On Sun, 31 Aug 2003 20:31:54 GMT, "Gordon Couger"
wrote:
"Torsten Brinch" wrote in message
.. .

On Sun, 31 Aug 2003 01:21:55 GMT, "Gordon Couger"

As I read the study it is compares actual costs snip

Oh,so. How come, then, that the author is apparently not in a position
to calculate true average actual costs, seeing he uses midpoints
in his cost ranges as stand-ins? (indeed without making a note of it)

Gordon, I tell you the author takes the midpoints from his cost
ranges, and calls that averages. He doesn't even make a note of it.
You find that OK?

..

Where do I say he discarded data points.

Nowhere, that's the strange thing. You can see the author reports
finding his data to be in the range 3.0-3.6, and you can see
the author has had a data point at 4.2. Dog ate it?

I said if we as the whole set of data point that the average were
calculated from we would probably see wide variations in them.

The author reports average 3.1, range 3.0-3.6 -- and you
need to see the whole set of data points to know how wide variation
is in them ???