School & District Management

Two Ways to Add ‘Computational Thinking’ to Middle School Science

By Sarah D. Sparks — April 09, 2019 5 min read
  • Save to favorites
  • Print

Toronto

The Next Generation Science Standards call for science teachers to bring more “computational thinking” into middle school science. Two pilot projects at the American Educational Research Association meeting highlight why that’s difficult, and two potential ways to do it.

Computational thinking draws on concepts from computer science—including organizing and analyzing data and modeling—to link science, technology, engineering, and math concepts and help students think about complex problems. As the use of technology and complex problem-solving becomes a bigger part of the workplace, education and business leaders alike have pushed for the concept to become a bigger part of STEM classes.

“This can be a very powerful strategy for students to learn science concepts as well as important problem-solving skills across various disciplines,” said Danielle Cadieux Boulden, a researcher with North Carolina State University.

But it has proven tricky to integrate, she said. “The problem is, even though computational thinking has recently been very widely embraced, there’s really no consensus pathway of how this is going to look in classroom practices,” Boulden said. “Computational thinking is not yet entirely embraced by the K-12 community and in particular with the teachers—they are not exactly sure what this is going to look like in practice with their students.”

At a symposium on the subject at AERA this week, researchers from North Carolina State University and from the University of Colorado, Boulder, highlighted two pilot programs to use computer modeling and environmental sensors to use computational thinking to enhance standard science units.

Telling Computational ‘Stories’

In Colorado, researchers Alexandra Gendreau Chakarov and Quentin Biddy are developing a series of “storylines,” scenarios based on real-life news events that set up students to collect data and solve problems using environmental sensors. The researchers, along with partners at Utah State University, are piloting the curriculum with three science teachers and one integrated STEM teacher of 200 students in grades 5-8. The teachers learn to frame science concepts through computational thinking and co-develop storylines over four summer workshops.

For example, in one storyline, students watched a video about a school being closed due to mold exposure. The video launched a discussion of mold’s life cycle, health effects, and how it could be found and dealt with in their own school. Students used digital environmental sensors to test for mold near bathrooms, drinking fountains, and other areas on campus, then collected and analyzed the data. At one school that actually found mold, the class developed a report and mold remediation plan which they presented to the principal.

While the curriculum is still in development, the researchers found after the first pilot, 82 percent of the students reported wanting to do another sensor-based project again, and 88 percent understood links between their computational activities and the science concepts in the standards.

The sensors cost about $100 for four students, but Chakarov said in the next iteration, the researchers are testing micro-bit sensors which provide a wider variety of information and cost about half that.

Modeling Diseases: ‘Bumping’ and Blocks

In a separate project, North Carolina State University researchers think teaching students to code and model before helping them understand what they could use the skills for is putting the cart before the horse.

In a curriculum unit exploring epidemics, students first modeled a disease in person walking around the class and bumping into one another to “transfer” an illness as the teacher tracked the number of students infected in each iteration. That exercise allowed them to explore basic data collecting and predictions before starting to model on computers.

“The embodied-cognition activity actually allowed the students to embody the science ... and so it allowed us to talk about the benefits of joining that within a modeling environment instead of having to keep track of this by hand,” said Jennifer Houchins of North Carolina State University. “So it allowed for that connection to why we would want to be doing this in a modeling environment.”

The team initially taught students to model from scratch using an open-source tool called blockly, but later modified the program to provide a set of example analysis code that students modify and take over throughout the class as they learn the scientific concepts and coding practices. They also embedded video-based tutorials to help students and teachers with coding specific pieces of the model, such as incorporating the time period in which different types of viruses are contagious.

“We’re offloading some of that overwhelming nature of coming into a blank environment and not really knowing how to get started,” Houchins said. “They use a prebuilt model to do some initial scientific exploration, get more comfortable with the environment and then they start modifying the code there. Taking over coding themselves allowed students to feel more ownership ... But also it allowed us to get the students into more complex scientific topics a little more quickly because they weren’t having to learn [coding the analysis] at the same time.”

Over the course of the unit, students learned to incorporate human behaviors and properties of different diseases to chart and predict different epidemics.

The researchers are still developing the pilot in the next year. “Creating a well-balanced curriculum is time intensive and requires many iterations in order to get it right,” Houchins said. “You have to make important design decisions to support students in both the [computational thinking] and science practices. Getting that balanced is really critical, so that you’re not striving to one side or the other.”

Image Source: Getty


Related:

Do you have a question about education research, or just want to know what the evidence says about that pesky instructional problem? Let me know! Drop me a line at ssparks@epe.org, or

Related Tags:

A version of this news article first appeared in the Inside School Research blog.