Structure and analysis cycles

[jeromelecoq]

Thanks all for preparing your analysis to present today.

I reflected on how to organize ourselves more effectively. I propose we follow the following ideas:

Written media is the best way to transfer information, allowing time for reflection and connection. Whenever possible, we should aim for deep interactions in written form.

First, we decide the agenda for the next meeting, then create the corresponding website (such as https://allenneuraldynamics.github.io/openscope-community-predictive-processing/meetings/2025-06-03/) and forum pages (such as #89). As people work toward results for our next meeting, they can post directly on the forum page, maximizing opportunities for deep interactions. Today, to help keep the forum organized, I improved the system to better link each meeting page with a single matching forum discussion page (see https://allenneuraldynamics.github.io/openscope-community-predictive-processing/github-actions-automation/#).

Verbal interactions are best for breadth, identifying problems, and aligning ourselves. During online presentations, we should limit ourselves to a few slides each. We should favor online meeting times for interactive discussions rather than direct information transfer.

After the online meeting ends, ​I propose we switch to a new forum page where we discuss the next meeting agenda. This way it will be easier to follow and keep structure.

[jeromelecoq]

In this scheme, we essentially design the next round of work at the end of an online meeting. The list of items is discussed there and we make a dedicated meeting page and forum page. Then as people work toward this goal, they write to the forum page with real-time results. Once we have the meeting, most people will know the general content and we can focus the online meeting to discuss points of interactions and reach agreements.
We should make sure we keep 5 min at the end of every online meeting to discuss the next meeting agenda.

[koweiss]

I agree with the above, and would add that sharing analysis before the meeting to allow sufficient time to review and think through before the meeting makes the discussions more fruitful and makes the best use of everyone’s time.
It’s also a good idea to focus on a small subset of open questions at a time in order not to diverge too much in any single meeting.

[lrudelt]

I think these are great suggestions!

After the discussion after the previous meeting, I was also wondering if it would make sense to open separate discussion threads for the three analysis directions that @jeromelecoq proposed in #79 . I think this is complementary to posting results, presentations etc in the threads for the weekly meetings, which could be the place where the different branches converge and interact. However, I think it could be difficult for starters to work their way through past meeting notes to find the relevant links and resources.

The discussion threads on the main analysis directions could have a structure where the research questions, hypotheses, TODOS and resources (such as analysis code) are summarized and updated in a top post. This could then be followed by in-depth discussions of analysis results.

If this is an idea that resonates with you, we could decide in the next meeting how to organize these threads and how to maintain the overview posts.

[maierav]

Great suggestions. I started new threads for each of the main analysis branches we have opened, so anyone interested can join them for more information and discussion.

The information on top provides a brief summary to the status quo as well as background information. They also provide the existing code to get started and build on. I also added links to the existing threads that were meant to coordinate and brainstorm, so they do not become obsolete.

I hope this will help de-clutter and provide a high-level overview for anyone new joining.

Please feel free to edit and update (e.g., add your own code, add analysis ideas, or take off what is of low priority or already had been accomplished).