Coding Like a Data Miner: Reflections on Epistemic Plurality in Sandbox Data Science With Youth (Poster 5)

• In Event: Exploring Critical Data Literacy in K–12 Social Studies

Abstract

Because it underlies most contemporary digital innovations, data science has been tantamount to national calls to support innovation and public literacy in the ways the field impacts daily life (Wolff et al., 2016; Rosenberg et al., 2020). Undermining these efforts has been disagreement in when and how to best support learning in these areas. While there has been notable recognition that learning data science is viable as early as with pre-secondary learners, the how has been an elusive challenge—and particularly in areas that emphasize situated inquiry as a context for learning both technical and humanistic aspects of data and the science used to interrogate and harness them (Lee, Wilkerson and Lanouette, 2021). Along this trajectory, there is growing critique, for instance, that data sets curated by others can, at best, be reductive instantiations of practice and, at worst, reify practices that undermine democratic ways of doing and knowing (Walker et al., in press). This can be observed in activities that overly simplify inquiry and problem solving heuristics inherent to authentic practice, or analyses that sever or misrepresent the relationship between data, its materiality, and society. We attend to these issues by offering an alternative approach to data science pedagogy that we call sandbox data science. We accomplish this in Coding Like a Data Miner (Barany et al., 2023), a set of learning activities that teach learners to access, curate and analyze big data sets for themselves that are drawn from social media platforms (e.g. Twitter) that are typically inaccessible to consumers due to the technical and expertise thresholds needed to communicate with these platforms.

In this project, we examine heuristic strategies a group of nine youth (ages 13-17) participants used when participating in an online pilot of Coding Like a Data Miner. Qualitatively analyzing participant interviews and researcher observations, we present three case examples of the myriad strategies participants leverage to plan, implement and troubleshoot inquiry projects they defined along their personal, cultural and sociopolitical concerns. We discussed the various affordances and constraints our sandbox approach provided with regard to epistemic plurality, or the freedom to engage with expanded and flexible agency. We also discuss the extent to which this supported learner productivity, resource use, and collaboration. We examine these observations in relation to the extant literature in data science education in contemporary pre-college teaching and learning.

Authors

• Michael A. Johnson, University of North Texas

• Amanda Barany, University of Pennsylvania

• Alan Barrera, University of Texas - El Paso

• Alex Acquah, University of Texas at El Paso

• Sayed Mohsin Reza, Pennsylvania State University - Harrisburg

• Justice Toshiba Walker, University of Maryland