Well, that wasn't really what I meant. But I don't think anyone would argue with that; that all these 'measurements' do, one way or the other, influence the way we perceive the sound that the equipment produces.
jan didden
Last edited:
There we have a prime example of what I was referring to in my earlier post (#9).
How do you know they are 'the nothing' ?
Exactly correct! And this always seems to be discounted & explained away as expectation bias & therefore worthless. When a large body of people still report their audio purchase as worthwhile after a long period of it's use, that this accounts for more than "the nothing of subjective anecdotes"
Here's another side to this expectation bias - what if you housed your audio device in a less than stellar looking box is there now a negative expectation bias?
Last edited:
Unfortunately in the early days of my now finished commercial endeavor I found that people's perception of the actual sound quality of a device correlates quite strongly to the appearance of the device making it. The nicer and more professional said device appeared, the better the perception of the sound quality. (At least if there were no egregious deficits in the performance of the unit in question.)
The best way to prevent this bias even in uncontrolled listening was to prevent the listeners from seeing the actual devices being compared whether or not the rest of the testing was blind in any sense.
Sadly this does lend merit to the need for blind testing in many circumstances as this tends to remove an unexpected personal bias from the equation.
My experience in fact was so negative that at one point I actually took two prototypes one nicely packaged and the other not that were otherwise absolutely as identical as I could make them and did a non-blind A/B test, the nice looking unit was described with glowing superlatives by all who listened, and the kludge was generally ranked as being significantly inferior again almost all of those who listened. (One stated categorically that there was no difference, and of course he was right.)
I then threw a wrench in the works and re-ran the same test only this time telling them that the nice looking unit was playing when in fact it was the kludge, and vice versa. Predictably the results again followed what was perceived to be the better looking of the two units.
I will note that these were non-technical audiophiles and random individuals from the apartment complex I then lived in, and that a few people actually were not fooled - all either very experienced audio engineers or hands on hobbyist types.
Listening experience I believe does play a role, but clearly there are surprising and unexpected biases that creep into comparative listening tests.
These days I rely almost entirely on a combination of measurements, with subjective assessments as I go, but my final assessment is always subjective, if I don't like the end result that is pretty much it, and fortunately I have only myself to satisfy - I am however notoriously fussy and relatively meticulous about what I do and how I do it. I'm no longer in a situation where I need to do tightly controlled experiments, but those trying to demonstrate significant changes or are in the process of vetting a product really do need to consider these and other bias issues in listening tests.
Yes, Kevinkr, that is my understanding of expectation bias from sighted auditions so my point is what if the auditioners in a sighted test prefer the inferior looking box? Have we now more than discounted the visual expectation bias? Can we now fairly deduce that there must be a real sonic advantage to said device as the sighted bias would be against it?
Last edited:
That would be the best outcome, but why risk it? I have had devices that were clearly preferred in listening tests where they remained unseen and under the same exact conditions where they could be seen the results were significantly different.
It would not be at all hard to devise sighted tests where there were, say two electrical systems to test and two enclosures thus giving four combinations. If the preference for one of the electrical systems over the other exceeded the preference for one enclosure over another that could be quite a significant result.
Just like a pretty woman wins regardless of her personality.
They like us in a suit too.
Pretty woman
I love that word
You're a naughty boy.
I don't think they are surprising or unexpected.
Thank goodness, hopefully you can relax now, enjoy the fruits of your labour and let the others do that sort of thing.
Hence the well positioned lights and white teeth sales persons. The vetting part is somewhat lost. Even the most hardened are not immune.
Kevin you're not only a good member you're a great moderator. I hope the members appreciate that.
Because, I've already done this & it has worked in favour of the lesser looking device. My point is that the using the bias expectation in sighted tests in this way can be a way of proving something is genuinely better sounding!
Last edited:
Like, today driving on freeway I read, "Expect terrific delays".
When I saw another display I "re-run the reading", and it was this time, "Expect traffic delays".
It happens...
During BAF 2008 many of us were surprised when tiny simple speakers in tall poles (BIB enclosures) sounded much better than all that complex and scientifically advertised exemplars.
Last edited:
That's my point, instead of sighted auditions always being given a blanket dismissal - there would seem to be results that can't be discounted on the basis of this bias!
Its not about them being "dimissed" its about understanding and controlling for that variable. The results of ANY test prove nothing until you understand all of the circumstances of the test.
Problem after that is that often the results of the test mean nothing because so many variables exist that have not been controlled for. Or even considered.
BTW, just because the "bias" is in favour of the unexpected doesn't make it any less a bias - it still needs to be understood and acounted for.
Do a search for the phrase "without peeking" on this forum as a phrase used to "dismiss" anecdotal evidence.
Of course & that is exactly what Thorsten is saying about calls for audio DBTs which on examination are no better than anecdotal evidence & are in fact a pretense at scientific analysis - pseudo-science yet again.
Exactly what the criticism of DBTs for audio is about
Never said it wasn't still a bias - so how would you now suggest you factor it into the results that favour the visually "less appealing" device?
The answer is simple: people who were trained to hear different kinds of distortions preferred the lest appealing one based on it's sound, despite of very humble appearance.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.