Exploring Real-World Phenomena through Interest-driven Data Investigations in a High School Data Science Curriculum

• In Event: How Data Science Education Can Enable Exploration of Real World Phenomena

Abstract

Objectives
This poster showcases how students’ interests can shape authentic data science learning experiences. We focus on the final project from API Can Code, an interest-driven curriculum that introduces high school students to the computational foundation of data science through exploring meaningful, real-world phenomena using public Application Programming Interfaces (APIs). This work shows how emphasizing students’ interests can promote deep engagement with data practices and strengthen the connections between data science and students’ lived experiences.

Theoretical Framework
This work is grounded in the Interest Development Theory (Renninger & Hidi, 2015) and builds on the Integrated Interest Development for Computing Education Framework (Michaelis & Weintrop, 2022), which provides a model for designing learning experiences that cultivate and sustain interest in computing. These frameworks contextualize interest as both a motivator for and an outcome of engagement in computing and data-rich inquiry. Throughout API Can Code, students have agency to engage with data practices in ways that interest them, either by querying provided datasets or by investigating phenomena they find intriguing.

Methods
The API Can Code curriculum was implemented in a public charter high school in the US. As part of the final project, 35 students were invited to complete the full data science cycle (IDSSP, 2019). They began by formulating a question about a real-world phenomenon they cared about. Students then identified a relevant data source and used EduBlocks, a Python-based block-based programming platform, to write programs to retrieve and manipulate the data. They then used CODAP to conduct statistical analysis and generate visualizations. The final step involved preparing and delivering presentations to communicate their findings to peers. We collected and analyzed student presentations, final project artifacts (including scripts and EduBlocks programs), and post-survey responses.

Results
Analysis of students’ final projects revealed a diverse range of topics related to real-world phenomena that students chose to explore. The most common themes included investigations related to animals (e.g., dog breeds), Music (e.g., Billboard Top 100), and games (e.g., Pokémon), each attracting multiple students. Other investigated questions relate to movie ratings (e.g., Top 100 IMDB movies), health issues (e.g., COVID-19), and affordable housing. These projects allowed students to pursue topics that were meaningful to them, such as one student investigating how popular her favorite music artist is, and a second student searching for affordable housing in the city, as her family was in the process of relocating. This distribution of topics and the associated motivations in pursuing them highlight how interest-driven activities can surface meaningful, authentic, student-centered inquiries that link data practices to lived experiences. The diversity also demonstrates the value of allowing students to pose their own questions and select datasets that resonate with their curiosities.

Significance
This work contributes to the emerging field of data science education by demonstrating the promise of interest-driven curricular design. In doing so, it helps students understand the ubiquity and importance of data in their world. API Can Code positions students as investigators of real-world phenomena around them.

Authors

• Rotem Israel-Fishelson, University of Maryland

• David Weintrop, University of Maryland