BILD 5 SP25: Course Retrospective

And that’s a wrap on quarter 2! This quarter, I taught BIPN 100 (Human Physiology I) a second time, and I taught BILD 5 (Data Analysis & Design for Biologists) for the first time.

BILD 5 is a practical introduction to programming, statistics, and experimental design, with a focus on asking and answering biological research questions. My colleagues Liam Mueller and Keefe Reuther co-designed the course and have spent the past four years refining it, and I was immediately sold on their vision for the class. Instead of focusing on disciplinary content knowledge, BILD 5 instead emphasizes the practical skills for actually doing science. In many ways, BILD 5 is the course I wish I had taken as an undergraduate. Indeed, early on in grad school, I quickly realized that I didn’t have any of the statistics or programming skills required to analyze the increasingly enormous datasets I was collecting. My undergraduate statistics course had focused on pen-and-paper calculations with t-tables and F-tables, and I had no idea how to contextualize statistical tools to my experimental research questions. Furthermore, my undergraduate programming-for-engineers course was focused on mathematical modelling and differential equations, not data analysis or visualization. Therefore, BILD 5 represented an exciting opportunity to fill a hole I had experienced in my own undergraduate education.

This year also felt like a good time to join the teaching team – starting in Fall 2024, BILD 5 was newly instituted as a required course for all 1000+ incoming biology majors. To me, this felt like an exciting opportunity to innovate and make an impact on a (fairly) new course as it underwent significant growth, whilst also knowing that I wouldn’t ever have to reinvent the wheel, since the core structure of the class was already established by Keefe and Liam. After many months of anticipation (and also after sitting in on Liam’s BILD 5 class in Fall 2024), I finally got to teach my own section this quarter! Now that I’m on the other side, I’d sum up my experience as follows: Teaching BILD 5 challenged me more than any course I’ve taught before – but it was also one of the most rewarding experiences I’ve had.

First, the good. More so than any other class I’ve taught, in BILD 5 I got to witness the most incredible transformations in my students’ abilities and self-confidence. Take coding as an example. The vast majority of the class had zero programming experience. In addition, many students entered the classroom carrying negative preconceptions about their ability to learn programming – on the first day of class, 75% of my students either agreed or strongly agreed with the statement “I am anxious about learning to code”. Despite this, by the end of the quarter, my students all completed an independent Final Project, where they proposed a research question and analyzed their own simulated dataset addressing that question. As I was scrolling through their beautiful Final Projects containing code, visualizations, and analysis, it felt like a real pinch-me moment. Just ten weeks ago, we had dedicated an entire class session on simply opening RStudio and typing a single print statement. It was incredibly affirming when students would tell me how proud they felt after submitting the Final Project (despite how onerous it may have felt in the moment).

Next, the challenging. Even though BILD 5 is a lower-division course, I found the material to be significantly more difficult to teach than my upper-division physiology course. For starters, there is so much ground to cover, and the material is incredibly varied – we begin the course by talking about the nature of science, then we pivot into statistical concepts like parameter estimation and hypothesis testing, and we wrap up with experimental design. In particular, I found many of the statistical concepts extremely difficult to teach. Since I was an engineering major in college, I took a calculus-based statistics course that assumed maths prerequisites. In contrast, BILD 5 has no prerequisites and teaches the statistical concepts in a more intuitive, application-based way. I often found myself having to ‘unlearn’ the mathematically verbose way I first learned the material, and instead ask myself: how can I distil the conceptual essence of each topic for my BILD 5 students? After I remove much of the math, how do I make sure that the narrative flow still makes logical sense, and that I haven’t dumbed anything down? Overall, I’m pleased with how this went, although some of my lectures (e.g. Confidence Intervals) will need revisions because they still teetered towards being too mathematically involved. Even though this process was challenging, I do think it actually helped me to gain a deeper conceptual intuition about foundational concepts like p-values, confidence intervals, and hypothesis testing.

Another challenge was simply having the stamina and patience to help 180 students navigate learning to code. Especially on Mondays, I had six contact hours (2 lectures and 4 discussion sections), and by the time I was showing up to the 5pm BILD 5 section, I was really quite tired. Of course, I didn’t want my 5pm students to see a crankier, more irritable version of Ming than my 8am students, so I’d try to remind myself: “even if I’m hearing a question for the 50th time today, for this student, it’s their first time asking it”. That said, it still isn’t easy answering the same question over and over again, especially because I eventually start questioning whether the problem is actually me as the teacher.

A final challenge is that some students would often approach their BILD 5 assignments in unproductive ways. In lecture, I would often demonstrate how to code, and then I’d set a coding assignment where students would practice on related problems. For example, I might teach five dplyr functions in lecture, then set students an assignment where they had to use those same five functions in different ways. Early on in the quarter, I had many, many conversations with students who would tell me they were completely stuck with the coding assignment. I would then ask them what they had tried so far to get unstuck, and whether they’d looked back at the example code from lecture. And to my complete befuddlement, many of these students hadn’t ever gone back to look at the lecture code at all. I found this genuinely frustrating—it felt like I was watching students unintentionally sabotage their own efforts. At the same time, I can also completely empathize. When I’m faced with a new challenge, especially when my self-efficacy is low, even a small obstacle can make me want to give up. I tried to gently nudge each of these students towards looking back at the lecture code, framing it in two ways: (a) In any other biology class you take, when you get stuck in an assignment, your first instinct is probably to look back at the lecture and check you understand the underlying material, and (b) If you don’t know what tools you have at your disposal, you won’t ever know to use them, especially not in inventive new ways. I’m very pleased that as the quarter went on, my students eventually got the hang of their coding assignments and would use all the resources at their disposal, including AI tools which I would demonstrate in lecture.

Despite these challenges, I am most proud of how we sustained a positive classroom environment throughout. I had four absolutely incredible instructional assistants who paved the way in doing this. Every time I overheard them speaking with students, they did so with kindness and patience, always gently nudging students in the right direction as opposed to outright giving the answer. As a team, we all tried our best to be maximally available to the students and to create an affirming, positive space. In particular, we really tried to celebrate all the small victories – when students would show us code that they deemed completely broken, we would make a point to highlight all the great things that they had already written.

At the end of the day, BILD 5 was about more than just teaching coding and statistics. More importantly, I wanted my students to leave the quarter realizing they were capable of learning hard things—even things they once believed were beyond their reach. And if that’s their takeaway from our quarter together, then I think it was a job well done.

Syllabus: Link