Context and Computation: A Relational Model for Highly Contextualized Data in Undergraduate Statistics and Data Science Education

Computational data analysis lacks sufficient frameworks for organizing and interpreting data that are specific to local contexts yet related to broader social phenomena. This article outlines a relational modeling schema for instruction in statistics and data science education. The relational model is informed by students and framed to assists both educators and students integrate computation and complex social reasoning when working with highly contextualized data. The model emphasizes the importance and value of connecting theoretical commitments with computational tools and methods to interpret socially specific data. Using archival and content analytic methods, this article presents findings from a two-year study of the relational modeling practices of secondary and postsecondary students (n = 23) in a research collective as they co-develop software in collaborative user groups. Particular attention was paid to how members engaged with a range of social issues and diverse data sources. This article uses these engagements to frame the organic coding practices that define a relational schema for working with highly contextualized data. Study findings highlight how interdisciplinary community-oriented research requires theoretically-grounded and contextually-aligned computational practices that attend to the various logic models, queries, and interpretations that emerge across users with diverse identities and domains of expertise.