ÿ�մ���

You can’t say that Associate Professor of Statistics Keshav Pokhrel’s transformation of his statistics courses was completely a pandemic thing. Pre-pandemic, he was already interested in changing up his teaching approach, especially in his intro statistics classes, which always brought in lots of non-math and non-statistics majors. Getting students to be genuinely interested in the subject was the core challenge, one that inspired lots of hallway conversations with colleagues, student feedback sessions, and some meetings with instructional designers at the Hub. Then the pandemic hit. And while he hadn’t had a chance to implement many changes yet, by that time, he was conceptually fluent enough in new teaching techniques to see that his traditional formula of lectures, homework, quizzes and exams was likely going to fall flat in a remote format.

Initially, he started with recorded lectures, following one of the Hub’s main suggestions to break the videos into short, easily digestible segments. But the real breakthrough came after a conversation with a colleague, who pointed him to a student-focused New York Times feature called “” The Times' graphs and data visualizations focused on all kinds of subjects, from COVID trends to global economics to sports. But the thing Pokhrel really loved was the type of questions the authors paired with the graphs. “It was always very general, like, ‘what do you find interesting about this graph?’ Or, ‘what does it make you wonder about?’” he says. “The beauty of that is you don’t need any advanced mathematical background to answer a question like that."

Pokhrel decided he’d throw something similar at his students. One of the first was a , an intriguing data set that prompts important socio-economic questions about aging baby boomers and today’s declining birth rates. He then asked his students to write about what they found personally interesting in the graphs, and he was blown away by their responses. “It was just intended to be a five or 10-point discussion post assignment,” Pokhrel remembers. “But they’d write pages and pages and they were commenting on each other's posts.” He tried again with other themes — the relationship between GDP and poverty, criminal justice, basketball. Again, he got these highly engaged responses. “And that’s when I started thinking, what if I made this my whole class? Like, what if I made my course project-based, instead of ‘remember this math, and put this formula here’? I became convinced that I had to connect the material to things that felt relevant to them.”

He began writing projects for his students — a time consuming but enjoyable task — and ultimately reorganized his course around a series of 10 of them. Now, the majority of the actual work his students do in the intro course isn’t formula-based statistics. It’s writing, interpretation and analysis of huge data sets. “Rather than asking them to memorize how to calculate a standard deviation, the focus was, ‘what does that standard deviation mean?’” Pokhrel explains. “My feeling was that if this is the only statistics course they ever take, they likely aren’t going to remember all the math. But I can teach them how to conceptually interpret and analyze data, and that’s something they can use in their lives and future careers.”

Not that Pokhrel has totally abandoned the math part. Statistics is a very quantitative discipline, and even if his students don’t memorize every formula, it’s important for them to understand conceptually how statistical computations work. That’s why he still has one assignment per week that requires some math done by hand, the old-fashioned way. And they spend a lot of time learning modern statistics software, which lets the computers do the complicated computations, but teaches students how to organize and model data in ways that reveal what it all means.

This approach has worked exceedingly well in his intro courses, but Pokhrel has also been inspired to similarly reorganize his upper-level courses. At the beginning of the semester, he posts six to nine projects from a diverse range of subjects, and then lets the students choose one that they’ll work on the whole semester. Teams of three self-organize based on interest. And then they basically start digging into their data sets, with the projects broken down into two mid-semester “mini-projects” so he can check their progress. A final presentation and 15-page report, worth about 25 percent of their grade, replace the traditional final exam. Notably, in his upper-level courses, the students are doing a lot more computational and advanced software work. But as in his intro classes, the context is real-world problems and relevant applications of theoretical concepts. The result is his students are way more engaged.

In a necessity-is-the-mother-of-invention kind of way, Pokhrel says the pandemic definitely accelerated changes in his teaching approach that otherwise would have taken him years. And he personally enjoys this version of his course so much more, though he notes that compared to grading quizzes, it is more work for him to review individual, often lengthy, examples of student analysis. (He’s not going to argue, though, with the fact that his students are so engaged they’re actually “overdoing” assignments.) He also finds that, particularly in a totally asynchronous format, he has to watch out a bit more for students getting hung up on certain concepts. “If they get stuck on something, it can really set them back. It’s like a virus. It’s small, but it can have a big impact.” His remedy is to make sure they have ample access to him during office hours and quickie Zoom sessions, where he can usually help get a student over the hump in “just a couple minutes.” Then it’s back to the action of a dynamic statistics course that’s so much more fun than it was before.

###

Story by Lou Blouin. Are you a faculty member who also has a story about how you've grown your teaching style or redesigned a course during the pandemic? If so, we'd love to hear about it. Drop us a line at [email protected].