HomeNews and blogs hub

Logistics of a hybrid Carpentries workshop

Bookmark this page Bookmarked

Logistics of a hybrid Carpentries workshop

Author(s)
Eli Chadwick

Eli Chadwick

SSI fellow

James Acris

Patrick Austin

Ivan Finch

Kyle Pidgeon

Posted on 17 May 2023

Estimated read time: 8 min
Sections in this article
Share on blog/article:
Twitter LinkedIn

Logistics of a hybrid Carpentries workshop

Posted by d.barclay on 17 May 2023 - 3:00pm Man taking notes from a computerImage by StockSnap from Pixabay 

By James Acris, Patrick Austin, Eli Chadwick, Ivan Finch, and Kyle Pidgeon.

This is part 2 of a 2 parts blog post. The first part can be found here.

Introduction

With the increase in “hybrid working” (part-time in the office, part-time from home) in recent years, the need for hybrid events – meetings, conferences, workshops – has become greater. This demand extends beyond COVID-related necessity; many workplaces have moved to a permanent hybrid working model, and remote access is also a huge benefit for those who have always struggled to access physical workshop spaces – including many disabled colleagues and physically distant collaborators.

STFC takes in a cohort of graduates, apprentices, and industrial placement students each autumn, with many of these taking up computing roles. Annual Software Carpentry workshops help these new starters fill in knowledge gaps. In 2020, a Unix and Git workshop was run remotely by other STFC colleagues. In 2021, during an STFC hybrid-working trial, we decided to experiment with a hybrid format for the course. In 2022, building on successes and lessons from 2021, we expanded our offering to include the Python and Intermediate Research Software Development courses, all in a hybrid format.

This blog post describes the logistics of these hybrid workshops in detail, as a reference for anyone who intends to organise one themselves. Our views on the experience of actually running the workshops are covered in the previous blog post.

The Workshop Format

The 2022 workshop series ran on the following dates:

  • 4-5 Oct: The Unix Shell & Version Control with Git
  • 11-12 Oct: Programming in Python
  • 1-2 & 7-9 Nov: Intermediate Research Software Development in Python (with VS Code instead of Pycharm)

All workshops ran from 9:00-13:00 each day. In 2021 we ran only Unix & Git, but in the same 2-half-day format. We ran half-days to reduce Zoom fatigue for online participants.

In 2021, we had 52 attendees – 16 in person and 36 online. There were 3 instructors (2 of which were co-organisers) and 14 additional helpers.

In 2022, each workshop had 36-39 attendees, split about 50/50 between in-person and online. In-person attendees were further split into two physical rooms with 8-12 attendees each, and one of these rooms had pre-configured computers for attendees who could not bring their own. There were 6 instructors in total (5 of which were also the co-organisers) and a further 13 helpers. Instructors/helpers were not expected to attend all 9 workshop dates.

Teaching

The teaching was done through a central Zoom room to which all attendees were connected simultaneously. Some instructors presented remotely, and others were in the two physical rooms, but teaching was always broadcast to all attendees. Questions and answers were also broadcast to all attendees, regardless of whether the question came through Zoom chat or from an in-person attendee.

Each day, one online helper acted as a Zoom host, who managed breakout rooms, communicated with helpers, and supported the instructor. They also kept an eye on the Zoom chat for questions and added useful content to the Etherpad (see Tools section).

Online attendees were allocated to 6-person breakout rooms for exercises, each of which had one online helper assigned. The Zoom host acted as a reserve if any room needed extra support.

The in-person attendees were evenly spread across 2 physical rooms, and each physical room had 2-3 helpers. These helpers were responsible for connecting to the central Zoom call and making sure any questions raised in their room during teaching were passed on to the instructor. This arrangement conveniently side-stepped the common issue of how to run small-group conversations in a hybrid format.

Organisation

The workshop format required a lot of consideration and planning to ensure everyone had a good experience. None of the organisers had prior experience organising in-person or online workshops before 2021, but we believe the hybrid format requires more work than either. With two sets of logistics to consider, it was helpful to have some organisers focused on physical logistics and others focused on online logistics, according to how often they worked on-site/remotely.

In 2021, given our lack of experience, it was also helpful to work through the CPACC’s PACT framework for event planning, considering carefully the differences between online and in-person participation and how to bridge between the two.

On registration, attendees specified whether they wanted to attend online or in person. In both years, we estimated the in-person/online split based on the first 20 or so registrations, and the final ratio ended up close to this.

Inevitably, as other commitments come into play, some attendees need to swap formats before the workshop starts. We considered setting a deadline of a week before the workshop for any changes - but a real bonus of the hybrid format is its flexibility to accommodate emergency changes in the final days/hours before the workshop starts. While a couple of people switched formats around a week before the workshop, many more changes came on the day due to sickness, childcare, or travel problems. These emergency changes typically involved a shift from in-person to online attendance, while advance changes could go either way. We were able to accommodate these last-minute changes, but this could pose challenges for organisers/helpers with less flexibility.

Helpers

Our helper-to-attendee ratio was about 1:3 in 2021 and 1:4 in 2022 (including instructors as helpers). This is more than would be necessary for either an in-person or an online workshop with the same number of attendees - The Carpentries recommend around 1:5 for online and 1:8 for in-person. The smaller ratio in our case reflects the extra bridging work involved and the desire to have at least 2 helpers in any physical room. In 2021, we over-recruited and ended up with some redundancy, while in 2022, the balance was about right.

We recruited helpers from among our colleagues on a volunteer basis. On sign-up, we asked them for their availability online and in person for each workshop date, plus the maximum number of days they were willing to cover (e.g. maximum 4 half-days).

With 9 workshop days and helpers limited to around 5 of those on average, it was a real challenge to recruit enough helpers to make sure all rooms were covered on all 9 days. This is something we’ll have to bear in mind for future years.

Tools

We used a few tools to facilitate interaction:

  • Zoom brought everybody together for teaching. We also recorded the whole workshop and used automated captioning. Recordings (alongside slides and any other material) were distributed to attendees following the workshop.
  • The Carpentries traditionally uses red and green sticky notes for attendees to show their progress in a physical room – we retained this for in-person attendees and used the Yes/No reactions in Zoom for online attendees (though usage was rare).
  • Slido was useful for “show of hands” type questions – we utilised the multiple-choice questions from the Software Carpentry material to check understanding. People could interact with Slido the same way whether in-person or online. We found Slido less useful for the IRSD course, where the exercises were more complex, and also for questions which required participants to submit code as an answer.
  • An Etherpad document was used for pasting notes, exercises, and links during teaching, and by attendee demand it also morphed into a copy-paste feed following along with the instructor. We also gathered daily feedback through the Etherpad during the workshop, as an alternative to writing on sticky notes which only works for in-person workshops.
  • Slack was used for back-channel communication between helpers/instructors – this was essential when trying to coordinate the different rooms. This worked well for the online components but in-person helpers found it difficult to monitor the Slack channel at the same time as helping attendees.

The workshops ran smoothly using these tools, despite technical issues on occasion.

For future runs of the IRSD course, there are tools that might be worth looking into which would allow participants to directly follow along with the instructor in an IDE, such as JetBrains’ Code With Me, or VSCode’s Live Share extension. These have their drawbacks though (e.g. connection issues, attendees not writing their own code), and would likely involve an overhaul of the way the workshop is delivered.

Conclusion

  • Distribute organising responsibilities among a few people - a minimum of 3, with dedicated focus on online, in person, and bridging components respectively
  • Use technologies that can be accessed by all attendees regardless of how they are attending (e.g. Slido, Etherpad)
  • Allow people to switch between online and in-person attendance close to the workshop, but set a deadline for this if you aren’t last-minute flexible with room capacity/catering

 

Share on blog/article:
Twitter LinkedIn