On 6/19/04 11:47 PM, in article aN7Bc.143806$Ly.57935@attbi_s01, "S888Wheel"
wrote:

From: Steven Sullivan
Date: 6/19/2004 1:50 PM Pacific Standard Time
Message-id:

Bromo wrote:
Care to provide examples where differences are not measureable?

When you don't know what to measure - or are measuring the wrong things.

And I predicted someone would retort in this fashion, several days ago.
Thanks for proving me right.

Are you suggesting we should not worry about people measuring everything that
matters or failing to measure everything that matters?


You should be a bit worried that if you set up a test that shows something
that observation shows otherwise - rather than assume that people are
deluding themselves (which may be entertaining and somewhat possible in some
cases) - you should entertain the notion that the test itself may not be
measuring the right things.

For instance - there is a lot of faith (yes FAITH) placed in ABX tests.
What are you measuring in an ABX test, really? It is repeatable, but is it
measuring the right things in the right method? These are the things that
should be bothering a true scientist. It is a bit like measuring the speed
of gravity and saying that the people who observe a feather falling slower
than a bowling ball are deluding themselves. Perhaps you aren't measuring
the right things .....