By Becky Purdy

The two abstract drawings (below) are more alike than different. Both depict colorful circles on a white background, some overlapping, some floating free. One drawing was created by hand, by students using colored pencils on a sheet of white paper. The other resulted from student-written code that instructed a computer to generate the art. Both pieces were based on the following set of instructions, which the students devised collaboratively:

“Eight bright randomly colored circles with radii of 15 percent of the length of the canvas. Intersecting areas of the circles are where the colors mix.”

The conceptual art project, a collaboration last spring among students in Drawing I and Introduction to Programming, tapped into the fascinating interdisciplinary possibilities of Digital and Computational Learning, a program formally launched four years ago at Loomis Chaffee. 

Digital and computational learning involves looking at data, algorithms, and programming as they relate to other things in our lives, explains Kate Seyboth, the school’s director of the program. Application of these skills and concepts to various disciplines has led to collaborative projects across the curriculum, from exploration of geometric probability to comparison of French- and English-speaking people’s feelings on a variety of topics. 

One of the primary goals of the program is interdisciplinary — to integrate the skills, concepts, and thinking surrounding data, algorithms, and programming into other parts of students’ academic lives. Another major goal from the outset was to offer courses in these areas, including expanding the school’s computer science and programming offerings. A third, overarching goal is to raise awareness and understanding of both the ubiquitous influence and the intriguing opportunities of data and algorithms. 

Algorithms are at work in nearly every aspect of our lives. What you see at the top of your Facebook feed is based on data and algorithms. The same is true of which ads pop up as you browse the web, which emails end up in your spam folder, and which songs show up on your Spotify app. The pattern of a traffic light at a busy intersection relies on data-driven algorithms, as do public transportation routes and schedules. Your Fitbit and other step-counting, heart rate-measuring, and sleep-tracking devices use data and algorithms to analyze and show your progress toward your health goals and to recommend next steps or even the need for rest and recovery.

The advent of generative artificial intelligence applications — programs that can create — takes the impact of algorithms to a new level in everyday life, and their presence makes informed awareness of them more imperative than ever in navigating the world with intention. 

“With the way our world is changing, it’s increasingly important that students understand the way these things work,” says Kate, who earned bachelor’s and master’s degrees in computer science at Tufts University. Before coming to Loomis in 2021, Kate taught for 15 years at Westover School, a girls school in Middlebury, Connecticut, where she also directed the Women In Science & Engineering (WISE) program.

Digital and computational learning involves three key concepts of problem-solving: abstraction, decomposition, and algorithmic thinking, Kate explains.

Abstraction describes taking a solution to a single problem and applying it more generally to other, similar problems. For example, you start with a triangle and its measurements and find the triangle’s area. Abstraction would explain how to find the area of any triangle based on its corresponding measurements. 

Decomposition is the process of breaking down a problem and its solution into smaller pieces. This is something we do all the time when faced with a complex task, such as traveling from one’s home in, say, Chicago to a hotel in Paris, France. You break down the solution into parts: First, drive from home to the airport. Next, fly from Chicago to Paris. Then, take a taxi from the Paris airport to your hotel.

Algorithmic thinking is the combination of abstraction and decomposition. It involves creating a step-by-step, or piece-by-piece, process by which you can solve a particular kind of problem. For example, suppose you need to figure out how to assign dormitories to all of a school’s boarding students with their chosen roommates. The problem involves multiple variables, including the number of boarding students, each student’s class year, the number of beds in each dorm, the class years housed in each dorm, each roommate pairing, and so on. The solution requires many steps whose outcomes affect the next steps in the process. Algorithmic thinking figures out how to accomplish this task. And computer programming expresses the solution in precise, unambiguous instructions.

To begin introducing these skills and concepts in subjects across the curriculum, geometry classes added a unit in which students learn to use a drag-and-drop programming tool called Scratch, which was developed by MIT. Math teacher Courtney Doyle’s geometry class, for example, is using this tool to explore points of concurrency, places where three or more lines intersect. Students in math teacher Allison Beason’s Geometry Advanced class are using a typing-based programming tool called Processing to learn about geometric probability. 

Environmental science classes have incorporated Arduino software to collect air quality data from sensors around campus as they researched such topics as the effectiveness of the air-quality infrastructure in the new Nichols Center for Theater and Dance, changes in the air quality in the pool when the swim team is practicing versus when the pool is idle, and other queries. And statistics classes, which traditionally have used spreadsheets to sort and analyze data, are turning to the computer language R for a deeper understanding of statistics and a less labor-intensive approach than using spreadsheets. R was designed primarily for statistics work, Kate says.

The usefulness of computational thinking is not limited to math and computer science fields, as some initiatives in Loomis humanities courses are showing. Adam Alsamadisi, a math and science teacher and a member of the Digital and Computational Learning faculty, is helping history and language teachers to use technology in ways that expand students’ cultural understanding of the past and present. 

Adam is working with World History teachers to help students decipher, analyze, and interpret the Catalan Atlas, a 14th-century map of the world. By bringing this ancient map to screens and applying computational tools to the map’s markings and symbols, Adam says, students can draw conclusions about the primary roles of communities in different parts of the world. One area might have been the spiritual center, for instance, and another might have been the crossroads of trade. 

Adam also worked with a History of the Present class as the students created digital stories using geographic information system, or GIS, mapping. The digital stories, often referred to as “storymaps,” bring together data, geographic information, and narratives to find meaning in complex subjects and present that meaning visually. For their final project, the History of the Present students used data, maps, and narratives to create storymaps related to social justice, Adam says. 

French and statistics classes collaborated on a project last year that compared attitudes toward various topics among English- and French-speaking populations. Students in language teacher Rachel Nisselson’s College-Level French 5 class and in Adam’s College-Level Statistics class engaged in the project, Adam says. They downloaded 500 tweets from the social media platform X (formerly called Twitter) in English and French and analyzed the expressed sentiments using a library that linguists employ to read tone from word choice. Then the students compared the sentiments that the French-tweeting people expressed versus those of the English-tweeting people on topics related to the environment, such as organic foods, climate change, and consumption. 

Importantly, technology is used in these projects not for its own sake but as a tool to enhance learning, Adam says. These approaches provide valuable, new insights that advance students’ learning. 

Usefulness is important. “We really want these projects to be authentic,” Kate notes. 

The conceptual art collaboration last spring brought together Kate’s programming students and art teacher Melanie Carr’s drawing students. To begin, the two classes took a field trip to MassMOCA in North Adams, Massachusetts, where they viewed conceptual art by artist Sol LeWitt. 

After the field trip, the students were divided into small groups, each with a mix of two art students and one programming student. Then, during a series of eight class meetings, each group developed an idea for a conceptual art drawing and wrote instructions. (Conveniently, the two classes met during the same time block.) The programming student in each group then wrote code designed to produce the art, and each group’s drawing students put colored pencil to paper to produce their own version. 

Melanie says it was interesting for her and her students to see the interpretive process of “the verbal world translated to the visual world.” Melanie had expected the two outcomes within each group to be more alike, but she was pleasantly surprised to see that “digital” and “analog” interpretations of a concept produced different results, beyond the difference in materials — a printed image on glossy paper versus hand-drawn art using colored pencil on uncoated paper. 

The project also gave the students an opportunity to collaborate across disciplines. “I like everything about the intersection of all things,” Melanie says. “I think that’s where life happens.” 

The programming students gained an understanding of how the skills they were learning in computer science class could be used in real-world situations. Kate says it is common for new programming students to see only the “tinkering” aspects of what they are doing, not the practical applications. The art project offered them a wider perspective.

As with the art, history, and language projects, ideas continue to brew for using digital and computational skills in other academic departments. Matt Johnson, an English, computer science, and math teacher and the assistant director for academic technology of the Kravis Center for Excellence in Teaching, is developing an interdisciplinary course with the working title “Algorithms, AI, and Us: Writing and Technology.” Matt says the course would explore what algorithms are and how their use influences various aspects of the world, and it would take deeper dives into generative artificial intelligence (ChatGPT and the like), issues of bias in algorithms, and the use of algorithms in such areas as social media, finance, and the health care industry. Students would write papers, articles, and persuasive essays or letters about the issues they explore in the course. In this way, students would develop both their writing skills and their understanding of algorithms.

“Part of what it should mean to be an engaged citizen in 2023 is to have a basic understanding of how algorithms are influencing things,” Matt says. 

Loomis computer science courses are interdisciplinary by design. It is not a coincidence that programming courses are taught in classrooms adjacent to the Pearse Hub for Innovation, whose theme is “make things and make a difference.” 

In College-Level Computer Science Topics, an advanced term course in programming, students work on projects with practical applications, applying their coding and computer language skills to real problems. The theme for the course last year was “game development for the common good,” a nod to Loomis Chaffee’s emphasis on the best self and the common good.

Students in the course last year worked with the school’s Alvord Center for Global & Environmental Studies to tackle several problems. They developed a game about what can and cannot be composted at Loomis. They wrote code to analyze data from the school’s solar array. And for the maple sugaring season, they wrote a program that scheduled student collection of maple sap from trees around campus, making sure the collection buckets were exchanged before they overflowed. After perfecting their programs, the students created video demonstrations to show their “clients” how to put them to use.

For their final project last year, the Topics students worked in small groups in collaboration with math and science faculty members who identified concepts in their disciplines that students struggled to master. Then the programming students designed games to help students understand these confounding concepts. One group tackled the physics concepts of motion and force in collaboration with astronomy teacher Steve Stewart. Another group worked with math teacher Joe Cleary to develop a way to explain algebraic fractions. 

And another group worked with physics teacher Julie Hinchman on the concept of conservation of momentum. The students in this group designed a game depicting a cannon inside a cave. The goal is to move the cannon out of the cave by shooting cannon balls, which propel the cannon backwards in the direction the player wants to move it, going around obstacles and ultimately exiting through the mouth of the cave. 

Kate explains that solving problems through digital and computational thinking synthesizes many of the skills students have been building as they advance through the computer science curriculum.

Several students each year also engage in programming-related independent studies and senior projects. 

For the last two years independent studies students have worked on developing a computer program that would help with the arduous task that the deans of students face each summer: assigning boarding students to dorms. The deans spend hours in a room trying to piece together this complex puzzle, sifting through the requests of the school’s 500 or so boarding students, considering the bed counts and configurations of the campus’s 13 dorms, and figuring out how to place each student in an appropriate dorm. This year, Kate says, the student who is taking on this challenge as an independent study is writing an algorithm that will take a first stab at the dorm assignments, giving the deans a starting point. The finished product will enable the deans to adjust the first configuration and see how various adjustments affect the big picture.

The CL Topics students this fall developed programs that would streamline the assignment of “work jobs,” the school-keeping tasks that all students and faculty do as part of the community work program. Hannah Hayes, the community work program coordinator, and Lillian Corman, associate director of the Norton Family Center for the Common Good, normally spend three days assigning work jobs manually. That means matching up the correct number of students and their free periods with the needed number of people for each task in each time block. Kate says the final student program will do it in seconds, and then Hannah and Lil will just need to make some additional tweaks.

As the Digital and Computational Learning program continues to gain momentum, one of the main things standing in the way of faster expansion is a universal one: time. Students don’t have enough of it to take all the courses that intrigue them and pursue all the interests they enjoy. It’s not a bad problem to have, and who knows? Maybe a Loomis student will write a computer program that squeezes more time out of each day or more days out of each term.