Overview: To assemble a team that can both author a high-quality grant proposal and complete the outlined research, faculty often need to reach across institutions. Having researchers from multiple institutions brings horsepower, but it also brings administrative headaches. Knowing this, many funding agencies ask investigators to describe how the research team will operate during the program—how it will be managed and how members will communicate and share data. However, there is little space in proposals for detailed plans and issues that arise often do not appear in these plans. For this post I brought some colleagues together to highlight some best practices for running a multi-institution project. The overarching suggestion was that investigators should expect the program to evolve!
Contributors: Marian Kennedy, Lisa Benson, Courtney Faber and Rachel Kajfez
If you were recently awarded a research grant, congratulations! You are about to engage in some challenging research and advance the knowledge of your field. In the contemporary funding landscape, there is a high probability that this new grant has a team of investigators working together, which requires effective collaboration. Thinking through what practices you want to put into place at the beginning of the program will minimize the impact of the minor bumps, such as personality conflicts, and larger barriers, such as investigators transitioning off and onto the grant (a situation I’ve encountered more than once in my career). Here are some key suggestions. Knowing this, many funding agencies ask investigators to describe how the research team will operate during the program—how it will be managed and how members will communicate and share data. However, there is little space in proposals for detailed plans and issues that arise often do not appear in these plans. Below are our six best practices to implement as your multi-site grant gets funded.
(1) Reread the proposal and it discuss as a team as soon as you find out about funding.
It has probably been at least four months since you and your team cowrote the proposal, so some of the finer points of the proposal such as the team management will be a little foggy. With limits on proposal length, some details are also not specified, and each contributor might have had different ideas about what those details should be when writing the proposal. Rereading the proposal should prompt discussion about those details as well as reminding everyone of the details that were spelled out in it. The discussion can be informal over coffee or one of the investigators can volunteer to give an overview presentation to the group. In either case, make sure that everyone feels comfortable in asking questions.
(2) Carefully consider the communication tools your team will use.
With the many options for communication (electronic, phone or in person) the flow of information can almost be constant on a project. While that is wonderful in some ways, it can also limit the time researchers have to do their deep work. Therefore, we suggest that you strategize a communication method that prioritizes “thinking time” in addition to speed and face-to-face meetings. On one project, the team had open discussions about the tools (email, meetings, project management trackers, cloud documents for technical memos and research details, video chat), the purpose of each tool and the frequency of communication.
Here are some keys as you start:
- As a team, define what is most important about communication. Based on this discussion, identify, and invest in a project management software. Some software is great for quick communication (e.g., Slack), other software is great for organizing and communicating deadlines (e.g., Trello), and some platforms are multifunctional, providing quick communications, organization of deadlines, and file storage (e.g., Basecamp, Microsoft Teams). Most academic institutions do not have a site license for these types of programs. A really good comparison of possible options was “21 Best Project Management Tools For Research In 2022”.
- Clarify the responsibilities for each contributor and situations each one is responsible for “leading.” This can as a baton is passed between the collaborating faculty. One option might be to use a spreadsheet to create a table ofresponsibilities for each of the principal investigators, graduate researchers, and undergraduate researchers. This will also help students know who to ask when issues arise.
(3) Don’t meet just to meet.
Every meeting held has a cost, so you want to make sure that there is a purpose and outcome for any meeting.
We have found that a monthly “organizational and big picture” was always a key component of our programs. These meetings have every person on the project in attendance. When these meetings are held for multi-site projects, we suggest that each site gathered in a conference room or office and then used a single laptop to connect into the virtual meeting and these meetings included both administrative (hiring, scheduling, documenting, etc.) discussions and technical conversations (limits of theoretical frameworks, selection of items, etc.).
It is important to identify the agenda and take minute for each meeting. These documents can become extremely helpful when creating reports for funding agencies or revising why decisions were made. One option is for the meeting organizer to email the agenda to all team members prior to the meeting and at the start of the meeting to ask for any changes needed. As the meeting flows, the organizer can take minutes so that the decisions are clear and action items are always distributed to team members appropriately (email attachments and post in a shared “Meeting Agendas and Minutes” folder). Another option is to use a Google sheet to create meeting agendas for either a semester or year. Lisa has tried this previously and found it helpful to have all the milestones and deadlines identified. As the project progresses, the team can add or moving things around as needed. During the meeting, one person takes the meeting minutes and tags the action items within a Google doc. When any meeting begins, the organizer (typically the investigator hosting the meeting) pops links to these documents in the chat window. This process eliminates the need to send emails with minutes or agendas.
Maintaining the general management meeting once a month will allow for focused research-intensive meetings, site meetings or journal club meetings that occur on a more regular basis (such as biweekly). I had not utilized a journal club meeting for a program before it was suggested by Lisa. The journal clubs can be organized around a subject pinning the project such as “researcher identity” or “dislocation multiplication” and used to help all the PIs delve into both recently published articles in this area and foundational literature. We would suggest that every researcher on the project affiliate with one of the journal groups.
As we all know, meetings in person and virtually can be different. Here are some tips for making sure that the quality of the virtual meetings is high.
- Ensure that everyone has the necessary tools to keep sound quality high (headphones, high speed WIFI and a good microphone).
- Send out all agendas in the calendar invites and keep a record within a shared folder on the cloud.
- Have one of the principal investigators take notes from the meetings, email them out to all participants, and then archive these minutes in a centralized data folder on the cloud.
(4) Take time to discuss what requirements you have when selecting software and data storage.
Selecting the right software for collecting and analyzing data can save time, reduce costs, and increase collaboration between investigators at each site. As you wrote your grant, you probably identified the software that you wanted to utilize, but you should revisit the selection process after funding to ensure you consider options that have just been released or that have incorporated new updates. During those selection processes, you will balance a long list of factors such as features, cost, and ease of use. One that we initially did not consider was “potential capabilities”—that is, capabilities that might be needed. When working on an engineering education grant, we had identified types of software that we wanted to utilize for the quantitative data (online survey collection) and the qualitative data (recording of interviews, transcription, and coding). While we made a list of immediate requirements for the software that were needed during the proposal development phase, it would have been prudent to make a list of potential capabilities. In this instance, our team initially only considered our need for qualitative software for coding of themes from transcripts. It would have been helpful to consider additional challenges or opportunities during data collection which would have influenced our software decision. Had we planned, we would have chosen a software option that allowed us to track the emergence of themes based on demographics related to the participants (gender, institution type).
Finally, make sure to identify what Cloud system all members of the group can use. A single cloud system was needed to allow everyone to have access to data collection protocols, data analysis methodologies, and meeting minutes. Discussions around selection included allowing for personalized synchronizing (controlling data on PI laptops), cost, compatibility, institutional support, and internal review board approval. On one project, we ran across a problem that while each institution had free cloud-based storage for its employees, these did not allow for access from users outside the institution. Therefore, the investigators paid for a cloud-based system that allowed access for investigators from all the participating locations for shared documents and a second local system for the encrypted storage of identifiable data.
(5) Formally address authorship expectations rather than leaving it to naturally emerge.
With large teams, authorship of journal articles and conference proceedings can be difficult to determine since it is not always immediately clear which researchers made an intellectual contribution to the manuscript. Courtney Faber suggested we write a publication charter like her team had done as part of another multi-institution project (“Collaborative Research: Supporting Agency Among Early Career Engineering Education Faculty in Diverse Institutional Contexts” funded by the National Science Foundation). This document suggested that it was the responsibility for all project team members to mindfully opt in or opt out of a publication during the development stage and that the minimum work to be a coauthor would include both intellectual contribution and revising of the manuscript. We also found their suggestion to write out the responsibilities by author order helpful (first author is the person “in charge of submitting and leading writing,” second author could be “substantially involved in the analysis for a particular paper,” etc.). If you have not had experience with this, make sure to read Contributor Role Taxonomy (CRediT).
A topic that is often overlooked is how to handle authorship after a PI has left the project due to a transition to another institution and change of roles. This has occurred on multiple large group projects. We have all participated in those that teams have struggled with how to list authorship of that investigator when they had been critical to the proposal development (research question development and initial methods) but had left the project prior to significant data collection and subsequent analysis. We would suggest that you have an open conversation about this once a year on large projects. (In addition to faculty, undergraduate and graduate students also transitioned on and off larger projects, but in these situations, it is typically clearer how to handle the authorship of manuscripts.)
(6) Don’t be afraid to let the grant evolve from the original proposal when needed.
This is the most important thing! Many of us want to meet the objectives that was initially outlined in a proposal and have a hard time shifting when our analysis of data collected in an initial phase suggests that the one or more of the hypothesis or experimental plans. We also want the plan to go as written and can get thrown off if our expectations do not meet with reality.
- Change will happen. Take time to reflect on the changes in project facilitation, data management, etc. to make sure that the framework meets the needs of the team members every six months. Reflect on whether scope changes are necessary to meeting project objectives or are outside the scope of the current project.
- When a team member leaves the project, it is important to find someone who can contribute to the project, but it is not necessary (or possible!) to find a carbon copy of the researcher who moved on.
- The regularity of the meetings should change as needed; in our experience it averages out to about once a year for each of these three types. Reasons for these changes tend to align with changing phases in the project.
About the contributors: Dr. Lisa Benson is a full professor within the Clemson University Engineering and Science Education Department. Her research focuses on the interactions between student motivation and their learning experiences, and she is the immediate past editor for the Journal of Engineering Education. Dr. Courtney Faber is a Research Associate Professor within the Engineering Fundamentals Program at the University of Tennessee, Knoxville. She explores how fundamental beliefs about knowledge and knowing create barriers. Dr. Rachel Kajfez is an associate professor at the Ohio State University Department of Engineering Education. Her research interests focus on the intersection between motivation and identity of undergraduate and graduate students.