Hi, I'm Xiangju. This is my wiki discussion page, feel free to leave messages. ^_^ Xiangju (talk) 13:48, 22 May 2013 (UTC)[reply]

A beer for you! Solomon7968 (talk) 16:33, 22 May 2013 (UTC)[reply]

Thanks, Solomon. Xiangju (talk) 18:40, 23 May 2013 (UTC)[reply]

Hi, I've never edited at WikiProject Chemicals as far as I'm aware and I'm certainly not a member... GiantSnowman 17:36, 22 May 2013 (UTC)[reply]

Hi GiantSnowman, it should be WikiProject Football. Sorry, I made a mistake, and correct it back now.Xiangju (talk) 17:48, 22 May 2013 (UTC)[reply]

Aha, thanks for clarifying. What is it you need from me? GiantSnowman 19:09, 22 May 2013 (UTC)[reply]

I would like to invite you to take a survey about the running of WikiProject and possible performance measures for WikiProject. Would you be avaible for the survey? If you are avaible for the survey, could you send me an email? I'm new to Wikipedia, I can't send too many emails to Wikipedia editors due to anti-spam measure. Thank you for your interest. Xiangju (talk) 19:12, 22 May 2013 (UTC)[reply]

Hi, yes please send me an e-mail. GiantSnowman 13:25, 23 May 2013 (UTC)[reply]

Hi GiantSnowman, I just sent you an email. Please check your mail box. Thank you. Best regards. Xiangju (talk) 15:56, 23 May 2013 (UTC)[reply]

Hello, Xiangju. You have new messages at Nick-D's talk page.
You can remove this notice at any time by removing the {{Talkback}} or {{Tb}} template.

Nick-D (talk) 10:10, 23 May 2013 (UTC) [reply]

Hello, Xiangju. You have new messages at Hoary's talk page.
You can remove this notice at any time by removing the {{Talkback}} or {{Tb}} template.

Hoary (talk) 12:03, 23 May 2013 (UTC)[reply]

I don't have a problem with completing a survey for your research.  BIGNOLE  (Contact me) 17:29, 23 May 2013 (UTC)[reply]

Hi Xiangju,

Forgive me for being nosy, but I could not help myself when I saw that you have contacted scores of Wikipedians with a request to participate in a survey regarding WikiProjects. I am keenly interested in this topic, so thought I would drop by your page and check out the survey.

I am curious to see that your survey is done in private via email, rather than the usual open approach used at Wikipedia. I don’t know about others you have contated, but someone like me would never be able to participate in your survey, since I do not communicate via email. I only communicate with editors through wiki-talkpages. I believe you can publish your survey directly on Wikipedia and avoid the need for email.

By the way, I noticed that you have not contacted user: Mabeenot and user: WhatamIdoing who may be able to give you insight into wikiprojects. XOttawahitech (talk) 14:58, 24 May 2013 (UTC)[reply]

Hi Ottawahitech, Thank you very much for your suggestions. I'm not that familar with how would other people or researchers conduct a survey in Wikipedia nor I know where should I post the survey in order to get reasonable response from editors. I sent an email to this wiki account: wiki@wikimedia.org enquiring about how to conduct a short survey in Wikipedia. But I haven't received response. One of the editors I contacted by emails suggested that I can do the survey by contacting people via personalized emails (which works well), posting requests on user talk pages, and post on the wikiprojects' talk pages (which might get fewer responses). To make the survey work well and get response from experienced editors, I choose to post invitation for the survey on user talk pages. In case, some editors are unable to finish the survey via email, I will post the survey questions on their user pages. I will contact the two editors you mentioned. Would you like to or would you be available to take our survey regarding the running of WikiProject and possible performance for it? Best regards. Xiangju (talk) 15:41, 24 May 2013 (UTC)[reply]

• Thanks for responding so promptly, Xiangju. You caught me just as I was getting ready to signoff. user: Mabeenot is a wp:Signpost reporter who interviews wikipedians involved in wikiprojects and I know he conducts his interviews on Wikipedia itself. I don't have the time right now to look into it, I wonder if anyone else can help you? Maybe someone over at the wp:teahouse? XOttawahitech (talk) 15:59, 24 May 2013 (UTC)[reply]

Wikipedia's new notification system informed me that someone was talking about me, so here I am. :) I've interviewed over 100 WikiProjects during the past three years. I typically create a single page for the interview questions and then post invitations on users' talk pages. Everyone's responses are on the same page, so people can view the other responses and they sometimes elaborate or disagree. For example, here's an interview I'm conducting with WikiProject Geographical Coordinates. Feel free to ask me any questions you may have about WikiProjects and peruse our archive of previous interviews. –Mabeenot (talk) 17:23, 24 May 2013 (UTC)[reply]

Hi user: Mabeenot, thank you very much for joining the discussion. We are talking about the survey I conduct in Wikipedia about WikiProject. To be brief, I analyze the Wikipedia data to analyse what communication patterns in WikiProjects are likely to be predictive of efficient operation. We assume that efficient projects manage to achieve promotions with a relatively small number of edits. Base on this study, we can provide insights into the health of online teams and share these knowledge and experience among teams for the benefits of whole community. We find in our analysis that some WikiProjects have been very efficient from the beginning (i.e., from 2006), while other projects have been less efficient in that the projects achieve promotions with a relatively smaller or larger number of edits. Our evaluation indicates a number of network measures are associated with the efficiency of WikiProjects. However, to make our results more convicing, we would like to conduct a brief survey with experienced Wikipedia editors about the running of WikiProject and possible performance measures for it.

Since different WikiProjects tend to have different level of efficiency, I would perfer to conduct the survey individually rather than create a page for all editors to answer the questions. In this way, I can ask slightly different questions for editors from different WikiProjects. You mentioned that, you've interviewed over 100 WikiProjects, I would like to ask you two questions. In your experience, (user: WhatamIdoing and XOttawahitech, you are also welcome to provide insight into these questions.)

(1) What factors make a WikiProject efficient or less efficient? Are there possible relationship between the efficiency (defined using the measure I mentioned above) of WikiProject and the communications in project talk pages?

(2) Which WikiProjects are well-organized, efficient and productive that deserve further investigation and more concentration for researchers?

Thank you very much for taking your time. Best regards. Xiangju (talk) 17:59, 24 May 2013 (UTC)[reply]

We assume that efficient projects manage to achieve promotions with a relatively small number of edits.

I don't think your assumption is valid. But you would first have to explain to me what "promotions" and "small number of edits" means. Do you mean a small number of edits to the articles, or to the WikiProject's pages? If the former, then I can tell you with a fair level of confidence what the major factors are: you need a subject area that isn't popular with either new or immature users (thus avoiding good-faith mistakes and vandalism), and you need editors who prefer to work as lone wolfs (very little collaboration, which increases overhead) and who like to do major editing by making a few large changes, rather than saving in between each sentence. The fastest way to "promotions", if by that you mean the WP:1.0 team's assessment ratings, is to have someone (like me) who regularly checks the status of pages and updates the WP:CLASSES.

IMO the most interesting projects are WP:MILHIST (formal organization, highest edits by humans in the project space [for communication with each other]), WP:MATH (big project with generally high-quality pages on subjects of little interest to the average user, thus avoiding vandalism), WP:HWY (small project with emphasis on getting pages through the WP:Good articles process), WP:FILM and WP:VIDEOGAMES (big projects with subjects of much interest to average users, to contrast with MATH) and WP:MED (teamwork).

To find the busiest discussion pages, see Wikipedia:Database reports/WikiProject watchers and sort by active watchers.

On a related point, I'd started a while ago to see whether I could discover any qualities that predicted success in a new WikiProject. I didn't get very far, but it seemed that having more experienced users involved in the proposal (e.g., WP:WikiProject Women's history) was a sign of success. Projects started by inexperienced users or by one or two users tended to be dead in less than a year. If that interests you, then I can tell you more about the idea. WhatamIdoing (talk) 18:30, 24 May 2013 (UTC)[reply]

Hi user: WhatamIdoing, thanks for your participation.

First, what "promotions" and "small number of edits" means in our research context: promotions mean promote a Wikipedia article from low quality level to high quality level (the same meaning as in Wikipedia). The issuse you just memtioned also mentioned by some other editors I contacted: project domain can have inpact on the number of edits per promotion. We have also considered for this issue. So we choose to normalize the number of edits per promotion in Wikipedia each year, and then use z-scores to measure the relative number of edits for each WikiProject. This solution should solve the issue to some extent.

Considering the fact you and other editos mentionthed the subject of WikiProjects can have influence on the relative number of edits for promotions, we will be careful about the selection of WikiProject in our analysis.

Regarding this point: "it seemed that having more experienced users involved in the proposal". Some researchers have suggested silmilar findings in open source projects, you may refer to this paper: " Network Effects: The Influence of Structual Capital on Open Source Project Success ".

By the way, how can you know a WikiProject created by which user? Is there an API for this purpose?

Regards. Xiangju (talk) 18:59, 24 May 2013 (UTC)[reply]

If it was formally proposed (recommended, not required), then you can see who proposed it at Wikipedia:WikiProject Council/Proposals. You can also see who created the WikiProject's pages at the Help:Page history. WPMED, for example, was created by a user named Ropers[1].

How will you decide whether the article improved as a result of the WikiProject? If, for example, I go improve the article on a film, will WPFILM get credit for it, even though they weren't involved? How will you handle new articles? If I create a B-class article for WPMED, will that count as "improved" (from a non-existent/below-stub level) or not?

You might want to track down some of the WT:DERM guys. They did a journal article on page creation that doubtless had to deal with some of the same questions. WhatamIdoing (talk) 19:16, 24 May 2013 (UTC)[reply]

Hi user: WhatamIdoing, you just memtioned another problem for research in Wikipedia. Regarding this problem: "How will you decide whether the article improved as a result of the WikiProject?" At the momemt, we only consider articles that are tagged by one WikiProject in our analysis. Again, the normalized z-scores we used to measure project efficiency should also solve this problem to some extent. Do you think is it reasonable to assume that articles tagged by single WikiProject would mainly contribute by members from this WikiProject?

Thank you very much for the useful link. Regards. Xiangju (talk) 19:29, 24 May 2013 (UTC)[reply]

No, I don't believe that it's always reasonable for us to assume that a WikiProject's members are the people who wrote the article. Look at In silico medicine, a new article written by a new editor. I hope that this editor will stick around and join WikiProject Medicine, but so far, I don't think that we can claim credit for him (or her). WhatamIdoing (talk) 23:28, 26 May 2013 (UTC)[reply]

In reply your two questions:

1) A WikiProject's efficiency is hard to gauge since there is no way to determine that a particular project was responsible for an article's promotion from a lower assessment to a higher one. The closest thing we have is an aggregate number based on changes in a project's workload over time, termed "WikiWork." Here is an article I wrote explaining WikiWork. The article also has links to the other metrics available to anyone researching WikiProjects. From my personal experience, I'd have to agree with User:WhatamIdoing that a single highly motivated editor will often make more improvements to an article than any collaboration between the other members of a WikiProject. WikiProjects tend to be a home base where editors go when they need help or want to discuss something pertaining to that subject. Unfortunately, some of the projects with the most talkative talk pages don't actually get a lot of work done.

2) The ideal project has always been WikiProject Military History. For whatever reason, a close-knit group of dedicated editors who care passionately about military history have been attracted to Wikipedia for years and they all gravitate toward WikiProject Military History. The members of the project have created their own goals and metrics to measure their productivity. They have an active talk page and publish a monthly newsletter. They even have contests, service awards, and other benefits that come with participating in the project. In addition to the projects mentioned by User:WhatamIdoing, there are a variety of other highly active projects that come to my mind: WikiProject Tropical cyclones, WikiProject LGBT studies, WikiProject Birds, WikiProject Doctor Who, WikiProject Football, WikiProject Geology, WikiProject Palaeontology, WikiProject Ice Hockey, WikiProject Cricket, WikiProject Trains, WikiProject Ships, WikiProject Oregon, WikiProject Russia, WikiProject Poland, WikiProject National Register of Historic Places, and WikiProject Visual arts. One very active WikiProject that would fall outside of your research methodology is WikiProject Guild of Copy Editors because their mission is to copyedit articles that users have tagged as needing copy editing, so the project does not actually assess articles. –Mabeenot (talk) 04:21, 25 May 2013 (UTC)[reply]

• Even though I am interested in this area I know little about it, so I am not sure that I can add value to your research. I am also not a member of any WikiProject. I am an outsider who frequently visits WikiProject pages, sometimes passively to find a certain piece of info, other times to ask for help or alert others through the WikiProject talkpage. Also, you seem to have made an assumption that the only purpose of WikiProjects is to "promote" wp: name-space articles to a higher wiki-rank, something I know litle about. I am more interested in preventing deletions (not only of articles) and in fostering a more friendly atmosphere amongst editors. In terms of your survey, the second is harder to quantify.

Here is a little tidbit that I noticed:

• WikiProjects do not truly understand the value of a continuous supply of fresh blood and put little effort into acquiring new members.

• Many WikiProjectects have a tendency to display slight (unconscious) hostility towards potential new members
• Many WikiProjects do not like to spend time bringing newbies up to speed. They like to have contributors who land running on their feet.
• Navigation of WikiProjects is non-standard making it difficult for WikiProject-outsiders and potential new members.
• Many editors sign on as members of a WikiProject, but seem to be inactive after signing on. XOttawahitech (talk) 15:20, 25 May 2013 (UTC)[reply]

See Wikipedia:Database reports (section "WikiProjects").

—Wavelength (talk) 17:01, 25 May 2013 (UTC)[reply]

Hi all (user: Mabeenot, user: WhatamIdoing and XUser:Ottawahitech), thank you for joining the discussion. Although "WikiWork" seems to be a good measure for a WikiProject's workload, I have some concerns about the measure: after some exploration on Wikipedia article promotion data, we find that on average, different level of (article) promotions require different number of (article) edits; and it generally becomes more and more difficult to promote articles as time goes by in that for the same promotion, it generally takes the editors more number of edits in recent years. For instance, we find in our dataset that on average, it takes less than 50 number of edits to promote an article from Stub-class to Start-class, takes about 60~230 number of edits to promote from Start-class to C-class, and takes about 170~330 number of edits to promote from B-class to GOOD ARTICLE. The more complete and rigorous rules and regulations (i.e., the more completeness of the Wikipedia article assessment system) in Wikipedia in recent years may explain this finding to some extent. The finding suggests that it would be more reasonable to assign different weights to different level of article promotions, while "WikiWork" treats all levels of promotion equally (i.e., assigns the same weight for different levels of promotion).

Moreover, three editors from efficient WikiProjects I contacted with responded that: "it is reasonable to assume that articles tagged by a single WikiProject are mainly edited/contributed by members of that project. Project members tend to know where the necessary resources are located, either on the Internet or in libraries. That is not to say that people from outside the project do not help, but project members tend to create more content than non-members. "

Based on the response I got from several experienced WikiProject members, I think as long as we focus our analysis on WikiProjects that belong to scientific and static subjects (e.g., WikiProject Physics, Mathematics, Medicine, Military History, Aviation, Plants), it's reasonable to use our current measure of project Efficiency.

Best regards, Xiangju (talk) 17:27, 27 May 2013 (UTC)[reply]

• Thanks so much for continuing to alert me to updates on your talkpage, Xiangju. As I said before, I don't have the knowledge to add to this discussion, but this is good education for me. For example, I just looked up the term "WikiWork" which I was unfamiliar with, and discovered a Signpost article on the topic which I missed. So, again, I really apprciate being a part of this discussion. XOttawahitech (talk) 19:30, 27 May 2013 (UTC)[reply]

  I have been thinking about this assertion that WikiProject participants contribute more than anyone else. I'm sure that some of them believe this, but I don't think it's true. In fact, I believe that the Wikimedia Foundation has done research that indicates that unregistered/IP users contribute more article content than registered users. Since unregistered users aren't members of WikiProjects (because they don't have an account), then that would pretty much disprove the assertion that WikiProject members do most of the work. I should go look for a link to the page that I'm remembering. WhatamIdoing (talk) 14:41, 6 June 2013 (UTC)[reply]

  I found one of the pages: http://www.aaronsw.com/weblog/whowriteswikipedia I think you should seriously look at the methodology used there. It would require fairly significant computing resources to replicate this on a large scale, but you would actually find out who wrote the article, rather than who fixed the most formatting or who removed the most contributions by other people. WhatamIdoing (talk) 14:49, 6 June 2013 (UTC)[reply]

Hello. If you are doing survey research on the Wikipedia community then please contact the Research Committee at meta:Research:Committee. Tell them about the nature of your research and how you are contacting users. I regret to say that there is not a well defined process for registering research projects, but still the community requests I think it is a good idea that human subject research be registered. Thanks. Blue Rasberry (talk) 16:30, 25 May 2013 (UTC)[reply]

It might be a good idea for you to draft an explanation of your research project somewhere in your userspace so that anyone can read it. Thanks. Blue Rasberry (talk) 16:32, 25 May 2013 (UTC)[reply]

Hi User:Bluerasberry, I'm doing a short survey in a small group of editors. I would think that the survey is a supplement to our current research rather than a survey research in which people analyse the hypothesis based mostly on the survey data. We study Wikipedia data, and get some results. However, to make the results more convincing, we would like to conduct a short survey with experienced Wikipedia editors. Actually, for those interested in the survey, they can know about the survey by reading the thread entitled “Your wikiproject survey” in my talk page. If you still think that we should contact the Research Committee at meta:Research:Committee, we can do that. Best regards. Xiangju (talk) 22:34, 25 May 2013 (UTC)[reply]

I looked again and there is no guideline. However, if you are affiliated with a university or research institution, I think that you should treat this by your institutions guidelines for any human subjects research. Surveys here take time away from the volunteer efforts which people are already doing and there is no certainty that your work is giving back to the community to the extent that the community contributes to this survey. I do think it is a good idea for you to inform the Research Committee that you are doing work but there is no guideline saying that you must. Good luck in your project and thanks for your interest in the Wikipedia community. Blue Rasberry (talk) 16:33, 26 May 2013 (UTC)[reply]

Hi User:Bluerasberry, thanks for reminding. I emailed the Research Committee about the survey I'm doing in the Wikipedia community. Regards, Xiangju (talk) 17:30, 27 May 2013 (UTC)[reply]

I guess the guidelines you are after can be found at meta:Research:Subject recruitment. -- Daniel Mietchen - WiR/OS (talk) 23:32, 28 May 2013 (UTC)[reply]

Hi Xiangju,

I noticed you stopped participating in Wikipedia on 27 May 2013. I hope you were not driven away by too much bureaucracy? XOttawahitech (talk) 23:32, 11 June 2013 (UTC)[reply]

Hi User:Ottawahitech, actually, I'm busy with analyzing the results for project at the moment. Like I memtioned earlier, I mainly analyse the whole Wikipedia data dump for research purpose from social network analysis perspective, so I would be not that active in editing Wikipedia articles or talk pages. I will continue working on Wikipedia data as my next project.

By the way, what do you mean "too much bureaucracy"?

Regards, Xiangju (talk) 17:53, 14 June 2013 (UTC)[reply]

• Thanks for letting me know you responded. What I was refering to is this. XOttawahitech (talk) 18:59, 14 June 2013 (UTC)[reply]

Hello, Xiangju. Your question has been answered at the Teahouse Q&A board. Feel free to reply there!
Please note that all old questions are archived after 2-3 days of inactivity. Message added by benzband (talk) 16:32, 23 November 2013 (UTC). (You can at any time by removing the {{teahouse talkback}} template.[reply]

Eric Corbett has also replied there (with a useful bit of info which may make things more difficult). benzband (talk) 02:16, 24 November 2013 (UTC)[reply]