User:Mu301/Learning blog/2016-01

Announcement
The Wiki Education Foundation is about to launch Wikipedia Year of Science 2016. This could be a great opportunity to expand science resources here at Wikiversity. Please share your thoughts at Year of Science 2016. --mikeu talk 14:05, 2 January 2016 (UTC)

Update
See projects and events at Wikipedia:Year of Science to get a sense of possible activities that we could work on during the year. --mikeu talk 12:22, 31 January 2016 (UTC)

Welcome and expand considered harmful
The template was called Whas which was short for "Welcome Header And Search." There was once a trend here that our greatest weakness was a lack of pages. Thus began a movement to bulk generate a very large number of stub articles. Most of them only contained a couple of links to Wikipedia and maybe a sentence or two. A very large number were empty, containing only the template and no content. There were a number of people who argued that if we just had enough of these it would be so enticing to new users that they would edit the stubs and flesh them out. Here's the complete listing of edits made to a typical example called Wormhole:


 * 2 September 2008
 * 2 September 2008‎
 * 10 January 2009‎
 * 17 June 2009‎
 * 15 July 2009‎
 * 25 August 2009‎
 * 15 December 2009‎
 * 27 December 2017 deleted (No educational objectives or discussion in history)

Most of the edits are wikignome maintenance like adding a category or removing a broken link. There's about 100 pages using this template that contain very little except for boilerplate. There are hundreds more that use we "Welcome and Expand" like Topic:Metaphilosophy. These pages haven't been edited since about 2009. The presence of these pages really didn't do much for development of content here and it makes searching very difficult. cf GOTO Considered Harmful --mikeu talk 02:14, 3 January 2016 (UTC)


 * Notes --mikeu talk 16:26, 5 January 2016 (UTC)
 * See my comment at the bottom of Wikiversity:Community_Review/User:Wikademia (17 July 2009)
 * A bot tagged the empty pages for speedy Notices_for_custodians/Archive/3 (5 September 2009)
 * A very small number were eventually turned in stub resources, many still remain devoid of content except for the template search buttons. See pages that link to   and

Survey
I really need to reexamine Why academics do and do not participate on wikis. There are slides from a presentation describing the results of a expert participation survey but it is a bit lacking in detail. --mikeu talk 23:59, 4 January 2016 (UTC)

A medium for scholarly publication
I found an interesting reference to the use of a wiki as a medium for scholarly publication. I've also been updating User:Mu301/Refs which lists research on or about wiki use. --mikeu talk 00:25, 31 January 2016 (UTC)



Public humanities
It was a pleasant surprise to (re-)discover The Crafting Freedom Project. It is a wonderful exercise to introduce teachers to the public humanities and involve students in discovering the relevance of history to their own lives. There are at least a couple of resources that fall into the Category:Public humanities but there is no organizational structure to bring them together. I'll need to put some thought into creating an introduction to the topic. I also learned that Public humanities is in sad need of references, so I've decided to adopt the article. --mikeu talk 21:27, 25 January 2016 (UTC)

Pending changes


I'm now a Wikipedia:Reviewer so I'm familiarizing myself with Wikipedia:Pending changes. The FlaggedRevs extension to MediaWiki could be very useful here to give stability to resources created by educators. For example, a grade school teacher who does not want to risk a student seeing age inappropriate vandalism. Another instance would be controversial topics like politics, sexuality, or religion which are frequent targets of vandalism. Unlike a permalink the "safe" reviewed page would be the default landing for any visitor until a reviewer approves the recent edits. (A click on the "View history" tab shows edits that are pending review.) Currently I'm watching some biographies that include:


 * Johannes Kepler
 * Carl Linnaeus
 * Samuel Morse
 * Ptolemy

Also Milky Way (the galaxy, not the w:chocolate bar). I'm guessing these are "test edit bait" given the prominence of the subjects in elementary education (or interest among students of that age for the last one;) Do school children look up the articles and scribble on them??? Other articles are an obvious source of contention such as Climatic Research Unit email controversy and Creation–evolution controversy. The system works quite well and is a better tool than semi-protection in some cases. --mikeu talk 22:48, 25 January 2016 (UTC)

An essay on the philosophy and practice of the "wiki way"
I feel like we need to have a conversation about anti-vandalism and the tools to prevent it. There seems to be some misconceptions about the use of page protection and other features.

The basic design of mediawiki software includes features like Undo and Rollback for occasional vandalism, and the Block for repeated vandalism from a single source. This is the appropriate response in the vast majority of cases. I've had unprotected resources that have only seen 3 vandalism edits in 9 years. The number of productive contributions to the same page from anonymous editors was greater than that. Should that page be protected just because I'm annoyed by pushing the rollback button once per three years? We have Curators, Custodians, and many other members of our community (both at Wikiversity and globally throughout the WMF who watch our recent changes) that look for this activity and remove it. The system works very well.

Page protection is primarily intended for situations where vandalism is high profile (like the Main Page) or where a single edit could affect a large number of pages (like a heavily used template.) Wikiversity does have some unique needs that differ from other projects. I can see a final accepted paper to one of our journals being "frozen" by page protection, perhaps with a second editable copy that is not marked "reviewed." This could also be accomplished by uploading a PDF file stamped "reviewed" and leaving the wiki page unprotected... There are a variety of ways this could be accomplished. Suggesting page protection as the first, and only, mechanism seems IMHO to embody a lack of imagination.

When I hear contributors casually suggesting page protection on a routine basis I shudder. This runs contrary to our WMF approved mission statement:

"'Page protection - there may be a need to restrict editing of pages or groups of pages to within groups of people (such as a research community, for example). However, this needs to be done with care, so as to minimise the exclusion of people to the work of that community.' Approved Wikiversity project proposal, 30 July 2006"

The Wikiversity Proposal (which btw is indef protected as it is both an historical page and in a sense a kind of legal document) places great emphasis on Learning groups and cautions against excluding anyone from joining these groups. The creation of Wikiversity was conditional on this premise. The WMF core values include the idea that the best way to generate knowledge is to ensure that participation is as broad as possible. You might be able to convince some of the local community that this is a good idea, but trying to sway the members of our governing Foundation that we are on the right track is like handing out panda sandwiches at a PETA convention. Many of them (myself included) are fanatically committed to the radical notion that "Information wants to be free" and that everyone can contribute to creating a world "in which every single person on the planet is given free access to the sum of all human knowledge."

I've been investigating the FlaggedRevs extension to Mediawiki. This solution could be the best of both worlds. Anyone could edit a page but the edits are held in limbo and not visible by default until a Reviewer approves or rejects the edit. There are a number of details that we need to work out before requesting this extension. There are certain switches that determine default behavior and we would need to define how we want it to work before submitting a request. I would like to suggest that others try out Wikipedia:Reviewing to learn the details of how it works. The page also has instructions on how to request Reviewer status at WP.

One other note is that Page Protection and Flagged Revisions are not an all or nothing solution. Both include an expiry time which should be used if we expand the use of these tools. There is no reason to indef protect a typical lesson beyond the term of a semester or school year if it doesn't repeat the following year. I can see indef protect on quiz questions or an exam study guide as this has real world consequences. In many other cases a temporary Flagged Revisions setting might be better suited to these tasks. Keep in mind the "indef" mean "for all of eternity." In essence you are protecting the page for a duration of decades or centuries. Is this reasonable? I've seen pages where an instructor swore that indef was necessary only to see them leave the project after 5 years.

On a personal note: as a scientist I find the idea of Evidence-based policy very appealing. Start with an objective look at the problem that we are trying to solve. Is there evidence that it really is a problem? How big of a problem is it? What methods have a proven track record of solving similar problems? Tossing out solutions that are in need of a problem (the shotgun approach) is an inefficient use of our time. --mikeu talk 19:22, 26 January 2016 (UTC)

Comment by User:Guy vandegrift
I am coming around to your viewpoint, and certainly retract the suggestion that we use "page protection" as a selling point to bring in Curators. At the time, I was looking at it from the recruitment angle, which asks, "Is there anything about new Curator status that will bring teachers into Wikiversity". The answer was page protection, and in retrospect, I overplayed that card.

I also explored page protection as a way to introduce a bit of "individualism". After you (User:Mu301) convinced me that page protection was an inappropriate way to achieve this, I decided to create an online journal. Then, I was delighted to see that Wikiversity Journal of Medicine had already pioneered exactly how to create such a journal. The Second Journal of Science has only one protected page, and if the community so requests, I could unprotect that page.

I like your essay, and think the next step is to categorize instances where page protection is required:
 * 1) High volume pages where even a few minutes of vandalism would do harm, and where it is plausible that page protection is a higher risk. (I emphasize plausible because I have no evidence in this regard).
 * 2) An active course where disgruntled students could use vandalism to hurt their fellow student's education in order to cover their own inadequacies. This is something Wikimedia should take seriously because there is no evidence that this won't happen (except that it hasn't happened to me in the past 2 years).
 * 3) Sensitive data where there is no need to ever edit. I am speaking, for example about this steam table This and the next item clearly represent weaker claims.  In the previous two, harm was done by not protecting.  But in this case, the argument is that since there is no reason to edit, we should forbid editing.  This is a weaker justification than one that points to actual harm (because I believe that steam tables are a low traffic item).
 * 4) Another "debatable" reason to page protect involves large amounts of text that the writer is too busy to watch carefully. I am trying to write an open source  Quizbank of exam questions.  Let's consider the extreme limit: Suppose, hypothetically that it was necessary for each question to have it's own page.  That would place almost a thousand of pages on my watchlist.  In the case of Quizbank, the solution is to many questions on each quiz, which can keep my watchlist to a managable size.

It is highly likely that I left out essential examples. For example, both of our Wikiversity Journals (WJM and SJS) store permalinks to the checked (accepted) versions of the submitted article. It would be tempting for a contributing author to make a well-intentioned edit to "upgrade" a checked version with corrections. But routinely allowing such practice could hurt the journal's reputation.

Another example of using page protection against well-intentioned edits involves my current effort to have students improve the quality of my exam questions. I have invited students to submit corrections, not to the current version of the quiz, but on a special subpage devoted to each question. With 65 students making such corrections, it is inevitable that one of them would get confused and edit the actual question. For an example of this, see question 7 of this page-protected quiz, and this sample of a recent student's effort to improve the quiz.

Recommendations:
 * 1) We need to carefully vet candidates for Curator status to verify that they understand that routine page protection needs to be avoided.
 * 2) Curators and Custodians who page protect should welcome audits of their protection policies.  Looking over my pages, it would be possible for me to reduce the number of pages that I protect.  I only recently realized that   permits permalink addresses that are easy to edit.  It would be possible for me to unprotect every quiz in Quizbank and secure my versions of them with a single page-protected page full of permalinks.  In fact, this index page could be in my userspace.  Updates are as simple as updating the permalink "oldid".  I would like to first beta-test having 65 student editors on Wikiversity using protected pages before I open everything up to edits.
 * 3) Keep in mind, that we at Wikiversity are following the spirit of Wikimedia's "make information free", but at the same time are exploring new ways to do it. It might be necessary to alter the traditional approach to page-protection, just a bit.

--Guy vandegrift (discuss • contribs) 00:38, 27 January 2016 (UTC)


 * I'm the fanatical open source zealot that your mother warned you about ;) But, I'm also very open to new way of doing things. FlaggedRevs is the tool of a Wiki Wizard. Not as clumsy or random as a permalink; an elegant tool for a more civilized age. I'm very interested in discussing ideas as to when and how certain tools can benefit the development of learning resources. I am in agreement that quizzes and raw data really shouldn't need to be edited, and the consequences of a student using incorrect information in an exam are serious. I'll discuss this at length later. --mikeu talk 03:03, 27 January 2016 (UTC)

Evidence that Wikipedia is loathe to page protect even when is is "reasonable" to do so

Both your "fundamentalist" open source viewpoint, and my "liberal reformist" take on this issue can be seen in the fact that Wikipedia routinely bypasses the issue with educational dashboards like w:Wikipedia:Wiki Ed/Wright State University/Introduction to Astronomy (Spring 2016). They do not page protect these educational pages. Instead the dashboard routinely uploads a refreshed version from wikiedu.org. There are many pages that use this backdoor approach. See: w:Special:Prefixindex/Wikipedia:Wiki Ed

I did a harmless test edit on my own page, and there really is no page protection. This refusal to protect essentially uneditable pages shows how much they are loathe to page-protect. I am not fond of this because you need to do "markdown" on the website, which then "marks up" to wikitext, and the markdown is just one more skill I wish I didn't have to learn. My favorite way to page protect is now that extension you propose which allows edits to be first checked and approved (forget what its called) -- PS: My calling you a "fundamenatalist" and myself a "liberal reformer" is meant to be taken very narrowly. I was not speaking about politics, of course. --Guy vandegrift (discuss • contribs) 13:06, 2 February 2016 (UTC)


 * I also feel that the FlaggedRevs extension could be an optimum solution. It was specifically created for Wikipedia articles that are not watchlisted by very many people and are at risk of test edits or vandalism going unnoticed for a long while. Many of our older resources fall into this category. I don't see any point in protecting resources that are actively being worked on as any unproductive changes (and they are few and far between) will get quickly noticed and reverted. There are exceptions, of course. Some of the Main Page linked pages are such frequent targets that it is a burden to revert often. In general I am quite surprised to see how rare vandalism is on this site compared to just a few years ago. Wikimedia seems to have taken some significant steps in squelching this activity globally which reduces our burden. Yes, I got the sense of the terms you used ;) --mikeu talk 15:29, 2 February 2016 (UTC)

Comments by User:Marshallsumter
I agree generally with a low need for protecting resources or talk pages. The Victor Hugo quote which was receiving spamming, apparently induced by a link from our main page, at a rate of about 1 every 3.6 months, was stopped by protection. Our Physics lecture was receiving IP vandalism requiring 3 reverts per hour. This persistence was stopped by indef protection for some months and has not reoccurred with protection removed. Indef is not infinity. From a negotiation point of view, a persistent vandal will use a def protection as a point of return, unless distracted in some other way. More importantly are tools for "educational moments" such as w:WP:AGF and not w:WP:BITE'ing the newcomer to turn uncivil, persistent, or deletionist IPs into positive contributors. I hope this helps. --Marshallsumter (discuss • contribs) 21:21, 1 February 2016 (UTC)


 * Thank you for joining the conversation.


 * Victor Hugo is part of the Main page learning project/QOTD which I have recently sought to revive. The idea was to create a kind of Honeypot to deflect vandal prone editors to a safe space that was on the watchlist of a number of contributors who volunteered to keep up with the traffic. This lapsed while I was on wiki break as no one else was watching it. The original plan was to engage anons in the way that I've been describing to educate them about what we do and entice them to make positive contributions. I won't claim that the success was overwhelming, but overall I was pleased with the contribs and it wasn't much work. There's no longer a need to protect those QOTD pages as I'm again watching them closely. As a precaution I've redirected (and protected) the mainspace pages to the talk to prevent search engines from indexing test edits or spam.


 * Slight point about wording. "3 reverts per hour" implies an average over a sustained period of time. I only see a total of 3 incidents of vandalism in 8 years which averages about 1 revert per 2.3 years. (not counting page protected time.) All 3 edits occurred within an hour indicating that the ip should have been blocked. I would've expected that to stop disruption to the resource without a need for protection. It is disingenuous to claim that the page protection solved the problem when it is obvious why the vandalism stopped; a single day block did the trick without the need for "some months" of protection. There were no vandalism edits here or globally from this ip after the block expired. Many vandal edits, like this example, are "drive by" scribbling. While there are some persistent attempts that I've seen it is more often the exception, than the rule.


 * I just don't see compelling evidence in the revision history or logs that support the idea that protection is necessary or useful. The examples that I've looked at indicate otherwise. --mikeu talk 22:48, 1 February 2016 (UTC)


 * Just FYI but the Victor Hugo quote resource page (now a redirect to the talk page) is still page protected like my Ice cores resource page (now a redirect) was. --Marshallsumter (discuss • contribs) 03:24, 2 February 2016 (UTC)


 * As I have already explained in detail above the QOTD pages are demonstrated to be at high risk of vandalism. To prevent spam or test edits from getting indexed by search engines I have redirected to the talk page and protected the redirect from editing by anons. The QOTD pages were created specifically to attract vandalism in a central location where it could be dealt with. Ice cores is not linked prominently from the Main Page, so it is not even close to being at the same level of risk as is evident by looking at the page history. Also, please include an edit summary when using the page protect. It make it difficult for others to review actions later if there is no rationale for why it was done. --mikeu talk 15:21, 2 February 2016 (UTC)

Version 1.0
I'm experimenting with a system for automatically uploading scientific data in near real-time to a wiki page for use in exercises where students analyze the data. The project uses a variety of instruments mounted in weather proof enclosures on the roof of Ladd Observatory:


 * 1) Davis Weather Station; live data at http://thuban.physics.brown.edu/weather/
 * 2) Boltek Storm Tracker for detecting lightning strikes; live data at http://thuban.physics.brown.edu/wasp.shtml
 * 3) SBIG All Sky "meteor" camera for monitoring sky conditions such as cloud cover or haze; live data at http://thuban.physics.brown.edu/skycam/
 * 4) Unihedron SQM-LE sky brightness meter for measuring urban light pollution; live data at http://thuban.physics.brown.edu/sqm.shtml

In the past I've manually uploaded some of this data to SkyCam or other pages. Now I'm trying to include dynamic scientific data into lessons. To implement this Mu301Bot runs on the same webserver that collects the data from these instruments and stores the images or generates real-time graphs/maps. I'm currently running a test where hourly samples of sky brightness measurements are added to User:Mu301Bot/nelm when new data is available. I'm also looking into the possibility of formatting the uploaded data points in mw:Extension:Graph format.

It is also possible to add image files from the sky camera automatically, perhaps for a special event like a Lunar eclipse or the May 9, 2016 Transit of Mercury. This could be used for a "real-time lab" where remote students work on the data while it is being collected or analyze it afterwards. Current images of the transit of Mercury could also be incorporated dynamically into a story or a  article.

Another possibility is weather events such as a severe thunderstorm. When a storm is detected the bot would edit a page that interested students have on their watchlist to alert them to a live data upload such as a map of lightning activity or a graph of wind speed during a hurricane. These meteorology uploads would be triggered by a threshold such as the Saffir–Simpson hurricane wind scale.

There are a number of possibilities that could be implemented. This could be of use to schools or colleges with limited resources to teach lessons using data from instruments that are too expensive to purchase for student use. It could also be used more generally for astronomical or meteorological w:Template:Current events. This is an extension to the Observational astronomy project that I started many years ago. --mikeu talk 19:36, 28 January 2016 (UTC)

Version 1.1
I've rewritten User:Mu301Bot/nelm to include an image that is dynamically updated once per day. I'm not sure what the ramifications are for a File: that will eventually have 365 revision updates per year. I'm not sure if anyone has ever dealt with this issue in wikimedia before. Another option, for a different purpose, would be to encode the timestamp with a prefix in the filename. But, here I want the page to automatically show the latest image and caption without editing the file link. I'm conducting a live test of near real-time uploads of images from a telescope. (Note: Check back tomorrow to see the most current image taken at about 2 a.m. local time. It will either show stars or cloud cover.)

I've also encountered the issue that appending data points to a file grows without bound. Currently the data in the subpage Mu301Bot/nelm/data only includes the last 50 measurements and is overwritten each hour, but only when new data is available. A possible solution is subpages of the form PageName/YYYY to create multiple archives by date. The Mediawiki graph extension is complicated. I'll have to rewrite the scripts to parse the data into the correct format.

I'll need to be careful with the particulars of lessons that use live data to take into account that there might be an interruption of uploads or long gaps where there is no new data. It is important to phrase the captions to avoid language like "today's image" given that for any page view the data might be stale. It will also be tricky to write lessons where I can't know ahead of time what the correct answer to an exercise is.

I'm also trying something new with my User:Mu301 page which now incorporates the "switch" feature used by QOTD to rotate a different image once per day. This could be used to feature one of 7 different learning projects that I'm currently working on each day of the week. Currently it just shows a gallery of images that I've contributed to Commons.

I'm excited about the possibilities of near live data in science lessons. We've come a long way since the mimeograph handouts of my childhood. Today's generation of students expect more from educational activities and we now have the technology to aide us in generating richer and more ambitious educational content. --mikeu talk 19:14, 1 February 2016 (UTC)

Version 1.2
The local copy of the image has been temporarily deleted to allow a Commons version of the file to appear here live. I've requested the bot flag there and I started running some automated upload tests today. The skies are clear tonight so I'm uploading twice per hour instead of just once per night. --mikeu talk 03:29, 10 February 2016 (UTC)

Version 1.3
The camera is currently offline and needs some work. It might be time to consider replacing it with a more modern camera. The images are showing artifacts such as "hot pixels" probably from exposure to the elements. Overall, I consider the project to be a success. I've learned a lot which will allow me to design a much better system the second time around. --mikeu talk 00:23, 3 January 2018 (UTC)