Can third graders learn data science? These researchers say yes

Tell me about your dog.

What type is it? What color is it? How much does it weigh? If you have a cat, the same questions apply. What is the dog-to-cat ratio in your classroom (or office or home)?

If you can answer that, that’s how you teach data science to third graders – taking what seem like complex or abstract concepts and applying them to tangible things in people’s lives. students.

That’s the approach Claire Bowen and her team took when they ran a week-long series of data science programs in 2020, when the pandemic first forced students and teachers to fall back on distance learning. Bowen, senior research associate at the Urban Institute, wanted to fill the lack of accessible educational resources with sessions on coding, mapping, and, of course, Bowen’s prolific workshop on data collection and visualization.

“There were kids saying ‘Dogs, dogs, dogs’ all the time, and that was an example of oversampling,” Bowen says with a laugh. In total, about 200 students participated. “It just showed how much the kids wanted to know more.”

A few years later, the Urban Institute turned their brainchild into Data4Kids, a series of “data stories” that educators can use to teach data science concepts in their classrooms. The project was launched in December in partnership with seven K-12 organizations and a $10,000 grant from the National Science Foundation’s South Big Data Hub.

“We wanted to collaborate with educators and figure out what they need, what they’ve learned over the past year about being in this remote environment, and the best ways to engage students. [in data science]”, says Bowen, and the concept of data stories was born. “We wanted to make sure it was as flexible as possible for instructors. Maybe they want to do something a little more advanced, or maybe they just want to immerse their kids in the data and start with something a little more basic.

Each data story toolkit comes with an instructor’s guide, data, and slideshow. They include activities based on students’ grade level – from junior high to senior high – that explore topics such as health equity, food insecurity and national parks. Each is designed for teachers to start using without a lot of extra legwork.

Jonathan Schwabish, a senior researcher at the Urban Institute, says today’s students will be tomorrow’s data consumers, grappling with all the security and privacy issues that come with it. Being able to discern what is true data from what is not will be part of the behavior of a responsible citizen, he adds.

“It’s not like we were born as humans who can read a bar chart,” says Schwabish. “It’s not in our DNA; we have to learn it. I think education can’t come soon enough, and we hope the materials are low enough for educators to use.

Data4Kids is about connecting students to data science in a way that feels familiar to them. This means drawing and coloring data visualizations by hand or collecting data such as height and weight.

“Third-graders don’t do multivariate analysis, but they can make some simple observations about the data,” says Schwabish. “One of the questions the team wrote down was, ‘Count the number of parks in your state.’ It’s a very personal thing that kids can do.

The project isn’t finished yet, says Schwabish, and he hopes the templates provided by the initial set of Data4Kids stories will encourage educators and data professionals to become contributors to data story collection. He’s already found a few people interested in creating data stories on topics related to sports and test scores.

“Our hope is that we can add more and more over time,” says Schwabish.

While the demand for data scientists and related professionals remains high, Bowen hopes Data4Kids will help interest more students in data science careers. She says collaborators have even talked about ditching the typical professional headshots that would go on the website in favor of photos of them taking part in hobbies, an effort to better connect with kids who might see them.

It’s part of a larger effort “to show that you don’t have to be a certain type of person to work with data,” says Bowen. “I’m a first-generation student and grew up in a field where women traditionally didn’t pursue careers. Having role models makes things much more accessible and easier to believe is achievable. »

Comments are closed.