my take on blind tests

The appearance of a product, its reputation and other factors can certainly influence how we perceive its performance. It has been well proven in science that it is very useful to evaluate all manner of things on a blind basis in order to eliminate or minimize subjective factors.

However, people in the audio world often have a very limited black or white view of blind testing as if there were only one kind. ABX is only one particular form of blind test. Double blind isn't the same thing as single blind. There is a continuum. How exacting one needs to be depends on what one is trying to prove or accomplish. (Think of it as a number. Sometimes "about 3" is good enough, other times you need 3.14159265 - or better - for a calculation.)

If I am buying something for my own use, I only need to compare until I am happy. That may not include an effort at blind comparison - or may not even include a comparison at all!

The catch is that many audiophiles like making declarations of universal scientific certainty from casual or poorly conducted blind tests.

The often heard "I had my friend swap wires and I could hear the difference so that PROVES there is a difference in cabling" is a good example. It is a perfectly valid test for justifying your purchase of a particular brand of something, but it is hardly something you base your next peer reviewed article on for a science magazine.

To be honest, well done, tightly controlled double blind tests in the audio world are very hard to set up, and as such rarely done. Almost every one I've ever seen has enough "wiggle room" in it to allow an out for any party that wishes to disagree with the results.

As I noted earlier, blind testing serves a useful function in audio as well as other sciences, but is hardly a mandatory requirement for an individual buying a piece of stereo equipment. People should buy what makes them happy, whether the underlying reasons are based on emotion, cold clinical analysis or some combination thereof.

That said, I think the people who completely pooh pooh blind testing on a wholesale basis are being short sighted. They are discarding an important tool that has been helpful in challenging old ways of thinking and developing new ones.

With any science it is always important to put existing beliefs under the magnifying glass. That's the way we can begin to separate the "old wives tales" from the traditions that have serious fundamentals underpinning them. Of course, any time a cherished belief is questioned, there will always be those who discount any results that conflict with their existing views.

The audio world is a tough case as it is at the intersection of electronics, art, psychology, the biology of hearing and chemistry of the brain. In general there are simply a zillion variables to track.

So, sure, you don't need a DBT setup the next time you component shop, but to belittle the usefulness of the practice at the levels of design and research is like throwing away your hammer and thinking all you need is a screwdriver and saw to build a house. You just end up cheating yourself.