[Wikipedia-l] Semi-protection

Wed Nov 15 04:13:08 UTC 2006

On 11/14/06, SJ <2.718281828 at gmail.com> wrote:
>
> On 11/12/06, theProject <wp.theproject at gmail.com> wrote:
>
> > Well, I know I personally don't have the patience to wait a few days and
> > then add my contributions. The way I see it, anon comes around --> sees
> he
> > can't edit without registering and waiting four days --> decides it's
> not
> > worth the hassle --> perfectly good contribution lost, maybe more -->
> > Wikipedia suffers more in the long run.
>
> Absolutely.

I don't know if semi-protection is generally discouraging to new editors or
not, and without good user studies none of us will ever know for sure. As
far as I know, we are very much lacking in good data both about the
experience of new editors and of "ordinary users" who don't edit, which is
something I hope the research community will be able to someday remedy. It's
all very well to have individual, qualitative impressions (sure, the place
seems harder to navigate and less friendly than when I signed up three years
ago) but those impressions don't mean much overall unless collected widely,
counted reliably and aggregated.

> > I also disagree with anything that presumes bad faith about anonymous
> > editors. Wikipedia built itself by being as open as possible. Can we
> diverge
> > from that path of openness as little as possible?
>
> And agreed again.   To expand on GMaxwell's comment about limiting
> visibility of newbie edits: we should be as liberal as possible in
> taking in new text and information from contributors, and as careful
> as possible in how we display it.  Each view should very clearly
> identify, filtered up to the first screen of the main display, the
> history, community trust, level of discussion and monitoring, and
> referencing put into an article (especially its last dozen edits).
>
> SJ
>

I agree, and with what Gmaxwell & Ray have said. Presenting the best
possible view of an article -- the vetted and referenced version, the
version that's flagged as being "ok" by three people, whatever it takes --
is a service to our users, who are using the site to find information. These
people are and should be our primary audience. It's important to remember
that Wikipedia is a complicated system, it is difficult to figure out, and
that most readers likely don't bother. I've talked to a number of people
this year who had used Wikipedia but didn't realize it was user-created;
many other people know what it is but are massively fuzzy on the details
(someone once asked me if I had worked on all the science articles). If
readers like this are in fact in the majority (more data to collect!), we
should behave accordingly, and work behind the scenes to ensure better,
unvandalized content (as well as mounting education campaigns).

This presentation of course needs to be balanced with openess; but if
something doesn't significantly impair the ability to edit but does
significantly improve content it seems like a good modification to make.
Again, we need data to know if semiprotection, for instance, significantly
impairs or significantly improves (or both, or neither).

It's also worth remembering that people's ability to read disclaimers,
whether related to liability limitations, article reliability,
degree-of-vettedness or anything else is not something to be relied on, and
hence presenting good content (not questionable content tagged as "this
content may be unreliable, read at your own risk, but here it is anyway")
should be our goal.

I also wonder what happened with the de: pilot study. Anybody?

-- phoebe