Re: QM support for a preferred frame

Surfer wrote:
On Sat, 30 Jun 2007 11:05:56 -0500, Tom Roberts
<tjroberts137@xxxxxxxxxxxxx> wrote:
the errorbars of Miller's original result [...] are HUGE,

In Fig. 5 in your paper they look about 20 times larger than the
variations of the data.

Yes.


Now according to Miller's paper, he processed a large amount of data
with a harmonic analyser. It seems to me this would have been
analogous to feeding a radio signal into a tuning circuit, so he could
well have successfully rejected the errors that concern you.

No. He did that after averaging, which completely negates the possibility you raise. The damage is done by the averaging. Had he applied such an harmonic analysis to the raw data, as I did with the 320-point DFT, it would have separated out the component with period 1/2 turn. But as I stated in the paper, by itself this is not sufficient to determine if that amplitude is signal or noise. For the radio and its tuning circuit, your EARS AND BRAIN separate the announcer's voice from the static, but there is no analogous method of separating signal from noise for Miller's data.


Miller estimated his errors [...]
I don't see any reason to distrust these estimates.

You need to improve your understanding of experimental science and learn how to "see": he did not account for the AVERAGING. The errorbar due to that is simple, indisputable, and is MUCH larger than the variation he found. We have learned something in the intervening 70+ years. You should learn it, too. <shrug>


You mention experimenter's bias, but Miller points out that owing to
the complicated relationships between orientation of the apparatus,
rotation of the earth, position of the earth in its orbit and
celestial direction etc., it would be impossible for an operator to
guess what fringe shifts to expect.

So operator bias is unlikely to be a factor in his results.

Miller wrote in 1933, before this had been understood. In the intervening 70+ years we have learned that even so this bias can affect results. That's why "single blind" and "double blind" experimental protocols are required in such cases today -- Miller's experiment is QUITE CLEARLY such a case.


There are no SIGNIFICANT correlations with Michelson and Morley,

Again the error bars you calculated might not be relevant.

You hide your head in the sand. <shrug>


Remember Trimmer had more air than solid, so he SHOULD
see the effect Cahill claims; he does not to much better
accuracy than any of the other experiments.

We know from the MM experiments that the air in gas-mode
interferometers can only produce visible fringe shifts if the path of
the light is very much longer than the path in the Trimmer experiment.

We do not "know" that at all -- this is merely yet another special plead to permit you to ignore an experiment with a result you don't like. Yes, his path is shorter, but his detector is more sensitive by a factor larger than the ratio of path lengths. To the rest of us that seems to establish that he did not see Cahill's "effect"; if you disagree, then you must present a QUANTITATIVE theory and perform a COMPUTATION, not merely spout guesses and special pleads.


So Trimmer's null
result is perfectly consistent with Cahill's claims.

You are GUESSING. Why is it that you don't provide a QUANTITATIVE theory for this "effect" and then apply it to Trimmer's apparatus (and to all the others)?


It is probably also unwise to trust single experiments, no matter how
good they look.

Hmmm. There are literally hundreds of experiments that confirm many different aspects of SR. You repeatedly mention a half dozen or so experiments in support of Cahill's claims, and ALL of them have clear problems with technique and/or analysis, or simply do not support his claims with any significance. <shrug>


Tom Roberts
.

Relevant Pages

  • Re: An Explanation of Dayton =?UTF-8?B?TWlsbGVy4oCZcyBBbm9tYWxvdQ==?= =?UTF-8?B?cyDigJxFdGhl
    ... degrees, or 1/6 the angle between his markers, so the variation from the average orientation is 1/12 the angle between markers. ... Miller wrong, any discussions as to the errors and features of his ... The full turn effect is also consistent with zero, because the entire variations of these data are consistent with zero. ... In averaging through summation, the ...
    (sci.physics.relativity)
  • Re: feedback please
    ... medical school. ... There are a lot of variations that could be used that ... would make it clear, such as: "The patient underwent CABG x 4 on July 1, ... probably still clear - hopefully someone would not need four separate ...
    (sci.med.transcription)
  • Re: Tadaima!
    ... sort of together and unseparated in time, ... time but separated in cyberspace. ... Though these two variations would have to be separate stories... ...
    (uk.media.animation.anime)
  • Re: Supressing page breaks at start of section
    ... I tried variations of this already and it didn't work. ... discover something. ... commands to read in each of my sections. ... that your sections were in separate files... ...
    (comp.text.tex)