Quote:
Originally Posted by RebelMan
In science the burden of proof lies with the entity making the claim, not with the entity that contradicts it. If the claim cannot be proven then all you have is theory.
|
Actually, that's not how hypothesis testing for making statistically valid inferences with respect to data analysis for scientific or engineering experiments works AT ALL.
Along these line, just curious if you've ever published a scientific or engineering paper in a peer-reviewed journal?...(I have, FWIW)
In fact, here how it
actually works: the default hypothesis, referred to statistically as "the
Null Hypothesis", Ho, is that there is "no difference" for the thing, system, whatever, being tested, measured or evaluated, etc. The Null Hypothesis the foundation for all statistical hypothesis tests. The other test case, that there "is a difference" referred to as the
Alternate Hypothesis, Ha.
And, you're making just this claim: that it (power conditioning, power cables, etc) doesn't matter. That's the Null Hypothesis, and its a claim
just as much as the Alternate Hypothesis is. Okay, well, I think we'd all welcome objective measurement data that
fails to reject the null hypothesis with alpha at 0.05 to provide a 95% statistical confidence.
Whenever I see these objectivist vs subjectivist "debates", usually involving cables, the Objectivists always say something along these lines: "Prove with data that power cables matter..."
I counter that with what should
actually be asked for a proper hypothesis test: Provide objective measurement data with the appropriate statistical analysis that
fails to reject the null hypothesis.
All I ever hear in response from the "objectivists" is...
crickets.