Wiki Survey Motivation

See WhoIsActiveOnWiki.

Ok, the first motivation is simply to get a "health" status of the wiki community. So C2 had a glorious past being considered one of the premier programming communities. The everybody agreed things went South for various reasons that are of no importance here, then things got quiet for a while. Undoubtedly some will like wiki to recover. But the big question is where is wiki now ? In absence of data, all we can do is useless speculation.

WR series of question tries to gauge the perception of wiki as seen by the WikiReader. Is wiki a good source of information, is it trustworthy ? Is it something readers care as much as some of them to fill in a survey ?

The WC series, tries to gauge the reliability of Wiki as a source of information. As with any online forum the quality depends crucially on the feedback loop. This is how WikiPedia works to achieve NeutralPointOfView, it's how other good online forums (comp.lang.functional for example) work. If somebody posts something incorrect or if somebody provides a partial picture on comp.lang.functional, or on WikiPedia, it will not stand without correction. Is this also the case on Wiki ?

Comment:

But, this is WikiWiki. We were the first and we do things "our" way.

Oh, but of course. Let me clarify that the survey was not designed with the bias that NeutralPointOfView was the best, not even the best that can be applied to C2 wiki. Hopefully it wasn't designed with the opposite bias (that NPOV has no place on C2), either. Whatever "our" way is, it'd be very interesting if we can find something about it.

If anybody thinks some questions are unrepresentative, biased, unuseful,that there could have been better questions, more important things to ask in a survey, let's hear about it. Anybody can add (within reason) something to the questionnaire, but I'm affraid that if it becomes too long people will not bother to respond.

As I said, the minimum amount of a response is adding your name weven if you don't respond to any of the questions. Why would you bother ? If not for the knowledge experiment (can a wiki community really find out stuff about itself ?), then as a token of appreciation for the work of WikiGnomes, Ward, etc. And possibly thank you for the work of contributors, but contributors can be under at least the theoretical suspicion of deriving other benefits -- like seeing one's name on a page.

If we end up not finding much at all about the process of wiki, it'd still be great if we find out at the end of that 100 readers still follow wiki and are motivated enough to sign their name there, it's a form of ThankYou. To make a risquee analogy, in US there are public radio stations (supported from public donation), you can hear pretty decent journalism on some or very good jazz selection on others if you ask me, but every quarter or so they reduce greatly their OnTopic programming and have a pledge drive where listeners are asked for their support through donations in order for the station to be able to continue its activity. They estimate that roughly 1 in 10 listeners contribute with donations and some stations raise pretty good money that allows them to compete and manytimes outperform commercial stations.

So if you don't think of it much as a survey, think of it as a pledge drive, let's raise some ThankYou notes, shall we ? Don't respond to any question, just add your name there. --CostinCozianu

Of course, I'd be grateful if you also respond to the questions, and the ThankYou note is not in any way diminished if you respond "negatively" to some of them, as CriticsAreYourBestFriends and information is extremely valuable.

Q&A (Brainstorming)

Q:Shoud the survey have been anonymous ?
A:That would have been ideal. Unfortunately I couldn't figure out how to have it anonymous while keeping it credible. I went with DoSimpleThings, in the spark of the moment. People should not try to add justifications for some of their responses either -- because this demotivates others, but could blog on their home pages. I'll see to it to erase all the names at the end and keep just the data (or if not enough people will respond, we'll have to disregard the data as well).
Q:I might contribute, but would like to stay anonymously, now what ?
A:You could try to sign with your IP. There's the danger that someone could try to use this mechanism to falsify the results, but we'd still have the data for which people sign with their name. If you do so, and nobody will find a problem in the data (like many open proxies, etc) maybe such data can be used by others (I'd still be reluctant to draw conclusion). I was really conflicted on this one, making it anonymous would likely have encouraged participation a great deal, so if anybody tries to find inventive solutions, let's do it. I'm thinking of a cryptographic scheme, but in the same time it'll have to be very easy to use, asking people to use openssl command line doesn't sound very appealing. Any ideas, please ?
Q:It is possible that the way questions are formulated, and with low participation, results could be used to push an agenda. And I might have a problem with the originator.
A:True. I'm sorry I cannot do anything about that other than plead my good intentions. It's just to find out some data, that I consider valuable. If you do not trust the originator, and you don't try to engage in improving the questions, you can still sign with your name without responding to the questions you don't like, or without responding to any for that matter. Let's imagine that at the end we get 1000 signatures from which 990 do not respond to any question. That'll still be great !!! Then the originator will have to take responsibility for being untrustworthy -- and any agenda real or imaginary will be doomed, but the community will get a big kick out of it. It'll still mean 1000 people who really care. Participating is a no lose proposition and most likely a WinWin.
Q:How much will the survey take, is it open ended?
A:I'd sure hope not.However I'd like to be sure that all who could particpate will be able to do so. Let's set the limit to 1 month (ends on 23rd september) so that there won't be excuses like people going on vacation, missing the RecentChanges/QuickChanges entry, etc. At the end of this if only 10 people will have responded, then the data for the answer will be irrelevant, but the fact that of all people only 10 will have participated will still be a valid piece of information, however important or unimportant people will find it.

Discussions

I suggest that a table layout is ergonomic, whereas the existing system above is non-ergonomic. Example below. -- DougMerritt

 Terse column headers; see above for explanation.
 Wiki Readers is WR*, Wiki Contributors is WC*:
 WR0    |WR1    |WR2      |WC0    |WC0.5    |WC1     |WC2      |WC3
 Wiki   |Insight|Insights:|Create |WikiGnome|Result  |Delurking|Seen
 Useful?|freq.? |pages    |pages  |pages    |bliss:  |expert:  |Utter
 4: Very|X per: |of age:  |per:   |per:     |3: bliss|4: yep!  |Crap?
 0: Yuck|day..yr|days..yr |day..yr|day..yr  |0: yuck |0: no way|y/n
 -------+-------+---------+-------+---------+--------+---------+--------
 Additional: Wiki Contributors: WR3: other online info sources URLS? (short)

E.g.:

 CostinCozianu
 WR0    |WR1    |WR2      |WC0    |WC0.5    |WC1     |WC2      |WC3
 Wiki   |Insight|Insights:|Create |WikiGnome|Result  |Delurking|Seen
 Useful?|freq.? |pages    |pages  |pages    |bliss:  |expert:  |Utter
 4: Very|X per: |of age:  |per:   |per:     |3: bliss|4: yep!  |Crap?
 0: Yuck|day..yr|days..yr |day..yr|day..yr  |0: yuck |0: no way|y/n
 -------+-------+---------+-------+---------+--------+---------+--------
 2       1/yr     1/yr     1/mo               1.5      3        1 (y)
 WR3: C2 2.5; comp.lang.functional 2; comp.object 1;
 GarryHamilton
 2+      1/mo     1/yr      2/mo    4/mo     1.75      2.5       1(ItDepends)

...which incidentally makes it easier to spot that CC was WC2:4 at times; his WC:3 is more recent. (his WC3 is historically unchanged ;-)

I would appreciate if we can keep discussions on a separate page. Maybe WikiSurveyMotivation. Yeah, and I am the next Lord Vader of wikis worldwide.

I'm not sure a layout that makes it easier to spot what a particular person thinks would be better. Like I said, in an ideal world this survey would have been anonymous, and while it's unavoidable to be curious how person X or Y responds, I urge everybody to forget what they see about any person in particular. When it's done, we'll erase individual responses and keep the aggregates. That's the point of surveys.

If anybody else wants to take over, provide summaries at the end, I'll be thankful to be relieved of more suspicions. --CostinCozianu

Hmm. I certainly won't fight you on this, I was just trying to be helpful about a narrow issue. But I truly think that your original format was completely unreadable. It's not that I am taking a position on non-anonymity or something, it was an observation about a write-only format versus a read-write format, that's all. I will now bow out and leave the resolution to others. -- DougMerritt

RE: "whether we are doing a good job or not, or whether it's a job worth doing." Makes you wonder why you ever worked so hard and caused so much fuss over a few OffTopic pages? In other words, there would have been something worthwhile to read here on wiki.

I will support Costin's study, even if I'm not clear what its good for. Please note, that I added WC0.5 to measure gnoming activity too. -- .gz