Could Data Science Diversify the STEM Field? Why Courses Designed This Century Feel so Relevant to All Students

You can listen to this episode of the MindShift Podcast on Apple Podcasts, Google Podcasts, NPR One, Spotify, Stitcher or wherever you get your podcasts.

There are many reasons students don’t like math: stressful timed tests, right and wrong answers, isolated work, math anxiety learned from adults around you. A 2012 PISA survey found that one third of high school students feel helpless and emotionally stressed when doing math. And if you don’t see people who look like you succeeding in a subject or a field, it can be isolating, especially for young people. 

“Something that’s really important, particularly for adolescents and high school students, is that they feel a sense of belonging inside STEM,” said Stanford maths education professor Jo Boaler. 

Researchers have found that a sense of belonging helps students succeed, in part because feeling like you’re a part of a community of learners is a powerful motivator to do well. “And unfortunately, a lot of students do not feel that they belong inside traditional high school maths classes,” said Boaler.

But there is one subject that students, including those who are math confident, enjoy learning: data science. As of 2020, data science is accepted math coursework for University of California and Cal State University’s A-G requirements, so students might see it offered in more schools.

Over the last decade, teams of teachers, researchers and academics have been developing data science curriculum and tools for the classroom, and having a modern approach to teaching is resonating with students. 

“It’s kind of a unique opportunity because there wasn’t a high school data science course before,” said Suyen Machado, director of the Introduction to Data Science program, which was started as a partnership between UCLA and the Los Angeles Unified School District nearly ten years ago. The program was funded with a National Science Foundation grant to increase the amount of students going into STEM careers and to bring computational and statistical thinking to underrepresented high school students, according to Machado. 

“Engaging lessons that are inquiry driven, student driven and collaborative are really well suited for underrepresented groups, and you will find all of that in our curriculum. And they’re good for students in general,” Machado said. 

REAL WORLD USES

The data science curriculum gives students opportunities to look at real data instead of abstract formulas.  

“It’s just so much fun,” said James Molyneux, a professor at Oregon State University who was involved in the development of IDS. For example, students can collect their data and compare themselves to larger government data sets, like the American Time Use Survey from the Bureau of Labor Statistics. Students can measure how much time they spend grooming, eating, being with family and consuming social media, according to Molyneux.

Among students, there’s a growing interest in data sets, such as pollution in school communities and which gender character is most likely to survive a horror film. For IDS participants, the most popular data project involves snacks. 

“It honestly made me more aware of what I was taking in and putting in my body,” said student Linda Solares of Leuzinger High School of the snack project. Not to worry, the unit is not about encouraging weight loss or anything. Students used the IDS app to track information like the amount of salt, sugar content, cost, number of ingredients or their reasons for eating.

“We’re in quarantine, we’re eating a lot more out of boredom and stuff. So honestly, it really helped me,” said Solares. “After I finished the survey, I was like, whoa,” she said, “I was really eating not so healthy.”

Surveys of IDS students in LAUSD found that coding was the most challenging part of the course, but also, the most important skill students learned. Using programming tools, like RStudio, they persisted by trying over and over again to get their code right. And that helped boost confidence in their ability to problem solve.  

“The lab is a lesson for us to learn about the codes and how we can implement them in certain situations,” said Leuzinger student Peter Tran, who would test different variables against one another, like finding the most common time of day students ate unhealthy snacks. 

An important part of the data science curriculum is understanding privacy matters, and knowing how data is collected about people and used against them. This knowledge can help develop a person’s media literacy.

“There’s a lot of misinformation out there,” said Boaler. “Having students develop a critical perspective – that’s one of the things we can teach in data science. Be skeptical of data that’s put in front of you, ask questions of it, think about who put that data together, what purpose did they have for it.”

LEARNING GETS MESSY

The messiness of the data sets is part of the appeal for students; it’s what engages them in learning and not shying away from unknown outcomes, according to Concord Consortium’s Chad Dorsey. 

“It’s almost sort of pre-chewed and preordained,” said Dorsey of traditional curriculum that doesn’t engage students. “And when we do that, we take a lot of the discovery away. We’re finding the value in putting students into the place of needing to ask and answer questions with data that might be ambiguous or that might have a missing value,” said Dorsey. As part of an NSF grant, the group developed the free CODAP tool so teachers can integrate data skills into their classes, such as science. The group also provides teachers with professional development.  

“We’re finding the value of putting students in the driver’s seat to do the exploration themselves, to uncover new things in the data that maybe the teachers didn’t understand was there in the first place and where students are finding something different than their neighbors,” said Dorsey.

For Leuzinger High School IDS teacher Ding-ay Tadena, that has meant giving students agency over the topics they want to investigate, such as sports. “They learn how to think deeper and then use these math skills and eventually they love it,” says Tadena, who has seen students of all math levels succeed in data science. She says that in data science class, students see themselves as more than the math track they’re in.  

“It taught them how to dream bigger rather than just being profiled as lower performing in terms of math,” she said. “And that is the beauty of it because you teach them how to code, how to do this data, how to scrape data from the internet and push it in R in the field that interests them.” Tadena, who has been teaching math for about two decades, says data science is in many ways a respite for math teachers like herself who are looking for ways to engage their students. 

“The students are so interested,” Tadena said. “They’re so into it.” 

For science teacher Emerlyn Gatchalian, having Concord Consortium’s CODAP tool makes understanding the periodic table easier for some of her students. “They’re looking at the different properties of elements in the periodic table using data like the atomic size, ionic size,” she said. “Because they’re using data using CODAP, it’s so easy for them to look for patterns and trends and make them feel that they can actually understand and interpret data instead of using all the equations that they’re learning in math.”

For high school special education teacher Michelle Murtha, students’ ability to graph their data using digital tools helped them understand it. “Sometimes, graphing itself is so hard for the students. But because the program helps them through it,” she said, “they’re able to actually see the graph. And for us, that’s more important, so they can actually analyze the data versus, ‘can you plot this point?’”

REDEFINING HIGH SCHOOL

When Emilio Jaime was a student at Phineas Banning High School, he was on track to take AP Calculus his senior year. He had been confident about math throughout school, but decided to take IDS based on a teacher’s suggestion. Plus, one less AP class would help ease his senior year course load. 

“I decided to let go of calculus and took on IDS, which I’m so glad I did, because I guess I was just scared because it wasn’t the norm that students were doing,” said Jaime, who graduated from UC Berkeley last spring.

What he liked about data science was the ability to play with formulas and not feel limited by right and wrong answers that were the hallmark of his math education. “This is how the formula is, and this is the answer, and there is a wrong answer,” he said of his earlier relationship to math. But data science was more fluid. “On our projects, I tried so many different graphs and so many different solutions to try to create so many different conclusions.”   

“I think IDS and data science really allows students to try different things without being scared to fail,” he said.   

IDS trains teachers across the country and abroad on how to teach data science as a course. It’s one of several programs, including ones operated by the Concord Consortium and Boaler’s YouCubed. The outcome of getting more underrepresented students in the STEM field has yet to be seen. But for now, these educators are shifting students’ experiences with STEM to increase the odds that they’ll stay.   

All of these skills will hopefully help students become better informed members of society. 

“I think that’s the biggest gift that we can give students right now – no matter how we’re doing it – is to help them understand that there are data all around them, that those data have answers, that they come from people, and that the things that they are doing are generating data all over, and to give them the ability to start to feel empowered to work with this data themselves,” said Dorsey. 

Subscribe to the MindShift Podcast in your favorite podcast app so you won’t miss a single episode. You can listen on Apple Podcasts, Google Podcasts, NPR One, Spotify, Stitcher or wherever you get your podcasts.