Grants:Conference/Celtic Knot Wikimedia Language Conference 2022/Report

Report accepted

This report for a Conference Grant approved in FY 2021-22 has been reviewed and accepted by the Wikimedia Foundation.

• To read the approved grant submission describing the plan for this project, please visit Grants:Conference/Celtic Knot Wikimedia Language Conference 2022.
• You may still comment on this report on its discussion page, or visit the discussion page to read the discussion about this report.
• You are welcome to Email conferencegrants at wikimedia dot org at any time if you have questions or concerns about this report.

Goals

Target Outcome	Achieved Outcome	Explanation
“We are hoping to have on average 10 active participants and 30 watchers in each of the 6 workshops. Overall, we hope to be able to reach out to 100 different people over the 2 days of the online event.”	Each of the 6 workshops had at least 30 participants, with the highest reaching 46. The survey responses provided a slightly different insight. The lowest number of self-reported participants was 42, and a high of 60. The videos of the workshops have been viewed 165 times on youtube after publishing	As explained further down we didn’t use the active participants / passive watchers distinction as strictly as planned; instead most people were invited to participate actively. In this sense we believe the outcome was met, with higher than planned active participation. The first figures reflect the highest number of participants in a workshop at any one time and the actual number of participants would be higher as some participants would have joined at different times, while others would have left midway through workshops. This is demonstrated through the survey responses. It is possible the survey answers may include some people who were unable to join synchronously.
“The number of participants for onsite gatherings will depend on local regulations for onsite gatherings in various countries - at that point, we can guess that we will integrate 5-6 local gatherings to the conference, involving 50 people in total.”	During the conference, there were live watch alongs in satellite locations. There were also in-person gatherings after the event, to discuss the themes of the conference.
“Our intentions for this event are to provide to participants: High quality sessions that can enable them to unlock new skills and apply them directly on their projects; Accessible content that can later be translated, reused and shared with other language communities to enable further capacity building; A spirit of connection, mutual support and exchange among participants who share similar goals and issues.”	Across six points relating to the quality of the conference, attendees gave the Celtic Knot a score of 4.27 out of 5. This encompassed the conference environment, accessibility, being able to participate, the content, and overall satisfaction. 71 out of 80 respondents (89%) reported that they left with new or different ideas about how they may work in the future. 44 out of 73 respondents (60%) reported that they talked to others about new ways of collaborating	One of the key aims of the conference was to improve people’s skills and help their community. The aim was to do this through a workshop format. That a high proportion of people left with new ideas about work, indicates we were successful in helping them develop skills and apply them to their work. In particular, the accessibility of the conference content was rated 4.15 out of 5. The conference sessions have been recorded and are available on YouTube and Wikimedia Commons. The most popular session, on Wikidata, is being given subtitles to make the video more accessible and facilitate translation by the community. When responding to the question about collaboration, the answers generally highlighted the benefits of talking to members of the Wikimedia community and sharing their experience, and generally providing a space for communities to talk to each other.

This is the video of the final session of Celtic Knot where we reflect on our highlights as organisers

Next steps

Priorities	Activities during the conference	2 months after the conference
Continuation of discussions - Satellite events	Before the conference we designed the programme to allow for pre/post conference in-person meetups. Support grants were advertised  and satellites chosen before the main conference. At the conference we promoted the satellites to boost their profile. Live watch along events were held for the Gungbe and Yoruba Wikipedia communities. Bringing the Gun Wikipedian community together offered the chance to build connections and led to the establishment of the Gun Wikimedians Working Group.	After the conference, various communities ran satellite events to continue connections established during the Celtic Knot. In July, Wikimedians in Kumasi, Ghana met to add Wikidata Lexemes in the Twi language, resulting in more than 3,000 edits. The following month, the Igbo Wikimedians User Group held an event to translate Wikidata properties in their language, and followed it up with an event about databoxes.
Capacity building of workshop leaders	We invested in the skills development of conference workshop leaders. Before the conference we run a series of prepping workshops (described below)	A month after the conference we ran feedback sessions, internally with the organiser team and optionally with an experienced trainer, to reflect on skills gained and new approaches tested. This should help in embedding the skills learnt, and could be applied at Wikimania if the leaders run sessions there.
Organisers reflect and plan on 2023 event	As the organiser team we did continuous evaluation as the event progressed, capturing things going well, and aspects to change.	Working on this evaluation report gave us further structure to reflect. Discussions on 2023 started.
Building wikis in other languages	During the Databox workshop on day 2 of the conference, Databox was installed on 6 new wikis! The list of wikis with Databox installed as a result of the Celtic Knot is documented on Meta-Wiki.	“In fact, a lot of participants reached out even on their homepage design, infobox et al. An example is the ig.wikipedia.org it has seen a facelift after the conference when they saw the dag.wikipedia.org and reached out for guidance”
Strategic Discussions: Were there problems or topics that your community focused on at the conference?	We built a session for open discussion into the conference programme. It took place on the first day of the conference, and participants suggested topics for discussion and then broke off into groups. These discussions explored the use of the Incubator; moving projects out of the Incubator; the role of Wikimedia in education and approaches to helping students and teachers learn about it; sharing technical tips; making global goals specific to local communities and their wikis.	The post-conference survey indicated an appetite for people to discuss with their communities what they had learned. There was further discussion on Telegram, and engagement with Wikidata aspects including people creating databoxes on their wikis. As well as taking place in naturally arising settings, this was supported through the event grant scheme we ran, which was designed to encourage people to meet and share their thoughts and experiences.
Capacity Building: Were there any important skills that your community wanted to learn at this conference?	The conference programme included four workshops to improve participant skills: Support your wiki's community to grow How to organize or join an editing campaign Set up your first Wikidata-powered infobox on your wiki with Databox Let’s translate Wikidata properties in our languages!	These sessions had a turnout of between 42 and 60 each, based on responses to the post-event survey; the actual number of attendees is likely to have been a bit higher. In the post-event survey, many participants indicated that they intended to Wikidata by translating labels into their language, and would be using Databox on their wikis. Others indicated that they wanted to train others to edit Wikipedia, and organise editathons with specific themes. One respondent also indicated that they wanted to organise a photowalk and editing competitions, relating to the community focused workshops from the conference’s first day. Encouragingly, the interest in Wikidata seems to be linked to community growth, with people wanting to pass on what they have learned and involve others. In September, the Ghanaian Pidgin Wikimedians Community organised an online workshop supporting new and experienced Wikipedia editors. During the session attendees were taught how to add Databoxes, building on the workshop at the Celtic Knot.
Working groups: Did participants spend time planning projects, or learning how to join existing projects at this conference?	With the focus on skills building and networking during the conference, the main collaborative project people worked on was setting up databoxes and exiting Wikidata property labels on the second day. The Gun Wikimedians Working Group was also formed on the first day of the conference.	In the post-conference survey, some attendees noted an interest in advancing the cause of Wikimedia in education and training others to edit. One attendee indicated they were considering organising an editing contest.

Learning

The Celtic Knot, a space to experiment on new formats and tools

Thanks to its relatively small but dedicated audience, and because of its changing theme and organising team every year, the Celtic Knot Conference is a great space to innovate and experiment with new program formats and ways to engage with participants. Especially since the Celtic Knot turned remote in 2020, we were able to try out tools, find the right balance between live and asynchronous content, and open up to more underrepresented communities outside of the Celtic European regions.

A light and curated program

Mindmap presenting the results of the community input survey prior to the Celtic Knot Conference 2022. The areas highlighted in blue are the ones that we decided to focus on for the 2022 edition of the conference.

For the 2022 edition, we initiated a reflection on the program design of the conference. The Celtic Knot always had a pretty minimalistic program, with one main track during two days, and a lot of breaks to accommodate informal discussions and community building. We wanted to push this concept even further and we articulated the program design around a key goal: the audience needs practical sessions, with hands-on elements to learn new skills that participants could immediately apply to their local context. Any reduction in live content could be addressed by providing a well curated ‘watch list’ to engage with asynchronously.

Therefore, we decided to build a very light and practical program, composed of four workshops on topics identified as crucial by the communities in the pre-conference survey; pre-recorded updates from the language communities; a guided discussion to keep the core identity of the conference (language communities sharing knowledge with others); and a social event. On top of that, we selected some content from previous conferences and other resources on useful topics to watch asynchronously in the “videos pool”.

For the workshops, the organisers decided on four practical topics, based on our knowledge of the communities needs and the replies to the community survey: two topics related to community growth and two related to Wikidata. Narrowing the choice down to four workshops was hard, but we wanted to give each session enough time in the programme, rather than having one-hour tasters with limited impact. Shorter sessions are often more didactic, with less scope for audience interaction than longer sessions.

We proactively contacted possible speakers to deliver our chosen workshops, looking for people with expertise on the topic and experience within their local community, and who would be able to deliver a workshop in a clear and understandable way. We aimed to represent various communities among the speakers: some of them were part of our “core Celtic” group, but the majority were from emerging communities who brought with them their own audience, increasing the diversity of the Celtic Knot participants.

A different way to build a collaborative program

When building this light program focused on the four workshops, we decided that we would not run the traditional “call for proposals” which usually leads to a selection of presentations proposed by community members. Since previous editions of the Celtic Knot and the majority of Wikimedia conferences have such a process in place, and people are used to being able to contribute directly to the program, this was a pretty bold decision.

We still wanted part of the program to be community-built and we wanted to offer a space for language communities to give an update on their activities. We decided to have a dedicated program, called “News from the Languages Communities”, where people could submit a short pre-recorded video to present their projects. This way of gathering content from participants required them to prepare content in advance and to record the video with their own equipment. In order to support them, we offered guidance, tips and suggestions.

Another part of the program that was collaborative was the live guided discussion session, where people could come up with any topic that they cared about, and propose it as a theme for a break-out room, where other participants could join and discuss. In the survey we asked attendees whether they talked to others about collaboration; people mentioned topics raised in the guided discussion, especially collaborating on activities related to education and mentorship, which was part of the guided session. Overall, this approach to incorporating collaborative aspects into the program was well received by participants and when selecting their favourite part of the conference several people specifically mentioned the guided discussion on day one. Removing the usual selection and scheduling process that was a heavy workload on previous conferences allowed organisers to dedicate more time and energy to preparing the workshops content with the speakers.

Coaching the speakers to deliver excellent content

When deciding to focus on a limited number of topics for the workshops, we were committed to provide practical, useful, and easy to reuse content to the participants. In running and attending previous Wikimedia conferences, we had noticed that although community members can have a deep understanding and expertise on a topic, they don’t necessarily have the communication and teaching skills to transmit this knowledge to a non-initiated audience. Knowing that we would record the workshops, we also wanted to make sure that the content would be structured and understandable by people who would watch the videos later. Therefore, we wanted to experiment on providing guidance and training to the speakers, not only so they could deliver high quality content, but also so they could gain professional workshop facilitation skills that they could reuse for future activities within the Wikimedia movement.

This preparation phase required a lot of time and commitment, both from speakers and organisers. We contacted the speakers three months before the event, and we ran three or four meetings with each speaker group (grouped by which workshop they would deliver) to help them define the target audience, the goals, and the outline of their workshop. We were accompanied by an experienced facilitator, Bhav Patel, who took the position of an advisor and provided useful methods and advice to the speakers to prepare their workshop and structure it in a way that would be participative and engaging. This focus on learning design was novel to many of our workshop leaders and was very useful. As a result the sessions were better designed for learning, and included much more interactivity than initially planned by the speakers.

For each workshop we put together a team of 2-3 speakers, and the preparation sessions were also used for them to meet each other and establish a rapport with the workshop host.

The aim of the community workshops was to draw on the experience of the presenters about community activities and to help the audience understand how it could be applied to their own communities. We decided on two broad themes: community growth and editing campaigns. Our thinking was that these topics will be relevant to a range of communities. Both of the community focused workshops took place on day 1, along with the launch of the updates from language communities, and followed by the guided discussion which brought through some of the themes of the community workshops.

Initially the format of the community workshops was quite similar, but we adapted the content as the presenters developed it with the conference team. The community growth workshop involved two speakers and three thematic questions from the host. The speakers spoke to the topics, and then we invited participation from the audience, and this process was repeated for each of the three themes. This structured approach worked well, but may have been harder to balance if there had been more speakers. With the editing campaign workshop, we decided that each of the three speakers would explain their topic at the start of the workshop, and then there would be discussion with the audience. When one of the speakers had to drop out, the other two were able to spend a bit more time discussing their topic. The different approaches gave the two workshops a different feel.

The audience interaction aspect was an experiment, and having different points of interaction throughout a workshop rather than all at the end was also a different experience for the audience. We had warm up questions in each session to get participants used to the approach; they may have been more accustomed to it in the second workshop. In the Wikidata workshops the audience interaction was quite free-flowing, especially when people were learning about how to set up databoxes.

We provided the necessary documents (e.g. slidedeck template) as well as equipment and internet connection to ensure the best experience for the speakers and the participants during the workshop. Speakers were encourage to put their slides up in advance of the conference, which meant that overall speakers did their conference preparation much earlier than usual.

As the conference organisers we were also the overall facilitators of each other workshops, supporting speakers to deliver their content. We had a good sense of the timing of each session and how much time was needed for each - this was also thanks to us attending the prep workshops. As a result, the workshops were well timed, and we could monitor the progress of each workshop, helping speakers time-manage. Sometimes this meant slowing the progress of the workshop down to make sure the audience is following and engaged.

As a result, the four workshops were of high quality. The shared preparation experience and facilitation approach led to a lot of coherence between the workshops, creating an overall feeling of the conference (which can be hard to achieve for an online conference, especially one that’s not run on a dedicated platform - we did ours on zoom). Participants indicated in the survey that they were very satisfied with the workshops content, scoring ‘the platform and conference content was technically accessible to me’ 4.33 out of 5 (participants especially enthused about the databox workshops), and during the wrap-up session we ran with the speakers, they indicated that they learned useful skills that they would certainly reuse for upcoming presentations. Some of the speakers actually already scheduled follow-up workshops with their local communities during the satellite events of the conference.

We met with speakers shortly after the conference to reflect on how their workshops went, to further embed learning gained. They all fed back that they learnt something new about delivering training through this.

• “Prep sessions was a game changer to learn how to present (Databox): “it was our own ideas but you helped us ship them” “
• “I’ve never invested as much energy into a conference, and enjoyed the energy of the organizers.”
• “I would recommend to keep providing this kind of training”
• “I enjoyed the training, enjoyed that it was more than just “how to use the tools”, how to structure and keep the session short, sticking to time, useful insights”

One presenter suggested we do more workshops of this sort. He received messages from participants saying that they benefited a lot from the practical workshops and that they were able to implement Databox on their wiki during or after the workshop.

A hybrid format to ensure community engagement on the long run

When we began designing the Celtic Knot 2022, parts of the world had already reopened after two years of pandemic restrictions. Wikimedians were eager to meet again, but we couldn’t ignore the fact that many countries were still facing strict COVID regulations, on top of the pre-existing challenges for people from many areas of the world to travel freely to Europe.

We faced a challenge: should we organise an onsite event in the UK, to accommodate our “historical” audience who could most likely travel easily to an in-person event, or run an online event, to allow people from emerging communities to more easily participate? We decided on a hybrid format with an “online-first” flavour: the main programme of the conference would take place entirely online, and we would support local communities to organise satellite events: small, onsite gatherings, before, during or after the main online event.

We launched an open call for satellite events, hand-in-hand with a scholarship program that would allow people to get funding for location, travel, accommodation and services they would need onsite. This led to considerable interest from the community, especially amongst audiences new to Celtic Knot. We had applications to fund events in Nigeria, Ghana, the UAE, and Wales. The events proposed were a mixture of watch-alongs so that people could meet and discuss the conference, and editathons so that groups could collaboratively work on their wikis.

The main challenge with the satellite events was transferring the funds overseas. Richard Nevell led on this aspect and was initially confounded by a lack of IBANs available for many of the target bank accounts; once this block had been cleared additional information was required for addresses of recipients and their bank branches which contributed to a slower than expected process. As a consequence, some events were delayed. Most satellite events were organised after the main online conference and allowed local communities to gather, to watch the content of the conference together and to exchange skills.

Non-streamed conference

With a somewhat complex participation design, we decided to keep the tech straightforward - we used Zoom for simplicity, and aimed to draw people to one conference space on Zoom. While it is possible to stream from Zoom to YouTube, we decided to record the sessions to share afterwards so that moderation of comments and ensuring a safe space was maintained was restricted to one online location (namely Zoom). However, this did mean that people couldn’t go back and rewatch if they were confused or joined late during any of the workshops. We did notice that the property translation workshop, for example, had a fast pace, and it would have helped to be able to go back.

We published all the workshops to rewatch on YouTube in the week after the conference and published them to Wikimedia Commons in July.

Watch the recordings

Accessibility

As organisers we were reflecting on programme accessibility while designing and planning the conference. As with previous Celtic Knot conferences, the content was primarily in the English language. But we aimed to make it accessible for speakers with English as a second language. The speakers were all invited to speak slowly. We had subtitles for talks in the live streamed videos. It was also really helpful that for a lot of our participants from African communities the conference run at a similar time zone to theirs.

Thinking about the programme content, we did take for granted that people will come with a basic understanding of what Wikidata is. This is how we prepared the speakers for Wikidata sessions too. But we had audience members who didn’t know what properties were, or were new to Wikidata all together. This was an assumption we made and didn’t make it clear that we are expecting some basic Wikidata knowledge for these sessions.

Reflections on the attendees journey

New audience vs historical Celtic Knot crowd

We had many people from emerging countries attend. It was great for us as organisers to welcome new language groups into the Celtic Knot community. Participants from emerging communities were eager to learn and exchange. It was clear to us that a lot of their participation was facilitated by the workshop speakers, many of whom also came from these communities.

Two-thirds of the audience were new to the Celtic Knot conference. In part this was a success in attracting new people to the conference, but did represent a decline in participation from other long-standing members of the community. The community itself is active on Telegram, and participated during the conference. In the post-event survey we asked whether there were barriers to participation to find out why there was proportionally less participation from returning Celtic Knot audiences, but this did not provide the insight we hoped for. In the pre-recorded language updates we featured news from the Language Diversity Hub, Welsh Wicipedia, Malta, Ireland, and Scots Wikipedia; these are important returning groups for the Celtic Knot Conference. The prominence in the programme of Gurene and Dagbani in particular may have drawn on new audiences and emphasised the conference’s relevance to those new communities. The inverse may have been true for language communities already involved in the Celtic Knot, with fewer opportunities for sharing their progress on the ‘main stage’ as we shifted to a workshop-led programme.

The audience’s level of knowledge and understanding was well suited to the conference; we felt that everyone was learning something and was starting from a similar level. If we had more traditional Celtic Knot participants, that might have made for interesting learning, but it could also be that it would be challenging to share across contexts.

In the end, we had a Celtic Knot - style learning event, with lots of exchanges, peer learning, connections -  just not for our usual participants.

Two levels of participation in the workshops

When designing the workshops of the online program, we faced an interesting challenge: how to ensure participation and engagement of attendees, while avoiding the possible mess created by a fully open format, keeping the workshop structured and understandable for people who would watch the video recording, and complying to data protection rules?

In order to solve this puzzle, we decided to implement two levels of participation:

• the “active participants”, a limited amount of people who would have the possibility to unmute their camera and microphone during the workshop, as well as sharing their screen, therefore interacting directly with the speakers and get some help from them if needed;
• The “watchers”, the majority of the audience, who would only have access to the chat to ask questions and get support, and would not appear in the video recording.

In the registration form, we offered participants to pick an “active participant” ticket for the workshops, informing them of the benefits and constraints (they would appear on the recording). Active participants would receive the same Zoom link as other participants, but the organisers would grant them some special rights at the beginning of the workshop.

This method rendered mixed results: participants didn’t necessarily understand the concept, some active participants didn’t show up or didn’t use their functionality to speak, while other participants requested to speak during the workshops. This way of splitting participants being as far as we know quite unusual in the online event formats of Wikimedia events, we can assume that people were not accustomed to this method. On top of that, we didn’t manage to communicate clearly about the two levels of participation and we encountered issues during the registration process. This meant that all ‘active participants' tickets got booked by the first few people who registered. We then got in touch with them to check which workshops people wanted to actively participate in, which was useful as it opened more slots to others. Eventbrite as a ticket environment wasn’t great to explain the set up we wanted to have.

In reflecting about the general design of the active/passive participation and how it worked in practice, we concluded:

• It was certainly useful for some people, knowing they can e.g. watch the conference without the stress of being ‘live’ and recorded. It would have definitely been very useful as a format with a lot of participants (100+). With the level of attendance we had (30/40 per session), one type of participant, and a simple zoom call, would have been ok to manage.
• It was unusual, people were not used to this kind of format at conferences - and it was a learning curve for us as facilitators of sessions too. It wasn’t easy to convey the design via the ticketing system/.
  • we could have put together a video/visual doc explaining how to take part
• We created the division so that people can participate without being recorded. But we didn’t see evidence that it mattered to people - what mattered is that people were able to speak when they wanted to contribute something

In retrospect, we could have tried a different approach: have all participants muted by default at the beginning of the workshop, and have them requesting to speak (in the chat or via the “raise hand” feature). We could have unlocked the possibility to enable their microphone, camera and screen sharing for a limited amount of time, allowing them to interact with the speakers, and after reminding everyone that the meeting was recorded.

Participants context and equipment

We didn’t anticipate that many of the participants would be attending by mobile. One of our workshop leaders shared with us afterwards that in their context, people would tend to join with their phone if they think it’s a presentation and they wouldn’t have to contribute actively. We should have been clearer that this will be an active session. Advice that was shared with us for next time was incredibly useful:

• if we use Jamboard again, we should advise people to download the Jamboard mobile app
• advise people to copy the useful links from the chat to save them for later (links not easily accessible on mobile zoom)
• Also, links and info shared in the comments on the Zoom call will disappear and will be difficult to find after the workshop, especially on mobile, where you can’t easily restore opened links => as good practice we should send list of links after the workshops
• Connected to this again, we expected people to take notes on etherpad (especially during the guided discussion session), but that would have been very difficult on mobile because of the issues above

Network issues were very common for our participants. It makes sense that most of our individual grant requests were for connectivity, and we were glad to be able to provide this.

Feedback from the post-conference survey

After the conference we distributed a survey to everyone who registered for the conference to find out about their experience of the Celtic Knot. The questions covered a range of topics, including quality of the workshops, new ideas from the conference, and what communities people represented at the conference. Some of the information has been used elsewhere in this report. 103 people replied to the survey.

We were interested in how people access the conference and asked that if people did not attend what barriers to participation there were. Of the 16 people who indicated they did not participate in the conference. The most significant barrier was access to reliable and cheap internet. Seven people noted that the high cost of data prevented them from participating in full; others indicated conflicting time zones and commitments or hardware issues.

As in previous years, many different language communities were represented in the audience. Between them the audience spoke 41 different languages. Interestingly, 60% of the participants had been active on Wikimedia projects for up to two years indicating that these are new communities to the Celtic Knot and new communities are seeking support. This was reflected in some of the discussions which touched on the Wikimedia Incubator. The survey showed that two thirds of attendees joined the conference for the first time this year.

The survey indicated that the audience were satisfied with the conference overall (scoring 4.4 out of 5), and they were safe, respected and comfortable (4.56). The survey also indicated that 89% people left with new ideas and 60% discussed new ways of collaborating. Encouragingly, 98% of people said they would consider attending the conference again.

Financial documentation

This section describes the grant's use of funds

All amounts in GBP.

Budget table

Category	Item description	Unit	Number of units	Cost per unit	Total cost	in kind contribution	£	actual expenses	Notes
Organisation	WMUK organising team, comms and financial oversight	hours	148	32.8	4854.4	-		4854.4
scholarships	administration (WMUK in house/contractor)	hours	20	15	300			300
Local events	admin support	hours	20	15	300	-		300
Merch	admin support	hours	10	15	150	-		150
Organisation	WMUK (in kind)	hours	200	32.62	-	6524			£3678.16 in kind in the end we spent less staff time total on the project than expected
Technology	used for zoom	per tool	2	100	200	-		200
						-
speaker support	training for workshop hosting	per training	4	500	2000	-		2000	working with external trainer
speaker support	training for video updates	per training	10	100	1000	-		1000	--> used for staff time - we provided training in-house
speaker support	equipment eg microphone, broadband	items	5	30	150	-		150	--> used for staff time, additional time to support trainers
Translation/transcription	Translation/transcription	hour	36	30	1080	-		1132.8
Comms	paid promotion i.e. boosting Facebook+Twitter	platform	2	125	250			250	--> used for staff time to support comms activities in house
scholarships	scholarships incl processing costs	scholarship	30	80	2400	-		-	(spent, the scholarship figure is given together lower down)
Merch	design	set fee	-	1000	1000	-		1585.57	italki giftcards
Merch	materials	item	50	10	500	-		109.2	used for italki giftcards
Merch	shipping costs	postage	50	10	500	-		500	--> used for staff time
Local events	scholarships for organisers (one per event)	event	10	1000	10000	-		11459.75	all of the scholarships together - individual and events
	international payment fees							600
					24684.4	6524		24591.72
10 % Unforseen costs					2468.44			2561.12	additional staff time was needed as we worked on some unfamiliar aspects of the conference
					TOTAL direct	TOTAL in kind		TOTAL direct	TOTAL in kind
					27152.84	6524		27152.84	3678.16

Summary of funding

Total project budget (from your approved grant submission): £33676.84

Total amount requested from WMF (from your approved grant submission): £27152.84

Total amount spent on this project (this total should be the total calculated from the table above): £30831

Total amount of WMF grant funds spent on this project: £27152.84

Are there additional sources of revenue that funded any part of this project? List them here.

• time in kind from Wikimedia UK - £3678.16 (in the budget this was estimated to be £6524 but we have spent less time)
• time in kind from Wikimedia DE, 100hrs (uncosted)

Remaining funds

Are there any grant funds remaining?

NO

Anything else

Documentation from the Celtic Knot

The introductory session, the workshops, and the concluding session were all delivered over Zoom; the workshops and conclusion were recorded and uploaded to Wikimedia Commons and YouTube. The introductory session consisted of some orienting remarks from the organising team and the pre-recorded updates from language communities, which are available separately.

Recorded workshops

The updates from language communities were an important way in which communities could share what they have been up to since the last conference.

Updates from the language community

Documentation and photos from satellite events

To give the Wikimedia community a new way of engaging with the Celtic Knot, we encouraged the community to run their own in-person events where possible either in-tandem (leading to a few meet ups with watch-alongs) or to serve their community directly in some way. We anticipated that the latter could take the form of skills workshops or meetups to help communities reconnect after a period of isolation. The result was that communities took the opportunity to edit collaboratively, with an especial focus on Wikidata as an area that can be improved with significant impact on language communities.

There were in-person meet ups during the conference involving the Gungbe and Yoruba communities. As well as watching the live workshops, new people were introduced to Wikipedia and there were editathons to write new pages. 40 people attended the Yoruba event, and of the more than 20 people at the Gungbe event eight were new editors.

At the event in Dubai, the organiser introduced 120 librarians and educators to Wikidata, laying important groundwork for understanding this growing project. The events in Kumasi and Awka also focused on Wikidata, getting people to edit and translate labels. Both events were attended by more than 25 people. The event in Kumasi, which allowed participants to share their experiences of working on minority and minoritized languages on the Wikimedia project, resulted in more than 3,000 edits to Wikidata, improving content related to the Twi language.

Gungbe Wikipedia Satellite Events (watch party, training, and networking), 1-2 July

Yoruba Wikimedians User Group satellite events (watch party, workshops, and editathon), 1-2 July

Celtic Knot Conference 2022 Kumasi (Wikidata editing), 23-24 July

Translate Wikidata properties to Igbo language, 4-5 August

Wikidata workshop at the Mohammed bin Rashid Library in Dubai, 14 September

• 
  A Wikidata editathon of translating properties for the Igbo Wikimedians User Group
• 
  An introduction to Wikidata workshop orgganised by the Dagbani Wikimedians User Group
• 
  Celtic Knot Conference 2022 Yoruba Remote Event

Photos from in-person gatherings