Risks section discussion
I agree and think this is a huge risk. I work in RC patrol from time to time and frequently come across well-meaning, but unsuitable, edits from (generally anonymous) first-time editors. Certainly, I do my best to be polite and provide advice and assistance, especially to new users, but this is always balanced against time and the fact that I have a queue of other articles to address. Some editors are even more aggressive. It's easy to see how many editors would be strongly discouraged by their initial attempts. Here's a few possible steps we could take to mitigate this problem:
- We've done a lot of study on the editing interface for new editors, but little on the social aspects. What are the most common non-vandalism errors made by new editors that cause reverts? What sorts of messages do they get back on their talk pages? Do they read their talk pages when they get messages? What messages encourage them to keep contributing and what cause them to give up and leave the project? Are the welcome messages helpful and how could they be more helpful? If their initial edits are rejected, do they try to make them again? Discuss with the editor involved? Give up? What happens when new editors are sent to lengthy policy documents? These are all important questions on which we have little data. Further usability research here would be quite useful.
- Work with the Welcoming Committee to revise welcoming templates. Perhaps we could have something that basically says "Don't get discouraged. We value your contributions. Most Wikipedians are great, but like anywhere else in the world, we are all different kinds of people. If your first efforts don't work out, stick around and try again." Better incorporate the Five Pillars and simplified ruleset to address the most common sources of errors made by new editors.
- Improve the messages received by new editors when their edits are reverted, especially the templates generated by tools such as Twinkle or AWB, to be as polite and encouraging as possible. In the case of non-malicious edits, encourage users to discuss (provide a link to the right place for this discussion) and not to give up on editing. The key point needs to be "Welcome and thanks for your contributions. Here's what happens now." Not "You've screwed up. Go read pages of rules."
- As KrebMarkt suggests, make new editors more identifiable at a glance. This should probably be done by edit count, rather than by time, since someone may have had an account for a long time but still be unfamiliar with editing. However, we should still be helpful and civil to all editors.
- Barnstars and other reward/encouragement mechanisms for those assisting new editors.
- Only if proposals like the above are implemented can a "Teach, Don't Trash" campaign be successful. Most of us are going to go for what's quick and dirty, which especially includes the automated tools like TW. The templates need to be improved before individual editor behavior is addressed.
I think you've made some really good points. We should gather information on common types of mistakes new editors make and create "in a nutshell" templates (and modify existing templates as you suggested) that can quickly and easily be used when a new editor makes a mistake based on this data. We should also make it easy for new editors to ask questions when the template isn't adequate and they don't understand what they did wrong. This can be something like the Wikipedia Help Desk but specializing in new users. I know there are helpful people on Wikipedia and things like barnstars and other recognition for this work would be nice.
I agree that this proposal has some serious risks and that the risks outweigh the benefits. This is based on my own experience in religion articles, where almost everyone on the planet has a different viewpoint. I suspect that religion and politics articles are the most likely to suffer from a blanket encouragement of everyone. §The situation is already bad enough in the articles I focus on. Awareness of Wikipedia rules about point of view, original research, and referencing is low to non-existent. These rules are continually being flouted. §People frequently insert their own personal philosophy or viewpoint, without references. Joker comments are frequent. §A “be bold” policy would encourage a great increase in the incidence of the above and of the work involved in keeping article well referenced and free of POV and OR. §Also, this proposal is to some extent in direct conflict with the “expert review” proposal. Experts are likely to be turned off if they find articles riddled with the above faults, and may retire from reviewing. The very next day after any changes resulting from their review, someone may come along and reverse any improvements made.
I think we are aware of potential the screw up.
This proposal will be probably the last one to be implemented from a set which would contains:
- Make others aware of new editors
- Make new editors aware of mentors, admin & "Senior Editors"
- Define & Demarcate editing areas where new editors can contribute and blunder
- Demarcate editing areas requiring more wiki-experience & culture
- ...
So yea we are going to be cautious on that one while doing as much prep works to maximize adhesion of new editors & limiting negatives side effects.
I'd be interested to hear some statistics on why this "be bold" proposal is even necessary - is there evidence of numbers of active editors declining?
- Is there evidence of articles' quality being affected for lack of editors?
- What exactly are the negative effects of not having enough editors?
- I note on the proposal page that the hope is that there will be more people to correct errors.
Isn't it just as likely - or even more likely - that a flood of new editors who are not well acquainted with the rules will create a lot more new errors than they correct in old errors? We know the negative effects of an excess of editors in those pages that are plagued with POV, OR and zero referencing.
- If the overall effect is to increase the workload of editors who are familiar with the rules, in correcting new errors rather than improving content, then the overall effect on quality will be negative.
- I would suggest doing a carefully controlled experiment on specific Wikipedia entries, with a "be bold" invitation only on those selected entries and not generally. After a few months these pages can be monitored to see what the increase has been in editing, and how much reverting and correcting of the new edits has happened versus how many corrections of previous errors. Then you can see the benefits and also the costs of the proposed change.
Thanks for these data KrebMarkt, they are fascinating. Just taking English for simplicity they do show a decline in NEW editors from a peak of 13689 in April 2007 to 8001 in Jan 2010. That might seem a matter for concern. HOWEVER, the TOTAL number of editors has gone on rising over this period from 233,573 to 558,878. That's an absolutely enormous number of people.
The total number of monthly edits and of new articles has declined. HOWEVER, that's not necessarily cause for concern. I think there's one overriding factor to consider here. As time goes by there is less to be done - all the big subjects are comprehensively covered and editors must be aware of that. Many of them are historical and don't necessarily need updating very often.
The rewards from building up the articles in the first place are considerable - the rewards from finding minor errors and adding detail are very limited. The need is also smaller - it's like the shift from 200,000 builders of the pyramids, to a smaller force maintaining and embellishing it.
I would still like to see evidence of any deterioration in the final product before we are launched into a measure that could actually increase the errors rather than reducing them. Sure, the number of small errors introduced will creep up if no-one attends to them - but is there evidence that no-one is attending to them? Is there evidence of any rise in the errors?
More stuff:
What is more important than the potential community decline, is the lack of Editors diversity in Wikipedia.
Do nothing and that bias won't change unless you are happy with that situation.
I see that women are very much under-represented, and single childless men are very over-represented. If there were any race data, it's probable that blacks would be very under-represented. There are probably social reasons for this - the people there are most of are those with most free time alone and those with most familiarity with computers. And because of that some bias will probably always be there. These biases probably affect content. I doubt if they affect POV much, but they may affect what subjects and aspects are better covered.
It seems worthwhile to try to remedy the imbalance. However, most efforts to correct diversity imbalances in other spheres address the imbalances directly - eg by positive discrimination, or appeals directed at that group only. A generic call to "be bold" is not likely to attract women and couples with children more than single guys, and so would not improve the diversity. Are there plans to direct the call to women Wikipedia users?
Re the number of editors with 5+ edits per month declining, I see that as a normal result of the content of Wikipedia having reached some kind of maturity. To take my own example, I became VERY involved in the set of articles that interest me most in January 2010. Over the first few months I made many many edits to get these articles into acceptable form. After the basic groundwork was done, however, my rate of edits declined and now has dropped maybe to 5 per month, maybe less. If you view this as magnified over the whole of Wikipedia, the fact that so many articles are already in place, and already large, is bound to reduce the need for edits.
I was thinking to use "be bold" aimed to areas where we are severely lacks stuffs with the idea that the underrepresented population to be most likely to answer the call in those areas.
You are lucky to not edit contents related to Fiction, it's a never ending works there. From my experience my project in the English is failing to recruit new active editors with the ultra majority of the project members close to the 10K contribs and beyond. This doesn't mean there is no work to do with ratio of over 400 articles per active project members, keeping articles in line from fanboys & vandals or avoiding GA/FA articles quality erosion is complicated.