greywolf42 wrote:
> [...] Tom applies this claim of "experimenter bias"
> against all experiments that he philosophically "disagrees" with.

That's another example of what I mean when I say you (and others around 
here) are unable to read what I write. I do not do as you claim. I apply 
the claim of experimenter's bias to experiments which are clearly 
subject to experimenter's bias, and to no others. It is not my fault that:
  1) many of the experiments greywolf42 wishes were true are
      subject to it.
  2) experimenters before at least 1960 or so were not aware of the
      pernicious and insidious effects of experimenter's bias, in spite
      of their earnest efforts to avoid any bias in their observations.
      Amateurs like Silvertooth and greywolf42 seem unaware of it still.
  3) technology before 1980 or so often required humans in the
      data acquisition, because computers and their interfaces were not
      available.

My "philosophy" has nothing to do with it; this is an objective aspect 
of experiments. greywolf42 seems to have no notion of objectivity and 
its relationship to science.

To be clear, an experiment is subject to experimenter's bias when:
  1) a human plays an essential role in data acqisition
  2) the human is aware of what constitutes a "signal"
  3) the human is also aware of the conditions of the experiment which
      affect the "signal" (e.g. orientation of the apparatus wrt the
      fixed stars)
  4) the human makes essential judgements that affect the data. In most
      cases (Esclangon, Miller, ...) this is rounding measurements that
      are then over-averaged to obtain a "signal" smaller than the
      round-offs; for Silvertooth this is in the search for regions of
      the table position meeting his preconceived idea of what "ought
      to be present", while he makes a non-exhaustive search.

For instance, the MMX does not suffer from this, because they did not 
over-average their observations (like Miller did); their reported 
error-bars are appropriate for their instrument. Miller's "signal" is 
smaller than his instrument's resolution, and is really a "measurement" 
of how his human observers rounded-off their observations. Ditto for 
Esclangon. Silvertooth is both doing a non-exhaustive search for 
preconceived patterns, and is also excluding observations for which his 
apparatus did not cooperate with his preconceived notions, so he has it 
TWICE.

> [...]
> Here we see Tom's litmus test. All "alternative" theories to SR must, in
> fact, BE SR. Any experiment that does not meet SR is not *viable.* 

Again you display your inability to actually read what I write. Whenever 
I have applied the word "viable" to theories, I have ALWAYS explained 
what I meant (as I have said probably hundreds of times around here, 
"viable" as I use it in this sense means "not already refuted by 
existing experiments").

I cannot help it if:
  a) SR remains a viable theory, within its domain of applicability,
      in spite of greywolf42's wish it were wrong.
  b) theories greywolf42 wishes were true are either not viable or
      are experimentally indistinguishable from SR
  c) the experimental record is diverse enough to refute essentially
      all theories in SR's domain that are not experimentally
      indistinguishable from SR.


The "essentially" in that last point is there to leave
open the possibility that someone will someday propose a
theory that "lives in the error bars" of existing
experiments, but is in principle distinguishable from SR.
This is VERY difficult, because of the tiny error-bars and
the diversity and scope of the experimental record.
To date nobody has presented such a theory, and Ensle, Seto,
McCarthy, greywolf42, and that crowd have made no attempt
to do so. They seem unaware of the necessity.... They also
seem almost completely unaware of the experimental record....


Tom Roberts tjroberts@lucent.com