How Elementary Teachers Design and Implement Inclusive and Interest-Based Data Science Units

• In Event: Data Science Education: Re-Envisioning for 21st Century Teaching, Learning, and Research

Abstract

This study examines how to design and integrate data science curricula in elementary classrooms by aligning students’ interests and relevant problems with data science practices. Through a research-practice partnership (RPP; Coburn & Penuel, 2016), we present descriptive cases of five elementary teachers who co-designed and implemented data science units with researchers and peers. Our central question is: How do elementary teachers co-design and teach data science units centered on student interests?

Preparing young students for data science is important for the health of society, yet it is underexplored in elementary schools (Adisa et al., 2023). Rosenberg et al. (2020) suggest that teachers provide students with hands-on experience working with real-world data and collaborative opportunities. Integrating learner-centered activities where students choose their own datasets for situated data storytelling with digital visualization tools as objects-to-think-with has been shown to increase students’ abilities to engage in data science practices and processes such as communicating results, identifying data patterns, and reasoning through and with data (Thompson & Arastoopour Irgens, 2022).

Drawing on connected learning theory (Ito et al., 2013), we helped teachers create data science units to support young learners’ data science practices. Connected learning draws on socio-cultural perspectives (Vygotsky, 1978) and considers the importance of students’ social relationships, their communities, and personal interests to pursue educational goals. The theory posits that students learn more effectively in environments where they are asked to produce artifacts that draw on their lived experiences and interests, and when supported by peers with similar interests.

We employ descriptive case study methodology (Yin, 2018), using three illustrative cases to exemplify the phenomenon of how elementary teachers develop and implement interest-based data science units. We focus on second, third, and fifth-grade teachers (see cases and Table 1). Data was iteratively analyzed from observations, reflective journals and interviews, using open coding. After close reading and memoing all data, we engaged in rounds of independent and consensus coding (Creswell, 2013). We triangulated data across sources, refined codes into categories, and collaboratively developed themes. Credibility and confirmability were ensured through prolonged engagement with data, data triangulation, consensus-building, and member-checking (Merriam & Grenier, 2019).

Case 1: Ms. Mackenzie designed a data science unit around her 2nd-grade students' love of Pokémon cards.

Case 2: Working as a team, Mr. Tims, Ms. Jones, and Ms. Aranda asked students to analyze dog traits, characteristics, and needs to choose the best dog for their families.

Case 3: A music teacher, Ms. Taylor, had her 5th-grade students explore and analyze publicly available datasets to answer the question, “What makes a song popular?”

Across the three cases, findings suggest inclusive data science instruction includes posing personally relevant and local problems for young learners, valuing and integrating students’ everyday discourse and lived experiences, and connecting research to practice through equitable partnerships (Arastoopour Irgens et al., 2023). Our examples offer elementary school educators’ inclusive approaches for deepening children’s data science practices through data storytelling, instructional units connecting children’s lived experiences and communities, and valuing their everyday language and ways of thinking.

Authors

• Danielle C. Herro, Clemson University

• Golnaz Arastoopour Irgens, Vanderbilt University