[Event]: Community Meeting Feb 2024

[jmanitz]

Event Summary

What does the term “validation” mean in different contexts?

Target Date

Feb 20

Tasks

Event Preparation Timeline

By target due date

28 Days Prior

• [x] [Planning] Title & abstract prepared
• [x] [Marketing] Pass along title and abstract info to R Consortium marketing team to advertise
• [x] [Marketing] Send out "Mark your calendars" email

7 Days Prior

• [x] [Planning] Event materials prepared
• [x] [Planning] Event trial run (as needed)
• [x] [Marketing] Send out reminder email

Event scheduling kickoff

• [ ] Create milestone for event
• [ ] Convert preparation checklist to issues
  • [ ] Assign all issues to milestone
• [ ] Assign appropriate target dates for all milestone issues

[jmanitz]

What does the term “validation” mean in different contexts?

• How do you define "validation" in context of your work/role?
  • How do you make sure package outputs are reproducible across systems and versions?
  • Why validation is important for health autorities?

Do we want to have slides?

• start with open with poll questions and have a discussions, then come back with pitch slides with working definition of "validation" in white paper/webpage
• have a lead in every breakroom with each 6-8 people, use tools to document conversation
• consider 10mins breakout rooms.

[antalmartinecz]

Title: Unraveling the term “Validation”: Join the Discussion at the R Validation Hub Community Meeting

Main text: 📣 Dive into the world of validation with us at the first R Validation Hub community meeting of the year! 🚀 Let's explore the diverse perspectives within our community on what defines a validated R package. Is it ensuring reproducibility across systems? Prioritizing bug-free and well-maintained packages? We want to hear YOUR take!

🗓️ Save the date for our February community meeting (February 20, 12:00 EST) and join the conversation. Let's collectively shape the understanding of validation in the R ecosystem! 🌐💬

@jmanitz @anujadas185 @jcabercrombie - What do you think? If it is okay like this (probably will add the exact date as well) we can send out the emails for the meeting

[anujadas185]

Looks good to me!

[jcabercrombie]

I agree with Anuja. I like this description, as it's to the point and even gives some sample opinions/definitions of validation like we mentioned. I also feel it has good enthusiasm to it when read and also reminds me of the good LinkedIn posts people interact with with emojies too. Nice touches!

[jmanitz]

Thanks for putting this together. Looks great. I might suggest to edit the title "Unraveling the Term 'Validation'" to avoid confusion

[antalmartinecz]

I have to admit, chatgpt helped a lot. I wrote a more basic abstract then just asked it to make it better for social media posts

[antalmartinecz]

@dgkf What email list do I use to send the "Send out "Mark your calendars"?

[dgkf]

Hey @antalmartinecz - the mailing list is managed at https://lists.r-consortium.org/g/RConsortium-Validation-Hub

I don't see you as a subscriber yet. Once you subscribe I can bump you to moderator so that you can send out events to the mailing list as well. Juliane is already a moderator, but it might help to have a bit of redundancy.

edit: And specifically, when sending invites I attach the invite.ics file (any calendar app will have this file for an event) to the topic. The R Consortium list serve tool will automatically identify it and add it to the group calendar.

[jmanitz]

@AARON-CLARK @dgkf @joseph-rickert @emilliman5 @Crosita Please share

The blog is posted here, and below are the links to the social media posts for you all to like and share.

Twitter https://twitter.com/RConsortium/status/1757067428125610158

LinkedIn https://www.linkedin.com/feed/update/urn:li:activity:7162833768989274112

Mastodon https://fosstodon.org/@RConsortium/111919288570172303

[antalmartinecz]

RvalHub_2024Feb_Community_meeting.pptx

@jmanitz @jcabercrombie @anujadas185 This is the current iteration of the slides, please feel free to add to/change it (stock photos from PowerPoint insert -> picture -> stock can make slides look nice for these kind of meetings). I will try to revise tomorrow, and day after

[jcabercrombie]

Hi @antalmartinecz! I looked over the slides and they seem good to me so far.

1. Did we have an approach for how we divide up people into their respective breakout rooms? I know we discussed having each room be dedicated to a perspective (health authority vs. individual contributor, etc.) or we can do an arbitrary split and have each perspective is addressed in each room.
2. I know we discussed providing some pre-formed opinions to folks who are running the breakout rooms in case we have quiet/shy attendees. Are these going to be along the lines of the questions on slide 3?
3. I was also curious about our allotted time. As we have a full hour for the meeting, does 10 minutes seem enough to discuss among 6-8 people and there be back-and-forth discussion? I wonder if something like 20 is preferable and that when we come together as a group again that we just

[antalmartinecz]

Hi @jcabercrombie,

1) I think having a random allocation is the easiest. It gives us the best chances of people ending up in a conversation they can contribute to. Given that we don’t really know the distribution of the target audience beforehand it’s hard to know what rooms we would need, how many and who to put in what room. Just randomly allocating 6-8 people per room would give us the best chance of 1-2 of them being extroverts and push a conversation forward since they have a freedom to choose a subtopic. 2.) The way I was thinking about it is that we could talk a minute or two about validation in general in advance to give some initial ideas on it. Partially because giving 3-4 separate and distinct preformed opinions on how to define validation is actually quite difficult. Having less than that probably wouldn’t help too much. 3.) I think 10 minutes is kind of a good length to avoid the potential for rooms with too long awkward silences. We could do however 2 rounds of this instead. That’s also a benefit of random room allocations - we can repeat it as many times as we want.

[jcabercrombie]

Got it, @antalmartinecz! Thanks for the clarifications. Random allocation, having something pre-planned to speak on validation, and keeping it to 10 minutes and re-randomize as we see fit all sound great to me!

[jmanitz]

@antalmartinecz @jcabercrombie I made some edits adapted from previous presentations and moved the slides into the correct location:

https://github.com/pharmaR/events/tree/main/community_meetings/2024-02-20

[jcabercrombie]

Thank you for the edits, @jmanitz!

[jmanitz]

For the blog post, a quick draft for a summary of the meeting:

• The community meeting had about 50 participants unraveling of the term “validation” in breakout rooms
• Different perspectives are have to be considered, including user, admin, package developer, and regulatory agency
• It is important that each organization is deciding on their own definition of validation, but the R validation hub can help to create a baseline
• It is crucial to understand that exact matches of results between programming languages may be hard to achieve due to (justified) differences in statistical methodology
• In the spirit of open source software, it was emphasized how important it is to contribute testing code back to the package developers
• The slides are available on github: https://github.com/pharmaR/events/tree/main/community_meetings/2024-02-20
• Unfortunately we forgot to record the meeting, but maybe we can advertise to sign up for the mailing list (https://lists.r-consortium.org/g/RConsortium-Validation-Hub/) to%20to) receive invites to the next community meetings on May 21, Aug 20, and Nov 19 (dates to be confirmed)