Perspective: The History and Future of Data Fluency

The Concord Consortium marks its 25th anniversary this year. That’s an admirable length by any measure. In technology innovation years, it’s practically eons. At our founding, in August 1994, the World Wide Web contained only a few thousand sites and the concept of ubiquitous powerful computers was barely more than a thought experiment. Although change is the only true constant in an organization founded on innovation, one of the most consistent themes running through our quarter century history is helping learners become fluent with data.

Decades before our founding, data was already beginning to play a central role in our work and vision. In the late 1970s, Bob Tinker adapted an early computer to collect and graph experimental data produced by a tiny temperature sensor. By demonstrating that physics students could do in minutes what had once taken them hours, and could see their results in real time as well, Bob inspired teachers across the country and launched a revolution in thinking about the role of data in STEM learning.

Today, the data revolution is fully upon us, engulfing industry and society alike. However, education is a very different story—the K-12 learner experience remains practically devoid of data. This poses a grand challenge for us as a society. Today’s fourth grader is tomorrow’s data scientist, worker, and voter. What are we doing to ensure that she can succeed and thrive? We’re committed to answering that question, and aim to bring data to the fore across topics, grades, and settings.

Setting the stage

While it’s been brewing for a while, we’re now indisputably entering the Data Age. Data has, of course, been all around us for decades (even centuries), but the degree to which it defines our current world is wholly unprecedented. In fact, we’ve become so accustomed to the ubiquitous nature of data in our lives, it’s easy to forget that the term “data science” was coined less than two decades ago and the practice itself has still barely been formalized. Standing on the brink of such changes demands recalibration across all sectors. Rethinking education is arguably the most urgent need of all.

There is precedent for such a tectonic shift, however. We stood at a similar inflection point with the rise of the personal computer. Over mere decades, computers moved from far-off novelty to fundamental business tool to everyday home appliance. As they did, it became clear that they would transform all aspects of modern life. This new age demanded new thinking, and a 1999 National Research Council report advocated for new paradigms: “Fluency with information technology … goes beyond traditional notions of computer literacy,” they noted.*

While literacy “might call for a minimal level of familiarity with technological tools,” they argued, a new age called for something deeper—learners now needed fluency with technology. Fluency with information technology represented “a deeper, more essential understanding and mastery … than does computer literacy as traditionally defined.” The report suggested that success required a change in definition that could properly signal—and catalyze—this essential change in mindset.

Over the past decades, data has followed nearly the same path, moving from a novelty to (in some cases, quite literally) a home appliance. Like computers, data is poised to transform everything we do. Not to acknowledge the necessity of an equivalent shift in our relationship to and understanding of data would be reckless. Just as the NRC’s call for IT fluency education was prophetic, it is time for an equally crucial change in the way we think about data.

So how do we proceed? What shifts in our view of education does this turning point require? In the same report, the NRC laid out some aspects of IT fluency. Individuals who were fluent with IT, they noted, would understand technology well enough to be able to apply it productively both in their work and everyday lives, “recognize when it would assist or impede the achievement of a goal,” and be able to continually adapt to its changes and advancements. Substitute “data” for “technology” in that thinking, and the implication is clear—today’s learners need precisely the same attitude and ability with data.

The notion of fluency is a valuable distinction. Most commonly played out in the context of language acquisition and reading, it connotes a sense of flowing grace and mastery. Timothy Shanahan described fluent readers as those who “can read like [they] speak.” Gaining fluency also affords benefits—Kylene Beers describes how fluent readers “move easily from word to word, spending their cognitive energy on constructing meaning.” In fact, in the realm of oral reading, fluency is among the best predictors of overall competence.

Changing our expectations

We can draw similar parallels with the world of data understanding. In other domains, literacy connotes only a basic level of comprehension or a simplified ability to read, translate, or decode. Arguably, that’s just where we’ve been stuck for decades with data in education. At the most basic level, learners see data as something to be read—values to be looked up, like player names on a roster or numbers in a telephone directory. Cliff Konold and his colleagues have noted that students new to data analysis see this task as the purpose of graphs.

And why shouldn’t they? It’s what we have valued in education. While educators have been hard at work making sure students draw properly numbered axes and remember graph titles, creators of standardized tests have reinforced this basic sense of literacy. In an analysis of 274 items from state tests, Cliff Konold and Khalimahtul Khalil found that 78% demanded nothing more of learners than encoding or decoding graphs or tables to identify connected pairs of values.

This approach to data is not enough. We have been underserving learners for decades, depriving them of the ability to truly comprehend data’s fundamental nature. We have inhibited them from seeing the world through the powerful lens of aggregated distributions, each with its own individual shape, center, and spread. We have protected them from grappling with the nature of measurement, structure, and complexity. But in a world of big data, continuing in this mode is unconscionable. Moving toward a more nuanced view of data understanding represents an acknowledgment that the world itself has changed. Data can no longer be simply looked at—it must be explored. Learners must view data as a medium they can filter, transform, join, and summarize, moving and navigating within it as an artist does with paints or a poet with words. The world of so-called Big Data, which is already exploding many of our existing notions of how to understand and work with data, demands such a fundamental shift.

The NRC recognized an analogous need on the eve of our last tectonic change. Information technology, they wrote, “is a medium that permits the expression of a vast array of information, ideas, concepts, and messages.” To be fluent with technology requires “effectively exploiting that expressive power.” We must view data in precisely the same way, preparing today’s learners to be fluent in the medium that will shape our—and their—futures. Properly prepared, they will stand ready and eager to embrace data as the fundamental building block underlying the solutions to our modern challenges. The Concord Consortium has been paving the way for this revolution since our founding, and we invite you to join us.

References

National Research Council. (1999). Being fluent with information technology. Washington, DC: National Academies Press.
Shanahan, Timothy (2006). Developing fluency in the context of effective literacy instruction. In T. Rasinski, C. Blachowicz, & K. Lems (Eds.), Fluency instruction: Research-based best practices (pp. 21-38), New York, NY: Guilford Press.
Beers, G. (2003). When kids can’t read: What teachers can do: A guide for teachers 6–12. Portsmouth, NH: Heinemann.
Konold, C., Higgins, T., Russell, S. J., & Khalil, K. (2015). Data seen through different lenses. Educational Studies in Mathematics, 88(3), 305-325.
Konold, C., & Khalil, K. (2003). If U Can Graff These Numbers—2, 15, 6—Your Stat Literit. Paper presented at the Annual Meeting of the
American Educational Research Association, Chicago, IL.
Erickson, T., Wilkerson, M., Finzer, W., & Reichsman, F. (2019). Data Moves. Technology Innovations in Statistics Education, 12(1).

Chad Dorsey (cdorsey@concord.org) is President of the Concord Consortium.