Linda Douw’s work is interdisciplinary by design. She works with neuroscientists, physicists, clinicians, engineers and philosophers to try to understand the various ways brain tumors affect the brain—and how to stop them. As associate professor of anatomy and neurosciences at Amsterdam University Medical Center, her specialty is network theory, a mathematical framework for studying complex, interconnected systems. She uses this approach to integrate data from different modalities, ranging from molecular analyses of brain tissue samples to brain imaging data on brain structure, activity and connectivity.
A few years ago, Douw and graduate student Marloes Bet realized that the same techniques they applied to the brain could help them analyze interactions among team members—and provide insight into team dynamics in the process. Their analysis of 70 lab meetings revealed six stereotypical roles. The Transmitter talked to Douw and Bet about these roles and how they used that insight to modify their lab meetings and improve interdisciplinary communication.
This interview has been lightly edited for length and clarity.
The Transmitter: What is network neuroscience?
Linda Douw: Network neuroscience uses network theory, a field of maths and physics, to understand the brain as a network. Instead of focusing analysis on particular brain regions or using task-based functional MRI paradigms, network neuroscience typically uses resting-state measurements or structural measurements of the brain and views the whole as a network. It’s a quantitative way to deal with the fact that everything in the brain is interacting most of the time.
TT: What inspired you to turn a scientific eye to your lab meetings?
LD: I’ve been in the field of network neuroscience since 2006, and my team has been applying network theory to the brain since 2013. As the team got more familiar with network theory and the way it can help to better understand systems—either man-made or naturally occurring—we wondered if we could use this quantitative method to better understand our own team.
At the time, I was part of a clinical neuroscience team that was very interdisciplinary, with clinicians, psychologists and neuroscientists, as well as physicists, engineers and philosophers. We noticed that it was sometimes hard to combine all of these different ideas into a meaningful lab meeting. We realized this is similar to what happens in the brain: It integrates across different regions that do specialized things. We better understand what is happening in the brain through network theory. Can we better understand our own lab meetings through network theory as well?
TT: Why is this such a timely issue?
LD: The major neuroscience problems of our time span disciplines. That means that not everyone working on a project starts off with the same assumptions or the same way to communicate, but everyone needs to use their expertise and their disciplinary knowledge to contribute to the problem at hand. We need tangible ways to have interdisciplinary interactions. That’s where this research comes in.
TT: How did you apply network theory to lab meetings?
LD: We recorded all of our lab meetings for two years, and we coded which team members were taking turns after each other in the conversation. We used network theory to build quantitative networks of how people were speaking to each other.
Marloes Bet: We manually transcribed 11 of those meetings—this was before reliable automated tools were widely available to do this. And we did a conversation analysis on the transcripts to extract actions in an objective manner. Say Linda is giving a presentation, for example. She can stop her presentation and ask if it’s clear for everybody. That’s an action because she wants to check if other people are following. But what it objectively does is stop the conversation and allow other people to give her feedback on whether or not they understand. My action in response might be to summarize what Linda just said and check whether Linda and the rest of the group agrees with my interpretation. In our analysis, we selected only those actions that we thought were important for our aim of interdisciplinarity.
TT: Based on your analysis, you identified six different roles that people play during lab meetings. What are they?
MB: We have “the chair,” which is basically the team leader, the person who manages the agenda and turn-taking. They typically encourage people to speak out and also remind the team of the bigger picture. Another role is “the clarifier,” a person who asks open questions. They facilitate a natural checkpoint for other people to assess whether they understood something correctly. Number three is “the skeptic,” someone who asks critical questions and encourages critical thinking and deeper analysis. Then there is “the expert”; they are similar to the skeptic in that they have a lot of knowledge, but they typically use it to teach the rest of the group something. If there is confusion in the group, the expert would step in and explain. “The connector” invites somebody to speak up, like, ‘Hey, I know this is your expertise. Could you tell us something about it?’ They also do a lot of connecting on the information level, coming up with metaphors or parallels between different fields, for example. The last role is “the practical person,” a supportive team member who wants to help people out. This is actually really important for interdisciplinarity, because it contributes to a sense of psychological safety. It helps build trust in the team.
LD: If you want to understand a problem from multiple disciplines, you need all of these roles. Asking open questions is important to estimate the broad sense of the concept. Experts and skeptics can fine-tune what it is you’re trying to look at. The practical people really help to then take the next step and go from the conceptual level to actually extracting information that you can then use to answer your research questions.
TT: How did this analysis change how you run team meetings?
LD: We have integrated the findings into our team by making people aware of these six roles. We also take explicit turns in these roles. For a while, we picked randomly which role we would play in the lab meeting, so that people would get a little bit more comfortable with roles that weren’t super natural to them. But now we have people pick a role that they want to practice.
TT: Do people tend to take on one role or play different ones?
LD: One of the beautiful things about these roles is that anyone can do them. In our analysis, people would take different roles depending on the topic and the meeting. I think that’s important to know; people don’t have to be stuck in one specific type of role. If you’re a master’s student, you don’t have to only ask clarifying questions. That’s something that we now explicitly share with all the team members.
In academia, roles are sometimes very hierarchical, so a student might feel like they can only ask clarifying questions and never be an expert or skeptic, whereas principal investigators might be more naturally tending toward those roles. And I think rotating really helps the conversation, because you get viewpoints that you didn’t have before, which for interdisciplinary work is super useful.
TT: Has this active role-taking had a practical impact on your team and your science?
LD: There’s more input from various people, not just the same people with an expertise on the particular topic. By including the viewpoints of people who would not normally speak up, I think we do a much better job at defining the concepts that we want to study and connecting different perspectives. In our team, we do a lot of multiscale network neuroscience. So we connect how oligodendrocytes function to large, long-distance structural connections in the brain and, ultimately, cognition. And I think by rotating these roles, we can better understand each other’s frameworks and come up with better research questions and also experiments to answer them.
MB: Being aware of the roles and the positive impact they can have is important. For example, if there’s a skeptical person in your group, you can better appreciate the positive impact that they make. Whenever I present this project to other teams, they are really engaged, because it’s something that they all recognize and that they can all readily implement. They start to recognize, ‘Oh, I’m usually this kind of role, and actually I might benefit my team better if I sometimes try to be more of something else.’ It’s so intuitive. Everybody recognizes it; everybody can apply it. It makes communication much smoother.
TT: Do you have an example of an insight that came from people trying different roles?
LD: I’ve noticed that when you assign the skeptic role to people without deep expertise on a specific topic, they generally offer general feedback that is very core to the project. In science, we tend to zoom in. The typical discussion of results is on a very specific jargony type level, which I think is good, and we should keep it. However, if you also ask people with not so much knowledge about the techniques and the details to be critical and skeptical of a result, they come up with very different types of questions, like: Why would you do this? Or, how does this relate to something that has already been done? Is it really new? I think that’s something that with overspecialization and increasingly elegant techniques sometimes gets lost. People are so set on doing experiments that use all the new tools that we kind of forget to talk about the basics and the conceptual background and the type of questions that you can answer.
The other thing I’ve noticed in the team is that if you want to work with different disciplines, you do need everyone to latch on in some way to the topics you’re discussing. Before we implemented this, I would try to check; does everyone get it? But sometimes people are just so far out of their depth that they don’t speak up. Assigning this clarifying role and making people aware that you’re trying to establish a common base of knowledge for the whole team has really helped everyone get on board.
TT: For people who want to try to implement this approach, what do you recommend?
MB: For team leaders, I would suggest bringing some awareness of the different types of roles to your team members.
LD: The preprint includes a lot of examples of the actions that people can take.
