I agree with John and would like to amplify his point.
This is my view from an industrial physics perspective. One of the biggest
shocks when I came to work at General Atomics was being asked to make
technical decisions based on incomplete information. This, I have found, is
the norm, not the exception. You must utilize available information and
your best judgment to make decisions about where you need to go to solve
some R+D problem.
School did not prepare me for this at all. Every problem I solved had a
nice tidy answer. Even graduate school basic research that I performed,
wrote papers on etc. were very logical and based on good data. Judgment was
not a factor and careful wording was always used such as "the data
suggest..."
Even now I have to confront an issue as simple as the index of refraction of
a thin film coating. What is the index? It is not necessarily what you
look up in a book. The index can vary by quite a bit depending on
fabrication conditions, density, where the material is made, etc. It can
depend on sputtering voltage, various sputtering gas pressures, sputtering
target composition, sputtering target density, type of substrate on which it
is coated, substrate temperature, etc. These are typical critical
uncertainties that have to be dealt with on a daily basis in my work. The
error uncertainties that are important for high energy physicists and those
who determine fundamental constants are not a major concern in my life. I
need to know how many digits I can trust - but more importantly how the
process I am using affects my measurements. And there is never time or
funding to investigate how every variable affects the data. We need to
determine what the critical factors are and which factor are sufficiently
under control that we don't need to worry about them.
Dr. Lawrence D. Woolf; General Atomics, 3550 General Atomics Court, San
Diego, CA 92121; Ph: 858-455-447; www.sci-ed-ga.org
John S. Denker
Sent: Thursday, March 08, 2001 10:09 AM
A huge part of real-world physics (and indeed a huge part of real life in
general) depends on making approximations, which includes finding and using
phenomenological relationships. The thing that sets the big leagues apart
from the bush leagues is the ability to make _controlled_ approximations.