volume 43 number 6 february 2011 focus on

VOLUME 43 NUMBER 6 FEBRUARY 2011

FOCUS o n

RTI School-Based Practices and Evidence-Based Models

Daryl F. Mellard, Amelia Stern, and Kari Woods

Response to intervention (RTI) is widely used as a framework lor providing highquality instruction and interventions that are matched to students' needs, as well as ameans of integrating important federal policies. "A multi-tiered system of interventions isrecommended as a means to integrate educational problem-solving across educational lev-els, consistent with federal legislation (e.g.. Individuals with Disabilities EducationImprovement Act of 2004; No Child Left Behind Act of 2001,) and scientific research"(Batsche, 2006, p. 3), including both general and special education. The concept of RTIdid not begin with these broad goals; rather, RTI was initially conceived as a preventionframework providing early intervention to students at risk of reading failure. Special edu-cators and others soon began to see that RTI frameworks could contribute important infor-mation to the identification of specific learning disabilities (SLD). This recognition waseventually encoded in IDEA 2004 (PL. 108-446), permitting RTI as a component of SLDidentification. The RTI idea continued to grow, with application made to not just readingbut all academic content areas, as well as behavior. Further, RTI processes extendedbeyond early identification to annual efforts to identify students at risk of academic orbehavioral failure throughout elementary, middle, and high school (Johnson, Smith, &Harris, 2009).

Initially, variability in the way that RTI was conceptualized, researched, and prac-ticed was expected, and even necessary, as new ideas transformed into current state andlocal educational agencies policies and practices. However, when IDEA 2004 permittedalternative assessments of SLD identification, such as including students' responsivenessin an RTI framework, this variability began to pose equity, quality, and efficiency issuesfor state education agencies, school districts, and local schools (Griffith, Parsons, Burns,VanDerHeyden, & Tilly, 2007). A review of nine state education agencies' RTI documents(Bocala, Mello, Reedy & Lacireno-Paquet, 2009) found a high degree of consistency witha model prescribed by National Research Center on Learning Disabilities (NRCLD; Mel-lard, Byrd, Johnson, Tollefson, & Boesche, 2004), yet within this framework statesallowed for substantial local discretion. Such place-to-place differences in local RTIimplementation have implications for equitable distribution of resources, SLD identifica-tion, disproportional representation of some groups, and the proper role of special educa-tors (Council for Exceptional Children [CEC], 2008; Fletcher & Vaughn, 2009; Fuchs &

Dr. Mellard is an associate research professor, Ms. Stem is a graduate research assistant, and Ms. Woods is aprogram assistant at the University of Kansas, Center for Research on Learning.

Copyright © Love Publishing Company, 2011

FOCUS ON EXCEPTIONAL CHILDREN FEBRUARY 2011

Deshler, 2007; Fuchs, Fuchs, & Compton, 2004; Hollen-beck, 2007; Mellard & Johnson, 2008, 2010; Reynolds &Shaywitz, 2009). Studies demonstrated how such local deci-sion making resulted in wide variability from district to dis-trict in tier structures, screening and progress measures, cut,scores and other decision criteria, and the potential forequity issues that challenge the viability of RTI (Mellard,McKnight, & Jordan, 2010; Mellard, McKnight & Woods,2009). Likewise, the CEC and Learning Disabilities Associ-ation (LDA, 2006) expressed concerns about RTI poten-tially causing delays in comprehensive evaluation referralsfor children suspected of having SLD and emphasized theneed for partnership of all school personnel and families toidentify and address the academic and behavioral needs oflearners.

In light of these issues, Fuchs and Fuchs (2009) arguedfor a unified model of RTI that "encourages shared under-standing among all school-based practitioners about inter-vention intensity, roles and responsibilities, and constructive

FOCUS o nExceptional

cnildrenISSN 0015-51IX

FOCUS ON EXCEPTIONAL CHILDREN (USPS 203-360) is pub-lished monthly except June, July, and August as a service to teachers,special educators, curriculum specialists, administrators, and those con-cemed with the special education of exceptional children. This publica-tion is annotated and indexed by the ERIC Clearinghouse on Handi-capped and Gifted Children for publication in the monthly CurrentIndex to Journals in Education (CUE) and the quarterly index. Excep-tional Children Education Resources (ECER). The full text of Eocus onExceptional Children is also available in the electronic versions of theEducation Index. It is also available in microfilm from Serials Acquisi-tions, National Archive Publishing Company, P.O. Box 998, Ann Arbor,MI 48106-0998. Subscription rates: individual, $50 per year; institu-tions, $68 per year. Copyright © 2011, Ixive Publishing Company. Allrights reserved. Reproduction in whole or part without written permis-sion is prohibited. Printed in the United States of America. Periodicalpostage is paid at Denver, Colorado. POSTMASTER: Send addresschanges to:

Love Publishing CompanyExecutive and Editorial Office

P.O. Box 22353Denver, Colorado 80222

Telephone (303) 221-7333

CONSULTING EDITORS

Steve GrahamVanderbilt University

Ron NelsonUniversity of Nebraska-Lincoln

Eva HornUniversity of Kansas

Carrie E. WattersonSenior Editor

Stanley F.Publisher

and effective relationships between general and special edu-cation" (p. 41). Fletcher and Vaughn, however, indicatedthat they were "less interested in promoting a unitary model... as long as schools use ongoing student data to informdecisions" (2009, p. 48).

Even as these debates continue, schools continue toimplement a variety of RTI frameworks. Thus, our purposehere is to demonstrate the variability in school-based prac-tices and to describe some prominent, evidence-based RTImodels. We discuss the implications of these differences inpractices and evidence-based models for learners, educators,and administrators. We anticipate that our descriptions anddiscussion will assist education leaders in making informeddecisions when planning, selecting, and implementing anRTI model that best fits their context.

VARIABILITY IN SCHOOL-BASEDRTI PRACTICES

From a survey of and interviews with schools practicingRTI, we briefly describe five elementary school-based RTIframeworks with the same overall purpose—to improveschool-wide reading aehievement (Mellard, McKnight,Woods, & Frey, n.d.). None of these schools reported fullyadopting an evidence-based model and did not considertheir model fully implemented at the time. Rather, eachschool implemented the RTI components that best fit its par-ticular context, value system, and available resources. In allcases, the school used RTI as a component of their compre-hensive evaluation for SLD identification.

School A. School A's 517 students were in grades K ^ .The school's RTI model had four tiers in which all except 5students with severe cognitive disabilities were screened forreading risk and placed in appropriate instruction. Seventy-six percent of students received instruction in Tier 1, whereSignatures reading series (Harcourt Brace) was supple-mented by repeated readings, sight word reading, pairedreading, and Blast Off readers (Bellwether Media). Screen-ing and progress monitoring was based on Dynamic Indica-tors of Basic Early Literacy Skills {DIBELS\ Good &Kaminski, 2002) oral reading fluency (ORF) benchmarksfor first and second grades, DIBELS ORF benchmarks andpercentile ranking (below 56th) on Terra Nova (CTBMcGraw-Hill) for third grade, and Hasbrouck and Tindal(1992) 50th and 25th percentile rankings in ORF for fourthgrade.

School B. School B enrolled 366 students in grades K-5.Its RTI model had three tiers in which all except a few stu-dents with low cognitive skills were screened. On average,80% of students received Tier 1 instruction with the follow-ing programs and materials; Literacy Place (Scholastic),Read Well K® and Read Well 1® (Sopris West Educational

' Services), Guided Reading, Corrective Reading and Cor-\ rective Decoding (McGraw-Hill SRA), Reading Mastery! (McGraw-Hill SRA), REWARDS® (Cambium LearningGroup), Peer-Assisted Learning Strategies'^ (Vanderbilt

' Kennedy Center), and Read Naturally (Read Naturally,I Inc.). Screening for risk and response to instruction in the': core curriculum was determined based on tbe dual discrep-: ancy data of level and slope (rate of growth) using DIBELS' measures at all grade levels and other evidence of respon-i siveness in reading components.

School C. School C had an enrollment of 977 students] located in three facilities with a four-tier RTI model. All stu-' dents were screened with the exception of a few studentswith severe disabilities who could not complete the assess-

; ment process. Sixty-three percent of students participated: solely in Tier 1 core instruction using Houghton MifflinI Reading (Houghton Mifllin) and The Language Tool Kit¡ (Orton Gillingham), and an additional 23% in Reading

P/M.S® (Taylor Associates/Communications, Inc.). Screening; and response to instruction in the core curriculum was basedI on level and slope using DIBELS benchmarks and belowI 70th percentile on local benchmark assessments in K-1: forI grades 2-5, CBM tluency levels, Houghton-Mifilin, and: Gates-McGinitie reading norms.' School D. School D's enrollment of 380 students was ingrades K-5. This school's RTl model had three tiers in

, which all students participated except those students whoseparents or IEP team opted for exclusion from the screening

i process. Core instruction in Tier I was provided to 68% ofstudents using Direct Instruction (MacMillian-McGraw),

j Read Naturally (Read Naturally, Inc.), REWARDS® (Cam-biutn Learning Gtoup), Open Court Phonemic Awareness

j and Phonics Kit (McGraw-Hill), and Waterford Early Read-ing Program'''^ (Pearson). Screening for risk and response to

i instruction in the core curriculum was determined by dualj discrepancy scores below the 25th percentile on level and'•• the slope on one or tnore DIBELS indices. In addition, thei school staff used confirtnatory data from CBMs, the 5/0/7-ford Achievement Test (Pearson), Qualitative Rectding Inven-j tory (Pearson), and Ekwall Reading Inventor}' (Allyn &! Bacon) in their decisions.! School E. School E had a total enrolltnent of 480 students¡ in grades K-5, with 257 K-3 students included in RTI data.At the time of these activities, the school reported only twotiers of intervention, with 47% of students in Tier 1, 19% in

¡Tier 2, and the remaining students receiving instructionI through English language learner and special education pro-I grams. Special education was a separate instructional set-S ting. Core instruction used Open Court (McGraw-Hill) cur-¡riculum. Students' risk status was based on a dual'discrepancy criterion indicated by the level and slope of! benchmarks.

EQUITY AND EFFICIENCY IMPLICATIONS

While we can describe what these five schools imple-mented, we cannot test whether students would receive thesame level of instructional intensity from one school to thenext (i.e., equitable treatment). Nor can we evaluate whetherschools are providing services in the most efficient manner(i.e., not overidentifying need for intensive instruction).However, in our view. Schools D and E are not "exemplary"implementations because they delivered intensive interven-tions to a high percentage of students. A well-developed RTImodel is intended lo prevent academic and behavioral diffi-culties and the need for extensive intervention.

Without a prevention focus, an RTI system cannot func-tion efficiently. Schools D and E's distributions of studentsamong the tiers and services suggest that instruction and thecore curriculum were deficient for a significant number ofstudents. These data provide programmatic evidence that,for far too many students, the core curriculum did not pro-vide a positive instructional experience and thereforedirected too many resources toward intensive services forthese students. An inefficient model is not likely to have arobust or strong implementation or even be sustainable overtime. Nor would we expect that such a model would gener-ate student outcome evidence (e.g., marked improvementsand positive trends from historical data) that would supportcontinuation of the system.

COMPARISON OF EVIDENCE-BASEDRTI MODELS

The various ways in which schools have implementedRTI are, in part, the result of having limited guidance fromtheory and research at the time they adopted their practices.However, several evidence-based RTl models have beenvalidated in the literature, and school staff could now adoptthem. We chose seven such models to describe and compare.We identified them by conducting a literature review usingsuch search terms as RTI, Response to Intervention, andResponsiveness to Intervention. We excluded publicationsfor which a complete framework of RTI (i.e., screening,progress monitoring, and tiered interventions) was not pub-lished. Many research articles have been written about aspecific component of RTl, bul our additional searches lorinformation on the other vital components of RTl from thoseauthors was inconclusive. The selected models wete sup-ported by varying numbers and quality of etnpirical studiesand were not typically available in a complete or integratedpackage for school staffs. At this time, research does notprovide a preponderance of evidence to support one RTImodel over another, and this review should not be tteated ascomprehensive but rather as a starting point for planning


and evaluating available resources prior to implementation.Proponents or designers of the seven selected models are:

Batsche, Curtis, Dormán, Castillo, and Porter (2008); Fuchsand Fuchs (2005); Johnson, Mellard, Fuchs, and McKnight(2006); Vaughn (2005); Shinn (2008); Chun and Witt(2008); and Sugai and Homer (2007a). We emphasize at theoutset of our comparisons that these models have morepoints of commonality than of difference. However, the dis-tinctions among models have varying levels of importance,depending on one's values on specific topics (e.g., ease ofimplementation, academic or behavioral emphasis, schoolstructures, research foundation, match to a school or dis-trict's professional development, and resource supports).

Overviews

Batsche et al. (2008) emphasized two unique features:exclusive use of a problem-solving approach and applica-tion to both behavioral and academic outcomes. Like manyother models, this model is primarily designed for elemen-tary schools and values early intervention, effective instruc-tion, and multi-tier service delivery in the context of prob-lem-solving processes.

In contrast to the problem-solving approach, Fuchs andFuehs (2005), Johnson et al. (2006), and Vaughn (2005) rec-ommended standard treatment protocols, that is, substan-tially uniform and well-established conventions governingthe instructional treatment of all students. These models dif-fer slightly from one another, in that Fuchs and Fuchsaddressed reading and math, while the other two includeonly reading; further, Vaughn extended beyond the elemen-tary level to eighth grade. Vaughn cautioned that herapproach is descriptive rather than prescriptive and thus val-ues flexibility to incorporate a variety of research-based pro-grams where appropriate. Like other models, Vaughnincluded validated interventions selected according to stu-dents' needs, delivered individually or in a variety of groupsizes and in a setting makes that the most sense for a givenschool.

Fuehs and Fuchs (2005) placed emphasis on early identi-fication through assessment of students at risk of academicfailure. Johnson et al. (2006) built specifically on this prin-ciple by designating essential components to accomplish thesame goal Fuchs and Fuehs articulated. The components ofJohnson et al. reflected IDEA and NCLB statues by requir-ing: (a) scientifically-based instruction for all students, (b)school-wide screening for students at risk of academic fail-ure, (e) research-based interventions appropriate to studentneeds, (d) continuous progress monitoring with the inter-ventions, and (e) fidelity to the process as prescribed.

Shinn's (2008), Chun and Witt's (2008), and Sugai andHorner's (2007a) models combine standard treatment proto-col with problem-solving methods. Shinn applied the

approach to secondary academics (middle school and highschool), while Chun and Witt applied it to academics at anygrade level, and Sugai and Horner used the combinedmethod to address behavior at any grade level. Chun andWitt valued empiricism, efficiency, and simplicity of inter-ventions and procedures. Schools using this model accom-plish these goals, in part, by defining problems in solvableterms, using very targeted interventions. Shinn emphasizedscientifically-based progress monitoring tools and twoissues particularly relevant in secondary settings: basic skillinterventions for content learning and consultation with theclassroom instructors. Shinn suggested combining his acad-emic model with Sugai and Horner's behavioral model,when needed. Sugai and Horner designed their model toachieve functional behavioral outcomes and prevent orreduce prevalence of behavioral incidents. They placedvalue on balance among data, practices, fidelity, and out-comes and emphasized redesigning instruction and environ-ment to improve behavior. Table 1 presents an overview ofeach model by the type of interventions the model employs,the applied content areas, grade levels for apptopriate use,and its core principles.

Tier Structure

Each model has a three-tiered intervention structure.Table 2 describes the models' three primary tiers and inter-vention features at each tier. Consensus exists around thedefinition of the first tier. Tier 1, as effective classtoominstruction with a viable general education curriculum. Con-sistent with an emphasis on efficiency, Chun and Witt(2008) allowed Tier 1 to include reasonable modifications tocurriculum or instruction, which could possibly avoid moreintense, costly, or disruptive Tier 2 intervention. Shinn(2008) included behavioral interventions and suppotts dur-ing academic instruction, recognizing the importance ofclassroom participation in middle and high school contentcourses.

Tier 2 in each model represents a context for providinginterventions to students who do not respond to generalinstruction or who were predicted to be at risk through ascreening process. Tier 2 is regarded as mote intenseinstruction, though not necessarily a different eutrieulum.Models that emphasize a standard treatment protocol aremore likely to impletnent a curricular change. Each modeldescribes Tier 3 as more intensive instruction than in Tier 2,and the Fuchs and Fuehs (2005), Johnson et al. (2006), andShinn (2008) models explicitly identified Tier 3 as specialeducation or highly individualized instruction.

Screening

Most RTI models operate from the principle that the put-pose of screening is to identify those students who may be

TABLE 1. Model Overviews

: Principle Author Model Name TypeContentArea(s)

TargetedGrades Core Principles

!: Batsche et al.

Vaughn

Shinn

Problem SolvingModel / RTI(PSM/RTI)

Problemsolving

BehaviorAcademics

Elementary

: Fuchs & Fuchs RTI

Johnson et al. RTI

RTI

Standardtreatmentprotocol



ReadingMath

Reading

Reading

Elementary

Elementary

K-8

RTI Combined Academics Secondary

I

1

, Chun & Witt

¡1!1 Sugai & Horner

1

1

System toEnhanceEducationalPerformance(STEEP)

PositiveBehaviorIntervention &Support (PBIS)

Combined Academics All Levels

Combined Behavior All Levels

Teach all children effectivelyEarly interventionMulti-tier service delivery modelProblem-solving methodology that

defines and validates the problem,identifies and implements interven-tion, and evaluates student responseto intervention

Assessment and identification linked toearly intervention

Scientifically-based instructionSchool-wide screeningContinuous progress monitoringResearch-based interventionsFidelity of implementation

Descriptive, not prescriptive modelFlexible to incorporate any research-

based programInterventions based on student needsInterventions delivered in whole class,

small group, pairs, 1:1Instructional groups can be vi/ithin

class, within grade, or across grade

Scientifically-based progressmonitoring tools

Basic skill interventions for contentlearning

Positive Behavior Intervention &Support (PBIS)

Mainstream consultation

Empirical, efficient, and simpleinterventions and procedures

Define student problems in solvableterms

Target interventions on academic skills

Balancing data, practices, fidelity, andoutcomes

Functional behavioral outcomesPreventing behavioral incidents and

reducing prevalenceRedesign teaching/environment to

improve behavior

j at risk or struggling to achieve academic or behavioral out-I comes. Table 3 presents screening and progress monitoring: methods. Batsche et al. (2008) used screening data to elimi-¡ nate two potential problems before designating students as

at risk of school failure. First, they suggested that if univer-sal screening yields less than an 80% pass rate, the generaleducation or core curriculum may not be effective and thusreflects a systemic problem to be resolved. In the previously


TABLE 2. Intervention Features by Tier

Principle AuthorTieri

Primary PreventionTier 2

Secondary PreventionTier 3

Tertiary Prevention

Batsche et al.

Fuchs & Fuchs

Johtison et al.

Vaughn

Shinn

Chun & Witt

Sugai & Horner

Core instruction for all students;school may make modifica-tions to general instruction toincrease effectiveness and/orto reduce disproportionality

Classroom instruction that pro-vides at-risk students evi-denced-based curriculum witha sound Instructional design

Effective, scientifically-basedinstruction in general education

Classroom instruction or corereading program for all students

High quality, research-based instruc-tion in content knowledge withbehavior interventions and sup-ports, school-wide pedagogy withteaching routines and strategies,effective classroom management,study and organizational skills,curriculum modification, and anobjective, open grading system

Core curriculum with problemsolving for individual studentmodifications

Stated behavioral expectations,teaching expectations andstrengthening expected behaviors,and consequences for behaviorsfor all students and staff in everysettings and with a data collectionsystem

Supplemental instruction, inaddition to core, for small-group instruction at intensityand frequency necessary toensure improvement

Supplementary, diagnosticinstructional trials for non-responders, provided in smallgroups of students with similarabilities for 30 minutes threetimes per week with a qualifiedteacher or aid; includes fre-quent progress monitoring

Specialized group instructionthat supplements generaleducation for at-risk studentsto remediate deficits; providedin small groups, 3 or 4 timesper week, in 30 to 60 minutesessions for 9 to 12 weeks

Supplemental instruction, inaddition to core instruction,for identified students; e.g.,K-3 small group, 20-30minutes daily, 10-20 weeks;middle school 50 minutes daily,12 weeks

Remedial programs to helpgeneralize learning to corecurriculum targeted towardsbasic skills and contentknowledge, such as remedialreading programs

Class-wide peer tutoring and peer-assisted learning strategies forwhole class, if needed, andstandard protocol interventionsselected from pool of interven-tions based on student needs

Specialized interventions forunresponsive students, includingstructure and predictability,positive reinforcements,and good home-schoolcommunication

Intensive instruction, inaddition to core, for a verysmall group or individualswith narrow focus ondefined skills instruction

Interventions that are gen-erally systematic andexplicit, with increased fre-quency and duration inhomogenous group or 1:1by highest expertiseteacher with students eligi-ble for special education

Special education or highlyindividualized instruction

Intensive instruction forstruggling students basedon needs, not necessarilyspecial education but typi-cally provided by a readingteacher, special educator,or trained interventionist

Intensive or special educa-tion matched to studentneeds and/or mainstreannconsultation and coachingto support students

Increased frequency andduration or change ofintervention based onproblem-solving methods

Function-based, person-centered, and intensiveinterventions for secondarilyunresponsive students pro-vided with consideration toschool, family, community,and mental health context

TABLE 3. Screening and Progress Monitoring Features

Principle Author Screening Features Progress Monitoring Features

i Batsche et al.

Fuchs & Fuchs

! Johnson et al.

1 Vaughn

Shinn

Chun & Witt

Sugai & Horner

Screen all students at beginning of school year withan 80% pass rate indicating an effective corecurriculum

Assess group identified as at risk for overrepresentationStudents below benchmark identified as at risk; enter

into individual problem-solving processFor behavior, consider student, teacher, peer, curriculum,

classroom/school, and family/community factors

Screen all students to identify at-risk statusAt-risk students monitored 5 to 8 weeks to reduce false

positives and identify those not responding toinstruction before placement in Tier 2

School-wide screening to identify at-risk students,three times per year

Overidentification is a worse error than under-identification

Use screening and at least 5 weeks of weeklyprogress monitoring to identify students needingintervention

Benchmark testing all K-3 students three times per yearScreening in word identification, comprehension, and

fluencySkill testing middle school students

Universal reading screening with grades 5, 6, & 9using Maze CBM probes

At high school level, screening is individual andtargeted rather than universal

Universal screening with validated CBMsAdministration of screening probes follows a scripted

implementation protocolDetermine effectiveness of core instruction, dispropor-

tionality, differences among classes beforeexamining individual student data

Performance motivation vs. skill assessment beforeplacement in Tier 2

Screening data examined at school, class, andstudent levels

CBMs and behavior measures at frequencysufficient to inform instruction and evaluatestudents' responsiveness to intervention

Chart progress vs. a baseline of peer per-formance (not standardized reference point)

At-risk students monitored 5 to 8 weeks usingCBM probes before intervention

Tier 2 progress monitored weekly for 8 weeksduring intervention

Progress monitor more frequently fornonresponders

Progress monitor at all tiers with screening probesused as progress monitoring of students inTieri

Progress monitor at-risk students for at least5 weeks before moving to Tier 2

Progress monitor students in Tier 2 and Tier 3at least 2 times per week

Formative assessment for instructional changesFrequency spectrum from three times per year for

those making adequate progress to weekly forthose with persistent reading deficits

Use standardized CBM tools and sufficientalternate forms of equal difficulty for repeatedtesting; course grades can be used with weeklyrates of improvement toward specified goals inhigh school

Tiers 1 and 2 nonresponsiveness determined after4 to 6 weeks of progress monitoring

Tier 3 progress monitoring more frequentdepending on intervention

Oral reading fluency, weekly, use median of threeprobes

Reading comprehension using Maze CBM every4-5 weeks

One math probe administered weekly

Collect and review data on individual students,classrooms, and school-wide on a weekly,monthly, quarterly, and annual basis.

Office discipline referrals, academic achievementscores, attendance/tardies, suspensions, andbehavior incidents monitored on a weekly,monthly, and annual basis


described examples, several schools' data reflected suchsystemic problems. Second, to eliminate any problem of dis-proportionality in placements, Batsche et al. called for eval-uating the group identified as at risk for overrepresentationof a particular demographic characteristic (e.g., racial or eth-nic status). If bias is indicated, individual student problem-solving processes should address it. Otherwise, the facultywould then begin a problem-solving process for each stu-dent judged to be at risk. In the case of behavior issues, themodel is designed to collect multiple perspectives for theproblem-solving process, including perspectives of the stu-dent, teacher, peers, curriculum, classroom, school, andfamily or community.

During screening, Fuchs and Fuchs (2005) emphasizedavoiding false positive predictions of students at risk. Theysuggested leaving the identified subgroup in general class-room instruction for 5 to 8 weeks while monitoring theirprogress to confirm nonresponsiveness to instruction. Thistwo-step confirmatory approach conserves resources foronly those students clearly needing secondary or tertiaryinterventions because of their lack of progress in the corecurriculum. Similarly, Johnson et al. (2006) suggested atleast 5 weeks of monitoring for students at risk in Tier 1before moving a student to Tier 2. They also suggestedscreening all students three times per year to identify addi-tional students who are not responding to the instructionthroughout the year.

Like Johnson et al. (2006), Vaughn (2005) recommendedreading assessments three times a year for all students inK-3. This more frequent screening is important at theseearly grade levels because of the multiple stages of devel-opment experienced by beginning readers in a relativelyshort time period. Vaughn suggests screening for word iden-tification, comprehension, and fluency skills. For middleschool students, the model recommends skills testing.

Chun and Witt (2008) suggested standardized curriculum-based measures (CBM) for universal screening. Because ofthis model's emphasis on efficiency, they recommended thatadministration of screening probes follow a scripted imple-mentation protocol. As with Batsche et al. (2008), Chun andWitt (2008) emphasized using screening results to determinethe effectiveness of core curriculum, test disproportionalityof minority groups in the at-risk category, and identify dif-ferences among classes before examining individual studentdata. This model introduces a second assessment with a per-formance reward to ascertain whether low-performing stu-dents simply lacked motivation during screening (a perfor-mance/skill or Can't Do/Won't Do assessment). Notice thatthis second assessment is administered only to those studentswho were judged as at risk on the initial screening.

At the secondary level, Shinn (2008) recommendedscreening with the selective use of CBM Maze reading tasks

(Fuchs & Fuchs, 1992) with fifth-, sixth-, and ninth-gradestudents. At high school level, they recommended individu-alized, targeted screening rather than universal screening.Conversely, Sugai and Horner's (2007a) behavioral modelprescribes widespread data collection and comparisons atschool, class, and student levels to support balanced deci-sion making.

Progress Monitoring

All models include progress monitoring to determine stu-dents' growth or improvement over time, which theninforms instruction or intervention decisions. As a generalrule, when the intensity of an intervention increases, the fre-quency of progress monitoring increases. However, the dif-ferences among models are in the details. Table 4 illustratesdecision-making criteria for movement between tiers.

Batsche et al. (2008) offered a loose guideline and indi-cated that the frequency of progress measurement should besufficient to inform instruction and evaluate responsivenessto intervention—leaving open the choice to ftt the specificsof the problem-solving process. The three proponents ofstandard protocol models offer more specificity about thefrequency of progress monitoring. Fuchs and Fuchs (2005),as previously discussed, suggested 5 weeks of preliminaryprogress monitoring in Tier I, weekly monitoring for 8weeks in Tier 2, and more frequent monitoring of nonre-sponders to intervention. Johnson et al. (2006) likewise sug-gested monitoring in Tier 1 for 5 weeks, and Tier 2 and Tier3 monitoring at least two times per week. Vaughn's (2005)frequency spectrum ranges from three times per year for stu-dents making adequate progress to weekly for students withpersistent reading deficits.

Consistent with their emphasis on efficiency and empiri-cism, Chun and Witt's (2008) monitoring frequenciesdepend on measured skills and standard protocol for theassessment (e.g., weekly oral reading fluency and mathprobes: after 4 or 5 weeks of instruction for reading com-prehension probes). Shinn (2008) recommended monitoringweekly rates of improvement among secondary students inTiers 1 and 2, with nonresponsiveness determined after 4 to6 weeks, and in Tier 3, monitoring with greater frequencydepending on the intervention. Sugai and Horner (2007a)recommended weekly, monthly, quarterly, and annual col-lection of student, classroom, and school-wide data to pro-vide balanced information for placement and interventiondecisions.

The models differ not only in frequency of administrationbut in the assessments they employ. Batsche et al. (2008)used CBM and behavioral measures compared to a baselineof peer performance rather than standardized referencespoints. The other six models use standardized CBM tools.However, the combined approach tnodels collect other data

such as course grades, rates of improvement toward a spec-ified goal, discipline referrals, attendance, suspension, andbehavior incident reports.

Decision-making criteria

Perhaps the greatest variability among models is criteriafor introducing an intervention or moving students amongtiers. Table 4 describes criteria for five decision points: (a)introduction of a class-wide intervention in Tier 1, (b) mov-ing individual students from Tier 1 to Tier 2, (c) movingindividual students from Tier 1 to Tier 3, (d) moving indi-vidual students from Tier 2 to Tier 3, and (e) moving indi-vidual students toward Tier 1.

Screening data may indicate the need for a school- orclass-wide intervention in Tier I. Batsche et al. (2008) sug-gested such intervention is needed when less than 80% of stu-dents do not score above predetermined benchmarks. Chunand Witt (2008) also compared whole class performance(i.e., median score) to an external standard (i.e., instructionallevel; Deno & Mirken, 1977) to determine whether a class-wide intervention is needed. Shinn (2008) had a more localor class-specific foeus and considered a class-wide interven-tion necessary when 80% to 90% of student needs are notmet by the general instruction. Sugai and Horner (2007a)proposed tbat all students participate in a core social behav-ior curriculum if school-wide data indicate more than 20%to 30% of students are demonstrating behavior problems.

Moving students from Tier 1 to Tier 2 is generally a func-tion of screening data, but in many cases additional data isconsidered. Batsche et al. (2008) indicated that for each stu-dent who scores 2 years below state or district benchmarks,a norm-based gap analysis of individual and class perfor-mance determines whether the deficit is specific to the stu-dent; if the deficit is student specific, the student moves toTier 2. Chun and Witt (2008) attempted to discern whetherstudents scoring in the bottom 16% of the screening assess-ment lacked motivation or skill by administering a secondassessment with a motivational reward. Students who im-proved their score by 20% or more remain in Tier 1 wherethey receive a motivation-focused intervention. Studentswho improved less than 20% are placed in Tier 2 and pro-vided an appropriate standard protocol intervention.

Fuchs and Fuchs (2005) and Johnson et al. (2006) bothcalled for collecting additional progress monitoring data inTier 1 before moving students to Tier 2. Fuchs and Fuchsrecommended monitoring students who were either pre-dicted as at risk based on nortn-referenced cut points orscored below the 16th percentile. A CBM slope discrepancyanalysis determines which of these students are placed inTier 2. The slope discrepancy refers to the difference be-tween the rate of progress evidenced by a student comparedto an aim line (i.e., rate of progress) based on local or

national norms or a slope greater than the standard errorestimate. Johnson et al. linked screening data to such crite-rion-referenced scores as state assessments to identify stu-dents to progress monitor in Tier 1 and a school-determinedcut point for movement to Tier 2. The intent is to improveprecision of decision making with additional objective, rel-evant student data.

Vaughn (2005) recommended placing students in Tier 2when they score below norm-referenced cut scores providedby the assessment tool(s) in one or more critical areas ofreading. Shinn (2008) suggested moving students to Tier 2when CBM grade level proficiency is lower than 75% oftheir peers. Sugai and Horner (2007a) suggested moving astudent to Tier 2, a classroom-level intervention, based onthe number of major behavioral rule violations under theexisting practices.

Most of these models are sequential, that is, students par-ticipate in Tier 1 and then Tier 2 instruction before movingto Tier 3. However, Batsche et al. (2008) allowed for Tier 1to Tier 3 movement in cases of clear evidence of a preexist-ing condition or history of SLD. Likewise, Shinn (2008)moved students who scored below the 10th pereentile on thescreening assessments directly to Tier 3 or special educa-tion, when appropriate. Sugai and Horner (2007a) alsoallowed for a student to move directly to Tier 3 through theuse of a functional bebavior assessment (FBA) process.

Slope discrepancy or gap analysis of progress monitoringdata drives decisions to move students to Tier 3 in mostmodels. For Batsche et al. (2008), student rates of progressare compared to peers. When the performance gap betweena Tier 2 student and his or her peers narrows, the studentremains in Tier 2; when the gap is still widening, but at aslower rate, the student receives a different Tier 2 interven-tion; and when the gap further widens with no change in rate,the student is moved to Tier 3. Other models prescribe thenumber of weeks of intervention or data points that must becollected for a slope discrepancy analysis compared to a goalline. For Fuchs and Fuchs (2005) and John.son et al. (2006),four data points collected during a few weeks of interventionare needed for a movement decision; if the data are consis-tently above or below the goal line, a change in interventionintensity or method is recommended, with the possibility ofbeginning a full SLD evaluation and movement to Tier 3after about 8 weeks. For Vaughn (2005) and Chun and Witt(2008), between 10 and 20 weeks of Tier 2 intervention areneeded to determine inadequate progress toward a bench-mark, that is, a slope discrepancy, at which time studentswould be moved to Tier 3. As previously indicated, Shinn(2008) simply used the below 10th percentile on local norm-referenced assessments for determining student placementin Tier 3. For Sugai and Horner (2007a) the characteristicsof individual students and specific circumstances related to

10 FOCUS ON EXCEPTIONAL CHILDREN FEBRUARY 2011

Need forClass-wide

Principle InterventionAuthor in Tier 1

Batsche et al. 80% of thestudents are notscoring abovethe benchmark

Fuchs & Fuchs —

TABLE 4. Decision Making Criteria

Tier 1 to Tier 2Movement

Students scoring2 years below state ordistrict benchmarkpredicted as at risk

Example benchmarks:Academic engage-ment 75%, high-stakes test 70-80%accuracy

Norm-based gapanalysis of individualand class determineswhether the deficit isspecific to that student

Students predicted atrisk based on norm-referenced cut pointsor below the 16thpercentile progressmonitored for 6 to 8weeks in Tier 1

Students scoring


Students sequen-tially progressthrough tiers exceptin cases of clearevidence (e.g., pre-existing condition.history of SLD)

Students sequen-tially progressthrough tiers


Student rate ofprogress comparedto peer rates; whenperformance gap isnarrowing remainin Tier 2; rate ofgap widening slowsbut continues towiden, change inter-vention; gap con-tinues to widenwith no change inrate, move to Tier 3

A minimum of 3weeks of interven-tion and four datapoints consistentlyabove or below thestudent goal lineindicates change inintervention inten-sity or method

Movementtoward Tier 1

Students respond-ing to interventionand achievingbenchmarksreverse placementsequence withprogressmonitoring

Students per-forming at classlevel return toTier 1 or Tier 2instruction from ahigher tier withprogressmonitoring

Johnson et al. —

below CBM cut points(local or national normsor slope greater thanthe standard errorestimate) placed inTier 2

Students predicted atrisk in criterion refer-enced screening areprogress monitored 5weeks in Tier 1

From this group,students performingbelow a school-determined cut pointor demonstrating in-adequate progresswhile scoring in the"at risk" range placedin Tier 2


Preset rules (e.g.,four consecutivedata points belowthe goal line)prompt an interven-tion change withinTier 2

Comprehensiveevaluation deter-mines SLD andTier 3 placement

Progress in rate ofimprovement orlevel in Tier 2 andTier 3 leads toréintégration intoa lower tierplacement

11

PrincipleAuthor

Vaughn

Shinn

Chun & Witt

1

;'

Need forClass-wideInterventionin Tier 1

Instruction doesnot meet theneeds of 80-90%

CBM probe dataindicate themedian studentscore is belowinstructional level

TABLE 4.


Students scoring be-low benchmark in oneor more critical areasof reading using pre-determined cut scoresestablished by theassessment tool(s).normed on comparablepopulation, placed inTier 2

Students performinglower than 75% ofpeers on CBMs con-sidered for Tier 2placement

Students scoring inbottom 16% of instruc-tional level on CMBsperformance/skillassessment

Students improving> 20% on perfor-mance/skill assess-ment and scoring ininstructional rangeprovided a motivationintervention in Tier 1

(continued)



Students scoringbelow 10th per-centile move tospecial education/Tier 3



After 10-20 weeks.students progress-ing at an inade-quate rate to meetyear-end benchmarkmay repeat 10-20weeks of Tier 2intervention

Students makingno or very limitedprogress may beplaced in Tier 3

Students scoringlower than the 10thpercentile on localnorm-referencedassessments con-sidered for Tier 3placement

Students placed inTier 3 when insuffi-cient progress inTier 2 is evidencedby an increasingslope discrepancybetween progressdata and a predeter-mined aim after10-15 interventionsessions

Movementtoward Tier 1

Students placed inTier 1 when makingadequate progresstoward year-endbenchmark andattaining instruc-tional goals

—

Students aremoved to lowertier based slopediscrepancyimprovements andperformance levelsimilar to the lowertier

Sugai & Horner School-wide dataindicate more than20% to 30% ofstudents demon-strating behaviorproblems

Students improving< 20% placed in Tier 2and matched by pre-determined criteriawith standard protocolinterventions

Students who have 0-1, 2-5, and 6 or moremajor rule violationsto determine who areresponsive to existingpractices and systems

Decision making rulesare made by a con-sensus of the staff afterexamining school-widedata

Decisions basedon weekly criteriaspecified in targetedinterventions (e.g..Check in/CheckOut, BehaviorEducation Pro-gram, Check andConnect)

Characteristics ofindividual studentsand specific cir-cumstances relatedto them (e.g., dif-ferences in theseverity of behavior,complexity of envi-ronment) gatheredthrough the func-tional behaviorassessment


them (e.g., differences in the severity of behavior, complex-ity of environment) gathered through the FBA is the basisfor movement to Tier 3.

When students are responsive to interventions and makeadequate progress toward goals, they may move down thetier hierarchy to less intensive services. For each modelthese reverse assignments are sequential, that is. Tier 3, toTier 2, to Tier 1 réintégration.

Specific Learning Disabilities Determination

With the exception of Sugai and Homer's model, whichis not used to determine SLD, the models comply with theIDEA requirements for comprehensive assessment andexclusion of other causes of underachievement, using datacollected through RTI processes to support the determina-tion. The differences, however, are at what point in theprocess a comprehensive evaluation of SLD determinationbegins. Table 5 illustrates SLD determination methods.Vaughn (2005) viewed SLD determination as taking placewithin the RTI framework. Likewise, Batsche et al. (2008)would begin the SLD determination process at any time theproblem-solving team observes more than a 2-year gapbetween student performance and grade-level expectations,along with a slow rate of progress during high-qualityinstruction. In contrast, Fuchs and Fuchs (2005) would

begin the determination process when a student is nonre-sponsive to 8 weeks of Tier 2 instruction. Johnson et al.(2006) described how RTI data ean contribute to SLD deter-mination along with a comprehensive evaluation, clinicaljudgments, and other relevant information. Their guidanceis that the intervention must be provided with sufficientdosage to judge student responsiveness. Chun and Witt's(2008) model does not begin the SLD determination processuntil after the RTI process finds insufficient progress in Tier3. So too, Shinn (2008) would begin determination after RTIinterventions have been exhausted and use the relativeachievement discrepancy (RAD) method of comparing astudent to local norms.

RESEARCH

Each model has documented research supporting eithercomponent parts or the entire model (Table 6). Overallresearch of RTI models or components at the elementarylevel is more robust than at the secondary level. The Batscheet al. (2008) study is supported by 2 years of positive quan-titative and qualitative outcome data from a school districtin Florida. Fuchs and Fuchs referenced research on desig-nating realistic and ambitious rates of growth using CBMmeasures in reading, spelling, and math in first and second

TABLE 5. RTI Role in Specific Learning Disabilities Determination

Principle Author Specific Learning Disability Determination

Bafsche ef al. More than 2 year gap between performance and grade-level benchmark, along with low rate of progresswith high quality instruction triggers SLD determination procedureComprehensive assessment is needed to investigate other causes of underachievement and exclusion fac-tors as per IDEA

Fuchs & Fuchs 8 weeks of nonresponsiveness fo intervention af tier 2 triggers comprehensive evaluation in compliancewith IDEA lawTier 3 is considered special education

Johnson et al. RTI data confribufes to SLD determination, along wifh comprehensive evaluation, clinical judgments, andother relevant information from interdisciplinary team in compliance with IDEA lawThe final tier of RTI is special education

Vaughn SLD identification takes place wifhin the RTI framework

Shinn Special education determination follows nonresponsiveness in RTIRelative Achievement Discrepancy (RAD), comparing a student to local norms, used fo determine SLD

Chun & Witt SLD determination process occurs after insufficient progress in Tier 3Problem-solving process wifh progress monitoring data used fo make referral and eligibility decisionsData and documentation collected through STEEP RTI on appropriate instruction, student motivation,fidelity, sustained lack of student response, and educational need is used in SLD determination process

Sugai & Horner Not used fo determine Specific Learning Disabilities

13

TABLE 6. Model Research Context and Findings

Principle Author Context Findings

Batsche et al.

Fuchs & Fuchs

Johnson et al.

¡ Vaughn

Shinn

Chun & Witt

ElementaryAcademicBehavioral

ElementaryReadingSpellingMath

ElementaryReading

K-3Middle schoolReading

SecondaryReading, Math, & Writingin content courses

K-12ReadingMathWriting

Sugai & Horner Middle school transitionBehavior

2 years of pilot site implementation in Florida school yielded positive results ingrades K-3, except for students with most intense instructional needs Insecond and third grades, who had negligible results (Batsche et al., 2008)

Research focused on the CBM component of RTI, specifically designatingrealistic and ambitious rates of growth using CBM measures in reading,spelling, and math in grades 1 and 2 (Fuchs & Fuchs, 1993)Issues affecting the development of decision-making rules for selectingstudents for Tier 2 interventions in grade 1 (Compton et al., 2006)Efficacy of math tutoring for students in grade 1 (Fuchs et al., 2005)

19 schools found to be engaged in one or more of the practices in this model(Johnson et al., 2006)Model developed from this study but not validated by it

RTI as a viable option in identifying second grade students with readingproblems (Vaughn et al., 2003)Middle school mixed results; some students demonstrated progress, othersmade little or no progress in either intervention or traditional reading class(Dentón et al., 2008)

CBM Reading Maze in high school using eighth grade norms for cutoffdecisions (Shinn, 2008)

Improved student achievement in Tier 2 and for Tier 1 in schools with pervasiveissues (VanDerHeyden & Burns, 2005: VanDerHeyden et al., 2007)Reduced overrepresentation in LD for ethnicify, gender, and achievement level(VanDerHeyden & Witt, 2005)Reduced evaluation referrals and associated costs and improved referralaccuracy (VanDerHeyden et al., 2007)Decision rules and benchmark standards validated for categorizing problems atTier 1 and 2 (VanDerHeyden, Witt, & Naquin, 2003; VanDerHeyden et al., 2007)Validation of one rather than three screening probes and benchmarks used fortier categorization (Ardoin et al., 2004)

Behavioral interventions at all three tiers supported by research(Sugai & Horner, 2007a)PBS approach supported by randomized control trials (Sugai & Horner, 2007b)

grade (Fuchs & Fuchs, 1993), issues affecting the develop-ment of decision-making rules for selecting students for Tier2 interventions in first gtade (Compton, Fuchs, Fuchs &Bryant, 2006), and the efficacy of math tutoring for studentsin first grade (Fuchs et al., 2005). Nineteen elementaryschools engaged in one or more commendable practices inthe Johnson et al. proposed model (2006). Chun and Witt(2008) referenced research indicating improved achieve-ment for both at-risk students and general education stu-dents in schools with pervasive Tier 1 issues (VanDerHey-den & Witt, 2005: VanDerHeyden, Witt, & Gilbertson,2007). Sugai and Horner refer to three randomized control

trials that support the PBIS approach (Sugai & Hornet,2007a). Vaughn is suppoited by a study of second grade stu-dents at risk for reading difficulties (Vaughn, Linan-Thomp-son, & Hickman, 2003) and a recent study of students at themiddle school level where tesults were mixed (Dentón,Wexler, Vaughn, & Bryan, 2008).

CONCLUSION

No doubt these models' developers and advocates willpursue additional development and validation work fromtheir respective viewpoints. Likewise, schools will continue


to refine their practices. An important and seemingly obvi-ous conclusion from this summary is that many of the com-ponents associated with RTI have been carefully researchedand evaluated. We note, however, that this research has hada narrow, explicit focus with defined research protocols. Incontrast, the literature is very limited on fully developedmodels and their scaling and sustainability across multipleschool districts, grade levels, content focus, and implemen-tation supports. Their application to SLD determination hasnot been tested in a comparison research study of alternativeassessment approaches. Currently, as state education agen-cies and districts pursue an adoption, adaptation, implemen-tation, and scaling up of RTI, they are encouraged to makea careful review of available research and supporting evi-dence. This description and comparison among the sevenmodels illustrates the many decision points that are inherentto RTI complexities. The irony is that while researchers, pol-icy makers, and advocates are pushing for rigorous, highfidelity implementation of screening, progress monitoring,interventions, and decision rules, this goal for practitionersseems to be a higher standard than has been demonstratedby the supporting research. Researchers and practitionersalike will wait to see how the current broad support of RTIwithstands the challenges that coexist with broad imple-mentation across the landscape of the US P-12 educationalsystem.

REFERENCES

Ardoin, S., Witt, J., Suido, S.. Connell, J., Koenig, J.. Reseter, J. et al.(2004). Examining the incremental benefits of administering amaze and three versus one curriculum based measurement readingprobes when conducting universal screening. School PsychologyReview 33(2),2\?,-2M.

Batsche. G. (2006). Response to Inlervention: Policy Considérationsand Impleineniarion. Alexandria. VA: National Association ofState Directors ot' Special Education, Inc.

Batsche, C , Curtis. J.. Dormán, C . Castillo. J., & Porter, L. J. (2008).The Florida problem solving/response to intervention model:Implementing a statewide initiative. In S. Jimerson, M. Burns. &A. VanDerHeyden (Eds.) Handbook of response lo inlervention:The science and practice of assessment and inlervenlion (pp.ÍTS-ÍC.";). New York. NY: Springer.

Bocala. C . Mello. D.. Reedy. K.. & Lacireno-Raquet. N. (2009). Fea-tiires of stale response lo inlervenlion initiatives in Norlheasi andIslands Region stales (Issues & Answers Report, REL 2009-No.083). Washington. DC: US Department of Education. Institute forEducation Sciences. National Center for Education Evaluation andRegional Assistance. Regional Education Laboratory Northeastand Islands.

Compton, D.. Fuchs. D.. Fuchs, L. S.. & Bryant, J. (2006). Selectingat-risk readers in first grade lor early intervention: A two-year lon-gitudinal study of decision rules and procedures. Journal of Edu-cational Psychology. 98(2). ."^94-409.

Council for Exceptional Children (CEC). (2008). CEC's position onresponse to intervention (RTI). Teaching Exceptional Children,40(3), 14-15.

Chun, M., & Witt, J. (2008). System to enhance educational perfor-mance. Retrieved from http://www.isteep.com/index.html

Deno, S., & Mirkin, P. (1977). Data-based program modification: Amanual. Reston, VA: Council for Exceptional Children.

Dentón, C . Wexler, J., Vaughn, S., & Bryan, D. (2008). Interventionprovided to linguistically diverse middle school students withsevere reading difficulties. Learning Disabilities Research &Practice, 23(2), 79-89.

Fletcher, J., & Vaughn, S. (2009). Response to intervention models asalternatives to traditional views of learning disabilities: Responseto commentaries. Society for Research in Child Development,.?(!), 48-50.

Fuchs, D.. & Deshler. D. (2007). What we need to know about respon-siveness to intervention (and shouldn't be afraid to ask). LearningDi.'iabililies Research & Practice, 22(2), 129-136.

Fuchs, D., & Fuchs, L. S. (200.5). Responsiveness-to-intervention: Ablueprint for practitioners, policymakers, and parents. TeachingExceplional Children, 38(\), 57-61.

Fuchs, D.. Fuchs, L., & Compton, D. (2004). Identifying reading dis-abilities by responsiveness-to-instruction: Specifying measuresand criteria. Learning Disabilities Quarterly. 27(4), 216-227.

Fuchs. L., Compton, D., Fuchs, D., Paulsen, K., Bryant, J.. & Hamlett, C.(2005). The prevention, identification, and cognitive determinants ofmath difficulty. Journal of Educational Psychology. 97(3), 493-513.

Fuchs, L.. & Fuchs, D. (1992). Identifying a measure for monitoringstudent reading progress. School Psychology Review. 2I(\). 45-58.

Fuchs, L.. & Fuchs. D. (1993). Formative evaluation of academicprogress: How much growth can we expect? School PsychologyReview. 22(1). 27-48.

Fuchs, L., & Fuchs, D. (2009). On the importance of a unified modelof responsiveness to intervention. Society for Research in ChildDevelopment, 3(\). 41-43.

Good, R., & Kaminski, R. (2002). Dynamic Indicators of liasic EarlyLiteracy Skills (6th ed.). Eugene. OR: Institute for the Develop-ment of Educational Achievement. Retrieved from http://dibels.uoregon.edu/measures/orf.php

Griffith, A., Parsons, L., Burns, M.. VanDerHeyden, A.. & Tilly, W. D.(2007). Response lo ¡nten'ention: Research for practice. Alexan-dria. VA: National Association of State Directors of Special Edu-cation. Inc.

Hasbrouck, J., & Tindal, G. (1992). Curriculum-based oral reading Hu-ency norms for students in grades 2 through 5. Teaching Excep-tional Children. 24(3). 41-44.

Hollenbeck. A. (2007). From IDEA to implementation: A discussion offoundational and future responsiveness-to-intervention research.Learning Disabilities Research and Practice 22(2). 137-146.

Johnson. E.. Mellard. D.. Fuchs. D.. & McKnight. M. (2006). Re.'ipon-siveness lo inlervenlion (RTI): How lo do it. Lawrence. KS:National Research Center on Learning Disabilities.

Johnson. E.. Smith. L.. & Harris. M. (2009). How RTI Works in Sec-ondary Schools. Thousand Oaks. CA: Corwin Press.

Learning Disabilities Association of America (LDA). (2006, Septem-ber) The role of parents/family in responsiveness to intervention.Author: Pittsburgh. PA.

Mellard, D.. Byrd. S., Johnson, E.. Tollefson. J.. & Boesche. L. (2004).Foundations and research on identifying model responsiveness-to-intervention sites. Uuirniiig Disabilities Quarterly, 27(4). 243-256.

15

Mèllard, D., & Johnson, E. (2008). RTl: A practitioner's guide toimpletnenting response to intervention. Thousand Oaks, CA: Cor-win Press.

Mellard, D., McKnight, M., & Jordan, J. (2010). RTl tier structuresand instructionat intensity. Manuscript submitted for publication.

Mellard, D., McKnight, M., & Woods, K. (2009). Response to inter-vention screening and progress monitoring practices in 41 localschools. Learning Disabitities Research atid Practice, 24(4),186-195.

Mellard, D., McKnight, M., Woods, K., & Frey, B. (n.d.). School-widestudent outcomes as evidential validity of response to interventionframeworks. Manuscript submitted for publication.

Reynolds, C , & Shaywitz, S. (2009). Response to intervention: Pre-vention and remediation, perhaps. Diagnosis, no. Society forResearch in Child Development, 3(\), 44—47.

Shinn, M. R. (2008). Implementation in secondary schools. In S. Fern-ley, S. D. LaRue, & J. Norlin, (Eds.). What do I do when: Answerbook on RTI. (pp. 8:1-17). Arlington, VA: LRP Publications.

Sugai, C , & Horner, R. (2007a). Response to intervention logic andpositive behavior support [PowerPoint slides]. Retrieved fromhttp://www.pbis.org/main.htm

Sugai, C , & Horner, R. (2007b). Evidence base for school-wide posi-tive betiavior support [PDF file]. Retrieved from htlp://www.pbis.org/main.htm

VanDerHeyden, A. M., & Burns, M. K. (200.5). Using curriculum-based assessment and curriculum-based measurement to guideelementary mathematics instruction: Effect on individual andgroup accountability scores. Assessment for Effective Intervention,30, 15-31.

VanDerHeyden, A., & Witt. J. (2005). Quantifying context in assess-ment: Capturing the effect of base rates on teacher referral and aproblem-solving model of identification. Schoot PsyctwtogyReview, 34(2), 161-183.

VanDerHeyden, A., Witt, J., & Gilbertson, G. (2007). A multi-yearevaluation of the effects of a response to intervention (RTI) modelon identification of children for special education. Journal ofSchool Psyctiotogy, 45(2), 225-256.

VanDerHeyden, A., Witt, J., & Naquin, G. (2003). The developmentand validation of a process for screening and referrals to specialeducation. School P.sydiotogy Review, 32(2), 204-227.

Vaughn, S. (2005). Interpretation of the 3-tier framework. Retrievedfrom http://www.texasreading.org/utcrla/materials/3tier_letter.asp

Vaughn, S., Linan-Thompson, S., & Hickman, P. (2003). Response toinstruction as a means of identifying students with reading/learn-ing disabilities. Exceptionat Children, 69(4), 391-409.

volume 43 number 6 february 2011 focus on

Documents