M
Martin Brown
Guest
Bill Sloman wrote:
the application of the tricky correction factors in such a way as to get
the "right" answer according to established experimental procedures.
Looking for *why* there is a difference is *very* important.
A graph of speed of light with error bars is a sobering one in this
respect. Initial estimates were good but as the techniques became ever
more sophisticated and based on evacuated waveguides one of the top
experimentalists made a tiny error in applying a dispersion correction
taking the true answer outside the formal error bars. Successive
experimenters then refined the technique still further narrowing the
error bars without spotting the fundamental systematic error. ISTR it
was noticed around 1945 when another method gave a new answer.
A version of Shewarts graph updated is online at
http://www.sigma-engineering.co.uk/light/lightindex.shtml
It was around 1945 that a new method with the corrections all applied
correctly came into play. I forget who the experimentalist that skewed
the distribution from 1922 onwards. During that period there was
considerable interest in the numerological conjecture that the fine
structure constant might be exactly 1/137. We now know it isn't.
Regards,
Martin Brown
You should always be a little bit wary of consilience. It can also meanOn Dec 20, 1:32 am, John Larkin
jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
On Sat, 19 Dec 2009 14:22:45 -0800 (PST),Bill Sloman
bill.slo...@ieee.org> wrote:
On Dec 18, 10:26 pm, "JosephKK"<quiettechb...@yahoo.com> wrote:
On Wed, 16 Dec 2009 00:57:15 +0000, Eeyore <rabbitsfriendsandrelati...@removethishotmail.com> wrote:
Bill Slomanwrote:
On Dec 11, 3:04 am, John Larkin wrote:
Science used to rely on experiment.
Newton's astronomical experiments are famous, as are Darwin's
evolutionary experiments.
Physics in particular also relies on repeatable OBSERVATIONS.
Name a single observation that the warmingists can show is even real,
never mind repeatable.
Graham
Kind of difficult when Hansen et al., keep adjusting the data from what
the satellite reported to what they want it to have reported.
Try reading the raw satellite data sometime. And note that the biggest
recent correction to satellite data was made by Roy W Spencer and John
Christy at the University of Alabama at Huntsville, under a certain
amount of external pressure
http://www.sciencemag.org/cgi/content/abstract/309/5740/1548
http://en.wikipedia.org/wiki/Satellite_temperature_measurements
What a mess! People seem to keep trying various corrections until they
get what they want.
Actually, the correct formulation is "until the data makes sense". In
this particular case, the University of Alabama at Huntsville had been
putting out data that was curiously out of whack with everybody else's
data for quite some time. Spencer and Christy weren't exactly
energetic about checking out their data-processing for possible
problems, and ended up with a certain amount of egg on their faces
when they finally got around to replacing verion 5.1 of their data-
processing package with version 5.2, which brought their data a lot
closer to consilience with the rest of the world.
Consilience - the process of getting a lot of different and
independent sources to fit together into a consistent picture of the
world - is a concept that appeals to me.
http://en.wikipedia.org/wiki/Consilience
It's a very low frequency word, but an absolutely fundamental concept.
the application of the tricky correction factors in such a way as to get
the "right" answer according to established experimental procedures.
Looking for *why* there is a difference is *very* important.
A graph of speed of light with error bars is a sobering one in this
respect. Initial estimates were good but as the techniques became ever
more sophisticated and based on evacuated waveguides one of the top
experimentalists made a tiny error in applying a dispersion correction
taking the true answer outside the formal error bars. Successive
experimenters then refined the technique still further narrowing the
error bars without spotting the fundamental systematic error. ISTR it
was noticed around 1945 when another method gave a new answer.
A version of Shewarts graph updated is online at
http://www.sigma-engineering.co.uk/light/lightindex.shtml
It was around 1945 that a new method with the corrections all applied
correctly came into play. I forget who the experimentalist that skewed
the distribution from 1922 onwards. During that period there was
considerable interest in the numerological conjecture that the fine
structure constant might be exactly 1/137. We now know it isn't.
Regards,
Martin Brown