Advising and Collaborating during a Pandemic and Sabbatical
The 2020–2021 academic year is shaping up to be drastically different from those that came before it, both in general and for me particularly. The aim of this blog post is to summarize where I am at and how I will be doing business for my collaborators and advisees during this time.
My Family’s Current Situation
This year will be my first sabbatical since starting my position at Columbia. We originally planned to spend a good chunk of our time in Italy, but circumstances redirected us instead to my home state of Vermont, a place whose exemplary handling of the pandemic has made for a relatively safe and comfortable environment in which to weather the storm.
Our family moved to Vermont in late march and, after a two-week quarantine in the basement, began living with my mother in Williston. We were very lucky to be able to rely on her help for about three days a week of childcare, allowing my wife and I to more-or-less continue to function in our jobs. Thanks mom! ❤️ But even with this help, working from someone else’s home with a four-year-old at around left our productivity significantly diminished.
Given our country’s largely failed response to the pandemic and the uncertain trajectory for the fall and winter, we have now decided to stay in Vermont through February 2021. We are renting a house in Burlington and, with some trepidation, putting our son in preschool three days a week. We will be Vermonters for a while. On a personal side, this is exciting to me. I’m eager to share all of the fun outdoor activities I grew up with (🚵♂️⛷🏕🏔) with my young son!
But I’m also nervous about how this new situation will affect my ongoing collaborations and particularly my relationships with my advisees. This concern is exacerbated by the fact that our new Climate Data Science lab, announced nearly a year ago, is finally spinning up with some exciting new hires!
Setting Priorities
Given this highly altered work situation, I think it’s important to enumerate my priorities clearly. Here they are, in order:
- My and my family’s physical and mental health. Fundamentally, we are in Vermont to stay safe from COVID-19. But the pandemic has also put mental and emotional strain on everyone. Working families with young children at home — like us — continue to face a unique set of challenges. (There is a non-zero chance we will have to pull our son from preschool if things get worse, in which case our work productivity will suffer a lot.) Beyond COVID, the past seven years on the tenure track have been emotionally exhausting for me and my family. I need to use my sabbatical to recharge, lower my stress, and seek a more sustainable relationship with my work. This means spending more time doing fun stuff with my family and less time working. I accept and embrace the fact that this will limit what I accomplish professionally during this time.
- Supporting my advisees. I am lucky the be a professional mentor to some truly awesome humans. These people — Ph.D. students, postdocs, software engineers, and research scientists — depend on my mentorship to advance their research and careers. Despite the unexpected circumstances, I will not let them down. I intend to continue regular meetings with my advisees and will remain fully involved intellectually with their projects. Doing this effectively in a remote-work context will require some trial and error, and I look forward to finding communication patterns that are effective and productive for each person’s unique personal style and needs. Furthermore, I hope that, as our group grows in number, we can build networks for research collaboration and mentorship that crisscross the group in different ways. (For example, research scientist Spencer Jones has already become a great mentor to Ph.D. students.)
- Making time for deep work. I have not written a first-author paper since 2018. Between advising, teaching, managing Pangeo, and constantly writing proposals, I am drifting away from what I love most about science — actually doing it! I have some exciting projects in mind that will hopefully help chart some new directions in the field of ocean mixing and transport. I really want to use my sabbatical to make some headway on these ideas. I also have some challenging software development projects in the pipeline. Making progress will require sustained periods of uninterrupted concentration, something that has been very hard to come by under normal circumstances. Hopefully the new location, the freedom from teaching, and some deliberate planning, will make this happen.
- Sustaining my collaborations. I consider myself extremely lucky to be a part of some exciting collaborative teams in oceanography and data science. These include Pangeo, the NSF-sponsored Poseidon Project, the NASA SWOT Science Team, and the NOAA/NSF Ocean Transport and Eddy Energy Climate Process Team. These collaborations allow me to learn new stuff from amazing colleagues, and to contribute to some ambitious large-scale scientific challenges. I will continue to participate enthusiastically in these collaborations during my sabbatical year. A particularly important goal for 2020 is to contribute to the launch of 2i2c, a new non-profit organization aimed at providing infrastructure and development for interactive cloud-based data science environments.
- Everything else. In order to prioritize the above items, I will have to let other things slide. One thing I will certainly be avoiding is writing any new proposals, since I am a good position with regard to funding at the moment. However, this category includes many important activities such as responding promptly and reliably to emails, reviewing papers, attending / presenting seminars, and generally being a busybody on Github / Twitter. If you see less of me in online forums, this is why. I apologize in advance to the folks I will let down because of these decisions.
Our Plan for Remote Work
I’ve done a lot of reading and research about how to help a remote team succeed. I’m particularly influenced by Google’s Remote Work Policy, which is driven by a huge amount of actual data. While not everything about this policy is applicable to academic work, it’s a great place to start.
The policy recognizes the importance of meetings for both the technical productivity and social well-being. On the other hand, having too many meetings can interfere with the need for continuous concentration / deep work. Meetings can also be difficult to schedule for teams spread across timezones. Below I outline our plan for both synchronous and asynchronous communication.
Meetings
We have traditionally held biweekly Monday morning group meetings jointly with Galen McKinley’s Ocean Carbon Group. These meetings focus on updates on research projects from group members. Given our focus on software , particularly with the CDS lab, we will be adding another biweekly group meeting focused primarily on the software development side. These will be the primary all-hands meetings for group members. In addition we will have several informal, socially focused check-ins throughout the week. The main one will be a Friday afternoon social hour, which is a chance to sum up the week’s progress and celebrate / complain as needed. On Tu/W/Th, we will also have a 15 minute “coffee break” style optional video check in, just to give everyone a chance to see a friendly face every day. These will likely be done jointly with the Pangeo coffee break, which Joe Hamman kickstarted back in March.
To summarize, the standing schedule will look like this:
- Monday, 10pm ET: All Hands Group Meeting. The meeting will alternate bi-weekly between science focused and software focused. The science-focused weeks will be joint with the Ocean Carbon Group.
- T/W/Th, 3pm ET: Optional Coffee Break. A chance to informally check in with colleagues.
- Friday, 4pm: Social Hour. Wrap up the week and get to know others better.
This is just a first draft — we will iterate and evolve this structure as needed.
Beyond this, I will continue to schedule 1:1 meetings with advisees as needed. The frequency of these meetings depends on the person and the project. However, I will strive to schedule all 1:1 meetings to Monday or Friday. This will leave T/W/Th largely meeting free, enabling me to do my much-needed deep work on those days. As always, group members or collaborators should feel free to book 1:1 meetings directly using my calendly.com page.
GitHub for Everything
Keeping track of many different projects and collaborations requires organization. This is especially important when team members are not regularly meeting face-to-face. I tend to struggle in this department. However, I have found that technology solutions can really help. In the past, I we used a mix of different tools, such as Nirvana for personal task tracking and Basecamp for group projects. In the end, I had tasks and todos spread among many different systems. 🤨 My resolution for this year is to track every aspect of my work in GitHub. Between Issues and Projects, I realized that GitHub is actually the most flexible and powerful project management system out there.
Since I’m already using GitHub for so many projects, it makes sense to just go all in. Here is our plan for how to use GitHub to track all aspects of our work.
- Software Projects. This is the standard use for GitHub and is pretty self-explanatory. We are already keeping all our formal software packages, such as Xgcm, in GitHub repos and using issues / PRs to track bugs and feature development.
- Science Research Projects. Every “project” will get a repo in the ocean-transport GitHub org. In this case, the main use of the repo is to use GitHub’s powerful issue tracker to keep tabs on the project progress. Each task in the project will get an issue. Of course, the repo can be used to store code, figures, text, etc., but this is optional and not necessarily appropriate for every project. Using GitHub in this way allows us to explicitly link scientific progress to software development needs, which is a central aim of the CDS Lab. Depending on the project status (published vs. unpublished), we can choose to make these repos private or public.
- Administrative Tasks. It would be wonderful if we could stop there, but unfortunately there is an ever-growing pile of administrative responsibilities that need doing. I have created a dedicated private repo where group members can add tasks to this queue in the form of issues. Examples include writing recommendation letters, signing forms, etc. No code will ever be added to this repo.
- Generic Inbox. I attempt to practice the “Getting Things Done” philosophy of time management. An important part of a GTD system is to have a place to quickly capture all ideas and requests that come across your desk. For this, I am now using a standalone private GitHub repo as my “inbox”. When it come times to process and plan, I review these issues, possibly editing, expanding, and transferring to a more appropriate repo / project. Like the administrative tasks repo, this one is completely code-free.
- Assigning Issues. Ok, so each issue represents a task that needs to be done…but these issues are now spread across dozens of GitHub repos and organizations!😱 Fortunately, GitHub has a fantastic feature in which issues can be assigned to one or more users. To see all of your work in one place, you just have to point your browser at https://github.com/issues/assigned. By using issue assignment effectively in this way, we can easily keep track of all our different tasks.
I am encouraging all group members to follow these same procedures as much as feasible.
Slack and Email
We have found Slack to be a very valuable tool for informal communication about research. It provides a place to chat casually, share random ideas, and generally reproduce some of the friendly, natural interaction that arises around the water cooler or at the lunch table, but in a remote way. Our group will continue to maintain a Slack for these purposes.
However, Slack also has some downsides for detailed technical discussions. Matt Rocklin summarizes these very well in this blog post.
The bottom line is that slack is great for casual chatting, but when it’s time to plan a project or share a specific results, this should happen in the GitHub project repo. (We may play around with GitHub / Slack integrations and see if they are helpful.)
If we follow the practices outlined above, we can basically avoid email as a communication medium within our group. That is indeed my goal. During my sabbatical, I really plan to minimize my email checking, as I find it to be a constant source of stress and distraction.
Summary
Thanks for reading through this long post. I hope it helps explain not only how we are going to be working for the next year, but also why. These practices are all evolving, and I’m looking forward to iterating on improving them throughout the year. Regardless of how the pandemic evolves, it looks like remote work will be with us for a long time. So let’s do it right! 💪
