Some additional thoughts on collective IQ: (01)
A lot of people have expressed discontent with the implications of
"IQ": (02)
http://collab.blueoxen.net/forums/tools-yak/2003-03/msg00052.html#nid061 (03)
http://collab.blueoxen.net/forums/yak/2003-03/msg00053.html#nid02 (04)
I don't disagree with these assessments, although I don't have as
strong a reaction against the term. I also don't quite agree with
Henry's interpretation of Doug's notion of collective IQ: (05)
Doug's use of the term "collective IQ" does not refer to any measure
of intelligence. It is used in the sense of "two (three, four ...)
heads are better than one." (06)
http://collab.blueoxen.net/forums/yak/2003-03/msg00029.html#nid07 (07)
I think Henry's right in that Doug doesn't think that there is any
single scalar value that represents collective intelligence (whatever
that is), but I also think the point of using that term is that we do
have to try and measure it. (08)
Henry's reference to the recent Pittsburgh study on spell-checking is
one indication why. (09)
http://collab.blueoxen.net/forums/yak/2003-03/msg00047.html#nid01 (010)
We make a lot of assumptions about what will work and how it will
work, and those assumptions are not always correct. We assume that
spell checking will decrease the number of error's in peoples writing,
and at least one study demonstrates that this isn't the case. We
assume that video conferencing is a superior way to collaborate
synchronously, and yet several studies seem to indicate that
teleconferencing is actually more effective: (011)
http://collab.blueoxen.net/forums/tools-yak/2003-01/msg00022.html#nid07 (012)
The point is that we can't just work on assumption and intuition. We
have to do our best to measure effectiveness. (013)
All that said, I'm a strong believer in the importance of qualitative
metrics as well. I've discussed one qualitative metric before -- the
Squirm Test. (014)
http://collab.blueoxen.net/forums/tools-yak/2003-01/msg00045.html#nid04 (015)
Another favorite is Chris's Less Dumb Test. At the end of each day,
you ask yourself if your collaborative work has made you less dumb.
If it has, then you're doing something right. It sounds obvious, but
ask yourselves: How often do you ask yourself this question? (016)
Qualitative metrics like these are important, because they bring
context to quantitative metrics. One metric that interests me is how
often we link to the archives in our e-mails, and how often do we use
purple numbers. I think those numbers would be interesting, but it's
questionable whether you could read anything objective into those.
Just because I write e-mails full of links (sort of like this one),
that doesn't mean I'm collaborating any more effectively than
otherwise. The question is, do those links make us Less Dumb? (017)
To summarize, I think we've been converging on a definition of
collective IQ: metrics for measuring group effectiveness. The
followup question is, what should those metrics be? Again, people
have proposed several already. The next step should be to apply some
of those metrics to this group to see if they tell us anything. (018)
-Eugene (019)
--
This message is archived at: (020)
http://collab.blueoxen.net/forums/cgi-bin/mesg.cgi?a=yak&i=20030318181547.GA16436@douge.blueoxen.net (021)
|