Response-to-Intervention Research: Is the Sum of the Parts as Great as the Whole?
How often do special education teachers hear that an intervention is "research based?" The answer is, just about every time a publisher tries to the sell them intervention materials. Fortunately, teachers have become more adept at recognizing the ones that are supported by research. This is an important shift because education has such a long history of fads that Ellis (2005) concluded that in education, "today's flagship is often
tomorrow's abandoned shipwreck" (p. 200).
The best way to avoid fads in educational practice is research. Too often in K-12 schools, interventions without a solid research base are widely implemented and interventions supported by compelling research are not. Thus, teachers are recognizing that when evaluating a potential intervention or educational practice, the first question should be "Is it a scientific, research-based intervention?" Currently, What Works Clearinghouse's (WWC) cites studies with participants randomly assigned to treatment and control conditions as providing the strongest evidence. That does not mean that practitioners cannot benefit from reading other types of studies; it simply indicates that the research about a particular intervention should include studies that represent strong evidence through use of randomized designs. In addition, practitioners should look for a meta-analysis, which is a study that empirically synthesizes several other studies, to be confident in the research (Ellis, 2005). A meta-analysis can provide important information about the effectiveness of an intervention because the conclusions are based on multiple sources of information rather than just one study.
Response to intervention (RTI) is a current topic about which teachers are seeking research information. RTI is the practice of providing quality instruction and intervention and using student learning in response to that instruction to make instructional and important educational decisions (Batsche et al., 2005). It is perhaps the most discussed educational initiative in the country today. School districts in every state in the country are currently implementing RTI and some have been doing so for over 20 years with impressive results. However, because of the nature of RTI as a schoolwide initiative, there are no studies that examine the model in its entirety using a randomized design. Although that is problematic from a research perspective, it makes perfect sense from a practical one. Imagine having to tell staff and parents that because they are in the control group for the next five years they are not allowed to implement a program that could help their kids. While at the same time, another school is fully implementing this beneficial practice. In the case of RTI, we have to examine the sum of the parts to evaluate the whole because components of an RTI model, unlike a schoolwide model, can be examined with rigorous research over long periods of time.
The purpose of this article is to describe the core components of RTI and the relatively limited research on RTI models in their entirety. Although RTI has many core components of RTI (e.g., data-based decision making, collaboration, and problem-analysis), the focus of this article is research that addresses Tier 1 (quality core instruction), Tier 2 (supplemental intervention), and Tier 3 (individualized interventions). Moreover, this article will identify and discuss relevant meta-analytic research for the topics listed above. Readers interested in syntheses of core components should read the summary available at www.nasdse.org provided by Griffith, Parson, Burns, VanDerHeyden, and Tilly (2007).
RTI Models as a Whole
Many articles provide descriptions of RTI models in their entirety and data to support their effectiveness (e.g., Marston, Muyskens, Lau, & Canter, 2003; McNamara & Hollinger, 2003) A recent meta-analysis of RTI research found large effects for both systemic (e.g., reductions in special education referrals) and student outcomes (e.g., increased increase reading scores) (Burns, Appleton, & Stehouwer, 2005). The data provided within this meta-analysis suggest that RTI is an effective practice, but the data were not disaggregated by study design primarily because of the lack of randomized studies for reasons mentioned above.
VanDerHeyden, Witt, and Gilbertson (2007) conducted an evaluation of an RTI model over multiple years. Instead of random assignment, their study used a multiple baseline design, which examines outcome data from an intervention with a staggered onset of implementation. The WWC (2008) recognizes research from multiple baseline designs as meeting standards for evidence base. Moreover, a multiple baseline design allows for data to be studied for groups of schools whose RTI implementation start dates are staggered across a few years. Data gathered before and after implementation of the intervention can then be used to determine the impact of the intervention. The results of this comprehensive study indicated that the RTI model reduced the number of students evaluated for special education services, essentially eliminated the disproportional rate at which ethnic minority and male students were referred for special education evaluations, and substantially reduced the amount of financial resources dedicated to unnecessary special education evaluations.
Researching the Parts of RTI
An effective RTI model should begin with quality core instruction that adequately addresses the needs of most of the students. If more than 20-25% of the students require additional support than what is provided in Tier 1, then the school will not have the resources necessary to address the needs of those students. Moreover, interventions should be highly and correctly targeted to be effective, but students cannot learn to read and do math if they are not receiving quality balanced instruction in addition to supplemental support. Fortunately, the National Reading Panel (2000) and National Mathematics Advisory Panel (2008) have both conducted meta-analyses to determine what constitutes quality instruction in those core areas. Moreover, individual studies of math (Crawford & Snider, 2000) and reading (Foorman, Francis, & Fletcher, 1998) have used strong research designs and found that the quality of the curriculum and the explicitness of the instruction led to improved student learning and reduced future student failures. It would go beyond the scope of this article to describe what these studies and meta-analyses found to be critical components of quality instruction and curriculum, but there is a well-confirmed research base that quality instruction leads to fewer students needing additional support. Readers are encouraged to examine the Florida Center for Reading Research website (www.fcrr.org) and the report of the National Reading Panel (2000) for more information about quality instructional practices.
According to many publications about the three-tier model of RTI, a school's goal is for no more than 20% of students to require additional support beyond good Tier I curriculum and instruction (Burns et al., 2005). For those students, an RTI model relies on supplemental interventions delivered in small groups for 20 to 30 minutes daily (Vaughn, Wanzek, Linan-Thompson, & Murray, 2007). The few meta-analyses of small-group interventions that have been conducted found moderate to strong effects of daily Tier 2 instruction (Elbaum, Vaughn, Hughes, & Moody, 2000). However, a panel convened by the Institute for Education Science (IES) found strong evidence for the effectiveness of providing small-group interventions as supplemental instruction to support the Tier I core curriculum (Gersten et al., 2009a). According to the panel, small-group supplemental instruction should a) target the components of reading instruction in which the student needs additional support, b) be implemented three to five times each week for approximately 20 to 40 minutes each session, and c) build skills gradually with high student-teacher interaction and frequent opportunities to practice the specific skill and receive feedback. It is also important to note, that the instruction provided within Tier 2 needs to focus on an aspect of reading (e.g., decoding) and that students need practice in that specific skill. Simply allowing a struggling reader more time to read, even if the text is carefully selected to provide an appropriate level of challenge, will likely not remediate the deficit in the long run.
After reviewing research on math instruction, the IES panel reached a similar conclusion regarding Tier 2 for math as did the reading experts. Tier 2 interventions are critical for success in math as well, and there is strong evidence to support the effectiveness of the interventions if they include explicit and systematic instruction and focus on common underlying structures of problems (Gersten et al., 2009b).
Schools implementing effective Tier 1 and Tier 2 instruction should find no more than 5% of students requiring more intensive interventions than those provided in Tier 2 (Burns et al., 2005). Students who need individualized support receive it from a Tier 3 intervention, the hallmark of which is the intensity of the intervention rather than how it is delivered. Some educators assume that Tier 3 interventions are delivered one-on-one, but that may not be necessary and is not the essential attribute of an effective Tier 3 intervention. Instead, Tier 3 interventions should be specific to individual student needs and involve sufficient resources to address those needs (Burns & Gibbons, 2008).
Meta-analytic research has found several effective interventions for students with severe learning difficulties and identified learning disabilities including mnemonic strategies, explicit reading comprehension instruction (e.g., vocabulary, pre- and mid-reading, and direct instruction of strategies), behavior modification, and direct instruction (Kavale & Forness, 2000). Moreover, several components of effective interventions for students with learning disabilities that could inform Tier 3 interventions were identified with meta-analytic research (Swanson & Sachse-Lee, 2000) and a research synthesis (Burns, VanDerHeyden, & Boice, 2008). Again, to fully describe the components would go beyond the scope of this article, but there is a strong research base from which to work that is summarized in Table 1.
TABLE 1: Research-based effective intervention practices for tiers 2 and 3
||se explicit and systematic instruction
||Gersten et al. (2009a)
||Use groups of three to five students
||Elbaum et al. (2000)
||Provide instruction in up to three foundational reading skills
||Gersten et al. (2009a)
|2 & 3
||Focus on underlying skills and structures for math and reading
||Gersten et al. (2009a;b)
|2 & 3
||Progress should be closely monitored until approximately eight data points are collected to assure sufficient reliability
||Use an intervention with at least three components that are different from Tiers 1 and 2
||Swanson & Sachse-Lee (2000)
||Potential intervention components include (a) highly targeted, (b) provides an appropriate level of challenge for the indvidual student, (c) explicitly teaches a specific skill, (d) allows many opportunities to respond, and (e) provides immediate corrective feedback for an individual student
||Burns et al. (2008)
It is difficult to make any summative statements about RTI research given that the research is ongoing and the research base is still not definitive for essential aspects of RTI. For example, most RTI models rely on a problem-solving team (PST) to identify interventions within Tier 3 (Burns & Ysseldyke, 2005). Meta-analytic research found strong effects associated with PSTs (Burns & Symington, 2002), but the quality of the research of the studies within the meta-analysis is unknown. Moreover, practitioners should take great strides to assure that the RTI model is implemented with fidelity, but we are unsure how to best accomplish this important objective. There is also little empirical support for various problem-solving and problem-analysis models implemented within RTI, and research on RTI models for middle and high schools is only just beginning.
Practitioners should implement a three-tiered RTI model with confidence that they are engaging in research-based practice that benefits students. RTI is a promising practice that already has positively influenced the lives of countless children. If we can take what research has already taught us, develop ways to directly translate that research into practice, and continue to evolve our practices based on cutting edge research, then RTI will be the accepted, ongoing approach to instruction rather than just another fad. RTI may be a collection of parts accumulated and pieced together over decades of research and practice, but the result of this compilation of parts is a sum that equals positive outcomes for kids.
Batsche, G., Elliott, J., Graden, J. L., Grimes, J., Kovaleski, J. F., Prasse, D., et al. (2005). Response to intervention policy considerations and implementation. Reston, VA: National Association of State Directors of Special Education.
Burns, M. K., Appleton, J. J., & Stehouwer, J. D. (2005). Meta-analysis of response-to-intervention research: Examining field-based and research-implemented models. Journal of Psychoeducational Assessment, 23, 381–394.
Burns, M. K., & Gibbons, K. (2008). Response to intervention implementation in elementary and secondary schools: Procedures to assure scientific-based practices. New York: Routledge.
Burns, M. K., & Symington, T. (2002). A meta-analysis of prereferral intervention teams: Systemic and student outcomes. Journal of School Psychology, 40, 437–447.
Burns, M. K., VanDerHeyden, A. M., & Boice, C. H. (2008). Best practices in delivery intensive academic interventions. In A. Thomas & J. Grimes (Eds.) Best practices in school psychology (5th ed.). Bethesda, MD: National Association of School Psychologists.
Burns, M. K., & Ysseldyke, J. E. (2005). Questions about responsiveness-to-intervention implementation: Seeking answers from existing models. California School Psychologist, 10, 9–20.
Christ, T. J. (2006). Short term estimates of growth using curriculum-based measurement of oral reading fluency: Estimates of standard error of the slope to construct confidence intervals. School Psychology Review, 35, 128–133.
Crawford, D., & Snider, V. E. (2000). Effective mathematics instruction: The importance of curriculum. Education and Treatment of Children, 23, 122–142.
Elbaum, B., Vaughn, S., Hughes, M., & Moody, S. (2000). How effective are one-to-one tutoring programs in reading for elementary students at risk for reading failure? A meta-analysis of the intervention research. Reading Research Quarterly, 92, 605–619.
Ellis, A. K. (2005). Research on educational innovations (4th ed.). Larchmont, NY: Eye on Education.
Foorman, B. R., Francis, D. J., & Fletcher, J. M. (1998). The role of instruction in learning to read: Preventing reading failure in at-risk children. Journal of Educational Psychology, 90, 37–55.
Gersten, R., Compton, D., Connor, C. M., Dimino, J., Santoro, L., Linan-Thompson, S., et al. (2009a). Assisting students struggling with reading: Response to intervention and multi-tier intervention in primary grades. Washington, DC: U.S. Department of Education Institute of Educational Sciences.
Gersten, R., Beckmann, S., Clarke, B., Foegen, A., Marsh, L., Star, J. R., et al. (2009b). Assisting students struggling with mathematics: Response to intervention for elementary and middle schools. Washington, DC: U.S. Department of Education Institute of Educational Sciences.
Griffin, A. J., Parsons, L., Burns, M. K., & VanDerHeyden, A. (2007). Response to intervention research to practice. Washington, DC, National Association of State Directors of Special Education.
Kavale, K. A., & Forness, S. R. (2000). Policy decisions in special education: The role of meta-analysis. In R. Gersten, E. P. Schiller, & S. Vaughn (Eds.), Contemporary special education research: Synthesis of the knowledge base on critical instructional issues,(p. 281–326). Mahwah, NJ: Lawrence Erlbaum Associates.
Marston, D., Muyskens, P., Lau, M., & Canter, A. (2003). Problem-solving model for decision making with high-incidence disabilities: The Minneapolis experience. Learning Disabilities Research & Practice, 18(3), 187–200.
McNamara, K., & Hollinger, C. (2003). Intervention-based assessment: Evaluation rates and eligibility findings. Exceptional Children, 69, 181–194.
National Mathematics Advisory Panel. (2008). Foundations for success: Final report of the national math advisory panel. Washington, DC: U.S. Department of Education.
National Reading Panel. (2000). Teaching children to read: An evidence-based assessment of the scientific research literature on reading and its implications for reading instruction. reports of the subgroups. Bethesda, MD: National Institute for Literacy.
Swanson, H. L., & Sachse-Lee, C. (2000). A meta-analysis of single-subject-design intervention research for students with LD. Journal of Learning Disabilities, 33, 114–136.
VanDerHeyden, A. M., Witt, J. C., & Gilbertson, D. A. (2007). Multi-year evaluation of the effects of a response to intervention (RTI) model on identification of children for special education. Journal of School Psychology, 45, 225–256.
Vaughn, S., Wanzek, J., Linan-Thompson, S., & Murray, C. (2007). Monitoring response to intervention for students at-risk for reading difficulties: High and low responders. In S. R. Jimerson, M. K. Burns, & A. M. VanDerHeyden (Eds.), The handbook of response to intervention: The science and practice of assessment and intervention (pp. 234–243). New York: Springer.
What Works Clearinghouse (2008). What works clearinghouse evidence standards for reviewing studies. Washington, DC: United States Department of Education.
This article was originally published in Perspectives on Language and Literacy, vol. 36, No. 2, Spring 2010, copyright by The International Dyslexia Association. www.interdys.org Used with permission.
Back To Top