[PhaedraCorruption]

Just wanted to share a project a was working on:

For my Psychology class in school, I was required to come up with a experiment to perform on a group of people in order to present results to the class, our group decided to test the effects perceived value plays on hearing. 
 
(Note: For the record, I am a believer of cables and this test was conducted completely unobjectivly, the only intention was to expose the Psychological aspect)
 
 
The Write-Up:
 
 
 
 

[size=medium]For the past few weeks we have been learning about the brain and it's amazing properties, we decided it would be a perfect time to test the brains capability to alter reality.  Our minds after all are very powerful in the ability to deceive us.   We decided to perform an experiment that we have devised to determine if the whether cost of something really scales perfectly with the performance, or does the mind play a greater role in determining this.   If the latter is true, then, our mind's perceived value of equipment will increase the fidelity of what we hear when we listen to audio.  This would reinforce the concept that our minds are powerful enough to change even the way we hear based on factors other than reality. In order for us to test this experiment we gathered a group of volunteers with the promise of an experience they would not soon forget.  Our procedure is as follows: We setup a test platform consisting of an audio set up with changeable variables the first setup used cheap and easily obtainable stock cables, in the second set up we swapped these out for far more expensive cables.  The aim of our experiment was to test whether or not the perceived value of the setup changed the way the listener heard the sound.[/size]
 
[size=medium]How:   After that we blindfold the participants and repeat the listening experience but switching the cables randomly each time.  After the test subject has listened to music 4 times, we ask the order in which the cables were used.  After they explain to us what they think they have heard in order we tell them what the actual order is.  The target of our experiment was people who listened to music.  To collect the data from these individuals we had invited them over to a house with our equipment set up and allowed them to be comfortable and sat them down.  From there we had let them start our experiment.  [/size]
 
[size=medium]Our experiment was conducted by letting participants listen to both the cheap and expensive audio and then letting them make an analyses on what the sound quality is with both. [/size][size=medium]After setting up the test station and gathering our test subjects, we began the test by first allowing the subjects to hear the cheap and the expensive cable separately. Each time we specifically told them which cable they were listening to. After a few tries, most subjects claimed that they were able to differentiate between the two. When the subjects were satisfied with what they heard, we blindfolded them and subjected them four random tests. Each one of the four consisted of either the expensive or the cheap set of cables. We did this randomly without sequence. After this, they [/size][size=medium]explained to us what they thought they had heard in chronological order and then we told them what the actual order is.  [/size][size=medium]All subjects we tested were unable to correctly identify between the cables despite having proclaimed a huge difference between the cables before the cable was put on. [/size][size=medium]We did this for 12 participants including ourselves, we felt that we too were also viable candidates for the test, as bias was nearly impossible in this kind of test.  This was easily proven by the fact that even though we tested each other before conducting the experiment, our results matched those of our test subjects. [/size]
 
[size=medium] Results analysis: The participants could tell a difference in sound quality when the participants knew which audio cable we were using.  This had changed dramatically when we did not tell them what we were using and were blindfolded; none of the participants were able to tell the difference between the two cables [/size][size=medium]consecutively[/size][size=medium].  These results tell us that the price of equipment affects our judgment due to the fact that our minds are powerful and that it can skew our judgment just by thinking that something is better than the other.  This would be due to the fact that our minds want to have a relation between both price and quality.  This might have to do with the fact that we do not want to be scammed into paying for something that is actually cheap.  In order for our minds to do this they will correct whatever we are missing or what we find is wrong.  A reason for this mental "safety net" could be to make us feel less regret when something bought that is expensive [/size][size=medium]underwhelms[/size][size=medium] us.  As these participants did not buy the equipment they might have subconsciously thought about how price will be matched linearly to quality thus altering what they think.[/size]
 
[size=medium]Due to the fact that before being blindfolded participants had stated they could tell a difference in sound quality between the two cables, most of which stated that the more expensive cable sounded better, we are able to conclude that our minds influence the way we hear things depending on the perceived value of what we listen to.  Therefore we can conclude that our hypothesis was correct and that perceived value really does change what we think is a better sound quality.  Some of the variables we may have missed though would be that our participants may have been subject to some type of hearing loss.  To prevent this, we would need to conduct this experiment on a larger scale then what we have previously tested.  Some other things that we can do is redo this test with a larger variety of cables.  This meaning that we could test new equipment to see if it really will make audio a better quality by using this same procedure.  Another thing that we could do is use the data to make things sound better to customers or the person receiving it by telling them that the equipment they are using is expensive.  This would revolutionize the way we see technology by learning that people will actively fix what our minds feel is missing.  This would mean that  doctors could prescribe "expensive" medication to amplify the effects of the medication being prescribed, a sort of "super" placebo effect. Our minds actively alters what the brain receives due to changes in expectations.  Despite it being only a minuscule or non-existent change it is amplified or created into something we perceive as a huge difference. Cognitively it can be said that does this due to expectation that have been set upon us.  We want to hear a difference because we expect a difference as a result of the price that we paid, therefore it can be argued that at times humans can easily be manipulated due to our expectation that value will scale linearly with performance. [/size]

 
 
Powerpoint presentation explaining more: https://docs.google.com/present/edit?id=0AYFxmkkTanFuZGNzenRnd3dfMGNkd2N0NmM2
 
 
 
 
 
I'm still very much a novice in this field, I expect there to be many errors feel free to give input.
 
Cheers

 

[Uncle Erik]

Thanks for sharing your results.

If you haven't already seen them, you'll find quite a few experiments with wine that corroborate your results.

Which is fine with me - I'm much more of a beer drinker.

 

[Prog Rock Man]

Thanks from me as well.
 
I have been reading up a bit and found these other links to how preception and our senses are related (an easy idiots guide from Wikipedia)
 
http://en.wikipedia.org/wiki/Perception
 
Here is Sennhesier's and Heriot Watt Universities write up of taste and sound and how they relate
 
http://www.sennheiser-annualreport.com/home/en/the_palate_has_ears.html
 
http://www.wineanorak.com/musicandwine.pdf
 
and the very interesting McGurk Effect and the relationship between hearing and vision
 
http://en.wikipedia.org/wiki/McGurk_effect.
 
Your tests further corroborates what us non believers have been saying for some time now, that the consistently different results you get when sighted and blind testing show that any preceived differences are in the mind and not the cable.

 

[TheAttorney]

Well written summary, you with the long name. I'd loved to have done such an experiment when I was at school. All we ever seemed to do was read books and listen to teacher.
 
The results tie in with other studies. 
 
As a cable believer, I have to say that I do agree with this in general: it is extremely easy to get fooled by expectation bias.  
 
Where I probably differ from most sound scientists here is that I also think it is extremely easy to get fooled by blind tests.
And that on balance, it is easier for an open minded person to overcome expectation bias, than it is to overcome the mind tricks of blind tests.
 
The wine analogies can be taken to back up either side, depending on your perspective.
Although people have pointed to evidence that the blind test issues can be overcome, I've never come across any evidence that answers my concerns.   
 
So when anyone confidently states that Theory X has been proven by a blind test, I say that they've only done half the analysis.
 
This only applies to what I call neutral sounding components that measure much the same. Obviously, a blind test will accurately spot significant differences in frequency response or distortion, as you can get with transducers, but what's the value of that?
 
PS. I should know better than to post in the Sound Science forum, but what the hell...

 

[Lorspeaker]

I love the opium of branded cables...lol.
Screw my mind by all means, I go in with my eyes open..and pockets opened too,
And I am a happy sucker with my cables.

That said, it took me a long while to learn about soundstage, width depth, height, punch, transient..Separation..etc etc
(good cables are supposed to *improve* those aspects...tats what the cablesellers say..hehe)
This experiment is best carried out on hifi enthusiasts...
An experiment like this, on my wife would churn out a 1000 different results,
Any sound louder than her tv program is..N..O..I..S..E.

 

[Prog Rock Man]

Quote:

Well written summary, you with the long name. I'd loved to have done such an experiment when I was at school. All we ever seemed to do was read books and listen to teacher.
 
The results tie in with other studies. 
 
As a cable believer, I have to say that I do agree with this in general: it is extremely easy to get fooled by expectation bias.  
 
Where I probably differ from most sound scientists here is that I also think it is extremely easy to get fooled by blind tests.
And that on balance, it is easier for an open minded person to overcome expectation bias, than it is to overcome the mind tricks of blind tests.
 
The wine analogies can be taken to back up either side, depending on your perspective.
Although people have pointed to evidence that the blind test issues can be overcome, I've never come across any evidence that answers my concerns.   
 
So when anyone confidently states that Theory X has been proven by a blind test, I say that they've only done half the analysis.
 
This only applies to what I call neutral sounding components that measure much the same. Obviously, a blind test will accurately spot significant differences in frequency response or distortion, as you can get with transducers, but what's the value of that?
 
PS. I should know better than to post in the Sound Science forum, but what the hell...

I think that we are reaching an agreement here, it is in the semnatics that we differ over. I think that cables sound different because of the mind rather than the cable itself. I do not think that is down to trickery, it is a very real and testable interaction between the senses.
 
I agree that blind testing is only half (if that) of the analysis. But this test backs up the results of these blind comparison and ABX tests
 
http://www.head-fi.org/t/486598/testing-audiophile-claims-and-myths
 
Combine that with the clear results of sighted testing where people can really hear differences and you get
 
Sighted tests - genuine audible differences, but not with all people and all equipment and there is some consistency with the listener and how the cable is made, its image and cost.
 
Blind comparison tests - genuine audible differences, the differences are now smaller and there is no consistency at all between the cable, its cost, image or how it is constructed.
 
ABX blind tests - all differences have disappeared. No one can reliably tell any cable apart from any other.
 
Put that lot together and you have a meta study of how our sense of hearing is affected by other senses and perceptions in a very real, testable, repeatable way. That people can really hear a difference in one situation and cannot in another is not trickery, it is not a fault in the tests, it is a discovery of how our sense of hearing interacts.
 

 

[JRG1990]

Its not just cables that are affected by tricks of the mind, dacs, amps , headphones , speakers nothing in your audio chain is immune.

 

[Prog Rock Man]

True, but since there are positive blind tests for amps and speakers, some are more different than others, so the influences that affect hearing are also backed up by a very real difference.
 
I also heartily disagree with calling this a trick of the mind. It is an affect, a phenomena. I am quite sure that when people hear a difference in sighted testing, they really are hearing one. I have heard it so often, most recently with DACs and a back to back test between a Firestone Audio and Musical Fidleity DAC. They sounded very different, with one sounding much better than the other.

 

[PhaedraCorruption]

IMO, and this is just my opinion, but I think cables do have an effect on a sound system, BUT, the effect is very small, nearly indistinguishable to the average listener, like 1-2% difference. 

However, if you make a tweak here, replace a cable there, use a anti-vibration stand here, and user a power cleaner over there, in the end(with a huge friggen hole in your wallet I might add), there will be a small, but substantial 10-20%(obviously this is subject to change depending on how many tens of thousands you spend) difference. And this only really applies to very very high end stuff.
 
Anyways, just my opinion. 
 
 
Thanks for the positive comments guys! 

 

[Vkamicht]

Quote:

IMO, and this is just my opinion, but I think cables do have an effect on a sound system, BUT, the effect is very small, nearly indistinguishable to the average listener, like 1-2% difference. 

However, if you make a tweak here, replace a cable there, use a anti-vibration stand here, and user a power cleaner over there, in the end(with a huge friggen hole in your wallet I might add), there will be a small, but substantial 10-20%(obviously this is subject to change depending on how many tens of thousands you spend) difference. And this only really applies to very very high end stuff.

Right, and after all that money and time spent trying to get a lowly 10% difference (if that) ... spend a fraction of the money on a parametric EQ and make your headphones sound multitudes better than they ever could otherwise

 

[Prog Rock Man]

Quote:

IMO, and this is just my opinion, but I think cables do have an effect on a sound system, BUT, the effect is very small, nearly indistinguishable to the average listener, like 1-2% difference. 

However, if you make a tweak here, replace a cable there, use a anti-vibration stand here, and user a power cleaner over there, in the end(with a huge friggen hole in your wallet I might add), there will be a small, but substantial 10-20%(obviously this is subject to change depending on how many tens of thousands you spend) difference. And this only really applies to very very high end stuff.
 
Anyways, just my opinion. 
 
 
Thanks for the positive comments guys! 

What you are doing there is keeping a gap open for those who think they have golden ears and the very best of kit to some how excuse themselves from the real science as presented by many, including you. Your test and the many other blind tests, which have been on some very high end kit with very high end cables with very experienced audiophiles have the same results. ABX and they canot tell a difference with cables.
 
 

 

[Argo Duck]

Some good responses here...
 
I have no view on cable differences per se. And the following is just my quick take, as there's a lot of science that feeds into this humble test (and I'm actually in the middle of a large writing project ATM, so this response has been dashed off with no particular thought and is not likely to be particularly coherent).
 
This was not a scientific study, as Phaedra shows he is well aware. We cannot treat it as such, and cannot consider it an entry in any meta-analysis. However, it was a nice piece of work for a school-level project - well done.
 
In any study, one has to estimate the effect-sizes of the various factors that influence the dependent variable or measure (DV). One then tries (by experimental design) to control for these factors. From my very quick once-through (sorry if I've missed or misread anything!), it seems to me expectation bias (EB) is one of these factors; and cable difference is a (possible) second factor. The literature shows EB is a strong effect in the right conditions. The literature is undecided (or, there is not much available - I don't know) concerning cable differences, and I know of no estimates of effect-size. However, I presume it to be either very small - like Phaedra - or nil.
 
Thus stated, the problem is we have a strong factor but we are interested in isolating a second factor of unknown but probably small magnitude.
 
It seems to me this did indeed test the first factor, not the second. This happened because prices were mentioned. As "price" is often an analog of "quality" in our culture, this sets the EB. This was reinforced here because the cables were introduced sighted (ST) - allowing EB to be introduced and a conformity/consensus effect - and then blind testing (BT). What should have been done was to have two conditions with equal numbers of participants: an ST then BT group and a BT then ST group. Had an "order effect" emerged between these two groups., i.e. listeners established a clear preference for one or other cable blind, which then was?/was not? overturned on learning of cost, this would have been interesting in itself. It would also have given greater scope to strengthen/weaken the finding as it is, thus producing a "more powerful" study. 
 
Although I realize this was an explicit "does cost influence perception?" project, an improvement for an actual cable-differences experiment would be to eliminate mention of price. In this case it would have weakened EB and allowed the hypothesized "cable difference" to have more experimental effect, presuming there to be any. Tajfel's minimal group paradigm employed labels such as the names of the artists "Klee" and "Klein". Better, in this case would be some random groups of letters and numbers, to eliminate associations of particular names with images and labels extant in our culture (here the semiotics literature is relevant).
 
A problem for the presumed cable factor is that cost and performance might not in fact be related. One cannot assume there is a positive correlation of the two. A truly interesting study would first seek two or more cables which listeners agree sound different (the basis of this agreement is unimportant at this point - it might not be evidence of objective differences., e.g. it might be a normative/group-labelling effect - consider Sherif, 1936) and then introduce cost into the experimental manipulation. We could then test "does awareness of cost alter perceived sound-quality differences of a set of cables?".
 
As Phaedra notes, the magnitude of the cable effect (again, presuming there to be one), could be small, in the order of 1-2%. This is not leaving a gap open, as effects of this magnitude are routinely measured (given sufficient study power and sample size) and allowed for in scientific models. They are not always meaningless, as small effects are sometimes decisive in predicting different outcomes. It is entirely a matter of the interactions within the model, and especially so when it includes non-linearities.

 

[Prog Rock Man]

It does not appear to matter exactly how or where an ABX or blind comparison or sighted test is conducted (assuming no faults in the ABX or blind test) nor who takes part, the results are consistent.

 

[PhaedraCorruption]

Quote:

It does not appear to matter exactly how or where an ABX or blind comparison or sighted test is conducted (assuming no faults in the ABX or blind test) nor who takes part, the results are consistent.

Yup, from what I have read and also experienced myself, I can confirm this. 
 
If the ABX test does not function as intended, I can't really think of any kind of test that can really test something such as the subtle sound variations given by audio cables.... I guess this explains why this debate has been ongoing for 50+ years in the Hi-Fi community. 
 
(Although, it has already been proven that cables produce a measurable difference on paper)

 

[Prog Rock Man]

The Null test will also show if there is an audible difference. But like blind testing, cable companies and cable believers are very reluctant to do the tests.
 
That there are measurable differences between cables is what cable companies rely to to suggest there is an audible difference. But none so far have show a link between the measurable difference and differences in sound quality. And from what I can find they have had 40 years of research to come up with a link.