Data Science Through Physical Computing in a Smart Greenhouse Project: A Case Study

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

Abstract

Objectives
This study aims to understand a teacher’s perceptions and experiences of teaching data science, particularly when it is embedded in a curriculum that features physical computing. We present the experience of a science teacher as an illustrative case.

Theoretical Framework
Data literacy involves using data to answer real-world questions through a systematic, multi-step inquiry process (Wolff et al., 2016). Another aspect of data literacy is the ability to analyze authentic data, which is accurate, qualitative or quantitative, and obtained from a real-life context/phenomenon (Kjelvik & Schultheis, 2019). Students can collect small amounts of (authentic) data manually, but sensors help them collect large datasets in a short amount of time (Lee & Thomas, 2011). Working with authentic, relevant data offers students unique advantages, like providing real-world context for the data and developing disciplinary concepts and connections (Kjelvik & Schultheis, 2019).

Methods
This study was part of a physical computing project in which middle school students learned to build a tabletop smart greenhouse. They then coded various environmental sensors using the micro:bit to collect data on environmental factors such as temperature, humidity, and light intensity within the greenhouse, and grew basil plants. In this case study, an experienced science teacher (Mr. M) implemented the greenhouse project in his Grade 8 classroom for two weeks (ten days) in Spring 2024. As a culminating project over the last two days, students conducted scientific investigations using the greenhouse, where they collected their own data, analyzed it, identified trends in the variables over time, and drew inferences. We conducted a post-implementation interview with Mr. M and collected classroom observations to gain a deeper understanding of his experience during the implementation process.

Results
We found that Mr. M did not face major challenges in implementing the project’s curriculum related to coding and data science, even when it was outside his area of expertise. In his interview, Mr. M described his primary learning goal for this project as introducing coding for a real-world, tangible purpose. This real-world context of the greenhouse was helpful for students to collect their own authentic data on temperature, humidity, and light intensity for their greenhouses by coding the sensors. Thus, students could also make interdisciplinary connections among plant science, data science, and computation. Additionally, during his interview, Mr. M placed a major emphasis on his coding-related objectives, but he did not even consider it an additional effort to implement data science concepts, as they blended seamlessly into the curriculum. In conclusion, the physical computing context made it conducive to implement data science concepts easily by providing a real-world, authentic context. Secondly, Mr. M could easily implement this curriculum due to his comfort level working with data science concepts.

Significance
A project like the greenhouse project can be beneficial for teachers with diverse expertise who aim to integrate data science with multiple disciplines. Additionally, such a project provides a context for teachers to adapt the curriculum according to their and their students’ interests, especially by developing real-world mini-projects where students collect authentic data.

Authors

• Jaai Uday Phatak, Boston College

• Sheikh Ahmad Shah, Boston College

• Avneet Hira, Boston College

• Helen Zhang, Boston College

• Michael Barnett, Boston College