Jump to content

Growth/Personalized first day/Structured tasks/Add an image/GLAM events

From mediawiki.org


Abstract

[edit]

The Growth team has developed and deployed a feature called Add an Image to several non-English Wikipedias. Recently, this beta feature was deployed to the Spanish Wikipedia, as Añadir una image, and in order to test its functionality and usability, four events were carried out with GLAM (Galleries, Libraries, Archives, and Museums) professionals who were also new Wikipedia editors. Despite some flaws, this feature proved to help GLAM professionals use their knowledge and skills to improve Wikipedia articles with minimal or no prior knowledge of wiki editing. The events helped not only to bring in new editors, but to increase awareness and visibility of the local affiliates’ work on the GLAM domains.

Introduction

[edit]

The Growth Team (GT) has developed a series of tools focusing on absolute newcomers, aiming to make the Wikipedia editing process easier and gentler so that new editors keep their initial motivation for longer and keep improving their skills as Wikipedians. These tools are conceptualized as mediating the very first edits of a new editor, answering the hypothetical question: «What would a new editor do on their first day as Wikipedian?»

To this end, the tools are aimed to make «small» contributions easy. In this context, «small» refers not to quantitative measures like byte count, but to incremental tasks that require no particular skillset in any profession and can be of help no matter how little they are. Examples of such tasks are: correcting typos, adding references to unsourced statements, linking orphaned articles, and adding relevant pictures.

It’s on this last task that the GT developed a tool internally called «Add an Image», whose purpose is self-explanatory. This tool has already been deployed in a few non-English Wikipedias, but when it came down to looking for deployment on the Spanish Wikipedia, there was a need for testing on a particular set of people: professionals in GLAM—Galleries, Libraries, Archives, Museums—and GLAM-adjacent professions.

Images, in general, are a powerful addition to encyclopedic articles, but in many cases, it’s not easy to discern which image should be present in an article, regardless of the problem of actually finding and/or sourcing this image. In the particular case of GLAM-and-adjacent topics, the questions run even deeper: any potential images an article would need to require context and information that are usually not present in the general audience: art history, knowledge of the interactions between artists, patrons, art movements and periods, techniques, materials and a long list of domain-specific knowledge that is hard to come by without advanced training in these subjects. Not only that, misinformation in such subjective areas can lead to misrepresentation and biases that affect the general public’s perception of art and its role in society.

It then becomes evident the need for Wikipedia to be accessible in its technical sense to specialists of all kinds—the GLAM areas are just an example. To borrow a certain movie’s quotation: it’s easier to teach art historians to be Wikipedians than to teach Wikipedians to be art historians.[1]

Objectives

[edit]

The set of four GLAM events, in partnership with the three Latin American affiliates, became known as the #1Pic1Article campaign. Its main objectives were to attract GLAM professionals from Latin America—focusing on Argentina, Mexico, and Chile but not excluding other countries—and teach them how to use Wikipedia in general and the Add an Image feature, in particular, to put their knowledge to use in Wikipedia, particularly in ways that a general-knowledge Wikipedian might not be able to.

The campaign centered on four events:

  1. A test event of lesser magnitude, focused on Argentina-based GLAM professionals;
  2. An event focused on Argentina-based GLAM professionals;
  3. idem for Mexico-based professionals;
  4. idem for Chile-based professionals.

This campaign aimed to teach a group of people the general works of Wikipedia and the Add an Image feature, in particular, asking them to make 3 additions during the event, as well as several more during the two weeks following the initial event. As these events weren't aimed at any particular organization, they were online meetings held over a teleconferencing platform.

The use of the hashtag in the campaign's name (#1Pic1Article) was an attempt at a possible future campaign, similar to #1Lib1Ref, but with images instead of references, as well as museum professionals instead of librarians.

Events

[edit]

Overview of events

[edit]

The general structure of events was the same for all events, namely:

  • A short introduction on Wikipedia;
  • A short introduction on the work between GLAM institutions and the Wikimedia projects;
  • A short introduction on the work done by affiliates in their respective countries (Argentina, Mexico, Chile);
  • A lecture by photojournalist and researcher Cora Gamarnik on the importance of images and their context, the relationship between informative, and artistic uses of an image;
  • A demonstration of the Add an Image feature;
  • A guided workshop for participants to try out the feature by themselves; and
  • A closing space open for participants to air reflections, questions, and suggestions about the feature.

Particular comments follow, with shared feedback in the next section.

Event 1: Argentina Test

[edit]
Date
Monday, March 7, 2022
Participants
~20

The first test event was focused on GLAM professionals who had already worked with Wikimedia Argentina on previous events and were aware of the general workings of Wikipedia. This event is framed as a «test» only in the sense that it was focused on a small group containing several participants who had already worked on Wikimedia events. It was also a test of timing and whether the first drafted schedule was realistic in time and subjects.

By and large, the event was a success, except for the fact that it went overtime. Participants were excited to try a new way to add to Wikipedia in short bursts and in a very visible way.

Event 2: Argentina

[edit]
Date
Friday, March 18, 2022
Participants
~30–40

The first «real» event was aimed at GLAM professionals based in Argentina, containing a few people already aware of the Movement as a whole but made up of mostly new editors. This was the first event when we scheduled time at the end of the session to gather feedback from the participants themselves. This proved to be a highly valuable time, as participants shared not only their opinions and suggestions of Add an Image, but on the process of contributing to Wikipedia and Wikimedia in general. This time was also valuable to let organizers know which other skills or topics they would like to learn, as well as further activities by local affiliates.

Event 3: Mexico

[edit]
Date
Friday, April 1, 2022
Participants
~40

The Mexico event aimed to bring in professionals from outside Mexico City, where Wikimedia Mexico itself is based and has worked before. This is to encourage participation beyond the largest urban centers of the country. The event attracted not only professionals but people working and pursuing a post-graduate degrees in several institutions in the country.

Institutions present
UNAM—The National Autonomous University of Mexico and the oldest university in the Americas;
ENCRyM—National School of Conservation, Restoration and Museography;
ENAH—National School of Anthropology and History;
Public universities (University of Guadalajara, Autonomous University of Tabasco, Autonomous University of Aguascalientes, Universidad Vizcaya de las Américas)
Public libraries (National Library of Mexico, Amado Nervo Community Public Library, Prof. Ma. Ignacia Martínez de Loza Community Library, Azcapotzalco Borough, Tulancingo Municipality)

Event 4: Chile

[edit]
Date
Wednesday, April 20, 2022
Participants
~30–40

The event had the difficulty of being scheduled originally during Holy Week, a time when many institutions close or alter their schedules. The ultimate decision was to delay the event for a week instead of risking an event with low participation. This decision proved to be for the best, as participants were not just present, but many of them were fresh out of a few days of vacation or reduced workload. Thus, with high spirits participants were eager to learn and try out new things.

Feedback

[edit]

Positive

[edit]

Across the four events, the following positive feedback was commonly shared by participants:

  • Add an Image lets editors see what is needed in a mostly clear way (in the sense that it’s clear which articles are lacking images);
  • AaI lets editors refine their topics of interest at least somewhat;
  • AaI lets editors be aware of and help in topics that are not immediately within their area of expertise, professional knowledge, and/or personal interest;
  • AaI lets editors jump right into the suggested task with clear instructions;
  • AaI lets editors use their judgment, common sense, reason, and/or professional opinion to ultimately decide whether a suggested image should be added or not;
  • AaI has a few steps to be completed;
  • Participants’ contributions are very quickly added and contribute in a measurable, visible way;

Negative

[edit]

Across the four events, the following negative feedback was commonly shared by participants:

  • Add an Image doesn’t allow for topics to be selected with the granularity they would desire;
  • AaI only lets people browse articles one at a time, needing sometimes many clicks to find a good candidate article to edit;[2]
  • AaI presented candidate articles in a linear way in the same ordering, which led to people seeing «the same few topics» every time they reloaded the AaI page;[3]
  • AaI did not present all image metadata present in Commons except when opened in a new tab;[4]
  • AaI made one and only one suggestion for an article, discarding the article entirely if the suggested image was considered to be not appropriate for the article. Participants made several remarks on their want for having more options to choose from and/or a way to search for more candidate images on Commons. A few participants wondered why AaI did not prompt them to contribute an image directly if the user happened to have one available;
  • In a few edge cases, AaI made erroneous image suggestions with very subtle differences (these cases were mostly on two people having the same or almost the same name and the image suggestion being pulled from Wikidata with no further validation)

Other

[edit]

Across the four events, there were other kinds of feedback adjacent but not necessarily related to AaI:

  • Several participants were eager to continue working on the target article (often because it fell within their area of expertise) and wished to learn the «general» way to edit Wikipedia;
  • Participants were eager to learn more on Commons, both in the technical aspects—i.e. how to upload and categorize images—and the cultural practices surrounding it—i.e. Commons as a repository of free media, rights and copyright, citizen participation, etc.
  • Participants were generally grateful for having a space where they could practice by themselves and have assistance if needed. They also were grateful to discover the reversible nature of mistakes—i.e. rolling back edits—so that they did not fear trying out features on their own;
  • Participants were eager to know what other GLAM-related activities were organized by the local affiliate so as to further their understanding of the Wikimedia movement in general; hoping to participate not only as individuals but as organizations.

Lessons learned

[edit]

Add an Image is overall a great feature to add to Wikipedia. As participants showed, it bypasses—at least partially—three interconnected hurdles in Wikipedia editing, which can and have discouraged new editors in the past, namely:

  1. Finding an article to add to,
  2. Finding a relevant image on Commons for said article, and
  3. Inserting the image onto the article (a bit easier to do with the Visual Editor than with code)

Having a clear-cut, atomic task at hand makes contributing easy, eliminating the overhead of deciding on such a task. It also allows for quick contributions, so that even a new editor can accomplish a visible addition in minutes.

The downsides include participants being «cut-off» from doing further work on their chosen article, namely in the form of:

  1. AaI not providing more suggestions and immediately prompting working on a new article after a decision was made;
  2. The perceived difficulty of accessing image metadata and images not having «complete» metadata;

In general, participants were eager to learn and left happy to have not only learned, but contributed to a visible project. Since their occupations deal in the intersection between academia and the general public, they found a way to put their knowledge to immediate good use. It also helped many of them even know of the work done between Wikimedia affiliates and GLAM institutions.[5] This further shows that outreach work with GLAM institutions is never over, in a positive way.

Notes for the future

[edit]

While the technical objective of these events was to test out the functionality of AaI among GLAM professionals, the event itself presented as a theoretical/practical workshop and this, in turn, allowed participants to feel not like a test subject, but as participant in a community that valued their expertise and knowledge. Tying the practical workshop with a theoretical lecture also helped ingrain the importance of outreach work between academia and the general public, a sector where Wikimedia is a key player.

The edits done by new Wikipedians were carried out and the grievances were never so bad so as to discourage participation. However, some of these grievances were only or mostly corrected by having the live assistance of other editors. Thus, the importance of guided mentorship for newcomers is proved once again.

Holding an event like this online made geographical location a non-issue (mostly). Many attendants were not located in major urban areas which in the past would have limited their participation. This also served to signal that the Wikimedia-GLAM work is not and should not be limited to the «big» institutions and that cultural organizations are welcome regardless of perceived importance on a regional, national or international level.

Holding events for newcomers to test out features should not be presented as a technical test, but as an opportunity for said newcomers to have a low-barrier entry point to Wikipedia and Wikimedia. This sends the message that their know-how and skills are valued (which should also serve to bring more underrepresented voices to the Wikimedia projects).

Bringing in new editors to Wikipedia and Wikimedia can be seen as having two interconnected problems:

  1. Wikipedians needing to have technical skills (editing, style, referencing, etc.), and
  2. Wikipedia having a technical barrier to overcome (learning how to edit a wiki, some inner workings of wikis, etc.)

Traditional editing events like editathons serve to address the first problem: training newcomers to overcome the technical barriers. Efforts to lower said barrier (the second problem) can only come from the technical teams working with the software itself. Both efforts are needed if Wikipedia is to welcome more editors and keep a fresh view on its content, practices, and policies. A live, online, cooperative project can only thrive if its base population is equally alive and cooperative, which necessarily means being actively welcoming to newcomers.

[edit]

Organizers

[edit]
  • Giovanna Fontenelle (Program Officer, GLAM and Culture, Wikimedia Foundation)
  • Angie Cervellera (Culture and Open Knowledge Program Manager, Wikimedia Argentina)
  • Patricia Diaz (Executive Director, Wikimedia Chile)
  • Andy Cyca (Contractor, Wikimedia Mexico)
  • Nicole GonzĂĄlez Herrera (Contractor, Wikimedia Chile)

Notes

[edit]
  1. ↑ The typical Wikipedian is curious by nature and this leads to them being very knowledgeable in many areas. This is not to say that a typical Wikipedian is not intelligent or cannot learn or be trained as an art historian, but GLAM-and-adjacent subjects represent knowledge that is very hard to gain without extensive training and/or investing a lot of time and guided practice into it; something that a professional tends to have done.
  2. ↑ This is in the sense that articles are not presented in a way that lends itself to browsing larger sets of candidate articles easily. Several participants mentioned having paginated browsing, in a way that showed 10 or 20 articles at a time.
  3. ↑ This also led to the events having editing conflicts: since participants were asked to use the same set of topics, the ordering for candidate articles was the same for everyone, leading to people tending towards the first few articles. This in turn led some people to try and add an image to the same article that someone else was working on (and more than one frustrated editor at working on an article only to find their contribution would ultimately be for nothing). Granted, this concerns mainly collective editing events like this, but it’s a behavior to be kept in mind.
  4. ↑ This, of course, is tied to the uneven quantity and quality of metadata across Commons media, but that is another discussion.
  5. ↑ More than one participant mentioned not knowing of the existence of this work, which had a two-part effect: First, they were pleased to know that the Wikimedia projects can be built with academic rigor—thus dissipating the misconception of it being misinformed work. Second, they were pleased to know that they and their institutions and their work could be added to Wikipedia. Many expressed hope in having a formal working relationship with the local affiliates.