Every semester I host a journal club, and sometimes I send some instructions beforehand to the students that are in charge of leading the discussion. I am trying to fight a pattern, and it is not really about the papers.
One of the most impactful experiences in my academic formation was a journal club hosted by Doris Chen at CeMM on bioinformatics papers. We had no slides, just literally the papers at hand and often a whiteboard. In the most intense moments we lost track of who was “presenting” because multiple people were giving their understanding of a method (e.g. SVD or PCA amiright) which you can understand from many different perspectives, or when we were collectively trying to figure out how the method worked, or what was the actual point (memory fails but I seem to remember some bayesian models in some 2010s papers from Oliver Stegle’s lab).
Nowadays (perhaps due to many things moving online during COVID?), journal clubs increasingly look like lectures: someone prepares slides, walks through the paper figure by figure, and the rest of the room passively listens until someone (often the PI) intervenes. I suspect part of this is just defaulting to the format students know best after spending years in lecture halls with mostly unidirectional flow of information (I studied mostly in Portugal where it was very much so - will certainly be different it other places).
The journal club I want to host (and attend if I “had more time”) is one where the hosts are more facilitators than presenters. The goal is not to “get through the paper”, but for everyone to flex the critical thinking muscle. Sometimes that means skipping figures. Sometimes that means spending twenty minutes on one panel because it turns out nobody is knows what the controls actually are. This is a feature, not a bug.
However, this takes deliberate effort from the hosts especially if the unidirectional mode is already established in the group. So I started being explicit about what I am hoping for, but also sharing some concrete examples on how achieve this rather than just saying “make it interactive”. Here are some of the things I suggest:
Ask the audience before revealing what the authors did. This is by far the most effective trick. E.g. after introducing the biological question and before showing how the authors did it, ask: “how would you have approached this? What experiments would you design?” Before showing a the figure ask: “what result would you expect? What would surprise you?” This forces everyone to commit to a prediction, which makes the actual result far more interesting whether it matches or not. It also may force some disagreement in the room which can tranform into organic discussion.
Stimulate critique explicitly. This may be the most common aspect of discussion in a journal club, and to some degree often happens: “What is the weakest aspect of this paper? Where would you push back if you were a reviewer?” But you can also use it go beyond and ask: “If this finding is true, what should also be true that the authors didn’t test?”
Check in with the group, often. E.g. after introducing the background and biological question, ask whether it is clear, whether anyone has comments, and what the impact would be if the authors address the question successfully; or after introducing a method or analytical step, ask whether it makes sense, and call on people working in the field: does this match your experience? These check-ins can help in two ways: making sure nobody is silently lost, and signal that participation is expected, not optional.
Prepare discussion points, but not a script. 2-3 discussion points that are actually discussed in the room are more valuable than polished critique. If it fits the structure of the paper, one can also break it in parts and have discussion in each section rather than at the end (when often by then the main points are clear). Also a very simple way to keep the discussino floor open.
None of this is revolutionary, but it’s important to say out loud. If you’re “attending” a journal club try to be conscious about the time you spend there. If you’re hosting, really make a deliberate effort to have everyone discovery and critique the paper together - ideally with a whiteboard ;)