At the Speed of Teaching: Data’s New Frontier

Big data has a place, but it's medium and small data that will inform classroom instruction the most

In the classic book Animal Farm, George Orwell described a society where all animals were equal, but some animals were “more equal” than others.

The same may be said of data — some data are “more equal” than other data. What are the most equal data? What should be the most equal data? It all depends on what we value.

When data are mentioned, many outside of education only think of what is now being called “big data.” These are data from state and national assessments, the kind we have received for years from the national testing companies and the statewide standardized tests under No Child Left Behind. Big data are the data that end up making headlines in the newspaper.

When standardized testing data rose to prominence as accountability measures in the 1990s, many data coaches like me gained some renown for conducting school improvement training using big state-assessment data sets, based on the premise that, if we could only parse big data sufficiently, we could find an instructional use for it.

If the past is any indication, it is unlikely that big data from Smarter Balanced, PARCC or new state tests will be any more helpful for teachers than data from the NCLB state tests have been.

What we forgot is that big data were never designed to be instructionally helpful in the short term. Big data answer accountability questions at the school, district, state, national and international levels and provide only a small amount of general guidance for school or districtwide curriculum modifications. If accountability is your only goal, big data will remain the most equal.

Medium Data

What schools get from common assessments administered to all the students in a grade or course is “medium data.” These data are the most equal to help teams of teachers collaboratively plan instructional units and lessons.

Because of scarce collaborative time, teams must be strategic about the medium data being analyzed. This kind of special treatment should be saved for learning objectives that are seminal to their content area, absolutely needed by all students for a complete understanding of the next unit of study and the most difficult for students to grasp.

The Howard County, Md., Public Schools offers a case study in the effective use of medium data. Superintendent Renee A. Foose is increasing the districtwide consistency of language about instruction and assessment through the use of a data protocol. “A well-developed, easily understood and user-friendly protocol is critical to making informed decisions about student progress in every school,” Foose said during a recent training session. “The data protocol supports our entire system in effectively leveraging data for the achievement of all.”

For example, while principal Kelley Hough and the leadership team at Howard County’s Bryant Woods Elementary School understand that PARCC is the state’s accountability measure, they don’t waste time fixating on the previous year’s results. Instead, teachers at each grade level use a protocol every few weeks known as the Classroom-Focused Improvement Process to analyze common classroom assessment results to identify students’ strengths and to craft consistent instructional strategies addressing essential weak areas. The key questions embedded in the protocol are: “What do the students know? What do the students not know? What are we (the staff) going to do about it?” The Measures of Academic Progress assessment is administered three times a year to document progress in real time.

The Durham, N.C., Public Schools also has committed to using medium data more effectively. The district’s professional learning community leaders have been trained to chair ongoing data dialogues using a consistent protocol.

“We heard over and over from teams that using a protocol to collaborate on medium data is eye-opening because it provides teachers with a nonthreatening way to talk about achievement, thereby de-privatizing the data and leading to a deeper understanding of all aspects of student learning,” says Jill Hall, Durham’s executive director of leadership and professional development.

In-the-Moment Data

This category of data includes the hundreds of cues that alert teachers discern every day from their students in the course of instruction. A classroom of blank stares, 15 hands flinging into the air with questions and massive confusion when students are asked to work on a problem independently all should inform a teacher that it is time to re-teach or to consider on the fly an alternate instructional approach. It is impossible to predict when these moments will occur.

Small Data

Somewhere between the unpredictable in-the-moment data and the medium data that require collegial collaboration lies data’s next frontier: the most-equal data with the greatest potential to make a significant difference in student learning.

Small data are gathered intentionally at critical, predetermined “hinge points” of lessons when teachers must decide immediately, “To what extent do the students in this class really understand this concept or can demonstrate this skill proficiently? Do I keep moving ahead with this lesson, or do I pivot to Plan B?”

Two areas must be addressed immediately if small data are to be used effectively. First, more advanced digital tools should be developed that can capture and summarize much more precise “data at the speed of teaching.” A few potentially useful digital products have come onto the market recently, ranging from entire software systems to standalone apps. Some are content agnostic and rely on teachers to enter standards and questions. Others come with preloaded standards, assessments and links to remediation and enrichment activities tied to instruction.

To be most effective, these products must be able to report data for each student on key standards, be aligned with other components of the school’s learning management system, operate on smart phones or hand-held tablets and, notably, be simple enough for teachers to navigate in classrooms as teaching is progressing.

Effective Integration

More importantly, professional learning for teachers should center on how teachers can integrate new digital technology into the fast pace of teaching and use the results to make immediate instructional decisions. This training should include identifying the most effective “hinge points” of lessons, writing or accessing short performance tasks that quickly and accurately differentiate proficient from nonproficient students on the focus skill, physically managing the technology amidst everything else going on in classrooms and developing Plan B instructional strategies to use if needed.

Baltimore County, Md., is a good example of a district moving forward intensively in this type of high-level professional learning. As a part of its Students and Teachers Accessing Tomorrow program, or STAT, the district has assigned in-house professional developers in every school building. “Their role is to … help teachers incorporate technology into their digital lessons,” says Superintendent S. Dallas Dance. “Using technology to gather data helps our teachers [identify] individual students’ strengths and areas for growth. This provides information to truly customize learning pathways, so all students are being challenged at the appropriate levels.”

Work To Do

The good news is that 93 percent of the 4,600 responders nationwide to the 2015 Gates Foundation study “Making Data Work for Teachers and Students” indicated they regularly use some type of digital tool to guide instruction. On the other hand, about two-thirds said they were not satisfied with the effectiveness of the available tools for working with data.

Through more advanced digital tools and high-quality professional learning, we can empower teachers to use standards-driven data collected at the speed of teaching in classrooms every day to increase student learning significantly. This is placing the most equal data right where they need to be — in the hands of teachers — and empowering educators to use these data effectively.