At 01:38 PM 4/29/2013, Ted Lemon wrote:
On Apr 29, 2013, at 1:08 PM, John C Klensin <john-ietf(_at_)jck(_dot_)com>
wrote:
If raising awareness and sensitivity
isn't enough to get people to think about and make decisions
differently
Statistical analysis shows that even when peoples' awareness is raised, biases
continue to exist, not because the people are bad people, but because
cognitive biases are simply not affected by "consciousness raising" alone.
So IMHO at least, what we are looking for here is not consciousness-raising,
but some method of determining if we are indeed suffering from cognitive
biases here, and if so, some method for actually addressing the problem.
Yup. The problem here is that the sample set of "leadership" positions is so
small its difficult to get any reasonable measure one way or the other. When
you start mixing and matching gender, race, citizenship etc into the pot as
possible determiners it just gets worse.
The normal measure for determining whether one population is distinct from
another appears to be the Chi Squared test.
Throwing in a matrix of
WG Chairs IAB/IESG Members
Male 187 25
Female 15 1
And running the calculation (http://www.quantpsy.org/chisq/chisq.htm) using the
Yates' values (because the sample size is so small), there is a 79.13% chance
that any observed differences in the composition of the two groups is solely
due to statistical variations.
And playing off of John's message, if you look around 2005 when there were 4
female members of the IAB and IESG (and assuming the same composition of WG
chairs), that calculation yields something 31.4% - or 2 chances in 3 that the
differences were due to something other than statistical variations.
When I look at this as a pure numbers problem, I'm unable to say there is a
cognitive bias in the selection process and in fact the numbers would argue
against being able to say that without a much larger set of IAB/IESG members.
Mike