Students learn or progress at their own paces. Each needs different amounts of support, at different points in a school career and at different times of the school year. Some need very little help to stay on track, while others are facing serious challenges in learning, in their behavior, or at home, and need significant interventions. It would not only be a waste of resources to give students who are on track the same intensive services as students who are struggling, it can actually be a counterproductive waste of on-track students’ time. So how can schools make sure that students get the help they need — and only the help they need?

Many schools are turning to tiered systems of support, usually implemented in three levels:

• Tier 1 represents the strong foundation that all students need, the whole-school structures that undergird a good education.

• Tier 2 interventions serve students who need moderate amounts of additional help, probably for a limited time.

• Tier 3, comprising the most intensive services, is designed for students experiencing the greatest difficulty.

Most tutoring is considered a Tier 2 service, for example, while more intensive case management, discipline and behavior management, and mental health services fall into Tier 3. These services can take place during the school day or after school. Part of the idea is that Tier 3 remediation is costly, and by investing in prevention — Tier 1 and to a certain extent Tier 2 services — schools can limit the number of students who need it.

The approach is gaining ground for a number of reasons:

• It has become more feasible. Schools have more frequent access to data about their students’ performance, and as a result they can establish quantitative rules to define which students are eligible for more intensive services. At the same time, they can adopt early warning systems to identify students who are just beginning to struggle. In general, districts need at least quarterly assessments of students to support a meaningful tiered support structure.

• Tiered support falls in line with other current efforts to integrate academic and behavioral interventions and to coordinate the services students receive.

• Federal, state, and district accountability policies have created strong incentives for schools to use data to guide their decisions. At the same time, many schools have faced budget cuts that have left them with smaller staffs and fewer resources for students. They need to use the resources they do have efficiently.

MDRC’s Studies

MDRC has evaluated and is currently evaluating several tiered approaches at the primary and secondary school levels. Most include some types of services or support for the whole school, and then two or three additional tiers of support for students who need additional help with behavior or academics.

Lessons from MDRC Evaluations

1. The tiered approach can be put into action on a large scale.

Schools and school districts are widely implementing tiered structures and using clear rules to identify students who need more intensive help. For example, the evaluation of Response to Intervention chose a random sample of elementary schools in 13 states, and found that in the 2011-2012 school year, more than half of them reported full implementation of the Response to Intervention framework. In a different sample of schools known to be attempting to implement Response to Intervention, 86 percent reported full implementation, and schools in that second sample did indeed adjust their reading services to provide extra support to students reading below grade-level standards.

Similarly, the MDRC evaluation of Communities In Schools focused on 28 of its schools; the study found that Communities In Schools formed an important part of the schools’ support environments, offering a variety of services to all students and providing case management to approximately 10 percent of students on average. The evaluation of Diplomas Now has found that the 31 schools randomly assigned to implement the model did manage to implement a majority of its 111 components. The most consistent implementation came in areas directly related to tiered support: using data to identify at-risk students and collaborating to plan and provide targeted interventions for those students.

2. Without clear guidelines, it is difficult to describe fidelity or hold schools to a standard.

There is evidence supporting the idea of tiered support, but little practical guidance to schools about how to coordinate the many decisions required to put that system in place — decisions involving scheduling, staffing, funding, and the use of data to identify at-risk students, among others. In its evaluation of Communities In Schools case management, for example, MDRC found variation among districts in how its case management process was executed. (In response to these findings, CIS has been strengthening its guidance regarding standards for implementation.) Diplomas Now does have a detailed model, but in part because the model was so new during MDRC’s evaluation, its guiding nonprofit organizations had not yet developed guidelines about what levels of implementation were acceptable, nor about what parts of the model were most important.

Yet in both of these cases nonprofit institutions created the models and can offer technical assistance to schools, to help them with implementation. Some models of tiered support — like Response to Intervention — are really only frameworks or sets of principles. They have no clear fidelity checklist, and no institution provides technical assistance in implementing them. MDRC’s evaluation of Response to Intervention found meaningful variation in how schools provided reading intervention services. Some schools provided them to students at all reading levels rather than only students reading below grade level, and many provided reading intervention during core reading instruction rather than in addition to it.

3. The nature of the impact findings depends in part on the experimental design.

It can be a source of confusion that some studies of tiered support models have found positive effects while other studies of models that seem similar have found no effects or even negative effects. It is important to keep in mind, however, that different evaluations may be making different comparisons. Some compare whole schools that implement a model with whole schools that do not (for example, the Diplomas Now evaluation and the Content Literacy Continuum evaluation), while others compare groups of students who receive specific services with other groups of students who do not (for example, the Response to Intervention and Reading Partners evaluations). In all cases, the “control group” still receives services, continuing with its previous practices or perhaps trying some other model. To figure out whether it is worthwhile to implement an intervention, therefore, it becomes very important to determine what difference the new model makes to the services students receive. How much does a new, tiered model actually intensify the services received by the students who are struggling the most? That question often turns out to be somewhat difficult to answer in practice, although in the Diplomas Now study MDRC was able to investigate how model implementation affected student services.

4. Using data to guide decisions does not necessarily get students better services or services more appropriate to their needs.

It is not enough for schools to have data systems in place to monitor students’ progress and identify those who need more help. They also need quality-control processes to make sure that the help given those students is appropriate to their needs and aligns with the general course of instruction all students are supposed to be receiving. The schools involved in the evaluations of Success for All and Response to Intervention all had school-wide data systems in place and conducted screening of students’ needs. Yet in evaluating Response to Intervention, which had no set model, MDRC found it very difficult to determine whether students were receiving instruction aligned with the overall curriculum. In Success for All, teachers actually reported that the curriculum they were given was not a good match for students’ needs. Finally, the Enhanced Reading Opportunities evaluation found positive effects on students’ reading comprehension and performance in core subject areas, but those effects were not sustained after the intervention concluded. If an intervention has to be continued to produce lasting effects, then it is not delivering on tiered support models’ promise of efficiency: short-term, intensive interventions are supposed to reduce the need for sustained services. Some schools and programs may consider phasing out intervention services gradually in order to create sustained effects.

For more information on tiered systems of support, e-mail Rekha Balu.