Adjust expectations, be flexible, support your groups
Research will be different from normal for a while and even in the best cases it will also be slower. The shift to working remotely will limit the kinds of work we can do and everyone doing research is experiencing a dramatic disturbance to their lives. This means the people in our labs will need flexibility and support.
Talk to your lab members to understand their needs: Recognize that these needs will be different for different people. Many will have new responsibilities and stresses that preclude working normally, but some may use work as a coping mechanism.
Make it clear that moving more slowly and delaying things is expected and 100% OK. Push back project timelines, understand that some folks will make little to no progress for a while, consider delaying stressful graduate activities like qualifying exams.
Provide financial reassurance: If true, ensure your team that their current funding won’t be cut. If possible, offer extensions on funding. This will help alleviate stress and uncertainty.
Recognize power dynamics when offering flexibility: Make sure that team members are comfortable opting-out of “optional” choices and don’t feel pressured to be productive, to work on-campus, or work at certain times.
Provide access to university resources: Inform group members about university programs related to mental health, expanded sick leave, and other forms of support for well being.
Give your team the resources they need to work from home: Encourage your group to move things from campus that they need to work at home including computers, books, and chairs. Universities typically allow this for remote work (there may be a form to fill out). If possible, purchase additional supplies needed for remote work (e.g., headsets).
Adopt & adapt tools and approaches for managing remote teams
Remote management recommendations focus on good communication, breaking projects into manageable pieces, keeping everyone on the same page with clear next steps, and tracking progress. This will make your group more efficient and inclusive when working remotely.
Use video conferencing to replace in-person interactions: Do this for any regular meetings you have (e.g., one-on-one meetings, lab group meetings) and also informal interactions (like popping into an advisors office or chatting science with labmates). Communicate your availability and how to set up meetings.
Use a group-based discussion tool (e.g., Slack or Microsoft Teams) : This supports asking questions and working on group projects and facilitates interactions among lab members with different work hours (important for those with responsibilities like child care). It can also provide an outlet for social interactions. Text does lose subtle social cues so video or audio is still best for delicate conversations. Check out the getting started documentation for Slack or Microsoft Teams.
Read up on managing remote teams: There are some unique skills to remote management, but there is lots of information on how to do this including: How to overcome your worries about letting people work remotely, How to oversee a remote team’s work, Ten simple rules for a successful remote postdoc.
Use project management and collaboration tools: These tools help you use good remote management practices. Most labs will benefit from a tool for writing and a tool for project management. Labs that write code (including for analyzing data) will also benefit from a code collaboration platform. Check out getting started guides for Google Docs (for writing) and Trello (for project management). Learning version control for managing code is a bigger commitment, but the Software Carpentry lessons are a good starting point.
Help identify research that can be done remotely, but understand the limits
It’s important to prioritize the safety of your team over research. This may mean changing your research plans to support social distancing and reduce or eliminate travel.
Focus on analyzing and writing up existing data and ideas: This is the easiest adjustment because it minimizes shifts in research area and need for new skills. Existing data isn’t just what a specific student has already collected, but can include previous data collection from your lab.
Synthesize existing knowledge: Writing reviews lets your team use their expertise to synthesize existing knowledge.
Conduct research on open data: There are increasingly large amounts of openly available data in many fields. There may be data that can be used to address questions similar to those you are studying using field or lab based approaches.
Collaborate when extending into new research areas: Computational research, working with large datasets, systematic reviews and metaanalyses all take expertise. To pivot into new methods or topics consider finding someone with the associated expertise to collaborate with. There may well be experts on your students’ committees or in your department or university.
Develop new skills/expertise to expand your groups’ research horizons: Instead of jumping into a new project requiring new skills support your team taking this time to learn new skills (e.g., computing methods or statistical approaches) or develop new expertise (reading up on new areas of the literature) to serve as the foundation for future research.
Initially prepared for the UF/IFAS Faculty Forum: Living, Working, and Adapting to the New Normal of COVID-19. Led by Ethan White (@ethanwhite) (who is responsible for anything bad) with contributions from SK Morgan Ernest (@skmorgane), Hao Ye (@Hao_and_Y), Brandon S. Cooper (@brandonscooper), JJ Emerson (@JJ_Emerson), Katy Huff (@katyhuff), Russell Neches (@ryneches), Auriel Fournier (@RallidaeRule), Jessica Burnett (@TrashBirdEcol), Melissa Rethlefsen (@mlrethlefsen), Eric Scott (@LeafyEricScott), Kathe Todd-Brown (@KatheMathBio), itati en casa (@itatiVCS), Alexey Shiklomanov (@ashiklom711) (who are responsible for anything awesome). A lot of the thinking in “Adopt & adapt tools and approaches for managing remote teams” was influenced by “Ten simple rules for a successful remote postdoc” by Kevin Burgio, Caitlin McDonough MacKenzie, Stephanie Borrelle, Morgan Ernest, Jacquelyn Gill, Kurt Ingeman, Amy Teffer, and me.
Zoom works great: I’ve seen up to ~50 folks attending the talk remotely and slides with video. Everything connection wise worked well except for a single committee member with some minor freezing during the private defense.
Have backup options: Give yourself time and backups in case things go wrong. Set up the connection early (15+ minutes) and ask the committee to show up early to check everything is working. Have one of more backups including a phone based conference call.
Call manager should not be the person defending: Have someone else, ideally a committee member, set up and manage the Zoom (or other system) call. This means that the student doesn’t need to deal with that on top of everything else and can focus on the defense.
Mute everyone for the presentation: Either ask all participants to mute themselves at the start or (better yet) have whoever is managing the call mute them all centrally. It’s easy for the audience to forget they aren’t muted and accidentally interrupt the presentation.
Audience, show you engagement: Leave your video on (unless bandwidth is an issue). If you’ve ever given a remote talk the lack of normal audience engagement is really challenging. A bunch of live video faces really helps. Also, consider exaggerating your positive responses. With lots of folks everyone is small so clear head nods, thumbs ups, and big smiles can all help mimic normal positive audience feedback.
Use a multi-monitor setup: Having two monitors will let you see folks attending the talk plus your slides and notes. Of course if it’s easier for you to not see the audience, then definitely take the opportunity of defending remotely to not have to see them. You’ll need to setup zoom to work with dual monitors for this to work properly https://support.zoom.us/hc/en-us/articles/201362583-Using-Dual-Monitors-with-the-Zoom-Desktop-Client Alternatively you can share your screen from one computer and join the call from another computer to see all the participants.
Manage bandwidth and adjust accordingly: Test streaming quality in advance in the same place and time of day you’ll be defending. If viewers notice bandwidth issues (blurry or dropping frames) try moving the laptop closer to the WiFi router or plugging directly into the router or a wired Ethernet port. If there are still bandwidth issues you may want to have the audience stop their video. Since the presenter often can’t tell if there are connection issues the person managing the call should either ask viewers to turn off their video via chat or turn off video centrally to avoid interrupting the presenter if possible.
Mimic “step out of the room”: The committee should have a plan for having the student “step out of the room. In Zoom this can be done by using a breakout room for the committee to talk and then return to the main room when done (which is made possible by having a committee member manage the Zoom call). You can also put the defending student “on hold” https://support.zoom.us/hc/en-us/articles/201362813-Attendee-On-Hold
Committee members should use video: Committee members should definitely use video if possible during the private portion of the defense. This is an inherently stressful activity and a lot of the usual positive encouraging social cues get lost with voice only communication. That said, if you’re freezing when asking questions it’s probably because of your local wireless/upload bandwidth and so you can probably help this by turning off your video so that you can communicate clearly.
Committee members should be kind and supportive: Frankly you should always be doing this, but it’s even more important now because everyone is under a ton of extra stress. This doesn’t mean you can’t probe the work, just do it in a positive way focused on helping the student. Also, consider minimizing required changes for the thesis. Most of us aren’t focusing well right now and revisions are often due on a tight timeline. Clearly distinguish recommendations for changes prior to submitting papers from changes required for the thesis.
Communicate excitement about a student passing clearly/effusively: This is a big deal even if you’re stressed and can’t celebrate it in the usual ways. Make really clear how big a deal this is to try to overcome the different feel of a video interaction vs. an in person congratulations.
Celebrate: New MS/PhDs – This may not be how you envisioned the conclusion of years work happening, but that doesn’t change that it’s a huge accomplishment. Celebrate in whatever (publicly responsible) way you can. One option is a video-based celebration. They’re surprisingly fun!
This post is based on a Twitter thread by Ethan White (https://twitter.com/ethanwhite/status/1240336385896316928) with ideas contributed to that thread by @echoechoR, @JosephLo16, @michaelhoffman, @kimpy79, and @ellelnutter. This document is released under the CC0 publication domain declaration (https://creativecommons.org/share-your-work/public-domain/cc0/) so that you can share and modify without restriction or need to provide credit. Thanks to @adanianscience for motivating me to put it up in formats useful for folks not on Twitter.
Understanding and managing forests is crucial to understanding and potentially mitigating the effects of climate change, invasive species, and shifting land use on natural systems and human society. However, collecting data on individual trees in the field is expensive and time consuming, which limits the scales at which this crucial data is collected. Remotely sensed imagery from satellites, airplanes, and drones provide the potential to observe ecosystems at much larger scales than is possible using field data collection methods alone.
We running the second in a series of data science competition where multiple groups attempt to use the same remote sensing data from low flying airplanes to infer the locations, sizes and species identities of millions of trees. This kind of collaborative data analysis challenge has proven highly effective in other fields for quickly improving methods for converting image data to useful information. This round of the competition focuses on exploring how methods generalize beyond a single forest.
There are two tasks in the current competition: 1) identifying individual trees in remote sensing images; and 2) classifying trees into species.
Teams (or individuals) can participate in either or both tasks. Task 1 requires working with remote sensing data (RGB, LIDAR, and Hyperspectral). Task 2 can either leverage this raw remote sensing data or use simplified tabular data provided by the organizers. Details of the different tasks will be available starting March 1st. To read more and sign up check out the competition website:
You can learn about the results of the first round of this competition in the summary paper and full PeerJ paper collection. We plan to write up the results of this round of the competition in a similar way, with a synthetic paper covering the competition, data, and comparison of different methods, and with each team given the opportunity to write up and publish associated short papers on the methods they used and results they produced.
This challenge is supported by the Gordon and Betty Moore Foundation’s Data-Driven Discovery Initiative through grant GBMF4563 and the National Science Foundation through grant DEB-1926542. It uses data from the National Ecological Observatory Network in addition to data collected by the organizers. It is being organized by the Weecology lab, Machine Learning and Sensing lab, Data Science Research lab, and Stephanie Bohlman and Aditya Singh‘s labs all at the University of Florida and the Environmental Observation & Informatics program at the Nelson Institute for Environmental Studies at the University of Wisconsin.
PH.D STUDENT POSITION IN COMMUNITY ECOLOGY
The Ernest Lab at the University of Florida has an opening for a Ph.D student interested in research in the area of community ecology, forecasting, and/or temporal dynamics to start fall 2020. The student will participate in the collection of small mammal and plant data at our long-term site in southeastern Arizona which will be used as part of our recently funded NSF grant to tackle challenges associated with ecological forecasting under novel conditions. While participation in the field data collection at the long-term site is expected, students in the Ernest Lab are free to develop their own research projects depending on their interests. The Ernest lab is interested in general questions about the processes that structure communities, with a particular focus on understanding when and how ecological communities change through time and how we can forecast those changes. Examples of research that students in the Ernest lab have pursued as part of their dissertation include: Does long-term change in communities occur through gradual species replacements or rapid reorganization events?, Are biodiversity patterns sensitive to changes in biotic interactions?, Do disturbances impact species populations and community-level properties similarly?, and How does the colonization of new species impact habitat patch preferences?
The Ernest Lab is part of the Weecology research group, Weecology is a partnership between the Ernest Lab, which tends to be more field and community ecology oriented and the White Lab, which tends to be more quantitatively and computationally oriented. The Weecology group supports and encourages students interested in a variety of career paths. Former weecologists are currently employed in the tech industry, with the National Ecological Observatory Network, at teaching-focused colleges, and as postdocs in major research groups. We are also committed to supporting and training a diverse scientific workforce. Current and former group members encompass a variety of racial and ethnic backgrounds from the U.S. and other countries, members of the LGBTQIA+ community, military veterans, people with disabilities, and students who are the first generation in their family to go to college. We work hard to create a supportive and inclusive lab environment and expect all members of Weecology to abide by the lab code of conduct.
More information about the Ernest lab and Weecology research group is available at: (https://weecology.org). You can also check out the blog for our long-term study (https://portalproject.wordpress.com) and our lab blog (https://jabberwocky.weecology.org; if you’re reading the ad here, you’ve found it already!).
Interested students should contact Dr. Morgan Ernest (firstname.lastname@example.org) by Oct 15th, 2019 to start a dialogue about the position and receive further information about the next steps in the application process.
It’s not uncommon to hear stories of mistakes resulting in graduates students missing paychecks. This is a major problem because most students live month-to-month and can’t wait for a missed check to be fixed in the next pay cycle. Despite the commonness and dramatic impact of missed pay in graduate school*, it’s common to see these issues written off as isolated incidents and not part of a more systematic problem. As a counterpoint to this idea, here are just some of the responses to a single tweet about this problem.
If you’d like to add your own experiences either post them to Twitter (responding to or quote tweeting the original tweet or tagging me @ethanwhite; I’ll keep updating this post as new experiences come in) or post a comment on this blog post.
*As a number of folks have noted this issue is also a real challenge for lots of folks both at universities (adjuncts, postdocs, and even professors) and beyond. I couldn’t agree more. That said I think there are some unique things related to graduate students that make the occurrence of these mistakes more common (like the fact that at many universities they are “hired” at least once and often 2-3 times/year; more on this soon) and I also think it’s OK to highlight and attempt to fix general issues within specific populations.
The weecology group is coming in force to the annual meeting of the Ecological Society of America which is being held in New Orleans next week. We’ve been up to a quite diversified list of things over the past year ranging from temporal dynamics of communities to forecasting and remote sensing. We also have people involved in a number of outreach or training events this year. We’re super proud of all the work everyone has been doing – both on the science end of things and the culture of science end of things. The Weecologists involved with each presentation/event are highlighted in electric blue with links to either their web page or google scholar profile so you can learn more about them. Talk or event titles link to the relevant listing in the ESA program.
10:15-11:30 Special Session: Path to being a strong ally in science. Location: Convention Center 238-239. (Juniper Simonis is involved in this session)
1:50-2:10 Weak impacts of climatic factors on intraspecific body size variation in endothermic species. Kristina Riemer, Narayani Barve, Brian Stucky, Stephen Mayor, Robert Guralnick, Ethan White. Location: Convention Center Rm 356.
3:20-3:40 The influence of data type, functional traits, and ecoregion on native bee phenology. Joan Meiners, Michael Orr, Kristina Riemer, Shawn Taylor, Terry Griswold. Location: Convention Center – R05
3:30-5:00 INSPIRE Session on Rapid Ecological Transitions: Synthesizing Concepts about Abrupt State Changes in Nature. Organized by SK Morgan Ernest, Christie Bahlai, Kendi Davies, and Sarah Elmendorf. Location: Convention Center – Rm 243
3:30 Rapid Ecological Transitions: A state space of concepts and ideas. SK Morgan Ernest. Location: Convention Center – Rm 243
4:40-5:00 Prediction and forecasting of Portal fauna via particle filtration. Juniper Simonis, Glenda Yenni, Shawn Taylor, Erica Christensen, Ellen Bledsoe, Ethan White, SK Morgan Ernest. Location: Convention Center Rm 355.
2:00-2:30 Evidence of rapid transitions in long-term community data. Erica Christensen, David Harris, Renata Diaz, SK Morgan Ernest. Location: Marriott – River Bend 1
5:00-6:00 during the poster session Juniper Simonis will hold an informal chat for those interested in knowing more about starting your own consulting business (likely in the poster hall, assuming there’s room). For more information, contact Juniper on Twitter @DapperStats
6:30-8:00 Long-term Studies and Paleoecology Mixer. Location: Rusty Nail Bar (SK Morgan Ernest is the chair of the Long-term Studies Section. If you like talking about temporal dynamics or want to learn more about paleoecology, come join us!).
11:30-12:30 during lunch Juniper Simonis will hold another informal chat for those interested in knowing more about starting their own consulting business (location not yet determined). For more information, contact Juniper on Twitter @DapperStats
3:30-5:00 Software skills for reproducible data-intensive research. Ethan White. Location: Convention Center Rm 243.
4:20-4:40 Scaling up remote sensing fundamental unit: from pixels to crowns. Sergio Marconi, Sarah Graves, Stephanie Bohlman, Jeremy Lichstein, Aditya Singh, Ethan White. Location: Convention Center 338
1:30-2:30 Getting a handle on your data with the dplyr R Package. Shawn Taylor . At the Data Help Desk in the Poster/Exhibit Hall.
4:30-6:30 Evaluating a near-term ecological forecast of plant phenology. Shawn Taylor, Ethan White. Location: Poster 110, Exhibit Hall
We are excited to announce the first release of a new Julia package that let’s you run our Data Retriever software with a native Julia interface.
For those of you not familiar with Julia it is a new programming language that is similar to R and Python, has a central focus on data analysis, and is designed from the ground up to be fast. It is an emerging scientific programming and data analysis language. Tim Poisot and his lab have been leaders in introducing Julia to the ecology community (thanks to them you can access GBIF data, analyze ecological networks, and more) and we’re pleased to start following his lead.
After installing the Python package getting your favorite dataset into Julia involves opening Julia and running:
julia> Pkg.add("Retriever") julia> using Retriever julia> iris_data = Retriever.install_csv("iris") julia> iris_data = readcsv("iris_Iris.csv")
Like the Python and R versions of the retriever the Julia version also lets you install into a number of different database management systems and formats to meet your needs including PostgreSQL, MySQL, SQLite, JSON, and XML. So if you need to install a large dataset and access if from the database you can do that:
julia> Pkg.add("SQLite") julia> using SQLite julia> Retriever.install_sqlite("breed-bird-survey", file="bbs.sqlite") julia> db = SQLite.DB("bbs.sqlite") julia> SQLite.query(db, "SELECT * FROM breed_bird_survey_counts LIMIT 10")
We use the PyCall package to directly run the Python code from the main retriever package. Cross-language support like this is really useful for letting difficult to develop core code be easily used in different languages and it’s great that this is a core feature of Julia.
This is our first Julia package and so there are sure to be lots of things to improve (starting with the documentation). If you use Julia, or are interested in experimenting with it, we’d love feedback, issues, and pull requests. We’re always enthusiastic to have new contributors and help everyone get started, especially if they’re just learning. For more information see:
Scaling-up ecological patterns and processes is crucial to understanding the effects of environmental change on natural systems and human society. We are piloting a Data Science Challenge where multiple groups attempt to use the same remote sensing data from low flying airplanes to infer the location and type of trees in forests. This will allow forests to be studied in detail at much larger scales than is currently possible. This kind of collaborative data analysis challenge has proven highly effective in other fields for quickly improving methods for converting image data to useful information.
There are three sets of tasks: 1) identifying individual trees in remote sensing images; 2) aligning ground data with remote sensing data; and 3) classifying trees into species.
Teams (or individuals) can participate in all of them or just pick the tasks they are most interested in. Tasks 2 and 3 can be accomplished using just tabular data. Task 1 requires working directly with spatial data. Details of the different tasks and links to the data are available at the challenge website:
We plan to write a general paper about the competition, the data, and the performance of the different methods used. Individual participants will be invited to write and publish associated short papers on the methods they used and results they produced. We already have a journal that has agreed to publishing all of these related contributions together into a collection (pending review of course).
The challenge is already open and the deadline for submissions is December 15th. Once you sign up on the website you will receive an email with some additional details. If you have any questions feel free to respond to that email or checkout the FAQ to see if they have already been answered.
This challenge is sponsored by the National Institute of Standards Technology as part of it’s Data Science Evaluation series and is also partially supported by he Gordon and Betty Moore Foundation’s Data-Driven Discovery Initiative through grant GBMF4563.It uses data from the National Ecological Observatory Network in addition to data collected by the organizers. It is being organized by the Data Science Research lab, the Weecology lab, and Stephanie Bohlman’s lab all at the University of Florida.
We are exited to announce a new release of the Data Retriever, our software for making it quick and easy to get clean, ready to analyze, data.
The Data Retriever, automates the downloading, cleaning, and installing of data into your choice of databases and flat file formats. Instead of hours tracking down the data on the web, downloading it, trying to import it, running into issues, fixing one problem, and then encountering the next, all you need to do is run a single command from the command line, R, or (now!!) Python:
$ retriever install csv iris
> portal_data <- rdataretriever::fetch('portal')
In : import retriever as rt In : rt.install_postgres('breed-bird-survey')
- Python interface: While the retriever is written in Python the package previously only had a command line interface. Now you can access the full power of the retriever from directly inside Python. See the full tutorial for more details.
- Conda packaging: The conda package manager has become one of the two main ways to install Python packages. You can now install the retriever using
conda install retriever -c conda-forge
- Command line autocomplete: As the number of datasets and backends supported by the retriever goes it can be difficult to remember specific names. Using Tab will now autocomplete retriever commands, backends, and dataset names. (Currently only available of OSX and Linux)
- We also made some changes to the metadata script system so if you’ve previously installed the retriever you should update your scripts using:
retriever reset scripts retriever update
Find out more
To find out more about the Data Retriever checkout the:
Ongoing work on the Data Retriever lead by Henry Senyondo is made possible by the generous support of the Gordon and Betty Moore Foundation’s Data Driven Discovery Initiative. This kind of active support for the development and maintenance of research oriented software makes sustainable software development at universities possible. Shivam Negi developed the Python interface as part of his Google Summer of Code project. You can read more about his time in GSOC at his blog.
Twelve different folks contributed code to this release. A big thanks to Henry Senyondo, Ethan White, Shivam Negi, Andrew Zhang, Kapil Kumar, Kunal Pal, Amritanshu Jain, Kevin Amipara, David LeBauer, Amritanshu Jain, Goel Akash, and
Parth-25m for making the retriever better.