User:Pelagic/Incubator/Classifying Teahouse/1123

This is an experiment with counting how many questions of different types are asked at the Wikipedia Teahouse, using Archive 1123. I'll see how it goes editing this on-wiki, but may have to switch to Excel for analysis or easier data entry.

A more elegant solution would be if there were some way to tag individual sections and then use a script to calculate totals for each archive.

As a first pass, I am using some keywords or phrases, then will look at other ways of grouping.

Possible abbreviations: nq = no question, cnp = create new page/article, htu = hard to understand, nt = no section title (added later by hosts), afc-no = draft rejected or declined (unspecified which) at AFC, chun = change user name, udel = delete user, svp = confusion over save versus publish (yes this comes up often enough to need its own category!), rff = request for feedback, ot = off-topic, ...

How-Why axis

 * Why-occurrence: something happened (deletion, reversion, afc decline) and I understand what it is that happened, but not why
 * Why-mistaken: OP is asking why something happened but have a misconception about what actually happened. E.g. article seems deleted but was actually draftified, edits reverted but user doesn't know about history pages.
 * How: how do I do X? Maybe consider subtypes like how-technical, how-procedural, ...
 * Request for: feedback, help, ...

What axis
What it is about: sourcing, notability, referencing, accounts, disputes, style, ... history pages, ...

Common problem areas: create article, afc delays, help getting draft to acceptable quality; blp subject, coi or paid...

Interaction
Question answered / unanswered. OP responded (satisfied, disputed, ...?) / not responded. Quantity of replies. Number of users in topic.

Type of user
IP / new account / established user. Subsequently blocked (for reasons related / unrelated to the post). Block reason.

Type of article
BLP (musician, businessman, academic, ...), business, school ...