So far science has only conducted tests devised in such a way to result in null results. What about the other side?
Oh man...how do explain it any clearer? Discussions with someone who has clearly a problem with grasping certain concepts is...
The Null hypothesis means one thing only: some claim is made for which so far evidence is anecdotal. I.e. that differences are audible between amps that measure within a percentage of the tested for parameters.
Now the Null hypothesis is that such differences are inaudible, because the human hearing is not able to differentiate at such a level.
A test methodology is established how to test the null hypothesis - it is about testing, about either confirmation of the Null hypothesis or its falsification, not TO PROVE ANYTHING.
After the tests the data is analyzed statistically within a p-value - a measure of confidence that the results could not have been achieved by chance alone.
At the end the conclusion says either of two things: We could not verify the claims made, the hypothesis is not falsified...or: We found EVIDENCE that the hypothesis has been falsified, and further investigation is necessary.
That is how science works (I had been a Lab tech for over eight years) and unfortunately what the typical press release tells you is usually NOT science but the spin a reporter puts on statements to enhance the ratings of his rag or the channel.
All conclusions come also with an error margin
- there is only confidence within this margin. Science does not provide irrefutable proof, science is contingent on better testing procedures, new observations etc.
This is science - everything else is religion.
Because scientists seem to be under the assumption their view is the only correct one.
No, not their view is correct, but the scientific approach is the ONLY correct one. Because science is not about a vault of knowledge, science is about to find out what makes the natural world tick, it is a process of investigating and looking for evidence.
Only a skeptical approach that avoids presumptions influencing
the final result (a hypothesis like the Null hypothesis is a presumption under test, but it is a negative assumption, therefore enhancing the necessary skepticism towards any claims and anecdotal observations) is likely to produce results that can be considered evidence - within the error margin.Edited by kraut - 3/11/13 at 12:58pm