May Institute 2020 finds a new kind of educational success using virtual format

July 28, 2020

By Hannah Bernstein

“Good morning, good afternoon, and good night.” That’s how Khoury College Associate Professor Olga Vitek opened every day of this year’s May Institute, which was broadcast over Zoom and YouTube livestream to hundreds of people across multiple time zones worldwide.

The Institute, which is designed to be a space for computer scientists and life scientists to come together to discuss data science, mass spectrometry, proteomics, and more, is normally held on Northeastern’s campus in May. But Vitek, who co-organized the event with research scientist Meena Choi, said the unexpected positives that came with the digital format made it all worth it.

Differentiating the two-week-long program from other online gatherings, Vitek said, “Conference organizers don’t teach like we do, so they didn’t benefit from that,” she explained. “We had to be reoriented so quickly, within that week [in March], we had to put everything online. We already had this experience.”

The Institute was one of the first large academic events to shift to the digital format rather than cancel. (ACM Collective Intelligence 2020, also held at Northeastern, was another.) Institute organizers returned registration and speaker fees, shortened each day’s material to avoid all-day Zoom meetings and accommodate different time zones, and reopened virtual registration for free.

“The quality of the video and the sound is much better than if you were sitting in the room with 300 people,” Vitek said. “People felt much more comfortable asking questions, because again, if you have 300 people, you have to walk up to a microphone and it’s very intimidating. But if it’s just a chat, you could ask questions anonymously.”

For past in-person Institutes, said Choi, there are about 180 registrations and 120 people who actually attend. But Choi’s analytics say there were 1,400 unique IDs registered online, and most Zoom sessions averaged at or above 300 attendees, with the largest reaching 670.

“One of our courses, we had almost 100 questions from participants, and we stayed within the time,” Choi said. “For things that involved some form of PowerPoint presentation or even a live demo, this was actually very effective.”

Innovations made large-scale learning effective

Beyond numbers, Vitek said one key aspect of the May Institute has always been the hands-on approach, where participants can use the software while the presenter is there to answer questions and work through problems. Replicating this on Zoom was challenging, but in one case, the presenter provided “homework” and simply remained on the call while more than 70 participants stayed and worked through the material.

The presenters found innovative ways to engage the audience through polls and interactive activities. For many of the technical talks on data analysis in R or software programs like Skyline, participants answered poll questions about their previous experience and knowledge. Brian Searle, a research fellow at the Institute for Systems Biology, was still answering participant questions on a Google doc more than three hours after his talk had ended, in a Q&A format that was seven pages long.

Along with Searle’s talk about data-independent acquisition, the two-week Institute featured presentations on a wide variety of topics. Scientific writing expert Alicia Williams led a workshop specifically focused on writing about proteomics, while software developer Brendan MacLean and computational biologist Lindsay Pino ran demonstrations of Skyline, a program used for large-scale protein analysis. Northeastern researchers also presented, including Vitek, Choi, lecturer and statistician Kylie Bemis, and computational biology PhD student Ting Huang.

“The event is structured such that the first week has more focus on biology and technology, and the second week has more focus on data science,” Vitek says.

Institute brings together two fields and diverse researchers

Choi highlighted that the purpose of the May Institute is to bring together people working in this very specific overlap between data science and life science. Younger participants can benefit from meeting and interacting with established researchers in the field, and the established researchers can mentor up-and-coming scientists and potentially find new collaborators. Although it felt different online, Choi and Vitek did what they could to retain that collaborative feeling.

“The May Institute is very unique because you have people who are doing very similar things, and these things are very interdisciplinary so it’s difficult to find people doing them within any given college or department,” Choi said.

As part of that goal, Vitek said they held a “future developers meeting” session to allow junior scientists a chance to present their research and get feedback from the Institute’s speakers. It was one of the best events of the Institute, and Choi said she even ended up in a long email chain with one of the speakers to provide feedback and give ideas.

Many researchers were also cut off from their wet lab experiments because of campus closures, sometimes losing all of their work. As a result, Vitek said the May Institute came at a really good time — many of the speakers who were able to participate virtually were more technically trained, and could teach lab scientists about the data analysis they could do from home.

“This was really good timing because people were actually looking for [computational training],” Vitek said. “Many were saying they’d always wanted to learn about this and they didn’t have time. Now they had an opportunity.”