This month, we look at computational tools and how they are being used and adopted at UBC. Our guest contributor is Jonathan Graves, assistant professor of teaching and majors program advisor at UBC’s Vancouver School of Economics.
Machine learning. Data science. In the last decade, these topics have gone from academia to the mainstream, from mainly living in textbooks and research papers to affecting our everyday lives.
This explosion in popularity has transformed disciplines across the social, physical, and applied sciences—and shows no signs of slowing down. Today, the use of data science tools and techniques is essential for success in many careers and fields. It provides powerful opportunities for individuals to engage in scientific knowledge creation and analysis in a way that has immediate relevance to their everyday experiences.
However, with these new opportunities have also come new challenges. The benefits of many of these techniques have tended to accrue within the parts of society most well-situated to adopt them. Powerful institutions have adopted these tools, often without due consideration, technologizing existing power structures—and creating systems which can be racist, sexist, transphobic, colonial, and ableist.
UBC, as an institution of higher learning situated on unceded Musqueam land, is grappling with these issues itself—and with training the next generation of leaders who will take on the task of realizing their transformative scientific and social opportunities.
In this short article, I will share with you some of these new tools and techniques that are being adopted and developed here, at UBC—and how you might incorporate them into your teaching or research practices. I will also share some examples and food for thought to accompany these tools, to hopefully help you as you explore the opportunities and challenges on your own. Good luck!
Tool 1: Jupyter
One key framework is the Jupyter project, which seeks to support open-source, interactive and cloud-based computational tools for both research and teaching. Using only a web browser, users write and interact with rich (e.g., text, media, code) documents called Notebooks. These can be connected with a wide range of computational tools, including statistical tools such as R or STATA, and more general programming tools like Python, JavaScript, or JULIA. This creates a powerful, integrated system for teaching and learning quantitative tools. By divorcing computation from hardware and specific software, Jupyter also provides a more consistent, inclusive, and accessible set of tools.
At UBC, we have recently developed a self-hosted instance, which is available to all students and faculty, and can be integrated into the Canvas Learning Management System (LMS) in a variety of ways. For non-UBC community members, there are a wide range of other Jupyter hubs online – including Google Colaboratory and the Pacific Institute for Mathematical Science Syzygy.
- One of the best features of Jupyter is how easy it is to get started; no software required. Try it in your browser, right now!
- UBC faculty members and the CTLT have also run an example of events, via JupyterDays. You can find their resources online, with examples for both teaching and practice.
- Learn more about teaching with Jupyter from EPFL or reach the open-source book Teaching and Learning with Jupyter.
Tool 2: Executable Books
Notebooks might be effective for research projects or notes, but how do we collect these into cohesive packages and share them with others? The answer is often via an executable book. Executable books allow you to combine documents—such as Jupyter notebooks—into an HTML-based “book” platform which can then be deployed online, compiled into a PDF, or integrated into a LMS like Canvas. When hosted online, they can include code or interactive elements that can be executed from within the book itself. If offline, they can link to a version that is executable.
While JupyterBooks, as part of the Executable Books Project, are one prominent example, many people have created learning resources using tools such as the R package Bookdown. These tools allow you to connect learning materials in a cohesive, connected way, while handing most of the hard work of implementing a website or creating a PDF to the software.
You can find a great example of how this is used in the QuantEcon project, which produces interactive lectures and teaching materials for economics classes ranging from third-year to advanced graduate level.
- Here is an example of the JupyterBook project, which is the core textbook for the Vancouver School of Economics' ECON 323
- The QuantEcon project also hosts a range of other books
- For a great example of a more static executable book, developed using Bookdown, check out Introduction to Econometrics with R
Tool 3: PrairieLearn
The final tool addresses learning assessment. While interactive notebooks or executable books allow you to share information, and can include questions, what other tools exist to help perform assessment? In many fields, enterprise tools such as Pearson’s MyLab or Cengage’s MindTap offer self-test and summative online questions to accompany their textbooks. However, what if you’re not using a textbook, or want to create an open alternative? What if you find these types of tools too expensive, or not flexible enough?
An emerging open-source alternative is PrairieLearn (PL). This is an online system for creating assessments using an easy-to-use HTML and JavaScript-based workshop—allowing you to create very flexible and powerful questions. These questions can include robust randomization and self-grading, and integrate with other systems to enable complex tools like symbolic algebra, mapping, code execution, and visualization.
As an open-source project, PL can be hosted and developed locally, but also offers a very affordable and centrally managed alternative to enterprise homework systems (from as little as $6 per student), if you want to let them handle the difficult parts of the project.
PL has also collected a wide variety of research projects demonstrating its use in real-world classrooms, and the effect it has on student learning in the classroom:
- To see a hands-on walkthrough of PrairieLearn you can watch these recorded workshops from the University of Illinois at Urbana-Champaign.
- PrairieLearn is highly engaged with the scholarship of teaching and learning; you can find more examples and papers on the PrairieLearn website.
Enjoyed reading about computational tools? Learn about other topics we covered in the July 2022 edition by reading the complete Edubytes newsletter. To view past issues, visit the Edubytes archive.
Are you interested in staying up to date on the latest trends in teaching and learning in higher education? Sign up for our newsletter and get this content delivered to your inbox once a month.
