidence to suggest that
future educational concerns, such as problem behaviors
CHAMPS can impact those who may be at risk of behav-
and academic decline, while supporting positive long-last-
ioral problems and potentially help them stay engaged with
ing outcomes, should be an essential component in future
the academic curriculum. Future research can help deter-
research endeavors.
mine the long-term impact of CHAMPS. Future research is
also needed to investigate the additional impact of embed- Declaration of Conflicting Interests
ding behavior support planning or additional supports
The author(s) declared no potential conflicts of interest with
within the framework of CHAMPS for students and teach-
respect to the research, authorship, and/or publication of this
ers. Past research that investigated the additive value of article.
including elements of behavior support planning within the
context of a universal classroom management intervention Funding
found it to be impactful (Reinke et al., 2014).
The author(s) disclosed receipt of the following financial support
Furthermore, this research has implications for provid- for the research, authorship, and/or publication of this article: The
ing support for teachers in a key area of need: classroom research reported here was supported by the Institute of Educational
management. If teachers use an intervention and it has a Sciences, U.S. Department of Education, through Grant R305
positive and recognizable effect on student outcomes, this A130143 and R305B150028 to the University of Missouri. The
may contribute to increased self-efficacy for teachers. opinions expressed are those of the authors and do not represent
Research has found that more than half of educators sur- views of the Institute or the U.S. Department of Education.
veyed did not have confidence in their interventions for
supporting youth with behavior problems (Stormont et al., References
2011). The lack of such empowerment can lead to increases Allen, J., Gregory, A., Mikami, A., Lun, J., Hamre, B., & Pianta,
in teacher stress and burnout. Teachers need to have access R. (2013). Observations of effective teacher-student inter-
to interventions that are sensitive to the contextual demands actions in secondary school classrooms: Predicting student
Sinclair et al. 29

achievement with the Classroom Assessment Scoring System- Harcourt Assessment, (2004). Stanford Achievement Test 10th
Secondary. School Psychology Review, 42, 76–98. Edition.
Balfanz, R., Herzog, L., & Mac Iver, D. J. (2007). Preventing Henry, K. L., Knight, K. E., & Thornberry, T. P. (2012). School
student disengagement and keeping students on the gradua- disengagement as a predictor of dropout, delinquency, and
tion path in urban middle-grade schools: Early identification problem substance use during adolescence and early adult-
and effective interventions. Educational Psychologist, 42, hood. Journal of Youth and Adolescence, 41, 156–166. http://
223–235. dx.doi.org/10.1007/s10964-011-9665-3
Brophy, J. (2010). Motivating students to learn. Routledge. Herman, K. C., Reinke, W. M., & Dong, N. (2018, July). Group
Buell, M. J., Hallam, R., Gamel-McCormick, M., & Scheer, S. randomized evaluation of a classroom management program
(1999). A survey of general and special education teach- for middle school teachers [Paper presentation]. International
ers’ perceptions and inservice needs concerning inclu- School Psychology Association Conference, Toyko, Japan.
sion. International Journal of Disability, Development and Herman, K. C., Reinke, W. M., & Dong, N. (2019, July). Does time-
Education, 46, 143–156. on-task mediate the effects of classroom management train-
Chafouleas, S. M., Kilgus, S. P., Jaffery, R., Riley-Tillman, T. C., ing on student achievement? [Paper presentation]. Specifying
Welsh, M., & Christ, T. J. (2013). Direct behavior rating as Mediators of Classroom Social-Behavioral Interventions, at
a school-based behavior screener for elementary and middle the International School Psychology Association Conference,
grades. Journal of School Psychology, 51, 367–385. Basel, Switzerland.
Christenson, S. L., Sinclair, M. F., Lehr, C. A., & Godber, Y. Hopson, L. M., & Lee, E. (2011). Mitigating the effect of family
(2001). Promoting successful school completion: Critical con- poverty on academic and behavioral outcomes: The role of
ceptual and methodological guidelines. School Psychology school climate in middle and high school. Children and Youth
Quarterly, 16, 468–484. Services Review, 33, 2221–2229. https://doi.org/10.1016/j.
Conroy, M. A., Sutherland, K., Haydon, T., Stormont, M., & childyouth.2011.07.006
Harmon, J. (2008). Preventing and remediating young chil- Koth, C. W., Bradshaw, C. P., & Leaf, P. J. (2009). Teacher
dren’s chronic problem behaviors: An ecological classroom- Observation of Classroom Adaptation-Checklist: Development
based approach. School Psychology Review, 46, 3–17. and factor structure. Measurement and Evaluation in
Darney, D., Reinke, W. M., Herman, K. C., Stormont, M., & Counseling and Development, 42, 15–30.
Ialongo, N. S. (2013). Children with co-occurring academic Lacour, M., & Tissington, L. D. (2011). The effects of poverty on
and behavior problems in first grade: Distal outcomes in academic achievement. Educational Research and Reviews,
twelfth grade. Journal of School Psychology, 51, 117–128. 6, 522–527.
Durlak, J., Weissberg, R., Dymnicki, A., Taylor, R., & Schellinger, Lane, K. L., Oakes, W. P., & Menzies, H. M. (2014).
K. (2011). The impact of enhancing students’ social and emo- Comprehensive, integrated, three-tiered models of preven-
tional learning: A meta-analysis of school-based universal tion: Why does my school—and district—need an integrated
interventions. Child Development, 82, 405–432. approach to meet students’ academic, behavioral, and social
Eccles, J. S., Midgley, C., Wigfield, A., Buchanan, C. M., Reuman, needs? Prevention School Failure, 58, 121–128. https://doi.
D., Flanagan, C., & Mac Iver, D. (1993). Development during org/10.1080/1045988X.2014.893977
adolescence: The impact of stage-environment fit on young Lassen, S. R., Steele, M. M., & Sailor, W. (2006). The relation-
adolescents’ experiences in schools in families. American ship of school-wide positive behavior support to academic
Psychologist, 48, 90–101. http://dx.doi.org/10.1037//0003- achievement in an urban middle school. Psychology in the
066X.48.2.90 Schools, 43, 701–712. https://doi.org/10.1002/pits.20177
Ferguson, H. B., Bovaird, S., & Mueller, M. P. (2007). The impact McFarland, J., Stark, P., & Cui, J. (2016). Trends in high
of poverty on educational outcomes for children. Paediatrics school dropout and completion rates in the United States:
& Child Health, 12, 701–706. 2013 (NCES 2016-117). National Center for Education
Finn, J. D., & Zimmer, K. S. (2012). Student engagement: What Statistics, U.S. Department of Education. https://nces.ed.gov/
is it? Why does it matter? In S. L. Christenson, A. L. Reschly, pubs2016/2016117rev.pdf
& C. Wylie (Eds.), Handbook of research on student engage- McIntosh, K., Flannery, K. B., Sugai, G., Braun, D. H., &
ment (pp. 97–131). Springer. Cochrane, K. L. (2008). Relationships between academics
Flay, B., Biglan, A., Boruch, R., Castro, F. G., Gottfredson, D., and problem behavior in the transition from middle school to
Kellam, S., . . . Ji, P. (2005). Standards of evidence: Criteria high school. Journal of Positive Behavior Interventions, 10,
for efficacy, effectiveness and dissemination. Prevention 243–255. https://doi.org/10.1177/1098300708318961
Science, 6, 151–175. McIntosh, K., Gion, C., & Bastable, E. (2018). Do schools imple-
Fredricks, J. A., Blumenfeld, P. C., & Paris, A. H. (2004). School menting SWPBIS have decreased racial and ethnic dispro-
engagement: Potential of the concept, state of the evidence. portionality in school discipline? OSEP Technical Assistance
Review of Educational Research, 74, 59–109. https://doi. Center on Positive Behavioral Interventions and Supports,
org/10.3102/00346543074001059 University of Oregon.
Greenberg, M. T., & Abenavoli, R. (2017). Universal interven- Missouri Department of Elementary and Secondary Education.
tions: Fully exploring their impacts and potential to produce (2015). Missouri assessment program grade-level assess-
population-level impacts. Journal of Research on Educational ments [Technical report 2015]. https://dese.mo.gov/sites/
Effectiveness, 10, 40–67. default/files/asmt-gl-2015-tech-report.pdf
30 Remedial and Special Education 42(1)

Morrissey, T. W., Hutchison, L., & Winsler, A. (2014). Familiy Sprick, R. (2013). Discipline in the secondary classroom: A posi-
income, school, attendance, and academic achievement in tive approach to behavior management. John Wiley & Sons.
elementary school. Developmental Psychology, 50, 741–753. Sprick, R., Garrison, M., & Howard, L. M. (1998). Library:
https://doi.org/10.1037/a0033848 Management, motivation & discipline: CHAMPs: A proactive
National Research Council and Institute of Medicine. (2009). and positive approach to classroom management for grades
Preventing mental, emotional, and behavioral disorders K-9. Sopris West.
among young people: Progress and possibilities. The National Stormont, M., Reinke, W. M., & Herman, K. C. (2011). Teachers’
Academies Press. knowledge of evidence-based interventions and available
Pas, E. T., Bradshaw, C. P., & Mitchell, M. M. (2011). Examining school resources for children with emotional and behavioral
the validity of office discipline referrals as an indicator of problems. Journal of Behavioral Education, 20, 138–147.
student behavior problems. Psychology in the Schools, 48, Stormont, M., Reinke, W. M., Herman, K. C., & Lembke, E. S.
541–555. (2012). Practical intervention in the schools series: Academic
Pavri, S. (2004). General and special education teachers’ prepara- and behavior supports for at-risk students: Tier 2 interven-
tion needs in providing social support: A needs assessment. tions. Guilford Press.
Teacher Education and Special Education, 27, 433–443. Sugai, G., & Horner, R. H. (2006). A promising approach for
Pianta, R. C., Hamre, B. K., Hayes, N., Mintz, S., & LaParo, K. M. expanding and sustaining school-wide positive behavior sup-
(2008). Classroom Assessment Scoreing System-Secondary port. School Psychology Review, 35, 245–259.
(CLASS-S). University of Virginia. Sweeten, G. (2006). Who will graduate? Disruption of high school
Reinke, W. M., Herman, K. C., Petras, H., & Ialongo, N. S. (2008). education by arrest and court involvement. Justice Quarterly,
Empirically derived subtypes of child academic and behavior 23, 462–480.
problems: Co-occurrence and distal outcomes. Journal of Tobin, T. J., & Sugai, G. M. (1999). Using sixth-grade school
Abnormal Child Psychology, 36, 759–770. records to predict school violence, chronic discipline prob-
Reinke, W. M., Stormont, M., Herman, K. C., Puri, R., & Goel, lems, and high school outcomes. Journal of Emotional
N. (2011). Supporting children’s mental health in schools: Behavioral Disorders, 7, 40–53.
Teacher perceptions of needs, roles, and barriers. School Wagner, M., & Newman, L. (2012). Longitudinal transi-
Psychology Quarterly, 26, 1–13. tion outcomes of youth with emotional disturbances.
Reinke, W. M., Stormont, M., Herman, K. C., Wang, Z., Newcomer, Psychiatric Rehabilitation Journal, 35, 199–208. https://doi.
L., & King, K. (2014). Use of coaching and behavior support org/10.2975/35.3.2012.199.208
planning for students with disruptive behavior within a uni- Walker, H. M., Nishioka, V. M., Zeller, R., Severson, H. H., &
versal classroom management program. Journal of Emotional Feil, E. G. (2000). Causal factors and potential solutions for
and Behavioral Disorders, 22, 74–82. the persistent underidentification of students having emo-
Rumberger, R. W., & Rotermund, S. (2012). The relationship tional or behavioral disorders in the context of schooling.
between engagement and high school dropout. In S. L. Assessment for Effective Intervention, 26, 29–39.
Christenson, A. L. Reschly, & C. Wylie (Eds.), Handbook of Wilson, S., & Lipsey, M. W. (2007). School-based interventions for
research on student engagement (pp. 491–513). Springer. aggressive and disruptive behavior: Update of a meta-analysis.
Rusby, J. C., Crowley, R., Sprague, J., & Biglan, A. (2011). American Journal of Preventive Medicine, 33, S130–S143.
Observations of the middle school environment: The context Zima, B. T., Hurlburt, M. S., Knapp, P., Ladd, H., Tang, L., Duan,
for student behavior beyond the classroom. Psychology in the N., . . . Wells, K. B. (2005). Quality of publicly-funded out-
Schools, 48, 400–415. https://doi.org/10.1002/pits.20562 patient specialty mental health care for common childhood
Sprick, R. (2009). CHAMPS: A Proactive & Positive Approach to psychiatric disorders in California. Journal of the American
Classroom Management (2nd ed.). Pacific Northwest Publishing. Academy of Child and Adolescent Psychiatry, 44, 130–144.