Monday, October 25, 2010

Same but different

I never got round to continuing with my same but different analysis.

The next step is to relax the simple binary classification of attributes & to assume that each follows a normal (bell shaped) distribution.

Most people would still come out not too far from the average on each of the measures, , sitting comfortably under the higher part of the curve amidst lots of people more or less just like them.

But a small proportion – say 5% - would be right out under the tail – noticeably different from the rest of humanity. And a tiny fraction would be in the tail of the distribution of all our twenty chosen characteristics – real freaks, though fundamentally no different, having no non-human characteristic.

But it must be very hard to maintain equilibrium & sanity in that situation.