Approaches to RTI
Selecting an RTI Model
Response to intervention (RTI) has a grassroots history with beginnings in multiple research areas, including applied behavior analysis, precision teaching, curriculum-based measurement (CBM), and effective teaching. Independent research efforts and practical applications in visionary districts like the Heartland Area Education Agency, St. Croix River Education District, and the Minneapolis public schools demonstrated the potential for universal screening, frequent progress monitoring, and systematic intervention to accelerate learning and enhance both the efficiency and accuracy of identification efforts in special education. Over time, general categories of RTI implementations emerged. Each type of model is described briefly below.
- Problem-solving models involve working with a school-based team to consider student performance data to identify and define learning problems, to develop interventions to solve those problems, and to evaluate the effects of the interventions on the defined problem or problems. Problem-solving models are a logical organizer to manage and evaluate system data, prioritize targets, and implement and evaluate individual intervention. Studies have demonstrated positive effects of the problem-solving model in schools, but this type of model is perhaps the most vulnerable to misapplication because the procedures and decision-making criteria are flexible and not well specified. So, for example, one problem-solving team may consider teacher verbal report of student academic performance and work samples to define the learning problem while another team may collect CBM data of student performance under different conditions (e.g., easier task, with guided practice, with incentives) to determine the cause of poor performance. The team who uses the second approach is likely to make more accurate judgments about what is causing a student’s learning problem and how best to resolve it. Hence, the outcomes of the problem-solving model depend heavily on the procedures used. Problem-solving procedures are not well specified, allowing for flexibility across sites but also causing variable or unreliable effects.
- Functional assessment models are typified by the work of Daly and colleagues (Daly, Martens, Hamler, Dool, & Eckert, 1999; Daly, Witt, Martens, & Dool, 1997) and are similar to procedures used in functional behavior assessment (Iwata, Dorsey, Slifer, Bauman, & Richman, 1982). Student performance is assessed before intervention (at baseline) and then conditions are arranged to test the effect of certain intervention efforts on student learning. Typically, these test conditions include providing easier material, providing practice responding, and providing incentives for improved performance. If a condition improves student performance (e.g., providing incentives), then that condition is used for intervention (e.g., incentives are provided for improved performance until learning has improved). While these procedures have obvious value for determining which interventions are likely to be effective for particular children experiencing particular problems, they offer little guidance about how to identify which students are in need of intervention and what decision rules to apply to resulting data to determine whether a problem has been adequately resolved with intervention or whether more assistance is required.
- Standard protocol models are typified by the work of Torgesen (e.g., Torgesen et al., 2001) and others who have implemented comprehensive interventions for reading in well-funded research trials. These studies have powerfully demonstrated the capacity for certain instructional procedures to enhance learning and have even led some scholars to characterize learning disabilities as a preventable diagnosis. The procedures are well specified and, therefore, open to replication across sites and amenable to research. However, these procedures can be difficult to replicate in applied settings because of the resources required to implement the interventions. Also, the literature on so-called standard protocol models of reading intervention has historically offered little guidance about operationalized procedures and decision criteria for determining initial risk status (or need for intervention).
- Hybridized or blended models have evolved that use some of the components of each of the three previous models. For example, as noted above, problem-solving models that use functional academic assessment procedures to define learning problems and build interventions will have better effects than problem-solving models that rely on other less-proven strategies such as verbal report. One example of a hybrid model is the STEEP model (Screening to Enhance Educational Performance, www.isteep.com), developed by Witt and colleagues (VanDerHeyden, Witt, & Gilbertson, 2007).
Research has demonstrated strong effects for RTI models in each of the above categories. Hence, there is no right or wrong model of RTI. Effects obtained under any RTI model depend on the quality with which implementation occurs. To attain strong positive effects on student learning outcomes, schools must focus heavily on the accuracy of decisions made at each stage of implementation. Effective implementation requires that schools implement procedures to do the following:
- Correctly identify students who need intervention
- Deliver intervention that effectively resolves the learning problem for the majority of students exposed to the intervention
- Monitor the effects of the intervention and troubleshoot to ensure intervention integrity and positive effects on learning
- Make decisions about the need for more intensive or less intensive intervention (e.g., progressing to higher tiers or lower tiers, discontinuing intervention)
- Link resulting RTI data to referral and eligibility decisions in special education
- Link resulting RTI data to system programming changes (e.g., resource allocation, professional development, program evaluation)
There is a direct and irrevocable relationship between how well schools do the above activities and their effects on student learning. That is, the effects obtained depend on the degree to which the above actions are correctly carried out. To select a model, districts or schools should evaluate existing practices and resources at their sites to determine the approach that will best help them accomplish the six activities listed above. In schools where core instruction in reading is actively being targeted, for example, instructors must evaluate how well they are accomplishing each of the activities above and determine which model will best help them address any activities they are missing. A guiding principle is to think of merging existing programs with RTI, enhancing effects of all programs, and substituting efforts and resources rather than adding. So consider a school in which core instruction in reading at the primary grades is being targeted. If the school has an adequate screening system in place to determine who should receive intervention and these screening data indicate that most students are successfully learning to read, but the school personnel are concerned that the special education referral team does not make use of the data to determine the need for evaluation and that the system does not make use of the data to make resource allocation decisions, then a model of RTI might be chosen that best helps implement actions 4, 5, and 6. In this case, a problem-solving model of RTI might be the best choice. On the other hand, if the district does not have an adequate screening process in place and existing data (e.g., year-end accountability scores) indicate that large numbers of children may be failing to meet expected learning standards, then a standard protocol model might be adopted to provide supplemental intervention to enhance core instruction effects on learning while initiating a technically adequate screening process for fall, winter, and spring to evaluate the effects of intervention on student learning.
Daly, E. J., III, Martens, B. K., Hamler, K. R., Dool, E. J., & Eckert, T. L. (1999). A brief experimental analysis for identifying instructional components needed to improve oral reading fluency. Journal of Applied Behavior Analysis, 32, 83–94.
Daly, E. J., III, Witt, J. C., Martens, B. K., & Dool, E. J. (1997). A model for conducting a functional analysis of academic performance problems. School Psychology Review, 26, 554–574.
Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., & Richman, G. S. (1982). Toward a functional analysis of self-injury. Analysis and Intervention in Developmental Disabilities, 2, 3–20.
Torgesen, J., Alexander, A., Wagner, R., Rashotte, C., Voeller, K., & Conway, T. (2001). Intensive remedial instruction for children with severe reading disabilities: Immediate and long-term outcomes from two instructional approaches. Journal of Learning Disabilities, 34, 33-58.
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.
Back To Top