Many educators continue to struggle with data literacy, in part because it’s not always clear what “data literacy” entails.
One of the primary reasons data-driven decision-making remains foreign to so many educators is that the educational community has struggled to come to a consensus on what “data” really means. Teachers and administrators often conflate data and “assessment information,” when in reality the latter is but one type of the former.
Organizations like the Data Quality Campaign (DQC) have attempted to clarify this distinction by issuing straightforward definitions of things like “data literacy.” According to the DQC, “Data-literate educators continuously, effectively, and ethically access, interpret, act on, and communicate multiple types of data from state, local, classroom, and other sources to improve outcomes for students.”
Formal definitions like these represent a step in the right direction, but they are generally intended for consumption by policymakers and even the general public, meaning they lack detail about the specific skills and knowledge required for classroom applications of educational data.
Researchers Ellen B. Mandinach and Edith S. Gummer highlight these shortcomings in a paper published in Teaching and Teacher Education, writing, “To fill that gap…we have laid out a definition and identified the skills, knowledge, and dispositions that comprise the construct, ‘data literacy for teachers.’” According to the paper:
Data literacy for teaching is the ability to transform information into actionable instructional knowledge and practices by collecting, analyzing, and interpreting all types of data (assessment, school climate, behavioral, snapshot, longitudinal, moment-to-moment, etc.) to help determine instructional steps. It combines an understanding of data with standards, disciplinary knowledge, and an understanding of how children learn.
Mandinach and Gummer argue that true data literacy is grounded in seven areas of knowledge, including content knowledge, general pedagogical knowledge, curriculum knowledge, and four others. Each of these areas falls under five domains of data usage: Identify Problems and Frame Questions, Use Data, Transform Data into Information, Transform Information into a Decision, and Evaluate Outcomes.
The knowledge and skills that fall under the “Identify Problems and Frame Questions” domain constitute the first phase of what Mandinach and Gummer call the “iterative inquiry process,” or, more simply, continuous data-driven decision-making.
Teachers should be able to identify, articulate, and explain a “problem of practice” regarding a specific student, group of students, or instructional vector. They must be able to contextualize the problem in multiple ways, grappling with it at both the student level and the broader school level. Once the problem is properly contextualized, teachers must solicit input from other stakeholders in the education process, including (where appropriate) administrators, parents, and even students.
The next domain, “Use Data,” is the most sprawling, and Mandinach and Gummer readily admit that their choice of terminology is insufficiently descriptive. Part of their difficulty stems from the fact that they identify a remarkable 27 skills that teachers must develop to effectively put data to use in the classroom. These skills include everything from “understanding the purposes of different data sources” and “using formative and summative assessments” to “using technologies to support data use” and “understanding elements of data accuracy, appropriateness, and completeness.”
The third and fourth domains — “Transform Data Into Information” and “Transform Information into a Decision” — detail the ways in which teachers can leverage their classroom applications of data into actionable insights. Transforming data into information involves eleven skills, including generating hypothetical connections between data and instruction, testing assumptions early in the inquiry cycle, and probing datasets for causality. Transforming information into a decision involves just five: determining next instructional steps, monitoring student performance, diagnosing what students need, making instructional adjustments, and understanding the context for the decision.
Perhaps most importantly, teachers must be able to examine — and understand — the impact that their data-driven decisions are having by reexamining the original problem, comparing student performance from before and after the iterative inquiry process, and monitoring changes in classroom practices. As Mandinach and Gummer point out, this is an iterative process, which “means that decisions do not have an endpoint. Data [is] collected, analyzed, interpreted, and acted upon. Then the cycles begin anew.”
The paper ultimately concludes that all of these skills and knowledge sets need to be grounded in “dispositions or habits of mind.” Mandinach and Gummer chose not to include dispositions in the already complex architecture of their domain framework — largely because they apply to all facets of teaching, not just data usage — but they go to great lengths to emphasize that they are just as important.
To foster these dispositions — which include “a belief that all students can learn,” and both vertical and horizontal collaboration — school administrators must provide teachers with the time and space to explore them (and their implications) at length. More often than not, this begins in the professional development (PD) space.
At Hoonuit, we offer anytime, anywhere access to a library of over 1,500 innovative PD courses, many of which directly relate to at least one of Mandinach and Gummer’s imperatives. Our goal is to build a nationwide community of learners who are eager to connect with their peers, share their insights, and give and receive feedback on their PD journeys.
The path to data literacy isn’t always easy or straightforward, but in today’s educational landscape, it’s an essential part of ensuring that every student gets the chance to realize their full potential.