04-28-12: GeoffkaitOn the other hand, positive results don't mean much either, unless:
There are obviously many reasons why someone might not get the expected results, including impaired hearing, lack of listening experience, faults somewhere in the system, failure to follow instructions, etc. Thus, negative results mean precious little, except to support claims by naysayers that the device under test disobeys all known laws of science, is fraudulent, or is simply a placebo.
A)The assessment was conducted in a sufficiently disciplined manner, and with sufficient understanding of the variables that might affect the results, such that there can be a reasonable degree of certainty that the result is being attributed to the correct variable.
For example, extraneous variables that could conceivably affect assessment of a fuse, especially one that is claimed by some to require 100 hours of breakin, would include ongoing aging, breakin, loss of breakin or rebreakin of system components; equipment being in a different state of warmup during the different parts of the comparison; differences in contact integrity resulting from removing and replacing the fuse, including scraping away of oxidation that may occur, as well as differences in contact pressure; changes in AC line voltage or noise conditions; changes in room temperature (temperature is a parameter that is fundamental to the physics of transistors and other semiconductor devices); and changes in RFI/EMI conditions.
B)There is sufficient understanding of the mechanisms by which the device works, if in fact it does work, to provide confidence that its effects are not just a quirk of its interaction with the particular system, that would not occur in many or most other systems.
For example, a finding that a fuse makes a difference with a Class AB or Class D amplifier, for which the AC current draw fluctuates dramatically as the volume of the music varies, IMO would say nothing about the likelihood that it would make a difference with a Class A amplifier, for which there is little fluctuation in AC current draw.
C)The assessment was conducted in a sufficiently disciplined manner to rule out the possibility of misperception, placebo effect, or self-reinforcing mass hallucination (a la the stock market, ca. 2000). As a minimum, that would mean going back and forth at least a couple of times between the devices being compared, to verify that the results are repeatable. In saying that, I am not necessarily referring to an immediate ABX-type back and forth comparison, since I recognize that not all differences will be perceivable in a short-term comparison.
04-28-12: BryoncunninghamAs with most things in life, IMO it comes down to judgment, hopefully judgment that is as informed as possible. Informed by a combination of technical understanding, experience, and inputs from others. While judgments will certainly differ considerably from person to person, it is all that we have to go on. In applying that judgment, we weigh what we consider to be the likelihood of significant benefit against the time, expense, and potential risks and downsides (see the link Clio provided) that are involved in pursuing it.
What is the standard by which Possible Unknown Parameters are distinguished from Impossible Unknown Parameters?
FWIW, my own "a priori," subjective, and certainly fallible judgment is that the possibility of a fuse making a sonic difference falls within the realm of a Possible Unknown Parameter, while its directionality does not. In saying that, I certainly do not exclude the possibility that cables (as opposed to fuses) may have directional sensitivity, the two situations being distinguished by the vastly different lengths that are involved, by the less direct relation between a fuse and the signal path, and by the fact that many cables are asymmetrical by design (with the outer shield being grounded at only one end).
Best regards,
-- Al