Crunching the Numbers: One School’s Solution to Digesting Student Data
Like most public schools across the country, Leasure Elementary School in Newark, Delaware, receives a lot of student achievement data. Numerous charts and graphs show various quantitative metrics, and compare Leasure students to others in the school district. And as at many of the other schools, the educators at Leasure didn’t know what to do with the piles of data.
Ideally, facts and figures, including the results of standardized tests, should inform educational instruction so that students learn more and perform better. Few teachers know how to generate relevant data points that can be examined rigorously to lead to measurable results.
Simple questions like “What is the point of all this data?” and “How do we create new data sets that help us be better teachers?” have complex answers.
Leasure’s reams of data did not lead to big changes. “We would just look at them,” says Deirdra Aikens, Leasure’s principal for the past six years. “There was no action, nothing to do next. As the school leader, I felt like that was a problem we had to solve.”
Using Data Wise to assess student learning
Aikens, who has an Ed.D. in educational leadership, had a hunch that a deep data dive could help her staff and students, but she wasn’t sure how to go about it. Over the past five years, Aikens’s team has gone from caterpillar to butterfly status when it comes to executing a meaningful data overview. Aikens came across a pamphlet describing Data Wise, an eight-step plan based on work done in the Boston Public Schools that helps educators collaboratively use data to improve learning. She applied for a grant for herself and seven colleagues to attend a weeklong training.
Kathryn Boudett, Ph.D., director of the Data Wise Project at Harvard University and author of “Data Wise: A Step-by-Step Guide to Using Assessment Results to Improve Teaching and Learning,” coached them in three sequential areas:
- Preparation: Building the foundation for collaboration
- Inquiry: Analyzing diverse data and identifying what problems to solve
- Action: Developing and implementing an action plan
That summer, the group began the transformation into a data team. Aikens likes having a blueprint for each step of the process.
“When you see the big picture at the end, you realize you’ve done the steps throughout your career,” though not necessarily as part of a systematic approach. “It’s not a program, it’s a process, a lens.”
For her dissertation, Aikens documented the first four years of working with data at Leasure. Part of the film has been incorporated into a new, free MOOC on EdX: “Introduction to Data Wise: A Collaborative Process to Improve Teaching & Learning.”
Knowing when to take action
Motivating teachers had not been Aikens’ challenge. Instead, the teachers needed to slow down and take the time to ask deep questions such as “What do we really want to know about ourselves?” and “Who are we really?” Aikens stresses the importance of creating data sets “to tell your story, not somebody else’s.”
In the case of Leasure, where about two out of five students are from low-income families, reading skills were OK, but math scores were not too high. Rather than looking at a school-wide overview, the educators began breaking things down at grade level to identify problems in math and reading.
“That’s where the narratives pop up,” says Aikens. “Special ed, regular ed, gender, race — we aggregated as much as we could. This is the kind of data we need to see. Knowing how we compare to the school down the street is useful, but not actionable.”
With more detailed grade level evidence, new questions arose. They asked a fresh round of questions:
- “How has our achievement changed over time?”
- “What are students doing well?”
- “What opportunities do we see for growth?”
They developed graphs and looked for trends that the original school district data could not provide, and which informed new instructional moves.
They noticed, for example, that a number of fourth-grade students struggled with problem solving. Digging deeper, they could see that the children were struggling with complex addition. As they observed one another’s classrooms, they began to see that they had made some false assumptions about what the kids had already been taught. Along with data points such as grade-wide tests, they created a data overview. At their weekly meeting, they analyzed the new spreadsheets, color coded in red, yellow and green.
As a result, they took action based on fresh analysis, prioritizing fact-fluency in the classroom by finding ways to insert it into the curriculum, and continued to review their hypothesis throughout the school year, adjusting as necessary.
That first year was slow and deliberate, given over to practicing the new protocols. Now, they spiral through three or four data cycles in a school year.
Leading the team
One of the key indicators of success for this collaborative approach to data is buy-in from school leaders — usually the principal.
Aikens is reluctant to characterize any of her teachers as being resistant, though she does allow that in the first year, there were “lots of questions, ‘What makes this different than anything else we’ve done?’ What we’ve gotten really good at, is [making sure] our personal preferences are aligned with professional responsibilities.”
Aikens continues to collaborate with Boudett. This June, she assisted at the training with a fresh crop of educators who want to return to their schools to train their colleagues. Pods of teachers took notes on a classroom exchange, and practiced putting descriptive, nonjudgmental observations on Post-it notes, which were then arranged to reveal trends. It’s a deliberative process, and not one that comes intuitively to most.
Lots of rich data cannot be found on any spreadsheet, but is found instead in classroom lessons and activities. Peer observations can tease some of this out.
For those willing to work hard and take risks, the impact on student learning can be tremendous.
Nowadays, Leasure teachers are consumers of data, requesting their own charts.
“I have become more of a coach than a leader,” says Aikens.
The 8-step Data Wise process
The Data Wise Improvement process has eight discrete steps, which fall into the three broader imperatives to prepare, inquire and act:
- Organize for collaborative work
- Build assessment literacy
- Create data overview
- Dig into student data
- Examine instruction
- Develop action plan
- Plan to assess progress
- Act and assess
Rebecca L. Weber is a journalist who covers education, the arts, the environment, and more for the New York Times, CNN, USA Today, and other publications. Visit her online at www.rebeccalweber.com or on Twitter @rebeccalweber.Learn More: Click to view related resources.
- Kathryn Boudett, "Introduction to Data Wise: A Collaborative Process to Improve Teaching & Learning," edX