As an increasing number of industries look to hire environmental data scientists, students looking to enter the field need a fast path to make their career shift. Bren’s Master in Environmental Data Science is a 11-month degree program that provides exactly that. MEDS students begin the program with an immersive, full-time curriculum to learn the essential quantitative skills they’ll need for their new careers.
A rigorous program like this can seem intimidating, but in the MEDS program, students have faculty like Allison Horst (PhD ’12) in their corner, dedicated to helping them succeed. Allison is one of the first professors students work with as part of the summer curriculum. She teaches Essential Math for Environmental Data Science (EDS 212) and Scientific Programming Essentials (EDS 221), courses she has designed to address rigorous subject matter while also welcoming students from a diverse range of backgrounds.
“One of the biggest misconceptions people have about data scientists is that they are people who were born as computer scientists—that somehow they emerged into the world coding,” she says. “But data scientists can and do come from all backgrounds. We need people who are creative, who are good writers, who pay close attention to detail, who have good design sense. Whatever your existing skills and talents are, they’ll probably be beneficial in data science.”
Preparing Students for Success as Data Scientists
Allison’s own career trajectory demonstrates that successful data scientists come from diverse backgrounds. She came to the field with an undergraduate degree in chemical engineering and a master’s degree in mechanical engineering. As a PhD student working with Professor Patricia Holden, she took on a teaching assistantship for a statistics class, and immediately discovered her career path.
“I’ve always loved math, and anything quantitative, and my engineering degrees gave me really solid quantitative skills. Then the first day I was teaching that stats class, it was a total lightbulb moment. I realized I loved teaching, and it was what I wanted to do with my life,” she says.
With this calling in mind, Allison took on additional teaching opportunities as a PhD student. The experience and successes she had led to her first job teaching statistics and data analysis at Bren when she graduated, a position that included the MESM program’s introduction to math and statistics. She quickly saw the need for a fresh approach to teaching this required course given the range of academic backgrounds students possessed.
If you’re considering a career in data science, reach out. The more people you talk to, the more you’ll see that you have skills to bring to the table that are valuable.
“I would literally get emails months before class from students saying how concerned they were about this class,” she says. “I started thinking, there had to be a better way than just showing slide after slide of equations and documentation.”
A life-long artist, Allison started brainstorming ways to illustrate mathematical concepts. She--and the Bren community as a whole--quickly found that combining art and mathematics made the subject matter not only more approachable, but also more memorable.
“A great example of how this works is when I’m teaching principal component analysis, where essentially you’re dealing with multivariate data and trying to simplify it, but retain enough of the complexity that you capture big patterns. When you put it in words, in a classroom, you can just see the students’ eyes glaze over,” she said.
“So, to help students understand what they’re about to do, I have them imagine being a whale shark, swimming along a horizontal axis. Then I tell them to imagine they come across this school of krill, and I ask them, ‘What direction would you tilt your face to get as many of the krill as possible in the first pass?’"
“We all actually make whale shark mouths with our hands and tilt our heads in class to mime gathering the krill, and students see they have an intuitive understanding of the process we’re about to learn about. It builds confidence, and I’ve actually had students come up to me years later and say ‘I remember principal component analysis because I remember being a whale shark!’”
Allison’s creative approach to teaching this foundational course--and others--has led to recognition beyond the Bren community. In 2019, she was awarded a UCSB Distinguished Teaching Award for innovations and excellence in teaching, and in the same year she became RStudio’s first artist in residence.
She joined the MEDS program as an assistant teaching professor in July 2020. After several years teaching in the field, she was ready to help students take the first steps in their careers as environmental data scientists, and train them with skills they’ll use throughout their careers.
“The whole point of me teaching is to make it so people can succeed in the career they’ve chosen,” she says. “That means I need to prepare them with skills that allow them to continue learning. I don’t just teach coding and math, but resourcefulness, and I want to inspire curiosity. Those are the traits that will allow graduates to work independently and in a way where they feel confident.”
Embracing an Open Approach to Education
In her quest to lower the barriers for students who are interested in data science, Allison draws on a variety of materials and methods. She notes that open science—a movement to make materials, code, and communication as open as is possible and appropriate for each project—has had a large effect on both how and what she teaches.
“I’ve learned so much from people who have been willing to share their analysis, their teaching materials, their talks, and their code,” she says. “It enables me and my students to see the way that people are working out in the world as environmental data scientists. Seeing that value has inspired me to share my own teaching materials, and I hope that can inspire students to see the value of sharing things if it can be useful to others.”
Among the resources Allison has made available are her artwork, which she provides as an open library for educational use, and the palmerpenguins data set. Allison developed the latter with Kristen Gorman, research faculty at the University of Alaska Fairbanks, and Alison Hill, a data science educator at RStudio—is just one example of the results of that inspiration. The data set, which was published as an R package under a Creative Commons license, has quickly become a popular option for data science educators.
“Kristin collected the data in Antarctica between 2007 to 2009—measurements like body mass, flipper length, and bill length. We bundled the information together, created a website, and published it, and it’s really taken off. It’s been added to Google’s TensorFlow data sets and Python packages, and it’s actually being used in data science education around the world,” she says.
Preparing for the Future of Data Science
As the first Bren MEDS cohort moves toward their capstone projects and then graduation, Allison is considering not just the future of her students, but of her field as a whole. To her, that means opening the doors of environmental data science still further.
“The future of environmental data science must include the empowerment and engagement of communities so that they have the skills and capacity to work with the data that impacts them,” she says. “These communities are already doing a lot of the work on the ground, and I hope that the next generation of data scientists will find ways to use the skills they are building to support those efforts.”
She points to recent Bren master's student group projects working with communities in Hawai‘i and Mexico as an inspiration for what is possible. “Bren students actually went to local fishing communities to learn what they were already doing, ask what they needed, and then produce that work product,” she says. “It’s an important strategy. We need to not just extract knowledge, but to support existing community-based efforts in order to effectively solve environmental problems.”
With this overarching goal in place, Allison sees a future for data science careers even more wide-ranging than already exists.
“The variety of data science jobs is just going to keep growing,” she says. “Data management, data wrangling and exploration, data visualization and communication, consulting—I think we’ll see students going into all of these types of jobs. If you’re considering a career in data science, reach out. The more people you talk to, the more you’ll see that you have skills to bring to the table that are valuable.”