RESEARCH DOSSIER
Multi-Tiered Systems of Support and Progress Monitoring
Abstract. This fundamental research monograph critically synthesizes the history, concepts, measurement models, empirical findings, and organizational challenges of multi-tiered school support and progress monitoring. Curriculum-based measurement, academic response to intervention, intensive individualization, and schoolwide behavioral prevention are assessed alongside screening validity, growth estimation, educational equity, and causal evidence. The synthesis differentiates demonstrated effects of specific instructional and behavioral interventions from unproven inferences about all implementations of a tiered framework.
Keywords: multi-tiered systems of support; MTSS; response to intervention; RTI; progress monitoring; curriculum-based measurement; universal screening; data-based individualization; school psychology; educational assessment; positive behavioral interventions and supports; implementation fidelity
Scientific Overview and Research Questions
Multi-tiered systems of support are institutional arrangements for responding to variation in student need. At the center of the arrangement is a recurring relationship among instruction, observation, inference, and action. A school identifies learning and participation goals, provides a broadly accessible core program, examines student outcomes, offers additional help when appropriate, and uses subsequent observations to evaluate whether its support has worked. The designation of tiers represents differences in support intensity and individualization, whereas progress monitoring supplies evidence about the consequences of instruction over time. The scientific object of this monograph is that combined organization of educational support and measurement-based decision making.
This object belongs primarily to educational and school psychology because it concerns learning, behavioral adjustment, assessment, prevention, and collaboration within institutional settings. It also intersects with learning and behavioral psychology, psychometrics, special education, developmental psychology, and implementation science. The relevant explanatory unit may be an individual student, a classroom, an intervention group, a school, or an education system. Confusing these levels produces errors: an improvement in the average school outcome cannot establish that every student benefited, while the trajectory of one student cannot determine whether an entire district framework is effective.
Five research questions organize the synthesis. How did frequent direct measurement and response-based intervention become a coordinated schoolwide framework? What constitutes valid screening and progress monitoring, including reliable estimates of change? Which academic and behavioral practices have credible causal evidence, and under what circumstances are they delivered? How should decisions about intensity, adaptation, and disability evaluation integrate multiple forms of evidence? How can schools evaluate their own implementation while protecting access, interpretability, and student rights?
The contemporary policy language of MTSS sometimes includes academic achievement, behavior, attendance, social-emotional functioning, and school mental health. These domains share a decision architecture, but each requires domain-appropriate constructs and instruments. A reading-fluency score, an office discipline referral, and a teacher-completed symptom rating have different causal determinants, measurement errors, and ethical consequences. Their placement on a common dashboard does not confer common validity. The scientific task is to explain their connections while maintaining the specificity of each outcome.
Historical Formation of Tiered Educational Support
From standardized testing to continuous instructional measurement
The intellectual history of progress monitoring includes school achievement testing, formative evaluation, behavioral observation, and special education measurement. Periodic standardized examinations were useful for describing relative achievement but often too infrequent or too distant from ongoing instruction to indicate when teaching should change. Educators needed measures capable of detecting meaningful progress during an instructional program rather than solely at its conclusion.
Deno (1985) gave a foundational account of curriculum-based measurement (CBM), describing standardized observations of academic performance within curricular activity. The methodological aspiration was to combine the practical relevance of direct performance samples with consistency across occasions and interpretable relations to broader achievement. This development established a distinction that still organizes MTSS: assessment for comparative status and assessment for instructional adaptation are related, but they answer different questions.
CBM emerged within research on monitoring students with disabilities and subsequently informed general education screening, intervention planning, and schoolwide accountability. Its historical significance rests on the idea that progress can be measured repeatedly and acted upon. The resulting evidence is still conditional on the quality of the task, the interpretation of the score, and the teacher's ability to change instruction in response. A graph is informative only when its measurement and decision procedures are defensible.
Response to intervention and the eligibility debate
The response-to-intervention movement joined early intervention to concerns about identifying learning disabilities. Historically, a severe discrepancy between intellectual ability and academic achievement played an influential role in eligibility procedures. Such comparisons generated controversies about delayed assistance, the explanatory status of discrepancy criteria, and the influence of educational opportunity on low achievement.
Fuchs (2006) described response to intervention both as a means of offering earlier assistance and as a potential alternative source of evidence for learning disability identification. The framework asked whether a student improved under instruction that was sufficiently explicit, appropriate, and documented. The resulting response pattern was intended to inform hypotheses about student needs; it could not by itself establish the cause of persistent difficulty.
Federal policy helped institutionalize response-based procedures in the United States. The Individuals with Disabilities Education Act regulations at U.S. Department of Education. (n.d.). 34 CFR § 300.307: Specific learning disabilities. Individuals with Disabilities Education Act regulations. https://sites.ed.gov/idea/regs/b/d/300.307 (n.d.) require states to permit a process based on response to scientific, research-based intervention when determining specific learning disabilities and prohibit requiring a severe IQ-achievement discrepancy. These regulations concern eligibility procedure, not endorsement of one fixed three-tier implementation model.
Schoolwide prevention and integrated systems
A partially parallel tradition developed in schoolwide positive behavioral interventions and supports (PBIS), combining behavioral principles, preventive environments, staff practices, and support of students with more intensive needs. Horner (2009) tested the implementation of schoolwide positive behavior support through a randomized wait-list design, establishing an empirical foundation for school-level implementation questions.
As academic response to intervention and behavioral prevention matured, institutional integration became a practical research problem. Separate teams, records, and referral systems could generate inconsistent decisions and consume resources that might otherwise support students. McIntosh (2016) described the integration of academic RTI and PBIS within a common MTSS organization. Their work concerns system design and implementation, rather than an independent proof that any arrangement carrying the MTSS label is effective.
The historical synthesis thus contains three distinct contributions: CBM provides longitudinal measurement; RTI places measured response inside early academic intervention and eligibility discussions; and PBIS brings schoolwide preventive organization and behavioral support. MTSS organizes these traditions into a broader model of coordinated school decision making.
Definition and Conceptual Architecture
An MTSS is a coordinated, multilevel approach to organizing educational opportunity and assistance through universal provision, systematic early identification, calibrated supplemental support, intensive individualized intervention, and evaluation of student response. Tiers describe an instructional continuum. They do not designate permanent student identities, standardized diagnoses, or natural kinds of learners. A student may need intensive reading instruction and universal behavior supports at the same time, or targeted support in one academic domain while demonstrating advanced competence in another.
The most analytically useful description is a set of linked functions. Universal core instruction sets the opportunity structure for the population. Screening estimates who may need closer examination or additional assistance. Intervention specifies a manipulable instructional condition with defined content and dosage. Progress monitoring estimates status and change under that condition. Data-based decision making compares observed performance with meaningful goals and evaluates whether to continue, adapt, intensify, or fade support. Implementation assessment determines whether intended instruction and measurement were actually delivered.
A functioning system also requires allocation of time, personnel, instructional materials, language access, family communication, referral procedures, and explicit decision responsibilities. These institutional variables are part of the causal process. If an intervention is theoretically effective yet delivered inconsistently or too late, a student's observed lack of progress may reflect limitations of the service rather than enduring limitations of the learner.
The tiers are best understood as intensities, not universal numerical prescriptions. Local systems differ in the duration of support blocks, student-to-teacher ratios, assessment intervals, and intervention content. Specific time or group-size choices should arise from the available evidence for a given program, the student's needs, and implementation capacity. A rigid schedule applied across unrelated domains may simplify administration at the expense of valid educational decisions.
Conceptually, MTSS contains descriptive propositions, normative goals, and causal hypotheses. It describes how schools organize help; it seeks timely and equitable educational access; and it hypothesizes that appropriate interventions delivered and adapted through reliable evidence can improve outcomes. These propositions demand distinct kinds of verification. An implementation checklist may show that a process exists, while randomized or well-controlled outcome studies are needed to establish effects on learning or behavior.
The Psychological Logic of Prevention and Response
Opportunity, instruction, and observed performance
The response of a learner is jointly shaped by prior knowledge, learning history, present instruction, attention, motivation, language demands, time, and context. A low score is an observed outcome of this interaction. Interpretation of that score requires questions about the learning opportunities and assessment conditions that produced it. This ecological perspective protects against the inference that all low performance reflects a fixed attribute of a student.
A prevention model aims to reduce the number of students whose difficulties become severe by improving the environment and offering early, targeted instruction. The aim is attractive because some academic problems respond more readily to timely instruction than to years of accumulation. Yet prevention must be evaluated against outcomes: earlier identification creates a benefit only when it leads to helpful educational action and does not unnecessarily stigmatize students or replace necessary comprehensive evaluation.
An intervention-response model is informative when it is genuinely responsive to new evidence. Student performance after a small-group reading intervention can distinguish among several practical possibilities: adequate growth, growth with a persistent low level, inconsistent performance, or continued stagnation. Each pattern requires a different examination of instruction, measurement, attendance, and alternative explanations. Changes in a score without knowledge of the intervention's implementation remain ambiguous.
Status and growth as separate quantities
A student may be below the current benchmark but improving rapidly, or may meet a present benchmark while exhibiting a trajectory that places future success at risk. Status is a measurement at a defined point relative to a criterion; growth is change across time; responsiveness is growth interpreted in the context of a specified intervention. These concepts should be reported independently before they are combined in a decision.
A complete decision therefore considers current performance, plausible growth under the present program, the importance and attainability of the goal, and the instructional alternatives available. This structure aligns with the National Center on Intensive Intervention. (n.d.). Data-based individualization: Intensive intervention. American Institutes for Research. https://intensiveintervention.org/data-based-individualization (n.d.): validated intervention begins the process, progress evidence evaluates its adequacy, and diagnostic assessment and adaptation follow when the response remains insufficient.
Universal Instruction and the First Tier
Tier 1 denotes the educational environment encountered by the entire student population. Its research significance is often underestimated because the more intensive tiers are more visibly associated with specialist services. In fact, the distribution of student difficulties is conditional on the quality and accessibility of core teaching. When a large proportion of students struggle with the same foundational skill, a schoolwide curriculum or instructional problem deserves investigation before interpreting each difficulty as individually unusual.
Universal instruction involves systematic teaching, adequate time, responsive classroom practice, and routine assessment of broad outcomes. In reading it includes developmentally appropriate instruction in alphabetic knowledge, decoding, fluency, language, and comprehension. In mathematics it includes coherent development of numerical understanding and procedural and conceptual connections. Behavioral universal prevention involves clear expectations, consistent staff practices, and accessible teaching of expected behaviors. These examples represent domain-specific programs within a common preventive function.
The What Works Clearinghouse. (2009). Assisting students struggling with reading: Response to intervention (RtI) and multi-tier intervention in the primary grades. Institute of Education Sciences (2009) combines screening with differentiated instruction and increasingly intensive intervention. Its evidence ratings vary by recommendation. This variation is critical: a comprehensive tier diagram cannot substitute for evidence regarding the particular instructional practice under consideration.
Tier 1 evaluation uses multiple indicators: achievement outcomes, participation, student experience, classroom observation, language access, and implementation conditions. Aggregate means may conceal pronounced gaps. A school can increase its average reading score while some groups lose ground, and a classroom can improve its referral count because staff altered referral behavior rather than student conduct. Interpretation must attend to measurement pathways.
Universal provision also establishes the reference point for fair escalation. If the base program has been delivered inequitably, subsequent high-intensity intervention may compensate for earlier systemic failures. Such support may still be valuable, but a school should treat the underlying access problem as a target for improvement rather than assigning its full causal burden to individual students.
Universal Screening and the Interpretation of Risk
Universal screening uses brief, standardized measures or other defensible indicators to identify students whose future educational outcomes may warrant support or further evaluation. Its central question is predictive: given a current measurement, how likely is an important later difficulty? The answer depends on the instrument, outcome criterion, timing, population, and decision threshold.
Screening decisions create true positives, false positives, true negatives, and false negatives. Sensitivity describes the proportion of students who eventually meet the defined risk outcome and are identified by the screen. Specificity describes the proportion who do not meet that outcome and screen negative. Predictive values additionally depend on the prevalence of the outcome in the population. Consequently, a threshold that appears successful in one school may have a very different operational cost in another.
The National Center on Intensive Intervention. (n.d.). Academic screening tools chart. American Institutes for Research. https://charts.intensiveintervention.org/ascreening (n.d.), updated in August 2026, reviews classification accuracy, reliability, validity, sample representativeness, and bias analysis. Its technical ratings describe submitted evidence, with inclusion on the chart explicitly distinguished from institutional endorsement. These dimensions establish a research agenda for selecting instruments rather than a simple commercial product ranking.
A screening decision has consequences beyond a statistical classification. False negatives can delay needed support; false positives may generate unnecessary testing, loss of instructional time, or stigma. The relative cost of these errors depends on the availability and burden of intervention. Low-intensity, broadly beneficial support may justify an inclusive initial threshold; consequential eligibility decisions require more comprehensive evidence.
Screening should be interpreted alongside instructional opportunities and linguistic accessibility. A student newly acquiring the language of instruction may perform differently on a language-heavy screener than on a measure in a familiar language. Conversely, multilingual status should never be used to dismiss a genuine difficulty. The relevant question is whether the instrument measures the intended skill adequately for the population and whether multiple sources of information converge.
Screening results are an invitation to investigate and act. Their predictive accuracy does not establish that an intervention is effective, and improvement after selection does not alone prove an intervention effect. Those claims require separate designs and evidence.
Supplemental Instruction and the Second Tier
Tier 2 provides additional, planned instruction to students for whom universal provision appears insufficient in a particular domain. The instructional intervention should state what content is taught, by whom, with what frequency, in what setting, and using which evidence-informed practices. A tier label without such detail cannot be evaluated or reproduced.
The What Works Clearinghouse. (2009). Assisting students struggling with reading: Response to intervention (RtI) and multi-tier intervention in the primary grades. Institute of Education Sciences (2009) assigns strong evidence to its recommendation for intensive, systematic small-group instruction on a limited set of foundational reading skills for students below screening benchmarks. The same guide reports lower evidence classifications for certain monitoring and differentiation recommendations. This demonstrates why the effectiveness of a model must be decomposed into components rather than inferred from the prominence of the overall framework.
Supplemental instruction is distinguished from more of the same poorly matched activity by a stronger relation between diagnosed instructional need and teaching method. A student with weak decoding may require systematic instruction in sound-symbol relationships and word recognition, while a student with adequate decoding but weak text comprehension may need language and comprehension-oriented intervention. Identical generic tutoring hours offer limited scientific guidance without content specificity.
Regular progress monitoring helps answer whether the student is receiving an adequate instructional opportunity. Decisions should consider the quality of intervention delivery, attendance, score variability, classroom language context, and the stability of the observed trend. When the response is insufficient, intervention adaptation may involve instructional content, explicitness, opportunities to respond, feedback, scheduling, or group arrangement. Altering multiple elements simultaneously complicates causal attribution but may be necessary when educational benefit is urgent.
Exit from Tier 2 is also a meaningful decision. An improved score may indicate that the student can sustain growth with universal support; it may also indicate dependence on the supplemental program. Follow-up monitoring should examine maintenance and transfer, not simply whether a single benchmark was crossed.
Intensive Individualization and the Third Tier
Tier 3 refers to intensive, individualized assistance for students with substantial or persistent needs. It is best defined by the intensity and responsiveness of the intervention process, with direct attention to what the student has learned, what remains difficult, and how instruction should be changed. Some students receiving Tier 3 support also receive special education services; others do not. These classifications and service rights must be handled according to their own procedures.
Data-based individualization (DBI) provides one explicit framework. The National Center on Intensive Intervention. (n.d.). Data-based individualization: Intensive intervention. American Institutes for Research. https://intensiveintervention.org/data-based-individualization (n.d.) describes a cycle that begins with a validated intervention platform, monitors the response, applies diagnostic assessment or functional behavior assessment when growth is insufficient, adapts support, and resumes monitoring. The logic is iterative: the adequacy of an intervention is tested against an explicit goal, and the intervention changes when repeated evidence indicates that the existing approach is inadequate.
Intensive adaptation should respond to an identified instructional hypothesis. For example, a student may require more opportunities for explicit decoding practice, a smaller instructional step size, or increased practice with feedback. In behavior support, individualized functional assessment may identify the contexts and consequences associated with problem behavior. Selection of a change should be explained through the relevant psychological mechanism, then assessed with outcomes suited to that mechanism.
The central measurement challenge is that students requiring intensive intervention may show small or irregular growth. Measures must remain sensitive at low performance levels and distinguish genuine change from floor effects, form difficulty, and rater variation. The National Center on Intensive Intervention. (n.d.). Academic progress monitoring tools chart. American Institutes for Research. https://charts.intensiveintervention.org/aprogressmonitoring (n.d.) examine evidence for slope reliability, slope validity, alternate forms, and empirically supported decision rules, including evidence drawn from intensive-intervention populations.
Educational and clinical referrals may coexist with intensive assistance. A persistent lack of response warrants careful consideration of comprehensive evaluation, potential disability, developmental and health factors, and systemic barriers. It should never be treated as an automatic diagnostic rule.
Curriculum-Based Measurement and Repeated Observation
What a progress-monitoring measure must represent
Curriculum-based measurement uses brief, repeatable, standardized tasks to estimate important dimensions of academic performance over time. Typical indicators include words read correctly in a defined oral reading task, computation performance, written expression samples, or other curriculum-relevant outcomes. A score is interpretable only in relation to its content, administration conditions, scoring rules, and empirical association with the broader skill the school wishes to improve.
Deno (1985) established the practical value of standardized repeated academic observations for instructional and program decisions. Subsequent research refined the key question: can repeated scores detect meaningful change accurately enough to support decisions? The value of CBM rests on evidence of reliability, validity, alternate-form equivalence, and responsiveness, not merely on the convenience of short tasks.
For longitudinal purposes, repeated forms must be comparable in difficulty. When a student's score rises because later passages are easier, the graph misrepresents learning. When a task produces a floor or ceiling effect, the scale may become insensitive to genuine change. Scoring discrepancies across teachers, changes in testing conditions, and inconsistent accommodation practices can likewise mimic or obscure a trend.
Measurement frequency should reflect both the desired decision interval and the instrument's technical properties. More observations can improve estimation of trends, but frequent administration of a poorly designed measure creates large volumes of ambiguous data. Administrative burden also matters: time taken from instruction or family engagement is part of the system's resource cost.
Measures and instructional decisions
The research reviewed by Stecker (2005) indicates that using CBM to guide instructional change can enhance achievement, with important variation associated with systematic decision rules, skill analysis, and specific instructional recommendations. This evidence supports an active decision process. It offers a reason to question systems that require frequent testing yet provide little mechanism for teachers to modify what they teach.
Schools need to document not only when assessments occurred but also what decisions followed and whether instruction changed. The essential causal pathway is measurement → interpretation → instructional adaptation → student response. Breaking this pathway reduces progress monitoring to an administrative ritual.
Learning Trajectories, Growth Models, and Decision Rules
A learning trajectory is a sequence of observations indexed by instructional time. An instructional goal specifies the meaningful outcome and expected timeline. Analysts commonly consider both present level and a rate of improvement. A simple linear slope expresses change in score per unit of time, but real learning can be nonlinear, interrupted by absences, shaped by seasonal effects, or affected by changes in measure difficulty.
A goal line and a trend line are communicative devices, not guarantees of statistical precision. A slope calculated from a small number of observations may be strongly influenced by an unusually high or low score. Valid inference requires consideration of observation count, spacing, noise, serial dependence, and the possibility that a student's growth process changes after instructional adaptation.
Decision rules can include comparing a series of recent points with a projected goal line, reviewing a fitted trend, or using validated decision algorithms. The best rule is one whose error rates and practical consequences have been evaluated for a relevant population. The National Center on Intensive Intervention. (n.d.). Academic progress monitoring tools chart. American Institutes for Research. https://charts.intensiveintervention.org/aprogressmonitoring (n.d.) explicitly considers evidence for decision rules in goal setting and instruction changes. Its technical review standards underscore the difficulty of treating a convenient graphical heuristic as universally validated.
Growth benchmarks are context dependent. A slope adequate for one grade or skill may be inadequate for another, and a student with a very low starting score may need substantial growth simply to reduce a persistent gap. Teams should consider both absolute attainment and relative trajectory. Changes to the measurement instrument during intervention require careful documentation to protect longitudinal interpretation.
The most consequential inferential error is to equate a flat line with a stable student deficit. Growth may appear flat because the task lacks sensitivity, intervention content is mismatched, attendance is irregular, or instruction has not occurred as planned. Conversely, an apparently steep line may be caused by practice with a narrow task while broader competence remains unchanged. A defensible decision integrates proximal and distal indicators.
Evidence from Early Reading Research
Reading research provides some of the clearest examples of staged intervention and progress-monitoring practices. Early studies examined whether systematic supplemental instruction could accelerate foundational reading skills and whether responsiveness offered information beyond a static achievement score.
In a treatment-response study, Vaughn (2003) followed 45 second-grade students at risk for reading problems through successive periods of supplemental reading instruction, with response criteria used to decide whether instruction should continue or be intensified. The study is historically important for illustrating a response-based classification procedure. Its small and specialized sample limits the conclusions that can be drawn about broad schoolwide effects or universal diagnostic validity.
In a randomized study with English language learners, Linan-Thompson (2006) evaluated supplemental reading instruction and reported that more intervention participants met established response criteria than comparison students, with gains maintained in second grade. The study adds evidence that well-specified intervention can benefit students learning in English or Spanish and complicates assumptions that multilingual learners form a homogeneous low-response population.
The What Works Clearinghouse. (2009). Assisting students struggling with reading: Response to intervention (RtI) and multi-tier intervention in the primary grades. Institute of Education Sciences (2009) provides differentiated recommendations for screening, small-group instruction, monitoring, and intensive intervention. The evidence classification is attached to each recommendation, and not all recommendations have the same evidentiary strength. This is a scientific advantage: effective instructional components can be identified without assuming that every school implementation is equivalent.
The evidence also includes an unusually informative counterexample. The large federal Balu (2015) studied Grade 1–3 reading practices in 146 schools across 13 states during the 2011–12 school year. Its regression-discontinuity analysis compared students close to school-defined screening thresholds. For Grade 1 students just below thresholds and assigned to intervention, spring reading scores were lower than for comparable students just above the thresholds; Grades 2 and 3 showed no statistically significant effects at those margins. This finding describes the marginal effect of assignment in the observed implementations, rather than the average effect of every possible reading intervention or every MTSS model.
Taken together, the studies establish an important methodological distinction. Efficacious, well-defined instruction delivered to selected students does not establish the effectiveness of the whole system that identifies those students, allocates resources, and coordinates classroom time. Selection thresholds can change who receives an intervention, and intervention placement may alter the instructional experiences available to selected students. Framework-level claims require framework-level evaluations.
For older students, the What Works Clearinghouse. (2022). Providing reading interventions for students in Grades 4–9. Institute of Education Sciences (2022) emphasizes decoding of multisyllabic words, purposeful fluency work, comprehension practices, and access to challenging text. These are domain-specific instructional recommendations that should inform the content of tiers. They also demonstrate why early-grade decoding screeners cannot simply be transplanted into every secondary-school intervention decision.
Evidence from Mathematics Intervention
Mathematics poses different screening and monitoring challenges because proficiency involves multiple interacting domains: numerical magnitude, calculation, procedural flexibility, conceptual understanding, word-problem interpretation, and mathematical language. A short computation measure may track fluency while remaining relatively insensitive to changes in mathematical reasoning. A screening instrument that predicts broad difficulty may be poorly aligned with the immediate instructional target.
The Gersten (2009) offers eight recommendations with distinct evidence ratings. These include screening, explicit and systematic instruction, attention to numerical representations, structured word-problem teaching, retrieval practice, progress monitoring, and motivation. Strong evidence was assigned to some forms of explicit teaching and word-problem instruction, while other components received weaker ratings. The relevant synthesis is that intervention content is experimentally consequential, whereas the use of a tier label offers little information about the quality of mathematical instruction.
The What Works Clearinghouse. (2021). Assisting students struggling with mathematics: Intervention in the elementary grades. Institute of Education Sciences (2021) provides updated instructional recommendations involving mathematical language, representations, number lines, word problems, and fluency. A school's monitoring plan should correspond to these instructional aims rather than relying exclusively on whatever short assessment is easiest to administer.
Mathematical progress may not be well represented by a single smooth growth slope. A student can consolidate a conceptual relation, exhibit a period of apparently stable test performance, and then gain rapidly when a prerequisite is mastered. In other cases, repeated practice produces rapid fluency gains with minimal transfer to complex problems. Interpretation therefore benefits from complementary measures and work-sample analysis, including task difficulty and transfer to new examples.
The broader lesson is measurement alignment. A tiered intervention process gains explanatory power when schools state the specific skill targeted, the expected mechanism of change, and the evidence that the outcome measure captures that change. Where these relations remain opaque, the program is difficult to evaluate and difficult to improve.
Behavioral Prevention and Schoolwide Systems
Behavioral tiers share the preventive logic of MTSS but require distinct outcome constructs. A school's objective may involve fewer disruptive interactions, improved attendance, greater engagement, stronger social participation, or reduced exclusionary discipline. These outcomes are influenced by student behavior, staff practices, classroom management, referral definitions, and institutional rules.
Horner (2009) reported a randomized wait-list effectiveness trial in elementary schools in Hawaii and Illinois, finding improvements in implementation of universal schoolwide positive behavior support following training and technical assistance. Implementation change is an important outcome, though the causal chain to individual students requires distinct measurements.
In a longitudinal randomized school trial, Bradshaw (2010) examined schoolwide positive behavioral interventions and supports across 37 elementary schools, assessing fidelity and outcomes that included office discipline referrals, suspensions, and achievement. In a related analysis, Bradshaw (2012) evaluated child-level functioning across the trial's schools and reported favorable effects on several behavior and adjustment outcomes. The latter publication describes a sample of 12,344 children and estimates that students in intervention schools were less likely to receive office discipline referrals. These findings support particular schoolwide behavioral prevention practices, with generalizability bounded by the participating schools, outcomes, and implementation conditions.
Behavioral data require special interpretive scrutiny. Office discipline referrals are records created through adult observation and decisions. Their frequency may change after staff training even when a student's underlying behavior changes little. For this reason, referral counts are useful institutional indicators but cannot be treated as direct, context-free measures of aggression, self-regulation, or clinical symptoms.
Behavioral progress monitoring can include direct observation, interval recording, frequency counts, rating scales, daily behavior reports, and other functionally defined measures. Instrument selection depends on whether the intended construct is an event rate, participation pattern, emotional symptom, or context-bound behavior. The National Center on Intensive Intervention. (n.d.). Behavior progress monitoring tools chart. American Institutes for Research. https://charts.intensiveintervention.org/bprogressmonitoring (n.d.), updated in April 2026, evaluates several technical characteristics including reliability, validity, bias evidence, and sensitivity to behavior change.
As with academic intervention, a behavioral tier requires a credible intervention hypothesis. Consequence patterns, classroom expectations, instructional engagement, peer interaction, and environmental stressors may all matter. Treating referral reduction as an end in itself risks ignoring whether students actually experience safer and more inclusive school environments.
Academic and Behavioral Integration
Integrated MTSS attempts to align academic and behavioral prevention within a coherent school system. Its organizational proposition is that shared teams, decision calendars, data governance, and resource allocation can reduce fragmentation and enable a more complete understanding of student needs. McIntosh (2016) provide a major practical formulation of this integrated approach.
Academic and behavioral processes often interact causally. Reading difficulty can affect classroom participation, avoidance, and emotional experience; disruptive conditions can reduce instructional time and learning opportunity. These relations justify cross-domain investigation. They do not justify combining different measures into an opaque single risk score without explicit validation.
Integration is methodologically demanding because teams must distinguish changes in a primary outcome from changes in the system that records it. A behavioral referral, a word-reading score, and an attendance count can each indicate a need for inquiry, but their numerical aggregation requires theoretical and empirical justification. Unweighted composite scores are especially vulnerable to arbitrary scaling and concealed tradeoffs.
A school can evaluate integration through the time between screening and instructional response, coherence of intervention plans, family engagement, distribution of specialist resources, reduction of duplicated assessments, and student outcomes. These service-process metrics should be reported alongside academic and behavioral results. Improving efficiency while leaving important outcomes unchanged may be institutionally valuable, but it should be identified as an efficiency effect.
Integrated models also need boundaries. A schoolwide framework may coordinate mental-health screening and referral pathways, yet formal diagnosis, treatment, and clinical safety decisions require qualified expertise and their own professional standards. Educational tier membership is an organizational designation rather than a medical classification.
Implementation Fidelity and Organizational Capacity
Implementation fidelity refers to the correspondence between intended and delivered practices. For MTSS it includes adherence to the specified intervention, quality of delivery, student exposure, progress-monitoring administration, use of decision rules, and the responsiveness of the system to results. Fidelity has both a scientific and practical role: without it, an observed outcome cannot confidently be attributed to the intended intervention.
A school may possess high-quality screening tools and evidence-based materials but fail to allocate staff time for intervention, experience chronic teacher turnover, or provide insufficient training. The resulting service can look formally complete in planning documents while delivering little meaningful assistance. The first research question in such cases concerns implementation capacity.
Implementation itself can be measured. For schoolwide PBIS, the PBISApps. (n.d.). Tiered Fidelity Inventory (TFI). University of Oregon. https://www.pbisapps.org/products/tfi (n.d.) describes structured team assessment across tiers. The current TFI 3.0 has a scoring structure distinct from earlier editions, which underscores why historical fidelity percentages should not be compared mechanically across versions without a validated conversion. Fidelity scores are evidence about implementation, not direct measures of student benefit.
Leadership, scheduling, intervention expertise, data literacy, and school climate influence whether progress monitoring creates action. A useful research design therefore measures the availability of personnel, actual intervention dosage, team decision latency, and barriers to access. These mediators can help explain why nominally similar MTSS systems yield different outcomes.
Fidelity should support improvement rather than punishment. When teachers report inadequate materials, impossible caseloads, or conflicting scheduling demands, these are data about the organization's opportunity structure. An accountability system that treats every deviation as an individual compliance failure may suppress the very information needed to improve services.
Psychometric Validity and Decision Error
Progress monitoring is consequential measurement. It may influence a student's placement, instructional time, disability evaluation, and sense of academic competence. The required standard therefore extends beyond producing a graph. Schools need evidence that scores are reliable for the intended use and substantively related to the construct being interpreted.
Several sources of error are particularly important in repeated measurement: form-to-form difficulty, scoring inconsistency, temporal variability, rater effects, practice effects, language accessibility, and floor or ceiling limitations. Errors in level and errors in slope are related but distinct. A measure can yield a reasonably stable ranking at one point while providing weak evidence about individual growth.
The National Center on Intensive Intervention. (n.d.). Academic progress monitoring tools chart. American Institutes for Research. https://charts.intensiveintervention.org/aprogressmonitoring (n.d.) evaluates level-score reliability and validity, growth reliability and validity, alternate forms, decision rules, and bias analysis. This framework is useful because it separates technical properties rather than assigning a generic quality label to an instrument. A measure that is adequate for universal screening may require additional evidence for frequent high-stakes instructional decisions.
Decision thresholds themselves require evaluation. If teams intensify support after a fixed number of below-goal points, they should know how often the rule generates changes for students whose true learning rates are acceptable and how often it misses genuine instructional failures. Statistical uncertainty around individual slopes can be substantial. A policy of rigidly escalating after every noisy dip may overburden students and staff; a policy of indefinite delay may deny needed support.
A defensible measurement plan records the target construct, assessment conditions, validation population, score scale, expected variability, decision rule, and relevant consequences. This explicitness allows later audits of whether decisions were psychometrically justified and equitable across student groups.
Learning Disability Identification and Legal Context
Response-to-intervention data can contribute to special education evaluation in the United States, but intervention participation and disability identification remain distinct decisions. Under U.S. Department of Education. (n.d.). 34 CFR § 300.307: Specific learning disabilities. Individuals with Disabilities Education Act regulations. https://sites.ed.gov/idea/regs/b/d/300.307 (n.d.), states must permit consideration of response to scientific, research-based intervention for specific learning disability identification and cannot require severe discrepancy. U.S. Department of Education. (n.d.). 34 CFR § 300.309: Determining the existence of a specific learning disability. Individuals with Disabilities Education Act regulations. https://sites.ed.gov/idea/regs/b/d/300.309 (n.d.) describes conditions including inadequate achievement and insufficient progress under specified procedures, together with other eligibility considerations.
These rules do not permit a school to equate the absence of a desired slope with a diagnosis, and they do not make a three-tier program the only valid pathway to evaluation. They require contextual interpretation, evidence that instruction was appropriate, and compliance with the full evaluation process and applicable state rules. A school must distinguish the question of what assistance a student needs now from the legal question of whether the student meets statutory eligibility criteria.
The historical appeal of response-based identification was to consider learning under actual instruction. Its conceptual difficulty is that response is simultaneously affected by learner characteristics, intervention quality, measurement, and context. A low observed response may contribute important information, but it is a property of a learner–instruction relationship rather than a pure test of inherent disability.
Families need comprehensible explanations of what was taught, how progress was measured, which decisions were made, and which rights and evaluation pathways remain available. Documentation should include instructional conditions and evidence of intervention fidelity. The clinical and legal stakes rise when a school treats cumulative educational delay as a reason to postpone comprehensive evaluation.
Multilingualism, Disability, and Educational Equity
An equitable MTSS requires valid interpretation across language, disability, socioeconomic opportunity, and educational history. These dimensions may affect baseline achievement, the accessibility of intervention, and the meaning of screen scores. The scientific problem is not solved by using a common benchmark if groups differ in the relationship between that benchmark and the intended outcome.
The Linan-Thompson (2006) randomized study is relevant because its participants were English language learners, and the investigators studied responses to supplemental reading instruction. It demonstrates that students whose home and school languages differ can respond to well-specified intervention; its sample and design do not establish a universal intervention protocol for all multilingual populations.
Fairness analysis should inspect the representativeness of the instrument's validation sample, potential differential item functioning, differential prediction, access to accommodations, and errors in referral classification. The National Center on Intensive Intervention. (n.d.). Academic screening tools chart. American Institutes for Research. https://charts.intensiveintervention.org/ascreening (n.d.) includes a bias-analysis category precisely because adequate average prediction does not settle questions of performance across populations.
Educational equity also involves the allocation of scarce support. Students in schools with fewer trained specialists may wait longer after the same screening result. A student may be referred more frequently because staff use different behavior thresholds. Districtwide comparisons should therefore consider opportunity measures, implementation variation, and distributions of services as well as mean scores.
An ethical system gives students and families intelligible reasons for decisions and avenues to revisit them. Data governance should minimize unnecessary disclosure, define who can access individual records, and avoid making a student's tier history a permanent educational identity. The proper object of protection is the learner's opportunity for growth and participation.
Research Designs and Causal Interpretation
MTSS research spans instrument-development studies, single-case designs, intervention trials, school-level cluster-randomized studies, observational cohort analysis, regression discontinuity, implementation evaluations, and qualitative inquiry. These designs answer different questions. Reliability and validity studies test whether an instrument works for a specified decision. Trials test defined interventions in defined contexts. Implementation studies assess organizational processes. Observational studies can identify associations and generate hypotheses about determinants of success.
Single-case experiments and repeated-measures designs can be particularly relevant to intensive individualized interventions, where a small number of students receive highly adapted supports. Their interpretive power depends on sufficient repeated observations, meaningful baseline conditions, defensible phase changes, and appropriate controls for history and maturation. A simple before–after graph without such controls supports weaker causal claims.
Cluster-randomized trials are important for schoolwide practices because the intervention is delivered at the level of the institution. Bradshaw (2012) evaluated a school-level randomized PBIS intervention while measuring child outcomes longitudinally. Correct statistical inference must account for similarities among students in the same school, attrition, and the number of randomized clusters. A very large number of student observations does not automatically substitute for an adequate number of independently assigned schools.
Regression discontinuity provides a different estimand. In Balu (2015), assignment near school-determined screening thresholds supported comparisons of students just below and above those thresholds. This method estimates effects near the cutoff under its design assumptions. It cannot automatically establish effects for the most severely struggling students, who may be far below the threshold, or for a differently organized implementation.
A mature evidence synthesis should ask who was studied, what was randomized or compared, how treatment actually differed, whether follow-up was sufficient, which outcome was primary, and whether observed effects are educationally meaningful. It should distinguish a well-defined intervention's effect from the effect of an entire framework that combines several interventions and administrative decisions.
Conflicting Findings and Evidence Appraisal
The evidence surrounding MTSS is heterogeneous because MTSS describes a system organization rather than one fixed therapeutic or instructional ingredient. Positive results from specific interventions, weak evidence for particular assessment rules, and null or unfavorable system-level effects can all be true. The scientific task is to explain how they fit together and what conclusions each supports.
Stecker (2005) reviewed experimental-contrast studies in which systematic CBM-supported instructional decisions were associated with improved achievement, especially where teachers had decision rules and instructional guidance. This supports the proposition that measurement can improve outcomes when it changes instruction.
The Balu (2015) reported unfavorable Grade 1 outcomes at assignment thresholds and no statistically significant threshold effects for Grades 2 and 3. The evaluation identified plausible mechanisms that included misclassification and mismatch of interventions with need, without experimentally isolating those explanations. The correct inference is that implementation and assignment can undermine anticipated benefit in practice.
Bradshaw (2012) reported favorable behavioral effects from a schoolwide randomized prevention program, with outcomes and implementation processes specific to PBIS. These findings have stronger causal implications for the studied behavioral program than for arbitrary academic interventions attached to the MTSS label.
The evidence base is therefore best represented as a matrix of research objects. Universal screening requires predictive validation. Academic programs require content-specific treatment evaluation. Behavioral prevention requires schoolwide outcome trials. Progress monitoring requires longitudinal measurement and decision-validity evidence. Whole-system MTSS requires independent evaluation of the combined processes, equity effects, and implementation costs. A favorable study in one cell cannot fill every other cell.
Evidence appraisal also considers publication bias, outcome selection, differences in intervention dosage, developer involvement, limited geographical sampling, and persistence of effects. A program may increase a proximal skill score without improving later classroom achievement. Long-term maintenance, student experience, and costs should be measured alongside immediate test performance.
Progress Monitoring as a Feedback System
The common organizing principle across these results is a feedback cycle in which educational action generates observations and observations inform revised action. In a functioning cycle, the student is not merely classified; the instructional environment is modified and assessed. This makes progress monitoring an element of school learning as well as student learning.
A transparent process can be described through six operations. First, a team defines a meaningful skill and goal. Second, it selects a measure with relevant validity evidence. Third, it delivers a documented intervention. Fourth, it gathers repeated observations under standardized conditions. Fifth, it evaluates level, growth, uncertainty, and implementation. Sixth, it makes an explicit instructional decision and records the reason. Subsequent observations test the adequacy of that decision.
Feedback systems can fail through delay, unreliable observation, poorly specified goals, or actions that are unrelated to the information collected. An assessment system that identifies struggling students but cannot provide intervention is an incomplete feedback system. A school that delivers intensive tutoring without reviewing outcomes may squander opportunities to adapt teaching. A school that changes instruction after every random fluctuation introduces instability into the learning environment.
The National Center on Intensive Intervention. (n.d.). Progress monitoring. American Institutes for Research. https://intensiveintervention.org/data-based-individualization/progress-monitoring (n.d.) emphasizes planning the tool, goal, and assessment frequency before intervention, graphing observations, and using the resulting evidence to determine whether to maintain or intensify support. Its sequence is a useful operational model, provided the selected intervention and measures meet relevant scientific standards.
The major theoretical insight is that an educational system's responsiveness should be evaluated as carefully as the student's responsiveness. When data reveal inadequate progress, the school must demonstrate that it has the capacity and willingness to revise its action. The quality of the system is expressed through its ability to learn from its own outcomes.
Digital Assessment and Artificial Decision Support
Digital platforms can reduce administrative burden, aggregate repeated observations, display learning trajectories, and flag patterns requiring human review. Adaptive assessments may provide more information within limited testing time. These possibilities connect school psychology with psychometrics, learning analytics, human–computer interaction, and the wider study of Artificial systems.
Technological automation introduces new measurement and governance questions. A predictive risk model may inherit biases from past referral practices. A dashboard can encourage overconfidence by presenting uncertain estimates as precise colors or categories. A recommendation engine trained in one population may perform differently in another. The legitimacy of an algorithmic intervention recommendation depends on documented construct validity, prospective evaluation, appropriate subgroup analyses, and actual improvement in educational decisions.
Artificial decision support can be assessed through testable hypotheses. Does a model improve the accuracy or timeliness of identifying students who need assistance? Does it improve intervention matching compared with existing teams? Are recommendations interpretable and actionable? Does it alter student outcomes, referral disparities, staff workload, or family trust? These questions require independent field evaluation, rather than inference from model benchmark accuracy alone.
The Artificial Era broadens the institutional environment in which human learning is observed and supported. Artificial systems can participate in records management, learning analytics, and instructional assistance as organized nonbiological agents within educational work. This philosophical horizon identifies a new domain of inquiry; empirical claims about a particular system must remain grounded in its measurable capabilities and outcomes.
Students should retain access to appropriate human professional judgment, and families should know when consequential recommendations rely on automated systems. Privacy, data retention, access controls, and opportunities to contest erroneous classifications become particularly important when longitudinal educational records are integrated across platforms.
Unresolved Questions and Research Agenda
Research on MTSS has matured in instrumentation and in several intervention-specific areas, yet major integrative questions remain. There is a need for multi-site, independently replicated evaluations of whole systems that measure implementation quality, intervention content, administrative burden, and long-term academic and psychosocial outcomes. Interventions must be compared with credible alternatives, rather than only with weak or absent support.
A second priority concerns the decision rules that connect progress graphs to action. Researchers should prospectively compare plausible thresholds, slope estimators, monitoring frequencies, and decision procedures across differing student populations. The key outcomes are not merely agreement among raters but changes in learning and the cost of mistaken decisions.
A third priority concerns individual heterogeneity. A student's response may vary by skill, developmental stage, instructional language, school conditions, and prior opportunity. Research should test how combinations of status and growth measures perform across subgroups and whether more individualized rules actually improve outcomes rather than overfit noisy records.
A fourth priority is equity and system accountability. Evaluations should examine who receives timely intervention, how long they wait, whether family communication is meaningful, and whether errors are distributed unevenly. Institutional resource inequality may determine which schools can deliver the support promised by identical framework documents.
A fifth priority concerns transfer and sustainability. Short-term changes in fluent task performance should be related to comprehension, broader achievement, participation, and subsequent educational transitions. Staff turnover and fluctuating resource allocations can erode implementation even after an initial successful trial.
Finally, artificial and digitally mediated systems require independent validation of automated predictions and recommendations. Comparative trials should test whether their use actually improves decisions, protects students' rights, and expands access. Responsible innovation will depend on measurement integrity and transparent institutional accountability.
Conclusion
Multi-tiered systems of support organize a schoolwide commitment to delivering instruction, detecting need, adapting assistance, and evaluating outcomes. Their scientific foundation is distributed across curriculum-based measurement, response-to-intervention research, educational treatment studies, behavioral prevention, school psychology, and organizational implementation. Progress monitoring makes the process longitudinal by showing how performance changes under a documented intervention.
The strongest conclusions are component specific. Repeated measurement can support better instructional decisions when scores are valid and linked to explicit adaptations. Well-specified reading and mathematics practices can improve targeted learning outcomes. Randomized studies support particular schoolwide behavioral prevention interventions. Large-scale evaluations also reveal that assignment and implementation can generate null or unfavorable effects. Therefore, the mere existence of tiers or screening instruments is insufficient evidence of student benefit.
A scientifically accountable MTSS evaluates instruction, assessment, decision rules, and access together. It interprets every student response within the educational conditions that produced it, protects comprehensive evaluation and family participation, and revises its own practices when outcomes are inadequate. The enduring research object is the coupled process through which schools transform evidence about learning into better opportunities for learning.
References
- Balu, R., Zhu, P., Doolittle, F., Schiller, E., Jenkins, J., & Gersten, R. (2015). Evaluation of response to intervention practices for elementary school reading (NCEE 2016-4000). National Center for Education Evaluation and Regional Assistance, Institute of Education Sciences, U.S. Department of Education. https://ies.ed.gov/use-work/resource-library/report/evaluation-report/evaluation-response-intervention-practices-elementary-school-reading
- Bradshaw, C. P., Mitchell, M. M., & Leaf, P. J. (2010). Examining the effects of schoolwide positive behavioral interventions and supports on student outcomes: Results from a randomized controlled effectiveness trial in elementary schools. Journal of Positive Behavior Interventions, 12(3), 133–148. https://doi.org/10.1177/1098300709334798
- Bradshaw, C. P., Waasdorp, T. E., & Leaf, P. J. (2012). Effects of school-wide positive behavioral interventions and supports on child behavior problems. Pediatrics, 130(5), e1136–e1145. https://doi.org/10.1542/peds.2012-0243
- Deno, S. L. (1985). Curriculum-based measurement: The emerging alternative. Exceptional Children, 52(3), 219–232. https://doi.org/10.1177/001440298505200303
- Fuchs, D., & Fuchs, L. S. (2006). Introduction to response to intervention: What, why, and how valid is it? Reading Research Quarterly, 41(1), 93–99. https://doi.org/10.1598/RRQ.41.1.4
- Gersten, R., Beckmann, S., Clarke, B., Foegen, A., Marsh, L., Star, J. R., & Witzel, B. (2009). Assisting students struggling with mathematics: Response to intervention (RtI) for elementary and middle schools (NCEE 2009-4060). Institute of Education Sciences, U.S. Department of Education. https://ies.ed.gov/ncee/wwc/PracticeGuide/2
- Horner, R. H., Sugai, G., & Anderson, C. M. (2010). Examining the evidence base for school-wide positive behavior support. Focus on Exceptional Children, 42(8). https://doi.org/10.17161/foec.v42i8.6906
- Horner, R. H., Sugai, G., Smolkowski, K., Eber, L., Nakasato, J., Todd, A. W., & Esperanza, J. (2009). A randomized, wait-list controlled effectiveness trial assessing school-wide positive behavior support in elementary schools. Journal of Positive Behavior Interventions, 11(3), 133–144. https://doi.org/10.1177/1098300709332067
- Linan-Thompson, S., Vaughn, S., Prater, K., & Cirino, P. T. (2006). The response to intervention of English language learners at risk for reading problems. Journal of Learning Disabilities, 39(5), 390–398. https://doi.org/10.1177/00222194060390050201
- McIntosh, K., & Goodman, S. (2016). Integrated multi-tiered systems of support: Blending RTI and PBIS. Guilford Press. https://www.guilford.com/books/Integrated-Multi-Tiered-Systems-of-Support/McIntosh-Goodman/9781462524747
- National Center on Intensive Intervention. (n.d.). Academic progress monitoring tools chart. American Institutes for Research. https://charts.intensiveintervention.org/aprogressmonitoring
- National Center on Intensive Intervention. (n.d.). Academic screening tools chart. American Institutes for Research. https://charts.intensiveintervention.org/ascreening
- National Center on Intensive Intervention. (n.d.). Behavior progress monitoring tools chart. American Institutes for Research. https://charts.intensiveintervention.org/bprogressmonitoring
- National Center on Intensive Intervention. (n.d.). Data-based individualization: Intensive intervention. American Institutes for Research. https://intensiveintervention.org/data-based-individualization
- National Center on Intensive Intervention. (n.d.). Progress monitoring. American Institutes for Research. https://intensiveintervention.org/data-based-individualization/progress-monitoring
- PBISApps. (n.d.). Tiered Fidelity Inventory (TFI). University of Oregon. https://www.pbisapps.org/products/tfi
- Stecker, P. M., Fuchs, L. S., & Fuchs, D. (2005). Using curriculum-based measurement to improve student achievement: Review of research. Psychology in the Schools, 42(8), 795–819. https://doi.org/10.1002/pits.20113
- U.S. Department of Education. (n.d.). 34 CFR § 300.307: Specific learning disabilities. Individuals with Disabilities Education Act regulations. https://sites.ed.gov/idea/regs/b/d/300.307
- U.S. Department of Education. (n.d.). 34 CFR § 300.309: Determining the existence of a specific learning disability. Individuals with Disabilities Education Act regulations. https://sites.ed.gov/idea/regs/b/d/300.309
- Vaughn, S., Linan-Thompson, S., & Hickman, P. (2003). Response to instruction as a means of identifying students with reading/learning disabilities. Exceptional Children, 69(4), 391–409. https://doi.org/10.1177/001440290306900401
- What Works Clearinghouse. (2009). Assisting students struggling with reading: Response to intervention (RtI) and multi-tier intervention in the primary grades. Institute of Education Sciences, U.S. Department of Education. https://ies.ed.gov/ncee/wwc/PracticeGuide/3
- What Works Clearinghouse. (2021). Assisting students struggling with mathematics: Intervention in the elementary grades. Institute of Education Sciences, U.S. Department of Education. https://ies.ed.gov/ncee/wwc/PracticeGuide/26
- What Works Clearinghouse. (2022). Providing reading interventions for students in Grades 4–9. Institute of Education Sciences, U.S. Department of Education. https://ies.ed.gov/ncee/wwc/PracticeGuide/29