Hi Aaron,
Is the data set from https://commons.wikimedia.org/wiki/File:Desirable_newcomer_survival_over_tim... available for correlation with the number of new articles each user created?
Aaron Halfaker ahalfaker@wikimedia.org wrote:
... I propose a project where we work together to generate a summary so that I can call it "work." I've started a stub here: https://meta.wikimedia.org/wiki/Research:New_editor_engagement_strategies
Sure! You'll find the hand-coded set of users here http://datasets.wikimedia.org/public-datasets/enwiki/rise-and-decline within the next half hour (cron job copies datasets over).
*Categories: *
1. Vandals - Purposefully malicious, out to cause harm 2. Bad-faith - Trying to be funny, not here to help or harm 3. Good-faith - Trying to be productive, but failing 4. Golden - Successfully contributing productively
-Aaron
On Wed, Sep 24, 2014 at 1:15 PM, James Salsman jsalsman@gmail.com wrote:
Hi Aaron,
Is the data set from https://commons.wikimedia.org/wiki/File:Desirable_newcomer_survival_over_tim... available for correlation with the number of new articles each user created?
Aaron Halfaker ahalfaker@wikimedia.org wrote:
... I propose a project where we work together to generate a summary so that I can call it "work." I've started a stub here:
https://meta.wikimedia.org/wiki/Research:New_editor_engagement_strategies
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
Hi, if we are analysing new editors creating new articles then there are two groups I would expect to see, individual spammers and corporate spammers.
The individual spammers are the kids writing about the band that will be the next big thing on the ....... Scene in their town, and their first gig is next Tuesday provided they can recruit a drummer.
The corporate spammers are writing about some business, but in a style that makes a corporate flyer look neutral.
There are some implicit assumptions that we make as new page patrollers, and which some research might help by proving or disproving.
1 people whose first article is an attack page or vandalism will very rarely turn into productive editors. I have met or heard of three former vandals who made good editors, there may be more, but I've never heard of a good editor who started out creating attack pages.
2 people who are paid to spam this site are unlikely to turn into good editors, but a large proportion of good editors made COI edits among their early edits. The classic wikipedian is a time rich altruist with Internet access that they can make personal use of, I'm not seeing much overlap there with corporate spammers, but people who make COI edits other than about their employment clearly have personal access to the Internet.
3 neutral point of view is something that new editors need to pick up, and POV editing amongst a new editor's early contributions is so common that if a new account writes neutrally they are assumed by some to be sock puppets or returning editors.
4 we live in a copy paste world, and while a student plagiarising to try to get better marks is clearly doing so in bad faith, a wikipedia contributor using copy paste is still a good faith contributor, they just need to be taught not to use copy and paste. Presumably these people are in Aaron's group 3, again we have a lot of former copyright violators in the community, and a lot of new articles get deleted as copyright violations.
My suspicion is that much of our tolerance of vandals is wasted effort. We would save a lot of time and not lose any good editors by moving from the current five strikes and you are out approach to vandals to one of blocking any new editor after their first edit if that edit was blatant vandalism. Copyvio by contrast is something where we could probably retain more of the editors by trying different approaches to explaining why their edits had to be rejected and how they could contribute.
Regards
Jonathan Cardy
On 25 Sep 2014, at 00:19, Aaron Halfaker aaron.halfaker@gmail.com wrote:
Sure! You'll find the hand-coded set of users here within the next half hour (cron job copies datasets over).
Categories: Vandals - Purposefully malicious, out to cause harm Bad-faith - Trying to be funny, not here to help or harm Good-faith - Trying to be productive, but failing Golden - Successfully contributing productively -Aaron
On Wed, Sep 24, 2014 at 1:15 PM, James Salsman jsalsman@gmail.com wrote: Hi Aaron,
Is the data set from https://commons.wikimedia.org/wiki/File:Desirable_newcomer_survival_over_tim... available for correlation with the number of new articles each user created?
Aaron Halfaker ahalfaker@wikimedia.org wrote:
... I propose a project where we work together to generate a summary so that I can call it "work." I've started a stub here: https://meta.wikimedia.org/wiki/Research:New_editor_engagement_strategies
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
This pretty much matches my own observations.
I think it's not uncommon for new editors to first write about something they know well, which often means some level of conflict of interest. But that doesn't necessarily mean the person is writing lies or being completely over-the-top with lavish praise; they can write quite reasonable content too. And generally new editors don't know that CoI editing isn't welcome. Similarly a lot get blocked for 3 reverts and get accused of edit warring, but I suspect it is often not the case. I suspect often they look back at the article and can't see the edit they did, so they just do it again because it didn't "stick". I am not convinced they do know that other people are reverting their edits because I don't think they know about the article history and the edit summaries and the user talk pages.
We make some awfully big assumptions that new editors know what we do; it's amazing how much stuff in a UI you mentally filter out. I did a talk about Wikipedia about a year ago for a group of librarians. They asked for some "tips and tricks" to be included, so for the first time I systematically scanned the article UI and found cute little features I never knew were there (even on menus I often used). And of course the UI changes over time and probably I didn't notice new things appear. But not noticing is consistent with "mission focus", which has been studied in retail settings. A customer "on a mission" (a specific transaction that must be dealt with, as opposed to a customer "just browsing") is oblivious to promotions until the "mission" is complete and then becomes aware of their surroundings as they turn to leave the store (which tells you where to put promotional signs relative to your "mission" products). I think Wikipedia editing is very mission-focussed (I must fix that error) so we see nothing but what we want to see on the way in and then have no "exit screen" to talk to the editor as they leave (unlike a bricks-and-mortar retailer). It's hard to communicate with new editors on-wiki.
I actually think doing a user experience experiment (ideally with eye tracking) with new users would be very informative. Given them a "mission" and watch what they look at, what they do and if they succeed in their task, and whether their edits survives over time (and if it doesn't survive, why is that). Indeed, seeing what they do when their edit disappears would be interesting too.
It is difficult for many of us as experienced Wikipedians to see Wikipedia as the new contributor sees it. I have some sense of the new editor experience because I do edit training and it certainly shows why nothing is fool-proof because fools are so ingenious. It's amazing how new editors dive into parts of the UI I've never even noticed and then tie themselves up in knots. They often get so interested in Preview that they never Save (and then suddenly oops, all their work is gone). They generate continual edit conflicts (often entirely on their own, I think by going back in the browser and editing the article in multiple browser windows, leading to their contributions being smeared across a number of edit windows, none of which can be saved, or so it seems). And if they are editing in mainspace, they panic if the article changes in some mysterious way and they don't know how they did it (it was another concurrent editor). In fact, if there is one thing I would like is for the watchlist notification to be slower than normal when a new editor is involved as having someone alerted by the watchlist coming and "fixing" things (even if it's well-meant) puts the new editor into a spin because they don't understand what's going on (they have no experience with collaborative document editing). And, a perennial favourite, why is the Reference section empty when they've added all those citations (they don't expect them to be in-line, they expect them in the References section). When I look at the constant questions and problems I have to deal with in edit training, I am not surprised that people trying to learn on their own give up. Seen through the eyes of the beginner, it's so much harder than we realise both technically and collaboratively.
Kerry
_____
From: wiki-research-l-bounces@lists.wikimedia.org [mailto:wiki-research-l-bounces@lists.wikimedia.org] On Behalf Of WereSpielChequers Sent: Thursday, 25 September 2014 2:18 PM To: Research into Wikimedia content and communities Subject: Re: [Wiki-research-l] FW: What works for increasing editorengagement?
Hi, if we are analysing new editors creating new articles then there are two groups I would expect to see, individual spammers and corporate spammers.
The individual spammers are the kids writing about the band that will be the next big thing on the ....... Scene in their town, and their first gig is next Tuesday provided they can recruit a drummer.
The corporate spammers are writing about some business, but in a style that makes a corporate flyer look neutral.
There are some implicit assumptions that we make as new page patrollers, and which some research might help by proving or disproving.
1 people whose first article is an attack page or vandalism will very rarely turn into productive editors. I have met or heard of three former vandals who made good editors, there may be more, but I've never heard of a good editor who started out creating attack pages.
2 people who are paid to spam this site are unlikely to turn into good editors, but a large proportion of good editors made COI edits among their early edits. The classic wikipedian is a time rich altruist with Internet access that they can make personal use of, I'm not seeing much overlap there with corporate spammers, but people who make COI edits other than about their employment clearly have personal access to the Internet.
3 neutral point of view is something that new editors need to pick up, and POV editing amongst a new editor's early contributions is so common that if a new account writes neutrally they are assumed by some to be sock puppets or returning editors.
4 we live in a copy paste world, and while a student plagiarising to try to get better marks is clearly doing so in bad faith, a wikipedia contributor using copy paste is still a good faith contributor, they just need to be taught not to use copy and paste. Presumably these people are in Aaron's group 3, again we have a lot of former copyright violators in the community, and a lot of new articles get deleted as copyright violations.
My suspicion is that much of our tolerance of vandals is wasted effort. We would save a lot of time and not lose any good editors by moving from the current five strikes and you are out approach to vandals to one of blocking any new editor after their first edit if that edit was blatant vandalism. Copyvio by contrast is something where we could probably retain more of the editors by trying different approaches to explaining why their edits had to be rejected and how they could contribute.
Regards
Jonathan Cardy
On 25 Sep 2014, at 00:19, Aaron Halfaker aaron.halfaker@gmail.com wrote:
Sure! You'll find the hand-coded set of users here http://datasets.wikimedia.org/public-datasets/enwiki/rise-and-decline within the next half hour (cron job copies datasets over).
Categories:
1. Vandals - Purposefully malicious, out to cause harm 2. Bad-faith - Trying to be funny, not here to help or harm 3. Good-faith - Trying to be productive, but failing 4. Golden - Successfully contributing productively
-Aaron
On Wed, Sep 24, 2014 at 1:15 PM, James Salsman jsalsman@gmail.com wrote:
Hi Aaron,
Is the data set from https://commons.wikimedia.org/wiki/File:Desirable_newcomer_survival_over_ti me.png https://commons.wikimedia.org/wiki/File:Desirable_newcomer_survival_over_tim e.png
available for correlation with the number of new articles each user created?
Aaron Halfaker ahalfaker@wikimedia.org wrote:
...
I propose a project where we work together to generate
a summary so that I can call it "work." I've started a stub here:
https://meta.wikimedia.org/wiki/Research:New_editor_engagement_strategies
_______________________________________________ Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
_______________________________________________ Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
This is a great discussion.
Kerry, I believe that Steven Walling and what was called the E3 team, now known as the Growth team, use eye tracking experiments. It would be interesting to hear about what they're learning, and tie those observations together with yours and WSC's.
Also, I am happy to discuss the idea of teaming up with one or more people in this discussion to create an Individual Engagement Grant proposal that would help us to explore some of the material that we're discussing here. Aaron and I are already putting together a proposal related to how editors interact with each other, and I'd be happy to work on concurrent projects related to research on new editors if I can make the scheduling and financing work.
Pine
Yes, training newbies is a great way to learn and to see the flaws that we mentally blank out. I also found that I need to keep a vanilla account for demonstrating things to newbies, if I use my WereSpielChequers account the various extra buttons confuse people.
I wouldn't worry too much about watchlisters making edits and causing edit conflicts, most of the time you aren't going to be in the same time zone, watchlisters even the most active ones are unlikely to check their watch list more than a few times a day. So as long as you don't start newbies on highly watched articles like Sarah Palin you should be OK with them. But edit conflicts are a real problem for newbies and especially those creating new articles. The new page patrol people need to look at articles as they are created in order to pick up attack pages etc, and when they find OK articles they tend to at least categorise them. Some of this could be fixed by improving the software for handling edit conflicts, for example it would be nice if adding a category and changing some text were not treated as a conflict. However this is the sort of Bugzilla request that gets closed as won't fix because of the implicit assumption that any editor should be able to handle an edit conflict in their stride. My preferred solution is to teach newbies to always start new articles in sandboxes and move them to main space when they are ready, alternatively creating a references section means that categories can be added without causing an edit conflict.
Re Pine's comment. When it comes to research topics It would be great to know how many editors are driven away by edit conflicts. But I don't know if that info is actually logged, and if you ask people why they gave up they don't necessarily know whether their edit conflict was with a machine or another editor. I once had to convince an angry editor that they hadn't been reverted by another editor, both edits had the same time stamp and theirs had been lost by the system without the other editor knowing. So if you ask former editors why they went and they blame other editors it may partly be that they don't know the difference between a conflict with the system and with another editor deliberately reverting them.
By the way, if anyone is around in London drop me a line and I will tell you if we have any editathons going on. Editathons have many roles in the community, one being as a focus group on the user interface.
Regards
Jonathan Cardy
On 25 Sep 2014, at 10:09, Kerry Raymond kerry.raymond@gmail.com wrote:
This pretty much matches my own observations.
I think it’s not uncommon for new editors to first write about something they know well, which often means some level of conflict of interest. But that doesn’t necessarily mean the person is writing lies or being completely over-the-top with lavish praise; they can write quite reasonable content too. And generally new editors don’t know that CoI editing isn’t welcome. Similarly a lot get blocked for 3 reverts and get accused of edit warring, but I suspect it is often not the case. I suspect often they look back at the article and can’t see the edit they did, so they just do it again because it didn’t “stick”. I am not convinced they do know that other people are reverting their edits because I don’t think they know about the article history and the edit summaries and the user talk pages.
We make some awfully big assumptions that new editors know what we do; it’s amazing how much stuff in a UI you mentally filter out. I did a talk about Wikipedia about a year ago for a group of librarians. They asked for some “tips and tricks” to be included, so for the first time I systematically scanned the article UI and found cute little features I never knew were there (even on menus I often used). And of course the UI changes over time and probably I didn’t notice new things appear. But not noticing is consistent with “mission focus”, which has been studied in retail settings. A customer “on a mission” (a specific transaction that must be dealt with, as opposed to a customer “just browsing”) is oblivious to promotions until the “mission” is complete and then becomes aware of their surroundings as they turn to leave the store (which tells you where to put promotional signs relative to your “mission” products). I think Wikipedia editing is very mission-focussed (I must fix that error) so we see nothing but what we want to see on the way in and then have no “exit screen” to talk to the editor as they leave (unlike a bricks-and-mortar retailer). It’s hard to communicate with new editors on-wiki.
I actually think doing a user experience experiment (ideally with eye tracking) with new users would be very informative. Given them a “mission” and watch what they look at, what they do and if they succeed in their task, and whether their edits survives over time (and if it doesn’t survive, why is that). Indeed, seeing what they do when their edit disappears would be interesting too.
It is difficult for many of us as experienced Wikipedians to see Wikipedia as the new contributor sees it. I have some sense of the new editor experience because I do edit training and it certainly shows why nothing is fool-proof because fools are so ingenious. It’s amazing how new editors dive into parts of the UI I’ve never even noticed and then tie themselves up in knots. They often get so interested in Preview that they never Save (and then suddenly oops, all their work is gone). They generate continual edit conflicts (often entirely on their own, I think by going back in the browser and editing the article in multiple browser windows, leading to their contributions being smeared across a number of edit windows, none of which can be saved, or so it seems). And if they are editing in mainspace, they panic if the article changes in some mysterious way and they don’t know how they did it (it was another concurrent editor). In fact, if there is one thing I would like is for the watchlist notification to be slower than normal when a new editor is involved as having someone alerted by the watchlist coming and “fixing” things (even if it’s well-meant) puts the new editor into a spin because they don’t understand what’s going on (they have no experience with collaborative document editing). And, a perennial favourite, why is the Reference section empty when they’ve added all those citations (they don’t expect them to be in-line, they expect them in the References section). When I look at the constant questions and problems I have to deal with in edit training, I am not surprised that people trying to learn on their own give up. Seen through the eyes of the beginner, it’s so much harder than we realise both technically and collaboratively.
Kerry
From: wiki-research-l-bounces@lists.wikimedia.org [mailto:wiki-research-l-bounces@lists.wikimedia.org] On Behalf Of WereSpielChequers Sent: Thursday, 25 September 2014 2:18 PM To: Research into Wikimedia content and communities Subject: Re: [Wiki-research-l] FW: What works for increasing editorengagement?
Hi, if we are analysing new editors creating new articles then there are two groups I would expect to see, individual spammers and corporate spammers.
The individual spammers are the kids writing about the band that will be the next big thing on the ....... Scene in their town, and their first gig is next Tuesday provided they can recruit a drummer.
The corporate spammers are writing about some business, but in a style that makes a corporate flyer look neutral.
There are some implicit assumptions that we make as new page patrollers, and which some research might help by proving or disproving.
1 people whose first article is an attack page or vandalism will very rarely turn into productive editors. I have met or heard of three former vandals who made good editors, there may be more, but I've never heard of a good editor who started out creating attack pages.
2 people who are paid to spam this site are unlikely to turn into good editors, but a large proportion of good editors made COI edits among their early edits. The classic wikipedian is a time rich altruist with Internet access that they can make personal use of, I'm not seeing much overlap there with corporate spammers, but people who make COI edits other than about their employment clearly have personal access to the Internet.
3 neutral point of view is something that new editors need to pick up, and POV editing amongst a new editor's early contributions is so common that if a new account writes neutrally they are assumed by some to be sock puppets or returning editors.
4 we live in a copy paste world, and while a student plagiarising to try to get better marks is clearly doing so in bad faith, a wikipedia contributor using copy paste is still a good faith contributor, they just need to be taught not to use copy and paste. Presumably these people are in Aaron's group 3, again we have a lot of former copyright violators in the community, and a lot of new articles get deleted as copyright violations.
My suspicion is that much of our tolerance of vandals is wasted effort. We would save a lot of time and not lose any good editors by moving from the current five strikes and you are out approach to vandals to one of blocking any new editor after their first edit if that edit was blatant vandalism. Copyvio by contrast is something where we could probably retain more of the editors by trying different approaches to explaining why their edits had to be rejected and how they could contribute.
Regards
Jonathan Cardy
On 25 Sep 2014, at 00:19, Aaron Halfaker aaron.halfaker@gmail.com wrote: Sure! You'll find the hand-coded set of users here within the next half hour (cron job copies datasets over).
Categories: Vandals - Purposefully malicious, out to cause harm Bad-faith - Trying to be funny, not here to help or harm Good-faith - Trying to be productive, but failing Golden - Successfully contributing productively -Aaron
On Wed, Sep 24, 2014 at 1:15 PM, James Salsman jsalsman@gmail.com wrote: Hi Aaron,
Is the data set from https://commons.wikimedia.org/wiki/File:Desirable_newcomer_survival_over_tim... available for correlation with the number of new articles each user created?
Aaron Halfaker ahalfaker@wikimedia.org wrote:
... I propose a project where we work together to generate a summary so that I can call it "work." I've started a stub here: https://meta.wikimedia.org/wiki/Research:New_editor_engagement_strategies
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
Australian outreach events generally edit Australian content. Other Australian editors are likely to be on the watch list and are likely to be in the timezone. And plenty of non-Australian editors are sitting in their pyjamas at all hours of the day and night waiting to pounce. Believe me, new editors encounter other editors very quickly (although sometimes they don't realise it). They often think it "rude" that other people are editing the article "while I am in the middle of working on it". Their mental model of collaborative editing is like a shared lawn mower. You have sole use for a while; then it is passed on to the next person.
Not sure I can help you with London editathons. But I do have a couple of edit training days coming up in Oakey, Queensland in a few weeks:
https://en.wikipedia.org/wiki/Oakey,_Queensland
Kerry
_____
From: WereSpielChequers [mailto:werespielchequers@gmail.com] Sent: Thursday, 25 September 2014 9:38 PM To: kerry.raymond@gmail.com; Research into Wikimedia content and communities Subject: Re: [Wiki-research-l] FW: What works for increasing editorengagement?
Yes, training newbies is a great way to learn and to see the flaws that we mentally blank out. I also found that I need to keep a vanilla account for demonstrating things to newbies, if I use my WereSpielChequers account the various extra buttons confuse people.
I wouldn't worry too much about watchlisters making edits and causing edit conflicts, most of the time you aren't going to be in the same time zone, watchlisters even the most active ones are unlikely to check their watch list more than a few times a day. So as long as you don't start newbies on highly watched articles like Sarah Palin you should be OK with them. But edit conflicts are a real problem for newbies and especially those creating new articles. The new page patrol people need to look at articles as they are created in order to pick up attack pages etc, and when they find OK articles they tend to at least categorise them. Some of this could be fixed by improving the software for handling edit conflicts, for example it would be nice if adding a category and changing some text were not treated as a conflict. However this is the sort of Bugzilla request that gets closed as won't fix because of the implicit assumption that any editor should be able to handle an edit conflict in their stride. My preferred solution is to teach newbies to always start new articles in sandboxes and move them to main space when they are ready, alternatively creating a references section means that categories can be added without causing an edit conflict.
Re Pine's comment. When it comes to research topics It would be great to know how many editors are driven away by edit conflicts. But I don't know if that info is actually logged, and if you ask people why they gave up they don't necessarily know whether their edit conflict was with a machine or another editor. I once had to convince an angry editor that they hadn't been reverted by another editor, both edits had the same time stamp and theirs had been lost by the system without the other editor knowing. So if you ask former editors why they went and they blame other editors it may partly be that they don't know the difference between a conflict with the system and with another editor deliberately reverting them.
By the way, if anyone is around in London drop me a line and I will tell you if we have any editathons going on. Editathons have many roles in the community, one being as a focus group on the user interface.
Regards
Jonathan Cardy
On 25 Sep 2014, at 10:09, Kerry Raymond kerry.raymond@gmail.com wrote:
This pretty much matches my own observations.
I think it's not uncommon for new editors to first write about something they know well, which often means some level of conflict of interest. But that doesn't necessarily mean the person is writing lies or being completely over-the-top with lavish praise; they can write quite reasonable content too. And generally new editors don't know that CoI editing isn't welcome. Similarly a lot get blocked for 3 reverts and get accused of edit warring, but I suspect it is often not the case. I suspect often they look back at the article and can't see the edit they did, so they just do it again because it didn't "stick". I am not convinced they do know that other people are reverting their edits because I don't think they know about the article history and the edit summaries and the user talk pages.
We make some awfully big assumptions that new editors know what we do; it's amazing how much stuff in a UI you mentally filter out. I did a talk about Wikipedia about a year ago for a group of librarians. They asked for some "tips and tricks" to be included, so for the first time I systematically scanned the article UI and found cute little features I never knew were there (even on menus I often used). And of course the UI changes over time and probably I didn't notice new things appear. But not noticing is consistent with "mission focus", which has been studied in retail settings. A customer "on a mission" (a specific transaction that must be dealt with, as opposed to a customer "just browsing") is oblivious to promotions until the "mission" is complete and then becomes aware of their surroundings as they turn to leave the store (which tells you where to put promotional signs relative to your "mission" products). I think Wikipedia editing is very mission-focussed (I must fix that error) so we see nothing but what we want to see on the way in and then have no "exit screen" to talk to the editor as they leave (unlike a bricks-and-mortar retailer). It's hard to communicate with new editors on-wiki.
I actually think doing a user experience experiment (ideally with eye tracking) with new users would be very informative. Given them a "mission" and watch what they look at, what they do and if they succeed in their task, and whether their edits survives over time (and if it doesn't survive, why is that). Indeed, seeing what they do when their edit disappears would be interesting too.
It is difficult for many of us as experienced Wikipedians to see Wikipedia as the new contributor sees it. I have some sense of the new editor experience because I do edit training and it certainly shows why nothing is fool-proof because fools are so ingenious. It's amazing how new editors dive into parts of the UI I've never even noticed and then tie themselves up in knots. They often get so interested in Preview that they never Save (and then suddenly oops, all their work is gone). They generate continual edit conflicts (often entirely on their own, I think by going back in the browser and editing the article in multiple browser windows, leading to their contributions being smeared across a number of edit windows, none of which can be saved, or so it seems). And if they are editing in mainspace, they panic if the article changes in some mysterious way and they don't know how they did it (it was another concurrent editor). In fact, if there is one thing I would like is for the watchlist notification to be slower than normal when a new editor is involved as having someone alerted by the watchlist coming and "fixing" things (even if it's well-meant) puts the new editor into a spin because they don't understand what's going on (they have no experience with collaborative document editing). And, a perennial favourite, why is the Reference section empty when they've added all those citations (they don't expect them to be in-line, they expect them in the References section). When I look at the constant questions and problems I have to deal with in edit training, I am not surprised that people trying to learn on their own give up. Seen through the eyes of the beginner, it's so much harder than we realise both technically and collaboratively.
Kerry
_____
From: wiki-research-l-bounces@lists.wikimedia.org [mailto:wiki-research-l-bounces@lists.wikimedia.org] On Behalf Of WereSpielChequers Sent: Thursday, 25 September 2014 2:18 PM To: Research into Wikimedia content and communities Subject: Re: [Wiki-research-l] FW: What works for increasing editorengagement?
Hi, if we are analysing new editors creating new articles then there are two groups I would expect to see, individual spammers and corporate spammers.
The individual spammers are the kids writing about the band that will be the next big thing on the ....... Scene in their town, and their first gig is next Tuesday provided they can recruit a drummer.
The corporate spammers are writing about some business, but in a style that makes a corporate flyer look neutral.
There are some implicit assumptions that we make as new page patrollers, and which some research might help by proving or disproving.
1 people whose first article is an attack page or vandalism will very rarely turn into productive editors. I have met or heard of three former vandals who made good editors, there may be more, but I've never heard of a good editor who started out creating attack pages.
2 people who are paid to spam this site are unlikely to turn into good editors, but a large proportion of good editors made COI edits among their early edits. The classic wikipedian is a time rich altruist with Internet access that they can make personal use of, I'm not seeing much overlap there with corporate spammers, but people who make COI edits other than about their employment clearly have personal access to the Internet.
3 neutral point of view is something that new editors need to pick up, and POV editing amongst a new editor's early contributions is so common that if a new account writes neutrally they are assumed by some to be sock puppets or returning editors.
4 we live in a copy paste world, and while a student plagiarising to try to get better marks is clearly doing so in bad faith, a wikipedia contributor using copy paste is still a good faith contributor, they just need to be taught not to use copy and paste. Presumably these people are in Aaron's group 3, again we have a lot of former copyright violators in the community, and a lot of new articles get deleted as copyright violations.
My suspicion is that much of our tolerance of vandals is wasted effort. We would save a lot of time and not lose any good editors by moving from the current five strikes and you are out approach to vandals to one of blocking any new editor after their first edit if that edit was blatant vandalism. Copyvio by contrast is something where we could probably retain more of the editors by trying different approaches to explaining why their edits had to be rejected and how they could contribute.
Regards
Jonathan Cardy
On 25 Sep 2014, at 00:19, Aaron Halfaker aaron.halfaker@gmail.com wrote:
Sure! You'll find the hand-coded set of users here http://datasets.wikimedia.org/public-datasets/enwiki/rise-and-decline within the next half hour (cron job copies datasets over).
Categories:
1. Vandals - Purposefully malicious, out to cause harm 2. Bad-faith - Trying to be funny, not here to help or harm 3. Good-faith - Trying to be productive, but failing 4. Golden - Successfully contributing productively
-Aaron
On Wed, Sep 24, 2014 at 1:15 PM, James Salsman jsalsman@gmail.com wrote:
Hi Aaron,
Is the data set from https://commons.wikimedia.org/wiki/File:Desirable_newcomer_survival_over_ti me.png https://commons.wikimedia.org/wiki/File:Desirable_newcomer_survival_over_tim e.png
available for correlation with the number of new articles each user created?
Aaron Halfaker ahalfaker@wikimedia.org wrote:
...
I propose a project where we work together to generate
a summary so that I can call it "work." I've started a stub here:
https://meta.wikimedia.org/wiki/Research:New_editor_engagement_strategies
_______________________________________________ Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
_______________________________________________ Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
_______________________________________________ Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
I don't doubt that Australian newbies editing existing well developed articles are going to find they are editing things on existing Australian editors watch lists. My experience of editathons is mostly about creating new articles or improving very neglected ones, usually by expanding stubs, or working from lists of articles without images. Such articles are unlikely to be on anyone's watchlist, or if they are they aren't likely to object to good faith expansion. But I would still expect that existing editors would be looking at edits as fast as they come in, not because of watchlists but because they are at recent changes or new page patrol watching new edits as they come in. Though if an Editathon is focussed on particular articles, we have one coming up at the Royal Opera house which will focus on some of the people historically associated with them, then inviting the relevant wiki project is one way to alert the existing editors. Thinking about this I will drop an invitation to the ROH editathon on the talk page of established articles we are going to focus on.
I think this would be an interesting topic for someone to do research, whether the problem is with watchlisters or patrollers should be easy to spot. When existing editors come into conflict with newbies at editathons, looking at the established editors editing are they past contributors to that article, and before their intervention were their edits to random new pages/random pages edited in the past few moments or were they to other articles they had also previously edited. If we can identify the types of conflicts going on we can be clearer about the types of changes that we need people to make. My editathons rarely hit problems of conflict with existing editors, and when they do it is usually that people insist on writing an article on a subject where they can't find the two independent reliable sources that I suggested they need before creating an article. Partly I avoid this by starting people off with easy steps, signing the event page may get them an edit conflict but also shows them how to resolve a simple one. Creating new articles in sandboxes, and warning them that others will probably edit them within minutes of their moving to main space; Encouraging people to save frequently means they don't have complex edit conflicts over multiple paragraphs and telling them always to leave an edit summary, "edit summaries are optional, so vandals rarely bother with them and even a one word edit summary if expand, pic or typo is code for I am not a vandal."
Regards
Jonathan Cardy
On 26 Sep 2014, at 00:31, Kerry Raymond kerry.raymond@gmail.com wrote:
Australian outreach events generally edit Australian content. Other Australian editors are likely to be on the watch list and are likely to be in the timezone. And plenty of non-Australian editors are sitting in their pyjamas at all hours of the day and night waiting to pounce. Believe me, new editors encounter other editors very quickly (although sometimes they don’t realise it). They often think it “rude” that other people are editing the article “while I am in the middle of working on it”. Their mental model of collaborative editing is like a shared lawn mower. You have sole use for a while; then it is passed on to the next person.
Not sure I can help you with London editathons. But I do have a couple of edit training days coming up in Oakey, Queensland in a few weeks:
https://en.wikipedia.org/wiki/Oakey,_Queensland
Kerry
From: WereSpielChequers [mailto:werespielchequers@gmail.com] Sent: Thursday, 25 September 2014 9:38 PM To: kerry.raymond@gmail.com; Research into Wikimedia content and communities Subject: Re: [Wiki-research-l] FW: What works for increasing editorengagement?
Yes, training newbies is a great way to learn and to see the flaws that we mentally blank out. I also found that I need to keep a vanilla account for demonstrating things to newbies, if I use my WereSpielChequers account the various extra buttons confuse people.
I wouldn't worry too much about watchlisters making edits and causing edit conflicts, most of the time you aren't going to be in the same time zone, watchlisters even the most active ones are unlikely to check their watch list more than a few times a day. So as long as you don't start newbies on highly watched articles like Sarah Palin you should be OK with them. But edit conflicts are a real problem for newbies and especially those creating new articles. The new page patrol people need to look at articles as they are created in order to pick up attack pages etc, and when they find OK articles they tend to at least categorise them. Some of this could be fixed by improving the software for handling edit conflicts, for example it would be nice if adding a category and changing some text were not treated as a conflict. However this is the sort of Bugzilla request that gets closed as won't fix because of the implicit assumption that any editor should be able to handle an edit conflict in their stride. My preferred solution is to teach newbies to always start new articles in sandboxes and move them to main space when they are ready, alternatively creating a references section means that categories can be added without causing an edit conflict.
Re Pine's comment. When it comes to research topics It would be great to know how many editors are driven away by edit conflicts. But I don't know if that info is actually logged, and if you ask people why they gave up they don't necessarily know whether their edit conflict was with a machine or another editor. I once had to convince an angry editor that they hadn't been reverted by another editor, both edits had the same time stamp and theirs had been lost by the system without the other editor knowing. So if you ask former editors why they went and they blame other editors it may partly be that they don't know the difference between a conflict with the system and with another editor deliberately reverting them.
By the way, if anyone is around in London drop me a line and I will tell you if we have any editathons going on. Editathons have many roles in the community, one being as a focus group on the user interface.
Regards
Jonathan Cardy
On 25 Sep 2014, at 10:09, Kerry Raymond kerry.raymond@gmail.com wrote: This pretty much matches my own observations.
I think it’s not uncommon for new editors to first write about something they know well, which often means some level of conflict of interest. But that doesn’t necessarily mean the person is writing lies or being completely over-the-top with lavish praise; they can write quite reasonable content too. And generally new editors don’t know that CoI editing isn’t welcome. Similarly a lot get blocked for 3 reverts and get accused of edit warring, but I suspect it is often not the case. I suspect often they look back at the article and can’t see the edit they did, so they just do it again because it didn’t “stick”. I am not convinced they do know that other people are reverting their edits because I don’t think they know about the article history and the edit summaries and the user talk pages.
We make some awfully big assumptions that new editors know what we do; it’s amazing how much stuff in a UI you mentally filter out. I did a talk about Wikipedia about a year ago for a group of librarians. They asked for some “tips and tricks” to be included, so for the first time I systematically scanned the article UI and found cute little features I never knew were there (even on menus I often used). And of course the UI changes over time and probably I didn’t notice new things appear. But not noticing is consistent with “mission focus”, which has been studied in retail settings. A customer “on a mission” (a specific transaction that must be dealt with, as opposed to a customer “just browsing”) is oblivious to promotions until the “mission” is complete and then becomes aware of their surroundings as they turn to leave the store (which tells you where to put promotional signs relative to your “mission” products). I think Wikipedia editing is very mission-focussed (I must fix that error) so we see nothing but what we want to see on the way in and then have no “exit screen” to talk to the editor as they leave (unlike a bricks-and-mortar retailer). It’s hard to communicate with new editors on-wiki.
I actually think doing a user experience experiment (ideally with eye tracking) with new users would be very informative. Given them a “mission” and watch what they look at, what they do and if they succeed in their task, and whether their edits survives over time (and if it doesn’t survive, why is that). Indeed, seeing what they do when their edit disappears would be interesting too.
It is difficult for many of us as experienced Wikipedians to see Wikipedia as the new contributor sees it. I have some sense of the new editor experience because I do edit training and it certainly shows why nothing is fool-proof because fools are so ingenious. It’s amazing how new editors dive into parts of the UI I’ve never even noticed and then tie themselves up in knots. They often get so interested in Preview that they never Save (and then suddenly oops, all their work is gone). They generate continual edit conflicts (often entirely on their own, I think by going back in the browser and editing the article in multiple browser windows, leading to their contributions being smeared across a number of edit windows, none of which can be saved, or so it seems). And if they are editing in mainspace, they panic if the article changes in some mysterious way and they don’t know how they did it (it was another concurrent editor). In fact, if there is one thing I would like is for the watchlist notification to be slower than normal when a new editor is involved as having someone alerted by the watchlist coming and “fixing” things (even if it’s well-meant) puts the new editor into a spin because they don’t understand what’s going on (they have no experience with collaborative document editing). And, a perennial favourite, why is the Reference section empty when they’ve added all those citations (they don’t expect them to be in-line, they expect them in the References section). When I look at the constant questions and problems I have to deal with in edit training, I am not surprised that people trying to learn on their own give up. Seen through the eyes of the beginner, it’s so much harder than we realise both technically and collaboratively.
Kerry
From: wiki-research-l-bounces@lists.wikimedia.org [mailto:wiki-research-l-bounces@lists.wikimedia.org] On Behalf Of WereSpielChequers Sent: Thursday, 25 September 2014 2:18 PM To: Research into Wikimedia content and communities Subject: Re: [Wiki-research-l] FW: What works for increasing editorengagement?
Hi, if we are analysing new editors creating new articles then there are two groups I would expect to see, individual spammers and corporate spammers.
The individual spammers are the kids writing about the band that will be the next big thing on the ....... Scene in their town, and their first gig is next Tuesday provided they can recruit a drummer.
The corporate spammers are writing about some business, but in a style that makes a corporate flyer look neutral.
There are some implicit assumptions that we make as new page patrollers, and which some research might help by proving or disproving.
1 people whose first article is an attack page or vandalism will very rarely turn into productive editors. I have met or heard of three former vandals who made good editors, there may be more, but I've never heard of a good editor who started out creating attack pages.
2 people who are paid to spam this site are unlikely to turn into good editors, but a large proportion of good editors made COI edits among their early edits. The classic wikipedian is a time rich altruist with Internet access that they can make personal use of, I'm not seeing much overlap there with corporate spammers, but people who make COI edits other than about their employment clearly have personal access to the Internet.
3 neutral point of view is something that new editors need to pick up, and POV editing amongst a new editor's early contributions is so common that if a new account writes neutrally they are assumed by some to be sock puppets or returning editors.
4 we live in a copy paste world, and while a student plagiarising to try to get better marks is clearly doing so in bad faith, a wikipedia contributor using copy paste is still a good faith contributor, they just need to be taught not to use copy and paste. Presumably these people are in Aaron's group 3, again we have a lot of former copyright violators in the community, and a lot of new articles get deleted as copyright violations.
My suspicion is that much of our tolerance of vandals is wasted effort. We would save a lot of time and not lose any good editors by moving from the current five strikes and you are out approach to vandals to one of blocking any new editor after their first edit if that edit was blatant vandalism. Copyvio by contrast is something where we could probably retain more of the editors by trying different approaches to explaining why their edits had to be rejected and how they could contribute.
Regards
Jonathan Cardy
On 25 Sep 2014, at 00:19, Aaron Halfaker aaron.halfaker@gmail.com wrote: Sure! You'll find the hand-coded set of users here within the next half hour (cron job copies datasets over).
Categories: Vandals - Purposefully malicious, out to cause harm Bad-faith - Trying to be funny, not here to help or harm Good-faith - Trying to be productive, but failing Golden - Successfully contributing productively -Aaron
On Wed, Sep 24, 2014 at 1:15 PM, James Salsman jsalsman@gmail.com wrote: Hi Aaron,
Is the data set from https://commons.wikimedia.org/wiki/File:Desirable_newcomer_survival_over_tim... available for correlation with the number of new articles each user created?
Aaron Halfaker ahalfaker@wikimedia.org wrote:
... I propose a project where we work together to generate a summary so that I can call it "work." I've started a stub here: https://meta.wikimedia.org/wiki/Research:New_editor_engagement_strategies
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
wiki-research-l@lists.wikimedia.org