Wikimedia Meta-Wiki

Research:Wikipedia Primary School SSAJRP programme/Process

< Research:Wikipedia Primary School SSAJRP programme
Wikipedia Primary School SSAJRP programme Timelines & progress

Access to working document

Method

edit

General structure of the process

edit
 
The four key steps of Wikipedia Primary School.

The project has defined four key steps to develop "Wikipedia Primary School".

  1. Identifying of the content gap from school programs.
  2. Triggering the production of content: new articles, improvements, editing of existing Open Educational Resources
  3. Peer review of content (on Wikipedia, with the involvement of scientific journals and with the involvement of experts
  4. Making new content available on Wikipedia or for Wikipedia.

Identification of content

edit

Articles have been identified based on topics presented in the national curriculum of primary school (South African Revised National Curriculum Statement Grades R-9). Here the detailed process.

From the topic a draft list of articles has been draft, created from two criterions: discoverability and impact of the article. Here the detailed process

Once defined the list of meaningful articles for the South African curriculum, they have been divided on:

  • Articles to review by experts
  • Articles to review by journals
  • Articles to review by Wikipedia community
  • Articles to re-write

Peer (aka expert) review of content

edit
 

Expert review of content proceeded along three phases

Phase I

  • Identification of the articles selected for peer review
  • Identification of the experts for each topic (detailed process described here)
  • Contact of experts to sollecitate review of one or several articles (detailed process described here)
  • Identification and contact of academic and scientific journals for each topic (detailed process described here)

Phase II (after an approval has been received from an expert)

  • Request for community review for about 2 weeks (detailed process described here)
  • After community review, retrieval of the article to be reviewed by the expert
  • Article content is sent to the expert along with a feedback form

Phase III (after a review has been received). Detailed process here

  • Upload of the expert review on Wikimedia Commons
  • Publishing of the expert review on the article talk page
  • Call for editorial activity by the community, using the expert review

Processed figures are

  1. Number of articles proposed for review by expert : 48
  2. Number of articles actually reviewed : 29 31

Whenever an expert produced the review of a Wikipedia article (see peer review of content), this review was

  1. uploaded on Wikimedia Commons and tagged with Wikipedia_Primary_School_SSAJRP_reviews category ,
  2. mentioned in the article talk page, and
  3. listed on the PSP project dedicated page.

Outcomes measured were

  1. whether the review published brought any comment from the community
  2. whether the review published actually led to improvement to articles

Making new content available on Wikipedia or for Wikipedia

edit

Full rewriting by an expert

edit

Experts were approached in asking them to review an existing article.
In some cases, experts approached to review an article, decided to actually

  1. either propose a entirely new text, sent as an attached document by email
  2. or directly edited the wikipedia article themselves

Process followed

  • in the case of articles rewritten directly on Wikipedia by expert, a light copy editing of the article was done, if considered needed
  • in the case of articles proposed in a document
    • The article was copyedited according to expert suggestions, consistently with Wikipedia rules and guidelines
    • An email was sent to OTRS with the document proposed by the expert, with mention of the licence of publication
    • OTRS agent confirmed having received appropriate documentation to confirm authorship and licence

Triggering production of content

edit

Several methods were used to trigger production of content, through creation of new articles or improvement of already existing articles.

Publishing of the expert review

edit

Whenever an expert produced the review of a Wikipedia article (see peer review of content), this review was

  1. uploaded on Wikimedia Commons and tagged with Wikipedia_Primary_School_SSAJRP_reviews category ,
  2. mentioned in the article talk page, and
  3. listed on the PSP project dedicated page.
  4. depending on articles, individual participants (known to be interested in the topic, or prior participant to the article) as well as specialized venues (such as portal talk pages) have been informed that the review had been published

The review process has been updated for the last 6 reviews published. The first reviews were published only in the form of pdf document mentioned in the article talk page. This system was possibly not very enticing for community members used to read comments and suggestions directly in a wiki page. As a consequence, the process was updated in such a way that the expert comments are directly copied on the article page, potentially facilitating their reuse.

Request for assessment or reassessment of an article by Wikipedia community

edit

The editor community sometimes organises itself in WikiProjects. A WikiProject is a group of contributors who want to work together as a team to improve Wikipedia. These groups often focus on a specific topic area, a specific location or a specific kind of task. The English Wikipedia currently has over 2,000 WikiProjects, each with varying levels of activity. Most WP projects select articles that belong to their area of activity, and evaluate the articles, both in terms of article quality and in terms of article importance with regards to the Wiki Project focus. Quality assessment is done using a scale :

  • Stub : A very basic description of the topic. However, all very-bad-quality articles will fall into this category.
  • Start : An article that is developing, but which is quite incomplete. It might or might not cite adequate reliable sources.
  • C : The article is substantial, but is still missing important content or contains much irrelevant material. The article should have some references to reliable sources, but may still have significant problems or require substantial cleanup.
  • B : The article is mostly complete and without major problems, but requires some further work to reach good article standards.
  • GA : The article has attained good article status by passing an official review.
  • A : The article is well organized and essentially complete, having been reviewed by impartial reviewers from this WikiProject or elsewhere. Good article status is not a requirement for A-Class.
  • FA : The article has attained featured article status by passing an official review.

Assessments are done on a voluntary basis. Lower grades are usually given by a single individual without specific discussions, whilst higher grades are given only after a more thorough review of the article by several individuals. Grades may be challenged over time (as the quality of the project increase, so do the requirements to be a featured article for example). It is possible for an article to be part of several wikiprojects... or of none. Grades are displayed on the article talk page.

Given that It is possible to request an assessment in a WikiProject discussion page and that the request is likely to bring attention on an article, the hypothesis was that such a request could result in the page being further edited and improved.

Accordingly, we have followed the following process, on 174 articles.

  1. review all articles in our list and record whether and when they have already been reviewed by one or several WikiProjects,
  2. compare assessments when several have been made for an article,
  3. evaluate whether the assessment was outdated (discrepancy between assessment provided and state of the article),
  4. list all articles in the following situation
    1. Case 1 : when the article belong to several projects, but quality grades given by projects differ significantly, ask for it to be reassessed by several wikiprojects : 7 articles identified;
    2. Case 2 : when an article does belong to one or several WikiProject, but has never been assessed, ask to get it assessed by a wikiproject (for the first time) : 8 articles identified;
    3. Case 3 : when the original assessment is obviously not representative of the current article situation anymore, ask for it to be reassessed the WikiProject it belongs to : 3 articles identified;
    4. Case 4 : when an article does not belong to any WikiProject, identify a relevant WikiProject that might adopt it and assess it : 7 articles identified;
  5. for each of the identified articles, decide which articles will actually follow the assessment or reassessment process, and which will not.
    1. In Case 1 : 5 articles were selected;
    2. In Case 2 : 5 articles were selected;
    3. In Case 3 : 2 articles were selected;
    4. In Case 4 : 2 articles were selected.
  6. Implement the requests. Requests were made in the talk page of WikiProjects, in the section dedicated to that purpose.
  7. Outcome were checked after 4 weeks and after 7 weeks. Three different types of outcomes were identified
    1. Feedback from community (such as a comment left on the talk page);
    2. Whether a new assessment was done;
    3. Whether the article was significantly modified in the weeks after the request.

Call for new article creation

edit

One of the WikiProjects of the Wikipedia community is meant to better deal with Requested Articles See here.

One hypothesis was that listing some of the topics identified during the analysis phase but with no equivalent article on Wikipedia on the Requested Articles would result in the creation of new articles. As a consequence, several articles were proposed to this WikiProject over the course of the project.

Call for participation to two online Writing Contest

edit

In occasion of the 15 year anniversary of Wikipedia (15 January 2016) the online “Wiki Loves Women” writing contest occurred. The goal of the contest was to create at least 15 new biographies of African Women in the following 15 days. The articles were created by self-organized teams in French or English. Names of biographies of South African notable women were suggested [1] for creation. Additionally, a list of articles existing in another langage but missing in the target langage was provided. But teams were free to cover any missing biographies they wished. Additionally, a list of articles existing in another language but missing in the target language was provided. Awards were attributed to teams, according to the quality and quantity of the content provided, and consisted in an official recognition (“barnstar”).

The contest was well attended and resulted in the creation of 234 articles overall. As part of the Primary School project, 29 names of biographies of South African notable women were suggested for creation to the participants.

In October-November 2016, another online contest was organized, the Africa Destubathon. The goal of this contest was to reduce the number of very short articles available on Wikipedia about Africa (over 37 000 articles were concerned). Volunteers were invited to expand and improve at least 2000 such articles over 6 weeks. Participation was proposed at the individual level. Awards were distributed to individuals, based on the number of articles improved within a certain geography (country level) or within certain thematic area. Awards were official recognition (“barnstar”) as well as financial benefit (with prizes distributed to winners as Amazon vouchers). The contest was well attended and successful with over 2000 articles improved. At the end of the contest, 18 people were eligible for a monetary price, the highest individual gift being 445 US dollars.

Whilst most articles improved were only worth 1 point in the challenge, some articles would provide double score, or triple score or even quintuple score. The very basic articles in the Wikipedia Primary School dataset were listed as double score at the country level and required only simple improvement to give the bonus. Other well developed articles in the dataset were listed as quintuple score in the general bonus page and would require the article to reach a much higher quality (“Good Quality”) level for the editor to get the bonus.

Listed as double score

  • Children's Act, 2005
  • Africa Day
  • Day of Reconciliation
  • Gana and Gwi people
  • National Women's Day
  • Makhonjwa Mountains
  • West_Coast_National_Park

Listed as quintuple score GA

  • Apartheid
  • Herero and Namaqua genocide
  • Saartjie Baartman
  • San people
edit

Portals are pages intended to serve as "Main Pages" for specific topics or areas. A portal may be associated with one or more WikiProjects, but unlike a WikiProject, it is meant for both readers and editors of Wikipedia, and should promote content and encourage contribution. Portals are created for encyclopaedic topics only. The idea of a portal is to help readers and/or editors navigate their way through Wikipedia topic areas through pages similar to the Main Page. In essence, portals are useful entry-points to Wikipedia content.
Accordingly, hypothesis was made that featuring some articles in portal pages could lead to further consultation and modifications of those articles.
Articles have been selected to be added on portal pages. Three months later, the articles have been studied to explore

  • whether they had received significant attention from editors (whether for improvement or in terms of vandalism)
  • whether the articles were more consulted (increase of daily page views, using https://tools.wmflabs.org/pageviews/#)

The observation of potential impact has been made up to 3 months after the set up on the porta page.

Edit-a-thons organised by the team in South Africa

edit

In the online communities of projects such as Wikipedia, an edit-a-thon (sometimes written editathon) is an event where editors get together to edit and improve a specific topic or type of content, typically including basic editing training for new editors. The word is a portmanteau of "edit" and "marathon".

3 edit-a-thons were originally planned. But the political contest was not favorable and events had to be delayed several times. Dates of workshops go as such

  • 31 October: Postponed due to University protest action (which also brought about a safety risk and we were not allowed to use University resources)
  • 7 November : Postponed due to University protest action
  • 14 November : Postponed due to University protest action
  • 5 December : Workshop took place but there were not attendees
  • 12 December : Postponed due to close of academic school year in the same week
  • 19 February : Second Workshop
  • 26 February : Third Workshop

The initial idea was to have two workshops in Cape Town and one in Johannesburg. But when the first edit-a-thon actually took place, no participants pitched up. We suspect that the reason for this might have been a lack of local advertising and not having partnerships with local organisations that could help to fill up spaces. We decided going forward to these two issues needed to be addressed and form the basis of the list of guests invited to the event. We also decided to stick to Cape Town.

The process followed in South Africa to organize the edit-a-thon has been:

  1. Identification of dates and suitable venue for the edit-a-thons (Wi-Fi access and in-kind contributed venue were the main criteria of selection): the American Corner, Central Library, in Cape Town.
  2. Identification of the suitable content on which the participants would work
  3. Organization of caterer for all events
  4. Communication and advertisement through various mediums
  5. An event page has been created for the edit-a-thon organized in December 2015.
  6. An event page has been created for the edit-a-thons organized in February 2016 in order to describe the objective of the events, tracks attendance and makes mention of the suggested content to be addressed.

Communication and advertisements has been done through the following channels:

  • Social media & websites - Facebook, Twitter (Kumusha, Wikimedia ZA, Joburgpedia, Wiki Loves Monuments, Wiki from Above, UCT IP Unit)
  • Mailing List - Wikimedia ZA, Africa Centre Mailing List
  • Various radio and tv stations and printed press (according to the latest Media List online)
  • Various wiki meta pages - SA Portal, Wikiproject SA, South Africa Wikipedia Notice Board, Wikipedia in South Africa, Portal Talk SA, Afrikaans Geselshoekie
  • Mailshots to: Historical Studies at UCT, African Gender Institute at UCT, Archeology at UCT, Centre for Higher Education, Library Association of South Africa
  • Wikipedia Geonotice for South Africa
  • Regular mails to the mailing lists and short blurbs on Facebook and Twitter. In addition, we advertised the event to our contacts such as LeadSA.
  • Creation of an event page on Eventbrite

Topics
The topics suggested to be covered during the edit-a-thons were all South-Africa related topics. All of them were either missing articles, or articles in need of extensive rewriting and expansion. They were generally rather focused topics rather than global in nature.

  • Africa Day
  • Children's Act (South Africa)
  • Freedom Day (South Africa)
  • Gana and Gwi people
  • Kaditshwene
  • List (or timeline) of South African inventions
  • Minaars Cave
  • Musical intruments of Africa
  • South Africa leaders

Edit-a-thons organised by other parties

edit

Regular (or more spontaneous) face-to-face meetings of Wikipedians take place in cities around the world. The Wikipedia:Meetup page is used as a starting point for Wikipedias organising meet ups (this page is primarily used to list meetings organised with English used as primary langage), and subpages are used for finalising the details once a meet up has been agreed. Meet-ups may feature different activities : training new editors, writing articles on a specific theme, workshops, photo hunts, informal get together etc. Many of those meet-ups are actually part of the regular activities of Wikimedia chapters or Wikimedia UserGroups. The most active groups (when it comes to face to face activity) are related to Washington / New York and UK groups. Some of the groups meeting face-to-face on a regular basis are also Tasks Forces on Wikipedia and collaborate to get specific things done. An example is the AfroCrowd, an initiative which seeks to increase the number of people of African Descent who actively partake in the Wikimedia and free knowledge, culture and software movements. AfroCrowd regularly meet in New York City (USA). Groups such as AfroCrowd typically maintain lists of articles to create or improve (aka todo lists).
The hypothesis was that it could be possible to get some of our targeted articles improved by

  • getting in touch with organisers of events and suggesting a theme for one of their future edit-a-thon (which could lead to the improvement of several of our targeted articles) or
  • adding some of the targeted articles to a todo list
  • joining an already planned edit-a-thon on a topic coherent with some of the targeted articles

The process followed has been

  • regularly track planned meet-ups on https://en.wikipedia.org/wiki/Wikipedia:Meetup
  • contact organisers of events listed on meet-ups when topic is relevant to some of the targeted articles
  • explore which active groups regularly organize face to face or online activities and contact their leaders to introduce them to Wikipedia Primary School Programme and seek help in improving articles
  • add targeted relevant articles to task forces lists

Bold reassessment of articles

edit

The majority of articles on Primary Project Articles list were attached to a WikiProject. In some instances, the articles were not assessed at all. In other instances, articles were visibly in a quality state higher than what was displayed (either because the assessment was quite old, or because the articles have been recently significantly improved).

The hypothesis was that in *updating* the quality assessment of the article for all WikiProjects listed in its talk page, it could foster attention from members of those WikiProject and lead to article modifications.

Accordingly, several articles assessment were updated in the article talk page. Impact was reviewed 1 week later.

Call for improvement of stub articles as part of a global challenge

edit

In fall 2016, a global challenge was set up, the Africa Destubathon. The general goal was a stub elimination contest, which essentially means a challenge meant to improve articles to a slightly higher level of quality. The challenge was organised for all African-countries. To support the challenge, a list of articles to improve was established and proposed to editors. 6 articles in the Primary School Program were still considered to be stubs and have been added to the Challenge suggestion list in South Africa and proposed to count double points within the challenge.

Set up of a WikiProject assessment system

edit

The Wikipedia community created a system to assess articles belonging to a WikiProject. A WikiProject is a group of people who want to work as a team to improve Wikipedia[1]. Wikipedia Primary School SSAJRP is a sub section of the South Africa WikiProject. One of the most common methods used by WikiProjects to monitor and prioritise their work is that of assessing the articles within their scope. A very small or less-active project can keep a hand-compiled table of assessments; as the number of articles increases, however, a specialised process becomes necessary and requires a more sophisticated approach: bot-assisted assessments. The bot-assisted assessment scheme works by embedding assessments in a WikiProject's talk page banner. Practically speaking, it requires the set-up of a WikiProject banner and the addition of this banner to the discussion page of each article the WikiProject wants to assess (See such an example here : Talk:Gana_and_Gwi_people).
The assessment may be done on two variables. The first is "Quality", the second is "Importance". Each article may be tagged with a range of options to describe its overall quality (per Wikipedia criteria) : stub, start, C, B, GA, A, and FA. Stub typically describes an article providing a very basic description of the topic. However, all very-bad-quality articles will fall into this category. An FA (featured article) is described as professional, outstanding, and thorough; a definitive source for encyclopaedic information. Importance may be estimated as well, but is to be considered as being the "importance of the topic in the point of view of the WikiProject.
Once articles are tagged according to their quality and their importance, a summary of information may be automatically produce by automated system. The assessment system will provide an instant view of the overall state of the articles included in the WikiProject, but can also provide further information, such as the changes to their quality and importance parameters over time, indicate the top most edited articles within a WikiProject over a course of time (typically a week) etc.

Hypothesis was made that setting up an assessment system could bring attention to articles, thus resulting to further consultation and modifications of those articles. This could happen

  • when the set up of the general assessment system is done (it requires creation of several pages to operate, call for help when troubleshooting is met, and attention for the WikiProject when registering to bot service);
  • when all the articles talk pages are tagged with the WikiProject banner
  • in the long run, when the assessment pages are up and working as it bring forward new global information on the topic.

The process followed was

  • study and understanding of the operational set up of a WikiProject Assessment and of all reporting systems proposed
  • creation of all generic assessment sub pages and templates for Primary School Project
  • addition of banners to all articles talk pages and removal of previous categories to avoid cluttering the space
  • registration to bot services and follow-up with bot operators
  • fine-tuning selection of bot services

Results

edit

Peer review of content

edit

Academic journals

edit

2 journals have answered to the call for review out of 11 contacted. Both of them have declined the invitation.

Analysis

The reasons of the declination include the degree of specialization of the journal (too specialized in relation to the articles requested) and the impossibility to share reviewers contact details.

Next

In the meanwhile, the following actions have shed light on the future directions and strategies to involve journals :

  1. We have contacted and informed the [Library] about our project and ask them for suggestions about scientific, open access journals who might be interested in contributing to our list of topics. They have informed us about a similar project leaded by the Wikiproject Med; and their partnership with Sabinet a South African platform which hosts a number of Open Access journals that may be interested in getting involved with our project
  2. We have been spontaneously contacted by a board member of the Wikiproject Medical Foundation which is running a similar peer-review project in collaboration with the British Medical Journal (BMJ) to help improve Wikipedia's medical content. They have showed their interest in getting the article Malaria (which is in our list) up to featured standard and then having BMJ review it.

During the 2016, we address the research of journal involvement on the Sabinet platform and we try to trigger and foster the collaboration with the Wikiproject Medical Foundation.

Academic experts

edit

136 academic experts have been contacted. 33 experts have confirmed their involvement and 25 effectively have effectively done their review, for a total of 31 reviews received and 6 pending.
Expert reviewers comes from 5 continents, including the following countries: South Africa, Switzerland, The Netherlands, Australia, Mozambique, Italy, United States, United-Kingdom, India.
Reviews

Analysis

Just one forth of academic experts (around 25%) accept to do the review. It has been noticed that experts's involvement strongly depend from the period when the requests are forwarded. A greater number of experts agrees to make at the beginning of the academic semester (February / March and September / October) rather than at the end (we receive just one reply for call sent in December and any from July).

Next

Increasing the number of requests allows to get more people involved.

Making new content available on Wikipedia or for Wikipedia

edit

Full rewriting by an expert

edit
Mapungubwe Collection
  • Mapungubwe Collection
  • two experts proposed a rewritten text. The text of the second expert was integrated by ourselves and substantially copyedited. The expert was contacted in a second instance to provide more illustrations for the articles, which she did gracefully. Images have been sent and followed the usual OTRS confirmation of licence process. Images were integrated in the article by ourselves. The original text written by the second expert has been uploaded to Wikimedia Commons then added to the talk page for reference.
  • the process worked smoothly. The article has benefitted from various edits since the complete rewrite. It is now B quality on Wikipedia quality scale.
Flag of South Africa
  • Flag of South Africa
  • One expert proposed a rewritten text in a document. We updated the article based on the expert suggestion, noting in the comment box the name of the expert as author. The document used as reference was sent by email to the OTRS system to record the authorship and the licence proposed. OTRS agent then pasted a OTRS notification of attribution on the article talk page. See the result. The article is now considered B quality on Wikipedia quality scale.
Saartje Baartman
  • Saartje Baartman
  • The expert decided to update the article on her own. She brought very important improvements to the article.
  • However, following her work, it was noticed that part of the original text (which was preserved by the rewriting) was actually a copyright violation (copy of another website). Per Wikipedia policy, all the former versions of the article displaying the copyvio have been removed from the public view. The copyright violation was added the 9th Dec 2013 and discovered on the 9th of July 2016. The expert edits were done between the 29th of June and the 8th of July. Accordingly none of the versions edited by the expert are visible to the public, but the final result is. The outcome of the work was a serious improvement to the article, with the indirect benefit of having brought attention to the article and allowed to identify copyrighted content in the article, now removed. The article is now considered B quality on Wikipedia quality scale.
San Rock Art
  • San rock art
  • The expert provided a new version, which has been added to the talk page. The content has not been integrated as of October 2016.
Domestic violence in South Africa
  • Domestic violence in South Africa
  • whilst not a full rewriting, the expert provided detailed lots of comments beyond the form proposed. The document has been added to the article talk page and editors have been invited to integrate the suggestions. The content has not been integrated as of October 2016.

Triggering the production of content

edit

Publishing of the expert review

edit

Processed figures are

  1. Number of articles proposed for review by expert : 48
  2. Number of articles actually reviewed and review published : 31 (in 26 cases, the review by the expert was made after community review. In 5 cases, the expert review was made with no prior community review)

Outcomes measured were

  1. whether the review published brought any comment from the community
  2. whether the review published actually led to improvement to articles

Results as of mid February 2016 are the following

  • 8 articles have been improved, with clear use of the expert review
  • 5 articles have been seriously improved, but with no community comment added in the talk page and no explicit use of the expert review
  • 1 article recently reviewed only received minor improvement, but the PSP team has been contacted by a team currently setting-up external expert review by a medical journal and interested in joint action
  • 4 articles have seen minor improvement, with or without community feedback about the review
  • 9 articles have not been improved

Results as of october 2016 are the following

  • 9 articles have been improved, with clear use of the expert review
  • 7 articles have been seriously improved, but with no community comment added in the talk page and no explicit use of the expert review
  • 1 article recently reviewed only received minor improvement, but the PSP team has been contacted by a team currently setting-up external expert review by a medical journal and interested in joint action
  • 7 articles have seen minor improvement, with or without community feedback about the review
  • 7 articles have not been improved

Analysis
The figures suggest that the publishing of an external expert review has actually a beneficial impact, either because the review is actually used to improve the article, or because the process brings attention to the article itself.
A careful look on the articles who have not been improved at all indicate that most of those are specialized South-Africa related articles (such as san healing practices or water supply and sanitation in South Africa) probably difficult to improve by the layman and not perceived to be impactful articles by non south african editors.
It should be noted that the process of seeking external expert review did not raise any negative feedback from the community. At worse, we got no comment at all. At best, we got enthousiastic comments and the expert reviews were actually used to improve the articles. In one case, we even met another team organizing expert review of an article by a medical journal.

Request for assessment or reassessment of an article by Wikipedia community

edit
  • Case 1 : to be reassessed by several wikiprojects : 5 articles proposed Dec 2015
    Outcome : One reassessment made. 3 requests got feedback from community. None of the 5 articles was significantly changed.
  • Case 2 : to be assessed by a wikiproject for the first time : 5 articles proposed Dec 2015
    Outcome : One assessment received. Neither feedback from community, nor article significant change was observed.
  • Case 3 : to be reassessed by a wikiproject : 2 articles proposed Dec 2015
    Outcome : neither feedback from community, nor assessment, nor article significant change was observed.
  • Case 4 : to be added to a wikiproject : 2 articles proposed Dec 2015 for 3 adoptions
    Outcome : neither feedback from community, nor assessment, nor article significant change was observed.

Analysis
Many of the WikiProjects were set up in the 2007-2008, as part of the global effort of the community, to prepare a print version. Whilst some WikiProjects are still active, many are obviously more or less inactive (this may be concluded by the very limited discussion on the discussion pages of the WikiProject. In some instances, WikiProjects are tagged inactive, but this is not the general case. It is likely that most of our requests were simply not considered by anyone due to the WikiProject page not being a target of interest by community members.
However, this analysis is somewhat contradicted by the fact that in some cases, our request was disregarded in spite of being posted in the dedicated space, even though the automated log show regular reassessments of the WikiProjects articles. This suggests that some WikiProjects communities are actually still active and regularly reassessing articles, but the assessment is being done more or less randomly, without a coordinated effort to follow up with requests posted.
None of the two assessments actually updated (one new assessment on a new article created as part of the PSP programme, and one reassessment on a very developed article) lead to any further improvement.

Call for new article creation

edit

2 articles have been proposed for creation in Nov 2014 and created since then, explicitely as a result of the call for creation.

2 articles have been proposed for creation in Jan 2016. As of end of 2016, they have not been created.

Analysis
The size of the sample makes it difficult to draw any sensible conclusion. It should be outlined that the 4 articles proposed for creation are rather specialized articles, all directly related with South Africa, hence of no obvious interest for the majority of wikipedia participants. Note: 3 additional articles were proposed for creation only on the Wikipedia Primary School portal page after the 4 articles above were proposed on AFC. None of the 3 have been created.

Call for participation to a Writing Contest

edit

During the course of the writing contest,

  • 15 English speaking teams (27 participants)
  • 18 French speaking teams (24 participants)
  • and 1 team (4 participants) from Armenian

produced no less than ...

  • 71 new biographies of African women in English,
  • 122 new biographies of African women in French
  • and 41 new articles of African women in Armenian.

Amongst the 71 new biographies written in English, 7 were about South African women.

Analysis
The contest was considered highly successful and raised the attention and interest of many wikipedians willing to feature more African Women on Wikipedia. Online writing contests seem generally to be a successful mean to stimulate the writing of new wikipedia articles.

edit

4 articles have been featured in South Africa portal pages on the 6th of November 2015 [2]. As expected, those articles are focused on South Africa.

  1. w:en:Flag of South Africa
  2. w:en:Steve_Biko
  3. w:en:San_people
  4. w:en:Water_supply_and_sanitation_in_South_Africa

The articles have been reviewed for improvement end of February 2016.

Article 1 has not been significantly improved since the addition in November 2015. However, it went under a collection of vandalism attempts all over January (removal of valid information followed by restoration of the original content). The number of page view has been studied over the past 120 days. It does not show any significant increase since it was added to the portal page (around 1100 pages views per day). A peak of consultation was observed on the 27th of January 2016 (it does not coincide with the few modifications, made prior to that date).

Article 2 has not been significantly modified since the addition in November 2015. The number of page view has been studied over the past 120 days. It shows around 1200 views per day. It shows a small increase in pages views since early 2016. A peak of consultation was observed on the 10th and 11th of February 2016 (it does not coincide with the few modifications, made prior to that date).

Article 3 has significantly changed and improved since the addition to the portal. The number of page views has increased, from around 900 per day in October to over 1000 in February.

Article 4 was only edited resulting from the publication of the expert review in 2016. No other changes were observed since the addition to the portal. The number of pages view changed over time. Around 120 on average before the addition to the portal, it jumped around 175 the week after the addition to the portal. Then dropped again to reach a very low during Christmas season. Since then, the figure increased significantly, with an average around 185 per day.

Analysis
The size of the sample makes it difficult to draw any strong conclusion. But no obvious peak of consultation was observed after the addition of the articles to the portal page.

Edit-a-thons organised by the team in South Africa

edit

Analysis
The edit-a-thon organized in December 2015 did not succeed to reach a significant attendance.
The edit-a-thons organized in February 2016 received 6 and 4 participants respectedly (not counting trainers/facilitator/volunteers)
During the edit-a-thons organized in February, edits and expansions were so far made during the edit-a-thons to the following articles:

Next
The results are rather modest both in terms of numbers of participants and in terms of edits made. So the next steps will be to discuss whether additional similar events should be further organized or whether it makes sense to focus on something different to get “better” results.

Edit-a-thons organised by other parties

edit

Three groups have been contacted, AfroCROWD in New York City, Wikimedia UK and Wikimedia Washington.

Per suggestion of the AfroCROWD leaders, several articles have been added to the Afro CROWD task list on the 5th of July 2016. Those may be edited either online, or during some of the edit-a-thons organized by AfroCrowd.

Wikimedia UK has been approached and project introduced. People approached confirmed they would keep the issue in mind and discuss it at further face to face meeting.

Some articles were proposed to an edit-a-thon Occupational Safety and Health - Washington DC.
A meet-up was proposed by Wikipedia:WikiProject Occupational Safety and Health : Wikipedia:Meetup/DC/Safe and Healthy at Work (July 15th 2016). The following articles were added to the suggestion list for the WikiProject and the meet-up

  • w:en:Fire safety: Needs style editing, sources, and has currently a limited geographical scope
  • w:en:home safety: Require expansion of this article. More global perspectives. Currently an orphan
  • w:en:Hand washing: Requires copyediting, review, and sources

Note:No edits were done on the 15th of July

Bold reassessment of articles

edit

12 articles were boldly reassessed in term of quality. The reassesment was applied to all WikiProjects in which the articles were categorized. The modification ranged from no assessment, or stub or start to C or B level articles. Some of the articles were very topic specific such Constitutional Hill, whilst others were quite mainstream, such as Gender role or Map.

The outcome has been evaluated after one week. Three observations were made

  1. none of the reassessments was reverted by anyone
  2. none of the reassessments raised any discussions
  3. no modifications were observed in the 12 articles in the week following the change of assessment

Analysis : reassessment to C or even B level is non controversial and does not raise anyone interest.

Call for improvement of stub articles as part of a global challenge

edit

7 articles were proposed as stubs to be improved, for double score

At the end of the contest, all 7 stubs listed have been improved, 3 of them very significantly, making it our most efficient strategy so far.

Listed as quintuple score GA

  • Apartheid (+)
  • Herero and Namaqua genocide
  • Saartjie Baartman
  • San people

Of the 4 articles listed in the quintuple score, only one has been significantly improved. It should be noted that all those articles are already pretty well developed articles, requiring a certain degree of expertise and in one case very controversial. The 5 points could only be granted if the article not only was improved but also be granted a Good Article assessment, which requires an additional effort.

WikiProject assessment system

edit

Analysis
The process was quite time-consuming to set up. The set up of the assessment certainly raised attention in particular due to

  • the fact Primary School was originally set up as a sub section of South Africa WikiProject rather than set up as a WikiProject on its own (the visual implication being that the page is primarily tagged as WikiProject South Africa rather than clearly tagged as Primary School Project) (example). Accordingly, some editors rejected the addition of the banner to the talk page of some articles, arguing that the article was unrelated to South Africa and consequently should not be tagged as such (examples of this situation are Outer space, or Settling, both of which are not clearly related to South Africa). This situation is still occurring from time to time, but the bot system allows to detect unwanted banner removal. However, this suggests that set up of project as of sub section of a larger WikiProject, should be considered carefully and generally avoided due to the confusion it creates.
  • Another issue met was related to the original category set up to "tag" articles belonging to the Primary School Project, which was considered non obvious (including "SSAJRP"), category consequently put up for deletion. Consensus was found to rename the category into a very obvious one Wikipedia Primary School articles.
  • A third issue was met due to the fact the articles had been first put in a maintenance category from their main space rather than their talk space, thus disrupting the community agreed practice.

The discussions around those issues were not always very friendly, but ultimately consensus was found and a volunteer editor helped fix the categories.

Results
The assessment set up can be found on Wikipedia_Primary_School/Assessment. Only quality is assessed (as it was estimated that all articles had the same importance within the Primary School Project). It provides information which helps tracking activity on articles and is a useful tool to do group work on articles. However, there is no evidence that it actually impacted editorial activity at the moment.

Documentation

edit
edit
  1. ↑ https://en.wikipedia.org/wiki/Wikipedia:WikiProject_Council/Guide/WikiProject
Retrieved from "https://meta.wikimedia.org/w/index.php?title=Research:Wikipedia_Primary_School_SSAJRP_programme/Process&oldid=17246406"