Understanding the Data in K-12 Data Science

Abstract

Our increasingly data-driven world is amplifying the need for everyone to develop foundational data literacy skills. In response, a growing number of K-12 data science curricula are being designed to introduce all students to data. These curricula define what data science is at the high school level and directly shape how students are introduced to and understand the discipline. Ensuring these curricula are effective, engaging, and, most critically, equitable is of paramount importance. This paper presents a qualitative analysis of four curricula, focusing on the data used to introduce learners to the field of data science. The analysis uses a series of analytical lenses to evaluate the 296 distinct datasets used across the curricula and identifies trends and best practices in dataset selection. The analysis includes using data collected from high school students about their interests and experiences with data to understand if and how contemporary data science curricula are tapping into students' lived experiences to situate data science learning experiences. The findings show that the curricula use relatively recent and small datasets covering a range of topics and that there is limited learner involvement in dataset selection. Further, the analysis reveals gaps between the datasets used and students' self-reported interests. This work highlights the importance of dataset selection, especially as it relates to supporting learners from historically excluded populations in technology fields. Finally, this paper provides practical implications to assess existing curricula and advances our understanding of