Stuck in the Cycle: What teachers really face when we ask them to use data in their classrooms

Educators are masters of working with scarce resources, including something most limited and valuable – time. Time is always in high demand but short supply, and teachers are constantly forced to make critical, and often tough, choices about where to focus their limited attention as well as students’ efforts. One of the promises of blended instruction is that teachers can use technology to free up this precious resource to take on high-impact activities that are often harder to execute in traditional approaches, like small group instruction, one-on-one conferencing and tutoring, and deeper learning.

Doing this well for every student means that teachers and students have to have access to data to inform how they should maximize their time together. While the advent of technology and instructional tools have helped to proliferate student data, as we alluded to in earlier posts, educators are finding that there isn’t enough time, or that data isn’t accessible in a timely manner.

So, what’s going on? When we ask teachers to “look at data,” what exactly are we asking them to do, and how can we make that process easier? By “unpacking” the task of data-driven instruction, we can better understand what it is we’re asking teachers to undertake, as well as identify specific pain points worth solving.

In a recent TLA project, we dug into this notion and talked deeply with educators to understand the steps teachers go through as part of their planning cycle to use data to inform instruction. Through this work, we were able to articulate a common series of actions teachers are undertaking to make data “work” for them as well as the common pitfalls they face. We uncovered some huge investments of time in data tasks that, frankly, have very low instructional value and often get in the way of high-value activities directly connected to instructional action. These time “sinks” are keeping our schools from realizing the promise of data-driven instruction: to better serve the needs of all students.

A Five-Step Process for Making Data “Work”

Educators are often introduced to data-driven instruction in a cyclical nature. Based on what they know about students and the instructional objectives, they develop a plan, implement that plan, and then analyze data to understand whether or not objectives were achieved and to identify a series of actions to take next. Hidden within this seemingly straightforward planning and analysis process is a pretty complex set of data actions, including:

1. Creation: educators design and implement tools and assessments, creating data around student behaviors, progress, and mastery.
2. Collection: once data are generated, educators grade assignments and extract, validate, and pool all data into accessible formats.
3. Organization: with numerous data sources, educators consolidate and format data into accessible and organized dashboards, allowing for easy manipulation and analysis. They also seek to validate through this process, looking across data from multiple sources to identify patterns of alignment and consistency.
4. Analysis: using a variety of data, educators identify trends and triangulate data to gain holistic insights into students’ strengths and growth areas.
5. Action: based on group trends and individual student needs, educators plan and adjust instruction for whole groups, small groups, and individual students, generating data to then restart the data cycle.

Each of these steps has different levels of value and difficulty. As educators move along in the process and get closer to being able to actually look at the data (ideally in partnership with students), the potential value of the effort they put in increases. Our efforts to support data-driven instruction should, therefore, focus on maximizing the time teachers have in analysis and action to understand progress and plan activities. Unfortunately, though not surprisingly, we’re currently asking most teachers to spend most of their time on lower level tasks, and important data gets “lost” along the way.

Stuck in Collection and Organization Modes

Due to the disparate, disconnected state of data systems, educators resort to becoming “human APIs,” manually collecting, extracting, combining, and formatting data sources in order to analyze and plan action. Teachers are often eager and empowered to adopt instructional and assessment tools but are then quickly confronted with the challenge of accessing and pulling student performance data across multiple platforms. In some schools, a major part of an instructional coach’s responsibilities is merely downloading, collecting, and organizing data for other educators, with the hope that taking this time-intensive lift off of classroom teachers allows those teachers to better analyze and plan action, though the better use of a coach’s time and resources would be directly supporting educators. We’re spending massive amounts of time and effort on parts of the cycle with the most limited value for students.

Interoperability Will Help Us Shift Efforts to Analysis and Action

The core of the issue isn’t a lack of the technical ability to support the organization of data; rather, technological tools producing data have proliferated without coalescing around common practices when it comes to exchanging and formatting data. Platforms are tied to their own dashboards, data formats, and ways of presenting and contextualizing data. But this holds us back from being able to harness the power of multiple data sources and personalize learning for students. We currently see educators trying to manage this data deluge manually, or with standalone platforms like Schoolzilla, Clever, or MasteryTrack. But in an ideal future state, where student data flow seamlessly among learning and practice systems, the ecosystem needs to align on data exchange practices and push for data interoperability.

Absent a perfect solution – when data is interoperable and seamlessly and securely transferred between multiple platforms – it’s tempting to push for simple solutions to this complex problem. Perhaps we should just reduce our expectations for teachers’ data roles, or limit the potential sources of data (e.g., instructional tools and assessments) available for analysis.

But rather than simplifying, and thereby limiting our opportunities to help educators and students more deeply understand the learning happening in their classrooms, it’s better to start with asking questions of our data-fatigued teachers. Once leaders and teacher supporters (coaches, technologists, professional service providers, etc.) truly understand where teachers are getting “hung up” in the data process, they can take specific, practical actions to add capacity and problem solve. By having these conversations, it’s much more likely we’ll be able to realize the potential of student data and personalized learning.

In the next few posts in this series, we’ll turn to some of the capacity-building and problem-solving actions we identified in our research. In the meantime, let us know what you think: where in the data process do you or your education peers struggle most? What are you doing (or wish you could do) about it?