Through a statewide initiative, Michigan has launched multitiered systems of supports as a framework to improve academics and behavior in more than half the state鈥檚 900 elementary and secondary schools. But educators and researchers have found that initial buy-in, financial incentives, and even early success don鈥檛 guarantee schools will sustain the model for the long haul.
Multitiered systems of supports, or MTSS, are intended to address all students鈥� academic progress and behavior in schools by providing research-backed instruction for all students in academics and social-emotional development. 69传媒 who do not progress based only on this core instruction, known as Tier 1, are assessed regularly and provided increasingly intense interventions, called Tiers 2 and 3. The framework is intended to combine elements of similar models, like response to intervention for academics or positive behavior interventions and supports for behavior.
The model 鈥渋s a recipe, it鈥檚 not a McDonald鈥檚 鈥榲alue menu鈥� where you go, 鈥業 like this part and this part and this part,鈥欌�� said Steve Netzel, the executive director of curriculum and staff development for the Holt public schools, a 5,600-student district south of Lansing. 鈥淚t takes a while to understand it鈥檚 a system and it all interacts with each other. You can鈥檛 pick and choose.鈥�
As similar models are tried out in districts nationwide, Michigan鈥檚 experience exemplifies the challenges of implementing systems of supports for general education students across grades, subjects, and schools.
More than 500 elementary and middle schools adopted the tiered-support-systems model between 2004, when the program started, and 2013. 69传媒 saw the percentage of students meeting state reading benchmarks rise from 52 percent to more than 60 percent in the first three years, and behavior problems also fell. But Steve Goodman, the project director of Michigan鈥檚 multitiered-systems initiative, found that after that initial strong beginning, implementation dropped off rapidly after the third year, when start-up funding from the federal economic stimulus dried up. Fewer than one-third of schools met state implementation benchmarks, according to a study in 2014. More than 4 in 5 schools simply stopped reporting student data by the fourth year of the program. That was a disappointment, Goodman said, since the schools that did fully implement the framework saw much bigger gains, particularly in student behavior.
鈥淲hat we learned is we could get good outcomes from behavior and reading鈥攚hen staff implemented it with fidelity,鈥� Goodman said.
Study of Implementation
Michigan鈥檚 results mirror other studies also finding mixed results for schools trying to use the framework.
鈥淭here鈥檚 this idea that a tiered framework can be applied in many different domains and that those can complement each other,鈥� said Rekha Balu, who studies response-to-intervention models for the research group MDRC. But, she added, 鈥渢hinking of next-generation support systems, I see huge coordination challenges ahead.鈥�
For example, this September, the Tennessee education department found fewer than 1 in 4 schools involved in the state鈥檚 mutltitiered-systems initiative actually had most teachers implementing all its core pieces.
Source: Michigan鈥檚 Integrated Behavior and Learning Support Initiative
Moreover, even among the so-called 鈥渉igh implementer鈥� schools鈥攖hose that showed the most improvement for students鈥攊mplemented the framework differently from those with smaller student gains. The schools that improved the most had strong leaders and specialized staff who prioritized the model. The schools also used several different data sources to make decisions about students, rather than relying on one screening test. Perhaps most importantly, the Tennessee study found, the schools arrange staggered schedules, so that teacher planning and student interventions don鈥檛 interfere with core instruction.
鈥淚n some of our small schools, there鈥檚 an assumption that communication will just happen,鈥� said Lynnette Borree, the director of response to intervention in the Copper Country Intermediate district, which serves three rural districts in the state鈥檚 Upper Peninsula. 鈥淏ut you need to be intentional about it. If you just assume it will happen without giving and dedicating time for those buildings to communicate, it just doesn鈥檛 happen.鈥�
Superintendent David G. Hornak said the Holt district 鈥渉ad some false starts.鈥� He was a building principal when Holt adopted the multitiered-supports system, and saw that implementation varied significantly from school to school, and changed when new principals came in. 鈥淎t a building level, you get a little possessive of your space and time and resources. You try to pivot and protect what you are already doing,鈥� Hornak said.
Whole-District Supports
The results in Michigan persuaded Goodman and other state-level experts to broaden the initiative鈥檚 implementation targets from schools to districts.
鈥淲e could only get so far with a school-based model鈥� of the MTSS, Goodman said. 鈥淲orking with schools, we [at the state level] provided training, coaching, evaluation, expertise. That is not scalable; you can only do it in small projects with outside resources coming in. To sustain it, you have to develop local capacity.
鈥淎t the district level, you have people with the authority and resources to say this is a priority,鈥� he added.
The Ingham Intermediate district was one of the districts that stepped up to help. A regional service agency representing 12 local school districts (including Holt) and 10 charter academies in and around Lansing, Ingham had been one of the original areas to pilot the framework. The district was quick to expand it using more than $11 million from the 2009 federal economic-stimulus money.
鈥淚 felt like all of us went back and got another master鈥檚 degree; it was a whole year of learning nonstop,鈥� said Roberta Perconti, Ingham鈥檚 director of student instructional services. The regional district has worked to coordinate training, planning, and budgets to help local schools stay on track with the model.
鈥淲e鈥檝e had districts who really embraced it, and then leadership changed, and they stopped,鈥� Perconti said. 鈥淎nybody who says it isn鈥檛 a challenge probably isn鈥檛 doing it right.鈥�
The federal money helped build some initial enthusiasm, but real understanding of the model took longer, Perconti said. Some school and district leaders thought of tiers as primarily a special education system, rather than for all students. Many schools did not implement the behavioral system at all at first, because they didn鈥檛 consider behavior something to teach.
鈥淚t all got rushed through at the end of the school year,鈥� said Netzel, Holt鈥檚 curriculum director, who was a special education teacher when his district started to implement the multitiered system. 鈥淣obody would say no to it because it sounded good, but that鈥檚 not the same as being committed to it. 鈥�
To keep schools on track, Ingham developed a detailed description of what implementation would look like鈥攁nd cost鈥攁t different phases. The district began to meet regularly with superintendents, curriculum and other specialists, and business officers at each district.
鈥淧eople get immobilized with there being so many things to work on and not knowing what to focus on. We needed to connect our implementation checklist to actual work,鈥� Perconti said.
鈥淲e asked a lot more questions [like], 鈥楧o you have evidence of implementing this? Can you show you鈥檝e allocated time and staff? Do all teachers have access to your data?鈥欌�� she said. 鈥淚n our districts, the offices didn鈥檛 always talk to each other. It was fascinating to see a superintendent hearing his staff report and saying, 鈥榃hat do you mean we鈥檙e not providing time for [professional development]?鈥� 鈥�
Hornak, Holt鈥檚 superintendent, said he worked with Netzel to build up training for all school leaders and staff. 鈥淲hen you are building systems ... It has to be slow and methodical so everyone understands the 鈥榳hy鈥� of what鈥檚 happening,鈥� Hornak said. 鈥淲ith strong systems in place, leaders can change positions without it all falling apart.鈥�
Structuring Time
In the neighboring 3,200-student Mason, another of Ingham鈥檚 districts, Lisa Francisco, the principal of the Alaiedon Elementary School, said her team also struggled to make the model match reality. On paper, the response-to-intervention 鈥減yramid鈥� describes strong, research-backed core instruction that is effective for about 80 percent of students, with increasingly intense interventions for the rest. But in the 475-student Alaiedon school, 鈥淲e had no real reading program, and our [RTI] 鈥榯riangles鈥� were really out of whack, almost upside-down,鈥� Francisco said. 鈥淲e were intervening with too many kids.鈥�
The district decided to bring in new reading, and later, math curricula to their schools, with training and data support from Ingham. Like the Tennessee researchers, they also found the need to provide more time and staff support to make sure interventions for students didn鈥檛 interfere with core instructional time. And Ronald Drzewicki, Mason鈥檚 superintendent, began regularly gathering school staff to compare data both within campuses and across the district.
鈥淪trategic planning is not just Lisa [Francisco] talking about the plan to her teachers; it鈥檚 educating a lot of staff members about district goals,鈥� Drzewicki said. 鈥淚t ... gives Lisa the ability to say, everyone in the district is doing this.鈥�
At Alaiedon Elementary, Francisco and her teachers agreed to shuffle all the students in each grade who needed extra practice in certain skills for a 30-minute literacy-intervention period each day, grouping students according to the specific reading skills that needed extra practice to avoid having to pull some students out of regular classes. As a district, Mason was also able to hire some retired teachers part time, to spread out students into smaller groups during intervention periods.
It seems to be working: In 2009, Holt and Mason were each referring 21 percent of their students for special education. Today, the special education rates have fallen to 9 percent at Mason and 11 percent at Holt, both below the national average of 12 percent.
鈥淚t鈥檚 still hard to be compared to other people, but we are seeing gains,鈥� Francisco said. The school has improved enough to leave its status as a 鈥渇ocus鈥� school since adopting multitiered supports. 鈥淚t felt for a while that we are learning this, and this, and this, and this ... but now teachers are all starting to see it come together and make a difference for all students.鈥�
