Unlocking the Potential of the Data You Have

Winter 2018

By Steve Wilkins

I approached the storage container with trepidation. It felt almost as if I was opening a closet in an old house’s deepest recesses—a closet full of cobwebs. Grime. Skeletons. Artifacts of times gone by. 

In fact, though, I was daring to open the file cabinet that stored our school’s student records. As I fearfully opened the cabinet, a puff of acrid air and bits of ancient detritus billowed out. But so did the opportunity to transform the way our school met students’ most pressing needs. 

In 2010, we at Carroll School (MA) came to the realization that we were sitting on a treasure trove of student data. Like many other independent schools, we had file cabinets—real and virtual—chock-full of information about our individual students. They contained data from previous schools, the admission process, outside testing, standardized tests of academic achievement, sets of grades and comments, and admission test scores for high school and college. 

Also like many other schools, we had never really maximized student records by analyzing the data they contained to guide our educational programming. Why not? The short answer is that most of us feel our curriculum, narrative reports, and grading systems provide sufficient data on student progress. But many of us are also scared, largely because we don’t know how to put what we might find in our student records to good use. When students struggle, when they fail to perform to standard, there is so much more that our schools can and should do to help. Yet, most teachers haven’t been educated to use data to improve student outcomes. 

There is an important story in data. But it takes institutional commitment to make the student data story understandable, usable, even lovable. 

In the past seven years since we came to that realization, Carroll School has made that institutional commitment. We now train our educators to harness data before deciding how and what to teach each student. It’s true that our mission is unique; Carroll is a school for children in grades one through nine who have dyslexia. But what we’ve learned from our experience with data can be useful for independent schools of all types and levels. 

Moving to a Data-Inspired Culture

Because our students are bright children with dyslexia, we at Carroll have always known that our job is to close the gap between their current performance and that of their peers who learn typically. Before coming to us, students generally make less than a year’s worth of progress in a given school year; they must make more than a year’s worth of academic progress or may never “catch up.” To help our students succeed, we have to understand and meet each one’s greatest need. 

The problem with doing this has never been a lack of data. If anything, children who struggle in school tend to get tested too much. Amid the babble and bluster of intelligence testing, academic proficiency assessments, formative assessments, progress monitoring, and constant dipsticking to check for understanding, our children come with barrels full of data. The problem is that typically these data sit in files, and even conscientious teachers read the files only when problems arise—if ever. 

At Carroll, our transformation began when we saw that we needed to change this dynamic. We needed to make data integral to our educational recipe. That meant taking the following crucial steps.

Creating a Database and Student Data Reports

Our mantra—our driving force—has become “Give each child what she or he most needs.” But to deliver on this lofty goal, we needed to understand deeply what exactly it is that a child most needs. And that meant having a means of making the child’s data useful for directing his or her educational program. This led us to create an interactive student information database that assigns cognitive development activities based on a student’s data profile. 

Sometimes the needs are obvious. When a child’s oral reading fluency score is more than a standard deviation below average, the response is to give the child more time with a teacher or tutor who can deliver high-quality reading instruction. But at other times, the needs are embedded and subtle. Why can’t this thoughtful child reveal the depth of her complex thinking in writing? Why doesn’t this great thinker know his multiplication facts? We shifted from wanting to understand what a student struggles with to understanding why a student struggles. This is a shift that heralds the arrival of more effective education. When education can identify and address the underlying causes of a child’s academic struggles, schools become lifesavers. 

Access to student information is complicated. We have a legal responsibility to protect student’s cognitive and academic data, just as we do their health records. Nevertheless, thanks to the permission guardians generally grant as part of the enrollment process, teachers have the right to access student data. This right is legal but also moral. We have an obligation to understand children’s academic data so we can determine which educational interventions are most likely to help them. It’s similar to doctors’ needs to have access to patients’ health records.

To significantly ease teachers’ access to student information, in 2016, Carroll School worked with a data science company, a programming firm, and in-house IT staff to create a relational database that brings together, into one student data report, the majority of each student’s cognitive and academic data. (Click here to see a data report.) The goal of the report is to give our educators access to information on student performance in a cloud-based, transportable, scalable, and affordable manner. This type of database isn’t all that complicated, but it addresses a sticky problem in independent school education: It helps us harness piles of student information in an accessible, user-friendly format. The resulting report depicts the relationship between cognitive skill growth and academic achievement. It also captures a child’s data over multiple years to provide a longitudinal perspective on a child’s trajectory. 

Through the student data report, both our families (who have access to their child’s report) and our school can understand more clearly the impact of a teacher’s work on a child’s outcomes. Again, the story is told in the data. To be sure, a child’s essence is more than the compilation of statistics. The report also captures a child’s “readiness to learn” (rated by the teacher) and a child’s self-reporting on social-emotional factors.

Conducting Relevant Research

If schools are to understand the impact of education on each child, there needs to be an active feedback loop to help assess the extent to which what we do in our classrooms is building better brains for learning. To make data as useful as possible for designing each child’s educational program at Carroll, we developed a research and intervention team. Led by Eric Falke, a full-time neuroscientist physician on our faculty, the team is made up of five other cognitive specialists who educate teachers and tutors to develop expertise in how to utilize data to inform instruction. The team is responsible for looking at each child’s profile and assigning specific remedial activities accordingly. A major part of this involves studying the impact of cognitive development interventions on the academic performance of our students. In effect, we are examining student data to determine the source of each child’s learning roadblocks. Once those obstacles are identified, teachers work to diminish them. 

As the cognitive work is underway, students are also learning fundamental academic skills in reading and math, among all their other content and arts curriculum. Our research question is: Does a decrease in cognitive weaknesses (e.g., processing speed, reaction time, working memory, cognitive flexibility, shifting attention) lead to an increase in academic skill levels and performance? It would be impossible to answer this question without careful data analysis and controlled scientific research. See the box on page 87 to learn more about how we’ve partnered with nearby university-based experts to conduct this high-level research using our student data.

Providing Teacher Training

It is reasonable to assume that teachers in all schools seek to help children in their care. Too often, though, a certain percentage of children have learning challenges beyond teachers’ level of training and expertise. That’s why a knowledgeable and skilled faculty is essential if a school is to make use of student data to design educational programs. At Carroll, we provide mandatory training in two areas: 
  • Harnessing data to guide decisions about how to help each student. All teachers and tutors take courses in how to read psychoeducational evaluations, interpret academic testing, understand the jargon and statistics of educational assessments, relate IQ and cognitive data to academic performance, and monitor the progress of each student. 
  • Using data to design lesson plans. The guiding principles behind the lesson plans are to give each child what she or he most needs and to move the student forward as quickly as possible and as slowly as necessary. Teachers learn to examine and communicate data associated with academic skills development, cognitive growth, social-emotional well-being, student work products, and features of a student’s self-concept. 
As a result of our romance with data, we have developed coursework that is required of all faculty during their first five years at Carroll. The data team and experienced faculty teach these courses. A partnership with Lesley University’s graduate program in education enables Carroll to provide master-level credits for this work.

Adjusting the Daily Academic Schedule

Understanding the power of student data led us to innovate with our daily schedule so we could create a more targeted, personalized education for each student. The new schedule enabled several major changes. 

1. We created focus blocks that meet all year long and as often as traditional classes do. Every child is given the opportunity to work directly on his or her area of greatest need. In a population of children with dyslexia, many already received reading tutorial or small group work. However, the data of many other children also revealed a more specific set of focal weaknesses in areas such as comprehension, fluency, math, executive function, and cognitive flexibility. We designed our focus block program to address each child’s greatest need. 

Here’s an example of how this works. For our 225 middle school students, focus block incorporates the following curricula: structured phonics for 33 percent of the students; reading fluency work for 17 percent; sentence, paragraph, and essay development for 17 percent; additional math fluency for 16 percent; comprehension and vocabulary work for 8 percent; and expanding oral and written expression for 9 percent. 

2. We also created a flex block in which we regroup students around common areas of strength. These include makerspace activities, focused clubs, additional arts, or service learning. The flex blocks run 40 minutes each day.

3. We heightened our faculty’s awareness that our job was to remove students’ learning roadblocks in all subject areas. We all are more aware of what each child most needs. As a result, Carroll is now a more focused institution. 

But perhaps the most important outcome is this: Our newly dynamic daily schedules— informed by student data—enable us to provide targeted cognitive interventions designed to remove obstacles to effective and efficient learning in each child. Although it is relatively easy to identify what a child struggles to do academically, it is enormously complex to understand why a child is experiencing such level of struggle. Collecting data on each student’s cognitive profile helped Carroll address each student’s roadblocks. Data enabled us to understand why children struggle and to enact interventions reliant on the neuroplasticity within each student.

Data’s Part in Your Value Proposition

Maybe you, too, feel trepidation at the prospect of exploring the deepest recesses of your student data. But Carroll’s experience shows that data doesn’t have to be scary. In fact, data can be liberating. 

Dare to open up the trove that contains your school’s student information. When you use data to provide each student with what she or he most needs, your students, teachers, parents, and administrators can more crisply understand the impact of their educational efforts. 

What’s more, you’ll be able to convey the value proposition of an independent school education more vividly. One of the strongest reasons to invest in an independent school education is so children can thrive. Data can help you make the case that your students can make more progress in a wide variety of areas because your independent school is able to understand their learning profiles and design their instruction accordingly.


Making MIT Experts Our Research Partners

Conducting research in schools is a complicated proposition because it would be unethical to deny any child our best efforts. But by working in concert with researchers at the nearby Massachusetts Institute of Technology, Carroll School has been able to find a balance between solid educational offerings and the scientific rigor necessary to assess the impact of our educational program for each child. It’s another way in which we’ve managed to put our student data to work, this time to help children of today as well as in the future. 

The research partnership began in 2014 as a result of our mutual dedication to scientific study of how the brain learns to read and a desire for quality research in schools that teach children with dyslexia. Eric Falke, Carroll’s staff neuroscientist, meets weekly with John Gabrieli, professor of brain and cognitive sciences, head of MIT’s Integrated Learning Initiative, and director of the Martinos Imaging Center at MIT, and Johanna Christodoulou, assistant professor at Massachusetts General Hospital Institute of Health Professions. This partnership has proven essential in helping Carroll determine whether a child is making meaningful progress and has focused Carroll School’s work on three areas:
  •  Developing cognitive skills in school-aged children
  •  Understanding how neuroscience can help improve educational outcomes
  •  Adding an external validation process to check the veracity of the data that we collect at Carroll
That last point is important because it’s one thing if we at Carroll School claim that our programs are substantially improving student outcomes. If MIT’s researchers confirm that Carroll students are making dramatic improvements, that is a far more believable and useful conclusion.
Steve Wilkins

Steve Wilkins is head of school at Carroll School with locations in Waltham (lower school), Lincoln (middle school), and Wayland (upper school), Massachusetts.