Lots of small-scaled, amateurish and negative trials don't add up to prove anything conclusive, only suggest a trend. For professionally conducted studies that are negative or a mixture of positives and negatives or inconclusive there are methods (meta-analysis) to evaluate them as a larger group to look for hidden positives.
Sample size IS important in statistics. Small differences that are statistically non-significant might become significant in a larger sample size, because the sample size is factored in determining the p value. You only have to read some clinical trials to see what sort of numbers (hundreds and thousands) are involved and the differences are often very small, not 80% vs. 20%, this large magnitude of difference doesn't happen. Usually when the sample size is down to 10 or 20 very very few studies have statistically significant results. Also note that 20 people each doing 10 trials isn't the same as 200 doing 1 trial in statistics.
As I said before some of you just use small and amateurish trials as concrete proof and this is not how proper science works in real. I'm not against DBTs but just the way you over-interpret their true meaning and significance. If you think some of these audio trials are well run you might also want to read some of the large scale clinical trials or studies (some are freely available) to get some idea of what's it all about at that level.
I think the big key is that, if there were noticeable differences between amps the companies making those better amps would show the hard data proving it. They would eagerly go into double blind tests and come out on top, repeatedly. I know I would be throwing my amp around in double blind tests if I ran a company that created one which was noticeably better - making sure ALL the professional magazines knew all about it. The dearth of these tests speaks volumes.
EDIT: I realize that in science the absence of evidence is not evidence of absence - but in capitalism it is.
Edited by cybrsage - 2/9/13 at 7:59pm