On 2015 Jul 25, Egon Willighagen commented:

David, you may want to read with J. Wales has to say about this [0]:

But if the Professor has a more nuanced view that Wikipedia should not be cited "as a source" by university students then I agree completely! I think the same thing about citing Britannica or any other encyclopedia. Citing an encyclopedia for an academic paper at the University level is not appropriate

0.http://www.quora.com/What-does-Jimmy-Wales-think-when-a-university-professor-discourages-students-from-citing-Wikipedia-as-a-primary-or-secondary-source

On 2015 Jul 23, James M Heilman commented:

You sort of missed the point. The point is that Wikipedia is incredibly frequently read. This includes by the lay public, medical students, and physicians. One can warn people until they are blue in the face but I doubt it is going to change how frequently they use Wikipedia for medical information. In my opinion time is much better spent trying to improve the site. It really is not that difficult.

On 2015 May 11, David Gortler commented:

I have to say that although this data sounds promising, I would never use wikipedia as an authoritative resource for anything patient-related, and firmly instill warnings regarding wikipedia to my students as well. I was given a moderator account at one point and took time and effort to carefully compose several specialty pharmacology articles, only to have those articles edited incorrectly, filled with obvious biases, or inexplicably "re-organized" to the point that they were essentially vandalized. I had offenders blocked or banned, only to have scores more crawl out of the bowels of the internet to do the same. I have since stopped trying to correctly edit or even view my articles for accuracy.

The last time I used wikipedia was years ago and I think it was to look up something about vintage automobiles. I also looked up Gilligan's Island on it around the same time. That's about as far as I would trust it. Anyone who relies on wikipedia for anything clinically related is placing patient lives at risk. Anyone who relies on wikipedia to look up anything or research related is wasting their time and money.

On 2015 May 10, James M Heilman commented:

Yes looking at top editors I know most of those listed. A proportion are those who write a great deal of content and mostly work on medicine. Another proportion are those who do maintenance and vandalism reversion Wikipedia wide. What we are capturing is only a proportion of their edits as we just looked at medical editors. If we were to look at Wikipedia as a whole they would far surpass all those listed here.

So I agree on a global scale the top editors are more involved in maintenance. However within a specific area the top editors are more content contributors (there are of course still maintence people aswell).

On 2015 May 08, Gwinyai Masukume commented:

Thank you for engaging with this and making the data publicly available.

Five users appear on both lists (Top ten by number of bytes changed and Top ten by number of edits) and only one user is in the same position.

Perhaps the situation is a hybrid of Jimmy Wales’ and Aaron Swartz’s thoughts, the degree differing depending on the context. Looking forward to further insights from your data.

On 2015 May 13, James M Heilman commented:

And now we have data for 2014 https://en.wikipedia.org/wiki/Wikipedia:WikiProject_Medicine/Stats/Top_English_medical_editors_2014

On 2015 May 07, James M Heilman commented:

We now have data looking at the top editors for 2013 based on bytes changed here https://en.wikipedia.org/wiki/Wikipedia:WikiProject_Medicine/Stats/Top_English_medical_editors_2013b

There is a fairly good relationship with top editors for 2013 based on number of edits here https://en.wikipedia.org/wiki/Wikipedia:WikiProject_Medicine/Stats/Top_English_medical_editors_2013

So my analysis is more in line with the thoughts of Jimmy Wales rather than Aaron Swartz.

On 2015 Mar 26, James M Heilman commented:

Often people with the highest edit count also have the largest number of bytes changed.

One would really want to look at the persistent bytes contributed by a user. One also needs to keep in mind that more text is not necessarily better text. Lots of improvements involved trimming material.

Agree that their is no perfect measure of "authorship" which can be carried out in an automated fashion. Will look at running the analysis you suggest and posting the results here.

On 2015 Mar 21, Gwinyai Masukume commented:

The authors use the number of edits made to measure participation. An edit, for example, can be adding a full stop/period “.” or it can be adding a whole paragraph. Each of these actions would be counted as one edit.

Another way to measure participation is by considering the amount of text added by a contributor. This approach of considering the amount of text added has in the past yielded different results and insights compared to considering the number of edits made. Some contributors with relatively low edit counts have been found to have added a relatively high amount of text.

I am of the view that only using edit counts is a limitation. This limitation could have been remedied by considering the amount of text added by contributors from a limited number of randomly selected articles from the authors' sampling frame.

Reference

Swartz A. Who writes Wikipedia? 2006. Available: http://www.aaronsw.com/weblog/whowriteswikipedia (accessed 21 March 2015).

On 2015 Mar 21, Gwinyai Masukume commented:

The authors use the number of edits made to measure participation. An edit, for example, can be adding a full stop/period “.” or it can be adding a whole paragraph. Each of these actions would be counted as one edit.

Another way to measure participation is by considering the amount of text added by a contributor. This approach of considering the amount of text added has in the past yielded different results and insights compared to considering the number of edits made. Some contributors with relatively low edit counts have been found to have added a relatively high amount of text.

I am of the view that only using edit counts is a limitation. This limitation could have been remedied by considering the amount of text added by contributors from a limited number of randomly selected articles from the authors' sampling frame.

Reference

Swartz A. Who writes Wikipedia? 2006. Available: http://www.aaronsw.com/weblog/whowriteswikipedia (accessed 21 March 2015).

On 2015 May 08, Gwinyai Masukume commented:

Thank you for engaging with this and making the data publicly available.

Five users appear on both lists (Top ten by number of bytes changed and Top ten by number of edits) and only one user is in the same position.

Perhaps the situation is a hybrid of Jimmy Wales’ and Aaron Swartz’s thoughts, the degree differing depending on the context. Looking forward to further insights from your data.

On 2015 May 10, James M Heilman commented:

Yes looking at top editors I know most of those listed. A proportion are those who write a great deal of content and mostly work on medicine. Another proportion are those who do maintenance and vandalism reversion Wikipedia wide. What we are capturing is only a proportion of their edits as we just looked at medical editors. If we were to look at Wikipedia as a whole they would far surpass all those listed here.

So I agree on a global scale the top editors are more involved in maintenance. However within a specific area the top editors are more content contributors (there are of course still maintence people aswell).

On 2015 May 11, David Gortler commented:

I have to say that although this data sounds promising, I would never use wikipedia as an authoritative resource for anything patient-related, and firmly instill warnings regarding wikipedia to my students as well. I was given a moderator account at one point and took time and effort to carefully compose several specialty pharmacology articles, only to have those articles edited incorrectly, filled with obvious biases, or inexplicably "re-organized" to the point that they were essentially vandalized. I had offenders blocked or banned, only to have scores more crawl out of the bowels of the internet to do the same. I have since stopped trying to correctly edit or even view my articles for accuracy.

The last time I used wikipedia was years ago and I think it was to look up something about vintage automobiles. I also looked up Gilligan's Island on it around the same time. That's about as far as I would trust it. Anyone who relies on wikipedia for anything clinically related is placing patient lives at risk. Anyone who relies on wikipedia to look up anything or research related is wasting their time and money.