Re: Request for feedback on statistical method

From: Shepherd Moon (shepherdmoon_at_yahoo.com)
Date: 11/04/04 

Date: 4 Nov 2004 12:03:22 -0800

Thanks to you as well for your reply. As I mentioned, I'll report back
with anything I find out in the sci.physics post that may bear on the
statistics.

Regards,
Shepherdmoon

Einar Andreas Rødland <e.a.rodland@labmed.uio.no> wrote in message news:<41890181.4020702@labmed.uio.no>...
> Shepherd Moon wrote:
> > There is one article that claims to have analyzed this claim
> > statistically and claims that there is a "95% confidence level" that
> > the speed of light has actually been decreasing over time. I'm trying
> > to find out if this trend is real or artifactual. I suspect it is
> > artifactual but am not a statistics expert.
>
> Ross Clement has already made some good points. There are more, though.
>
> As methods and technology has improved, we have been able to make more
> and more accurate measurements. Thus, more recent measurements are
> more accurate than the older ones. The statistical method used,
> however, assumes (as far as I can see) that the estimation errors are
> random and of constant magnitude, which is clearly not the case.
>
> Errors in physical experiment may be of several kinds: they may be
> statistical errors (related to random measurement errors) or they may
> be systematic errors (related to the method or experiment). The random
> measurement errors may be assumed to be independent from experiment to
> experiment, but systematic errors can not be expected to be
> independent. E.g., some of the early methods used stellar aberation
> http://scienceworld.wolfram.com/physics/StellarAberration.html
> which depends on knowing the speed of the Earth around the Sun; if the
> estimate of this speed is too high, it would consistently give too
> high estimates of c no matter how accurate the measurements are; other
> methods will have similar kinds of systematic errors. Thus another
> basic assumption is unfounded...or at least lacking a proper motivation.
>
> If you take a look at the light speed estimates at different times,
> the difference from the presently known value is typically in the same
> order as the stated uncertainty of the estimates. This is typical of
> certain 'remarkable claims': they tend to be based on deviations that
> are just about the same magnitude as the uncertainties...and these
> deviations tend to decrease as the methods improve. The author,
> however, seems to favour the explanation that the speed of light has
> changed gradually until around 1960 when it converged very quickly
> towards the value know today: it seems somewhat arbitrary for a
> physical constant, after 6000 years of change, to suddenly stabilize
> just as we start to measure it accurately.
>
> Another thing is that today the speed of light is constant by
> definition: it is the conversion factor between metres and seconds,
> where one (don't remember which) is defined in terms of the other
> given the defined value for c. The only fundamental dimensionless
> constant is the fine structure constant:
> http://scienceworld.wolfram.com/physics/FineStructureConstant.html
> If this changes, it will not be possible (or meaningful?) to tell
> which of the underlying physical constants have changed.
>
> If the fine structure constant (or speed of light) had changed
> dramatically in the last 6000 years, I'd think that should have given
> very clear effects on the spectra of moderately distant stars...but
> probably sci.physics would be the right place to ask about that. I
> recall some time ago there was a report, from serious reasearchers,
> indicating that it may have changed; but they were talking about the
> early hours of the big bang, not the 'creation of the universe 6000
> years ago'. I'm not certain what the final verdict was; but, again,
> sci.physics should be the right place to ask.
>
> > The statisitical article is here:
> >
> > Is the Velocity of Light Constant in Time?
> > by Alan Montgomery
> > Mathematician
> > 218 McCurdy Drive, Kanata, Ontario K2L 2L6 Canada
> > and
> > Lambert Dolphin
> > Physicist
> > 1103 Pomeroy Avenue, Santa Clara, CA 95051
> > http://www.ldolphin.org/cdkgal.html
> >
> > The original article is here:
> > The Atomic Constants, Light, and Time
> > by BARRY SETTERFIELD
> > and
> > TREVOR NORMAN
> > http://www.setterfield.org/report/report.html

Relevant Pages