Capstone project for completion of Data Science minor
Capstone project for completion of the Data Science minor (Fall 2025)
Pomona College / Claremont Colleges
R for Everyone is a four-lesson, interactive micro-curriculum that helps middle school students (grades 6–8) build foundational data literacy and ethical reasoning using R as the engine behind the scenes.
Each lesson is built as a learnr tutorial in R and deployed as a shinyapp, so teachers and students can access it through a simple web link—no local installation or coding background required.
The curriculum aims to:
- Introduce core ideas in data, chance, and variability in an accessible way
- Give students practice asking questions, reading graphs, and telling data stories
- Embed ethics prompts about bias, fairness, and representation directly into math tasks
- Provide teachers with plug-and-play lessons that fit into an existing middle school course
Recent work from Data Science 4 Everyone (DS4E) and NAEP highlights a worrying trend:
- Between 2019 and 2022, 8th grade performance on Data Analysis, Statistics, and Probability fell significantly on NAEP.
- Only a small fraction of high school students take a course that looks like modern data science, and access is uneven across districts.
At the same time, students live in a world where data drives newsfeeds, recommendations, policing, health, and finance. This project responds to that gap by designing short, realistic, ethics-aware data lessons that can slot into a middle school classroom without requiring teachers to become programmers.
The four lessons are aligned with the GAISE II statistical problem-solving cycle:
Ask → Collect → Analyze → Interpret & Communicate
Each lesson has a specific role:
-
- Distinguish between opinion questions and statistical investigative questions.
- Sort example questions into “can use data to answer” vs. “cannot.”
- First ethics prompt: What happens if you only survey your friends? Who is left out?
-
- Work with a small, simulated classroom survey on minutes read last night and device used (phone/tablet/laptop/paper).
- Explore the dataset in a simple table and view plots of minutes read by device.
- Discuss typical values, variability, and limitations of such a tiny sample.
-
- Reuse the same reading dataset to compare device groups.
- Interpret a boxplot of minutes read by device type.
- Practice careful language:
“In this sample, tablet users read more minutes on average than phone users”
versus overstrong claims like “Tablets make kids read more.”
-
- Introduce a new, simulated sleep dataset (hours of sleep, device in bedroom, sports team, etc.).
- Students interpret a provided graph and summary table instead of coding from scratch.
- Study a model data story that combines:
- a graph
- a key number (mean/median)
- a clear statement of limits (“in this sample”)
- Critique over-claiming blurbs and write their own short data story.
Across all lessons, students interact by clicking, choosing answers, and typing short responses—no prior coding is assumed.
The project is grounded in several existing frameworks and resources:
-
GAISE II (ASA, 2020)
Provides the Ask → Collect → Analyze → Interpret & Communicate structure used to organize the four lessons. -
Data Science 4 Everyone (DS4E)
Supplies national context (NAEP trends, course-taking patterns) and motivates the push for modern, technology-supported data experiences in K–12. -
Chiodo et al. (2023/2025), “Teaching Resources for Embedding Ethics in Mathematics” (arXiv:2310.08467)
Informs the decision to embed short ethics prompts directly inside everyday math tasks rather than as a separate unit. -
youcubed “Big Ideas” in Data & Statistics
Supports the emphasis on patterns, sense-making, and communication, especially in Lesson 4’s data storytelling.
All lessons are built in R and delivered through the following tools:
learnrfor interactive tutorialsshiny/ shinyapps.io for deployment as web-accessible lessons- Simulated toy datasets (reading & sleep) created inside the R code
Key design choices:
-
No real student-level data is used.
All “classroom” datasets are small, simulated tables designed for instructional use. -
No automatic logging of student responses.
Quizzes and free-response questions run locally in RStudio or in the web app; logging is disabled in the prototype to avoid storing identifiable data. -
Teacher-controlled storage (optional).
If teachers want to keep student work, they can use their own tools (e.g., Google Forms or an LMS) to collect written reflections or screenshots.
This repo contains both the curriculum code and the written capstone. Key items include (names may evolve slightly):
-
Final_Writeup.Rmd/Final_Writeup.pdf
Full written thesis / project report (motivation, design, pedagogy, ethics). -
Project_Draft.Rmd
Earlier draft of the write-up. -
Lesson1_*.Rmd,Lesson2_*.Rmd,Lesson3_*.Rmd,Lesson4_*.Rmd
Source files for the fourlearnrtutorials. -
slides/(if present)
Presentation slides for the final capstone talk. -
README.md
You’re here!
Faculty Advisor: Prof. Jemma Lorenat (Pitzer College)
Capstone Instructor: Prof. Jo Hardin (Pomona College), GitHub: @hardin47
This project was also informed by:
- Data Science 4 Everyone (DS4E) policy briefs and reports
- The GAISE II report (American Statistical Association)
- youcubed’s data & statistics resources
- Chiodo et al.’s ethics-in-math teaching materials
Status: Core curriculum + written report completed in Fall 2025.
Next steps:
- Pilot the lessons with middle school teachers
- Collect structured feedback using a short validation rubric and survey
- Iterate on timing, contexts, and teacher supports based on classroom use
This curriculum is intended for educational use only.
Feel free to adapt the ideas and code for your own classroom, with attribution.
Inspired by:
- youcubed and Jo Boaler’s work on data science in schools
- NAEP and DS4E analyses of student performance and course access
- Chiodo et al.’s work on embedding ethics in mathematics