I am applying for a summer student to do a Wikipedia Medicine research project through my department at UBC. One potentially project I am looking it is having them review all the edits made to Wikiproject Medicine articles. The student will go through each edit and a) determine if the edit is okay and revert it/fix it if it is not b) determine which edits are made from IP/new users verses long term edits c) calculate the percentage of positive/negative edits from each group d) they will be going over edits from more than one day old thus we will be able to determine how good Wikipedia is at repairing itself. I am thinking of collecting a weeks worth of edits.
While we have a list here http://toolserver.org/~tim1357/cgi-bin/wikiproject_watchlist.py?template=Wik... if multiple edits are made to the same page in a single day it only shows the last one. Is it possible to get a list of all edits? If should be possible to work with this list if another is not available.
If I am able to get approval and funding from UBC I am hoping to run a second round collecting the same data but with "pending changes" turned on for a week on all medical articles. This students would be required to handing all pending changes to all medical articles and will be collecting the same data as before. This will allow us to determine 1) if pending changes affects the numbers of IPs editing 2) if and to what degree pending changes reduces the visibility of poor quality content. The proposed student will be either between first and second year or second and third year medicine and will be working 40 hours per week for 6-8 weeks during the summer. If of course the last part of the project does not get approval I will still try to go ahead with the first part and will have the student join me on the "Medical Translation Project" as discussed here http://en.wikipedia.org/wiki/Wikipedia:MED/Translation_project
I am applying for a summer student to do a Wikipedia Medicine research project through my department at UBC. One potentially project I am looking it is having them review all the edits made to Wikiproject Medicine articles. The student will go through each edit and a) determine if the edit is okay and revert it/fix it if it is not b) determine which edits are made from IP/new users verses long term edits c) calculate the percentage of positive/negative edits from each group d) they will be going over edits from more than one day old thus we will be able to determine how good Wikipedia is at repairing itself. I am thinking of collecting a weeks worth of edits.
While we have a list here http://toolserver.org/~tim1357/cgi-bin/wikiproject_watchlist.py?template=Wik... if multiple edits are made to the same page in a single day it only shows the last one. Is it possible to get a list of all edits? If should be possible to work with this list if another is not available.
For the first topic on that list click on "hist" and you'll get the editing history for the article:
http://en.wikipedia.org/w/index.php?title=Psychopathy&action=history
You can take that back, usually, to the first edit which created the article.
This is a wonderful project. There is a study that 50% of doctors sometimes consult Wikipedia and that 5% edit, probably the largest percentage of any professional group; and you can't say they're not busy.
I suspect that ips in this area are more often responsible editors than is usual, simply doctors who do not have an account.
Fred
If I am able to get approval and funding from UBC I am hoping to run a second round collecting the same data but with "pending changes" turned on for a week on all medical articles. This students would be required to handing all pending changes to all medical articles and will be collecting the same data as before. This will allow us to determine 1) if pending changes affects the numbers of IPs editing 2) if and to what degree pending changes reduces the visibility of poor quality content. The proposed student will be either between first and second year or second and third year medicine and will be working 40 hours per week for 6-8 weeks during the summer. If of course the last part of the project does not get approval I will still try to go ahead with the first part and will have the student join me on the "Medical Translation Project" as discussed here http://en.wikipedia.org/wiki/Wikipedia:MED/Translation_project
-- James Heilman MD, CCFP-EM, Wikipedian _______________________________________________ foundation-l mailing list foundation-l@lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/foundation-l
Experimentation that relies on the availability of Pending changes would be better carried out of those language Wikipedias that have adopted the system, such as the German Wikipedia.
The English WP decide not to use Pending Changes; I cannot for certain predict what would happen if the request to revive it on a basis such as you propose were to go to the English WP community, but I think most of the interested people there would be extremely reluctant to revisit the issue, after having debated in so extensively for so long. This would I think apply even to experimental use: the initial implementation was proposed as an experiment, and was continued without authorization for a while even after the experiment was supposed to end, before community protest put an end to it. I don't think the editors in any one subject area would consider themselves to have the authority to reverse this overall decision for their area, and I doubt the community as a whole would be prepared to grant such authority.
There is no evidence that the medical sections of Wikipedia have a particularly high incidence of problems; indeed, recent reviews from various sources found the precise opposite. Another audit of the editing would certainly be useful, of course and, as Fred points out, the interface is designed to facilitate such audits.
On Sat, Jan 21, 2012 at 9:28 AM, Fred Bauder fredbaud@fairpoint.net wrote:
I am applying for a summer student to do a Wikipedia Medicine research project through my department at UBC. One potentially project I am looking it is having them review all the edits made to Wikiproject Medicine articles. The student will go through each edit and a) determine if the edit is okay and revert it/fix it if it is not b) determine which edits are made from IP/new users verses long term edits c) calculate the percentage of positive/negative edits from each group d) they will be going over edits from more than one day old thus we will be able to determine how good Wikipedia is at repairing itself. I am thinking of collecting a weeks worth of edits.
While we have a list here http://toolserver.org/~tim1357/cgi-bin/wikiproject_watchlist.py?template=Wik... if multiple edits are made to the same page in a single day it only shows the last one. Is it possible to get a list of all edits? If should be possible to work with this list if another is not available.
For the first topic on that list click on "hist" and you'll get the editing history for the article:
http://en.wikipedia.org/w/index.php?title=Psychopathy&action=history
You can take that back, usually, to the first edit which created the article.
This is a wonderful project. There is a study that 50% of doctors sometimes consult Wikipedia and that 5% edit, probably the largest percentage of any professional group; and you can't say they're not busy.
I suspect that ips in this area are more often responsible editors than is usual, simply doctors who do not have an account.
Fred
If I am able to get approval and funding from UBC I am hoping to run a second round collecting the same data but with "pending changes" turned on for a week on all medical articles. This students would be required to handing all pending changes to all medical articles and will be collecting the same data as before. This will allow us to determine 1) if pending changes affects the numbers of IPs editing 2) if and to what degree pending changes reduces the visibility of poor quality content. The proposed student will be either between first and second year or second and third year medicine and will be working 40 hours per week for 6-8 weeks during the summer. If of course the last part of the project does not get approval I will still try to go ahead with the first part and will have the student join me on the "Medical Translation Project" as discussed here http://en.wikipedia.org/wiki/Wikipedia:MED/Translation_project
-- James Heilman MD, CCFP-EM, Wikipedian _______________________________________________ foundation-l mailing list foundation-l@lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/foundation-l
foundation-l mailing list foundation-l@lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/foundation-l
On 21 January 2012 13:42, James Heilman jmh649@gmail.com wrote:
While we have a list here http://toolserver.org/~tim1357/cgi-bin/wikiproject_watchlist.py?template=Wik... if multiple edits are made to the same page in a single day it only shows the last one. Is it possible to get a list of all edits? If should be possible to work with this list if another is not available.
This is slightly clumsy, but it works:
Produce a list of all the articles you want to "watch" (which is the most tricky part of the process) and format them as a long list of wikilinks. Drop this into a userspace or projectspace page - it should just be a sea of blue links to individual articles. Then, use "Related Changes" from the sidebar - this will produce a list of all the edits made to every article linked. It isn't treated as a watchlist, so it shows all edits and not just the "top" ones.
One caveat, though - it doesn't handle redirects or pagemoves very well. You need to make sure the links go direct to the target page, not via a redirect, or you'll end up watching for changes to the redirect page; and if a page linked from there is moved, you may miss any changes to it.
If I am able to get approval and funding from UBC I am hoping to run a second round collecting the same data but with "pending changes" turned on for a week on all medical articles. This students would be required to handing all pending changes to all medical articles and will be collecting the same data as before. This will allow us to determine 1) if pending changes affects the numbers of IPs editing 2) if and to what degree pending changes reduces the visibility of poor quality content. The proposed
As David says, I fear you may have more trouble getting agreement from the community! I'd love to see this part of the study done (and make a small bet as to what the results might be...), but it's going to be a hard sell. If you're planning to get this running in the summer, you might want to start the negotiations quite soon...
On Sat, Jan 21, 2012 at 5:42 AM, James Heilman jmh649@gmail.com wrote:
I am applying for a summer student to do a Wikipedia Medicine research project through my department at UBC. One potentially project I am looking it is having them review all the edits made to Wikiproject Medicine articles. The student will go through each edit and a) determine if the edit is okay and revert it/fix it if it is not b) determine which edits are made from IP/new users verses long term edits c) calculate the percentage of positive/negative edits from each group d) they will be going over edits from more than one day old thus we will be able to determine how good Wikipedia is at repairing itself. I am thinking of collecting a weeks worth of edits.
While we have a list here
http://toolserver.org/~tim1357/cgi-bin/wikiproject_watchlist.py?template=Wik... if multiple edits are made to the same page in a single day it only shows the last one. Is it possible to get a list of all edits? If should be possible to work with this list if another is not available.
If I am able to get approval and funding from UBC I am hoping to run a second round collecting the same data but with "pending changes" turned on for a week on all medical articles. This students would be required to handing all pending changes to all medical articles and will be collecting the same data as before. This will allow us to determine 1) if pending changes affects the numbers of IPs editing 2) if and to what degree pending changes reduces the visibility of poor quality content. The proposed student will be either between first and second year or second and third year medicine and will be working 40 hours per week for 6-8 weeks during the summer. If of course the last part of the project does not get approval I will still try to go ahead with the first part and will have the student join me on the "Medical Translation Project" as discussed here http://en.wikipedia.org/wiki/Wikipedia:MED/Translation_project
-- James Heilman MD, CCFP-EM, Wikipedian _______________________________________________ foundation-l mailing list foundation-l@lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/foundation-l
Hey James,
Interesting initiative! I just wanted to say that reviewing of edits by your student is the kind of qualitative coding that we've done a lot of here in the research end of the Foundation, so if you want to any assistance on either the technical end (we have a couple software options) or with anything else, we'd be happy to help.
-- Steven
wikimedia-l@lists.wikimedia.org