The Inner Workings of Updating a Lab Manual Chapter

Dr. Mikelle Nuwer’s favorite OOI Lab Manual Chapter is the one that focuses on primary production in the Atlantic Ocean. She prefers it over Lab 3, which focuses on the plate tectonics in the Pacific Northwest, even though it’s near where she teaches at the University of Washington. She also prefers it over Lab 5, which focuses on developing an understanding of density stratification, even though she co-authored that chapter. She even prefers it over Lab 8 with its murder ‘mystery’ about crabs off the Oregon coast.

For Nuwer, a Teaching Professor in the School of Oceanography at the University of Washington, Lab 7 is her favorite because it is “an amazing example of integrative oceanography, the interconnected relationships between various physical, chemical, biological, and geological processes in marine environments. You need to understand all of it to understand one part of it.”

So, when Nuwer was invited to participate in the Data Labs Design workshop last fall, she immediately volunteered to help update and expand her favorite lab. At the workshop, she recruited Kathy Qi (also at the University of Washington) and Dr. Amanda Kaltenberg (Savannah State University) to help her.

The University of Washington Data Labs team brainstorm the different audiences they teach at the Princeton 2024 Design workshop. (Left to right: Dr. Mikelle Nuwer, Kathy Qi, Atticus Carter)

One of the key changes they are making is adjusting the data range the lab focuses on. In the original version, developed when the OOI was only a few years old, only a limited period of reliable nitrate data was available. As such, students had difficulty verifying their predictions for the seasonal cycles of nitrate and chlorophyll. This incompleteness in the dataset was frustrating to her students, who often expressed dismay over the years at feeling like they were missing the complete picture. “Even though one of the things we want to teach them when they’re working with data is that there are gaps and that’s real,” she said ruefully.

The updated lab will also be expanded to include a Python component. “It’s an extension in terms of using Python and an extension in terms of understanding phytoplankton variability based on the factors that they explored in chapter seven,” Nuwer explained. Using preloaded Python data, students will be able to go beyond looking at the preselected year of data at one location, and instead they will be able to compare several years of data from multiple places.

A coding hand

Nuwer has a lot of pedological knowledge, which has been an asset to updating the lab activities, but she admits that her experience with coding is limited. Thankfully, she can lean on her fellow colleague, Kathy Qi, to help with the coding expansion.

Qi is a fourth-year graduate student in biological oceanography, currently researching cyanobacteria and primary productivity. She picked up coding while working on a lot of large datasets for her research, like those collected by an Imaging Flow Cytobot (IFCB), which was introduced to her, in part, by Data Labs team member Stace Beaulieu. The IFCB can generate thousands of images of phytoplankton from a single water sample, which can then be classified and identified for further analysis. She learned how to create semi-automated workflows with Python to help her navigate these large data sets and now she co-teaches a Python course at the University of Washington. It’s through this course that Qi also gained experience teaching, which she now relies on to work with Dr. Nuwer and Dr. Kaltenberg.

Kathy Qi shows off her creative interpretation of data literacy at the Princeton 2024 Design workshop.

Pulling the code and data together

At the workshop, Qi imagined creating Python packages that allow instructors and students to write 1-2 lines of code to get a neat dataset, which could bolster their confidence in coding and data literacy. On the back-end, Qi and the team would have already written the multiple lines of code needed to get that neat dataset. “The students don’t have to look at it if they don’t need to understand the process of loading and cleaning up data,” she explained. “You could just use the module and be like ‘oh, just run this line of code here.’ You get your data, put it in a nice graph and the notebook will spit out a figure for you.”

Figuring out the best way to approach this challenge was a key discussion during the Design workshop. Qi and the other Python aficionados in attendance brainstormed about the manual, step-by-step work it would take to get the data in the right shape for student analysis. “We talked about scaffolding it,” she recalled. “If an instructor teaching an introductory class is using it, maybe they would rely more heavily on the module versus an instructor who’s teaching an advanced coding based class. They might not even have to use the module at all.”

In the end, Nuwer asserted that while developing new activities, the team’s most important goal is to design tools that can be used by students and instructors at any coding skill level. “One of the things really important to us is that it’s so user friendly that even people like me can do it.”

The revised Primary Production Lab 7 and its related new coding activity will be published later in 2025.