Authentic Data in K-12: How DIFUSE Engaged High School Students
At the 2024 AI Connect K-12 Data Science Conference, Sheri Johnson of Essential Analysis Solutions and a Dartmouth College researcher described a high school lesson built on real COVID-19 and pollution data. Students liked presenting findings more than analyzing them, a result the presenters said points to teaching communication.
By The AI Enterprise Magazine Team4 min read
Key takeaways
- Process over answers. The GAISE II framework treats data science as asking questions, collecting and cleaning data, analyzing and communicating results.
- Real data engages. High school students worked with a documented COVID-19 and environment data set and said it felt like an internship.
- Communication stood out. Students grew more excited about presenting data even as interest in analysis fell, suggesting a need to teach presentation skills.
- Free tools exist. CODAP, Tuva, Desmos and online applets analyze data without coding.
- Open resources. DIFUSE modules are in a public GitHub repository for teachers and parents.
At the 2024 AI Connect K-12 Data Science Conference, a virtual event, Sheri Johnson, Data Science Educator at Essential Analysis Solutions, and her colleague, a research associate at Dartmouth College, described how authentic data can pull high school students into data science. Their 20-minute session covered the state of K-12 data science education, a Dartmouth-led project that built data tools for classrooms, and the reactions of students who used one of its modules.
What does data science look like in K-12 education?
Johnson said that when she was in school there was no probability or statistics in K-12. Common Core, adopted by most states, added some to middle and high school, and many states are now revising it, she said, in the hope of including more data science.
The framework behind that change is the GAISE report, first published in 2005. A second edition, GAISE II, was updated in 2020 and is endorsed by the National Council of Teachers of Mathematics and the American Statistical Association. It has levels A, B and C, which broadly line up with elementary, middle and high school.
The levels are sequential, Johnson said, because concepts such as standard deviation take time to develop. An adult who has only ever measured the center of a data set and not its spread, she noted, would need to start at level A.
"Statistics and data science is more than math and analyzing numbers," Johnson said. Students have to ask questions, collect and clean data, consider it and communicate the results. She described the four-step statistical problem-solving process in the report as non-linear: a student who starts a graph often ends up with new questions and has to go back and collect more data.
What is the DIFUSE project?
The Dartmouth researcher presented DIFUSE, a project funded by the National Science Foundation. She said the name stands for data science infusion into undergraduate STEM education, and that it aims to bring data science into introductory courses, including social science courses.
The team partners with instructors who may lack time or data skills to build analysis tools themselves, then develops modules and evaluates how students respond. Some tools need only a spreadsheet, she said, while others use programming in R or Python. The resources are stored in a public GitHub repository, and teachers, instructors and parents can use them.
How did high school students use authentic data?
For the high school pilot, the team chose a module whose mathematics was accessible to younger students and built a lesson around it. The module, from an environment and society course, examines links between COVID-19 death rates early in the pandemic and other variables. It focuses on "Cancer Alley," which the presenter described as a heavily polluted corridor along the Mississippi River.
The data set combines demographic and environmental variables from many documented sources, and she said it took a great deal of effort to build. Students explored it with three tools: an interactive map shaded by variable, a correlation matrix, and scatter plots with linear fit equations and R-squared values.
Working in groups, students followed the statistical problem-solving process and made posters to submit to a high school data contest. Before and after the lesson, the team surveyed them and interviewed three.
What did students say about it?
Feedback was positive overall, the presenter said. Most students valued data science, were interested in working with data and saw it as important for their careers. They enjoyed multivariable data sets and said their confidence grew, moving from single-variable to two-variable analysis.
One result stood out. Between the surveys, students' excitement about presenting data rose while their interest in data analysis fell. Many preferred to focus on communicating results over analyzing them, which she said points to a need to teach presentation skills and adjust the approach.
Interview comments were upbeat. One student said they felt engaged, and "I feel like an intern."
Which free tools and resources can teachers use?
Johnson pointed teachers to the American Statistical Association's K-12 education resources, which include lessons, tool links and student competitions, and to the ASA's Statistics Teacher site, with a photograph-based lesson and a color analyzer tool.
She said spreadsheets are the most widely used tool but cannot draw a dot plot with an overlaid box plot, now part of the middle school curriculum. Students do not need R or Python for that, she said. CODAP and Tuva are free and built for data analysis, and can map data and cross-link graphs with a data table. Desmos and free online applets also help students visualize statistics.
Where to hear more
AI Connect continues to run free live sessions on data and AI. The 2027 schedule includes 24 free live webinars across four series: the AI x Enterprise Panel Series, AI Demo Days, Agentic AI Showcases and the Plug & Play Series. Registration for 2027 events is coming soon. See the webinar schedule.
Frequently asked questions
What is the GAISE II report?
It is a 2020 update of the 2005 GAISE guidelines for statistics education, endorsed by NCTM and the American Statistical Association, with levels A, B and C that roughly match elementary, middle and high school.
What is the DIFUSE project?
Presenters described it as an NSF-funded Dartmouth College project that brings data science modules into introductory undergraduate STEM courses, with resources shared on GitHub.
How did high school students respond to authentic data?
Feedback was positive. Students valued data science, gained confidence moving from one to two variables, and enjoyed presenting results more than analyzing them.
Which free tools can teachers use to analyze data?
The presenters pointed to CODAP, Tuva, Desmos and free online applets, which require no R or Python programming.
This article is based on remarks by Sheri Johnson, Data Science Educator, Essential Analysis Solutions; Dartmouth College research associate on the DIFUSE project (name not confirmed) at AI Connect K-12 Data Science Conference 2024 on July 25, 2024. More coverage: More AI Connect Events.
Filed under: AI Use Cases & Applications, People, Culture & Leadership
