New Users creating New Articles

It might be enlightening to study the retention rate of new users who begin their editing career by creating a new article, compared to new users who begin with more tentative attempts at editing existing articles. From my own observation at en.wikipedia, those who start by editing existing articles have a relatively smoother learning curve, with fewer surprises. Many (most, perhaps) new articles that I've seen created as a user's first edit have little chance of being kept, so their introduction to editing is bound to be negative.

Based on that, a large part of the problem seems to lie with new users creating new articles. Solutions include: education before a brand new user can create an article, in the form of a better 'new article by a new editor' template/window; live chat for the new user creating their first article (as suggested above); allowing article creation only after a certain number of edits.

The problem could be new users creating new articles, or new articles by new users being deleted. I would not say from the outset that it is the former and not the later. If Wikipedia wants to grow and attract new users it has to accept, that more people will have more diverse interests. If existing users delete everything they personally deem unimportant, then they will remain among themselves.

That would have to be addressed by changing the deletion policies on en.wikipedia — to overhaul them and allow articles that are currently considered non-notable, advertising, spam, neologisms, etc. to be kept. But simply allowing a new editor's article to remain for a longer time, until it were deleted under current policies, would only add to their frustration since they would likely put a great deal more work into it before their project is deleted.

No! The project does not improve by welcoming new editors who write crap so they will continue writing crap. There is already too much advertising (a lot of it well written by corporate PR flacks), spam, and articles about trivia. The goal is to help new editors understand and conform Wikipedia's content standards, which are already much looser than the standards of traditional, professional publishing. A lot of experienced editors treat well meaning newbies badly, with nasty edit summaries on their reverts, nasty talk page posts ("Welcome to Wikipedia, thanks for your contribution, but what you wrote is utter nonsense ..."), and overly aggressive use of warning templates.

I disagree with you about "articles about trivia". One man's trivia is another woman's hobby, and if it meets notability, it's not trivial. Moreover, in some cases it seems to me the notability guidelines are too strict; many academics only meet them once they die, when obituaries become available, even though they are major figures in their fields.

I agree about the rest though.

One question is what is perceived as "nasty edit summaries". I recently observed a non-native english speaker edit warring on english wikipedia about grammatical correction made by a native speaker, who as it happens writes extremely well. I don't remember the edit summary used by the corrector, but it was probably "copy edit" or similar.

Re: edit summaries - when I correct what I think is a grammatical error, I try to link to a Wikipedia article that explains it, such as Dangling modifier from the edit summary. Links are especially necessary in edit summaries with Wikipedia jargon or shortcuts. Unlinked jargon annoyed me when I was a newb, especially after I learned how easy it was to link. I realized there were other editors who couldn't be bothered to make themselves understandable by typing a few extra bracket characters. So I vowed not to make the same mistake too much.

It's not possible to create an article without first doing some number of edits under one's new user ID - even if one is an experienced wiki editor, using wikimedia software routinely on other projects, and has furthermore made many anonymous contributions to wikipedia. One must also wait several days.

On the English Wikipedia, an editor can open an account and create a new article as their first contribution a minute later. I just looked at one to be sure.

When did that change? I distinctly remember having to accumulate 10 edits under my spiffy new user ID, and wait 3 days.

The 10 edits and 4 days is required only "to move pages, edit semi-protected pages, and upload files or upload a new version of an existing file", not to create an article. That can be done by any registered user, with their very first edit. See en:WP:AUTOCONFIRM.

That depends on what the new editor brings to Wikipedia. Wikipedia is not the only place, and is very far from the best place, to learn how to write. Although it is not a large percentage of newbies, some newbies come here already equipped to write an article that is better in content and writing style than the average Wikipedia article. A paper that earns a B at a decent college is better than the average Wikipedia article. Good newbie writers will eventually learn whatever they don't know about wiki markup and Wikipedia style conventions; experienced editors can fix these kind of mistakes until the newbies learn.

True. Which is why I'd prefer to encourage newbies to look at similar articles before creating one. I suspect most people bright enough to write a decent article are bright enough to figure out how to use the formatting code which already exists in a similar article. The problem is with us because instead of suggesting that obvious approach, we tell them to read endless material about how to do anything and everything from scratch. I'm not sure I could create an Infobox from memory myself - I always search-cut-paste-change one of the sort I want. Doesn't everyone? (I don't memorize baseball statistics either, thanks for asking.)

The thing is, most of the 'training material' might be suitable if one were training for a (paid) career, but we shouldn't self-limit Wikipedia to be the province of only those who can and want to do this as a full-time occupation. imo, trying to come up with a WYSIWYG interface misses the point - someone still has to know what it is they want to create. Which generally means seeing something in an existing article. So copy that article's code! The first thing on the to-do list should be to replace the current nonsense with useful descriptions for the hover-cursor messages in the wiki markup box. You may as well hang out a sign: We Hate Non-Nerds! and be done with it. Fix the obvious first instead of looking for endless ways to spend more money and time. This. Isn't. Rocket. Science. A few descriptions would solve the problem for 90% of the people. (I really can't believe no staff or Admin has ever noticed that. That tells you a lot, right there.)