On 10/29/07, phoebe ayers <phoebe.wiki(a)gmail.com> wrote:
Hi Greg,
I know as research officer you are well aware that the
results from such an
experiment will be of interest not just to the en: community itself, but
also to the wider wiki research community. Is there a page detailing the
metrics you have in mind, and listing possible studies that could be done to
determine "evidence of harm" from the switch? It seems like this is a good
chance for brainstorming on-wiki with both the research community and the
newpage patrol folks about possible ways to measure quality, etc., of new
articles, a discussion that seems overdue anyway given some general
unhappiness about deletion practices.
Also, before taking on such an experiment, it seems like it would be
worthwhile and sensible to run any intended metrics & studies on the current
state of affairs *first*, so there is something to accurately compare to.
I'd suggest not making such metrics public until after the experiment
is over. It'd be way too easy to manipulate the experiment if you
did. As it stands now it's probably already too easy to manipulate
the experiment.
AFAIK our understanding of what gets deleted, how many
pages get deleted
versus kept; how many articles are speedied a day out of these, etc. is
imperfect; feel free to correct me if I'm wrong. I'd also be interested in
seeing which studies you're referring to that inclusively suggest that "the
change has been harmful to the quality of Wikipedia"; I'm not familiar with
that work and it seems like a tough thing to measure given overall explosive
growth in this timeframe.
This would be useful. The recent studies I've seen suggest to me that
this change has had virtually no effect on anything. Of course, I'm
biased, that's the effect I expected it to have.