Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Medic. study—placebo tends to lack important effect—relevant to audio? (Read 48875 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #50 –
Well. Expectation bias also hasn't a clear-cut universal definition, has it? E.g., http://en.wikipedia.org/wiki/Expectation_bias points the finger at the researcher. I guess what is commonly referred to as expectation bias in this forum, is this effect: http://en.wikipedia.org/wiki/Subject-expectancy_effect

This is a forum for scientific discussion, so one shouldn't discard a discussion of the taxonomy of biases, but one should also beware that if those terms are not necessarily the same as used in everyday life, then the inferences and interpretations of the studies should be adjusted accordingly. Just because a study doesn't find any “placebo”, it doesn't mean it isn't there – it could very well be a proven (significantly nonzero) bias, but not sure to be within that particular definition of “placebo”. In this case, the definition requires not only that patients report the sham medicine to work because they think it does, but because sham medicine actually did improve compared to a baseline of, say, no treatment.

How to translate that latter effect to audio quality (as opposed to the effect of pleasure from music, where it should be obvious to think of an exampe)? It would be something like an artifact training scheme that where the bias from the expensive price tag on the equipment actually has the effect of improving your learning. Not that you start reporting differences that aren't, but you become truly better at discovering true differences. Well. If nobody can find that effect to be significant, it still doesn't mean that those sighted reports of .wav superior to .flac are unbiased, does it?

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #51 –
Quote from: krabapple on
I'd be willing to be you have experienced expectation bias in audio at some point in your life.  You surely haven't tested every instance where you thought something sounded different/better, have you?

As children most of us learn that imagining what it would be like to get a gift can be more enjoyable than actulally getting it. As adults we sometime reliaze that reliving the past makes things seem better than they really were. That's one of the ways human beings deal with the past and present, but their actual experiences of the events remains untouched. If I thought that I heard a difference that I later failed to substantiate in a test, I'd rather presume that my memory is unreliable, instead of my hearing. This is not so for the McGurk effect and other illusions - there you can be certain, after blind testing, that you really heard what you thought you heard.

Quote from: krabapple on
The part you keep missing is that your preference is not necessarily due to sound...even if you believe it is only due to the sound.

But I'd be perfectly fine with that! If the difference is not due to sound, then the analogy with the medical findings would follow perfectly. Did you read that part of the paper in which they raise the possibility that the kind of service-with-a-smile health care could benefit patients, not because of any direct effect but rather because it could motivate patients to take their medication regularly, follow other medical instructions and change to a more health conscious lifestyle?

These indirect benefits would be like listening to music in a leather sofa instead of wooden chair - it wouldn't significantly alter the way the music sounded, but it would improve the experience. If the listener later reported that leather sittings make music sound better that would just be a misreport.

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #52 –
Quote from: gnusmas997 on
If I thought that I heard a difference that I later failed to substantiate in a test, I'd rather presume that my memory is unreliable, instead of my hearing.


If there was evidence that your hearing was unreliable, would you still rather presume this?

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #53 –
Quote from: drewfx on
If there was evidence that your hearing was unreliable, would you still rather presume this?

No. But the question was related to my personal experience.

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #54 –
Here's a discussion related to "auditory memory" (or whatever you would like to call it, I don't care; check your nitpicking at the door), in it you can find links to other similar discussions:
http://www.hydrogenaudio.org/forums/index....showtopic=71595

There are many many discussions about this type of thing that are tangential to this topic.  If you haven't searched the forum and read through a few heated debates you're missing out on a lot.

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #55 –
Quote from: gnusmas997 on
These indirect benefits would be like listening to music in a leather sofa instead of wooden chair - it wouldn't significantly alter the way the music sounded, but it would improve the experience. If the listener later reported that leather sittings make music sound better that would just be a misreport.



 

Try telling an audiophile that his claims about the *sound* of his new cable/power conditioner/DAC sound are probably just *misreports*, and see what happens.

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #56 –
Quote from: gnusmas997 on
Quote from: krabapple on
I'd be willing to be you have experienced expectation bias in audio at some point in your life.  You surely haven't tested every instance where you thought something sounded different/better, have you?

As children most of us learn that imagining what it would be like to get a gift can be more enjoyable than actulally getting it. As adults we sometime reliaze that reliving the past makes things seem better than they really were. That's one of the ways human beings deal with the past and present, but their actual experiences of the events remains untouched. If I thought that I heard a difference that I later failed to substantiate in a test, I'd rather presume that my memory is unreliable, instead of my hearing. This is not so for the McGurk effect and other illusions - there you can be certain, after blind testing, that you really heard what you thought you heard.


Both hearing and memory are far from perfectly reliable.  Yours included.  This is why no *experts* (and I seem to *remember* you saying you defer to experts on things) on hearing or memory would rely on your report on its own.

It need not be a distant memory.  Even as you are taking an ABX text, it is quite possible you will report 'hearing' differences that aren't real.  The memory lag there can be  fractions of a second.

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #57 –
Quote from: Porcus on
Well. Expectation bias also hasn't a clear-cut universal definition, has it? E.g., http://en.wikipedia.org/wiki/Expectation_bias points the finger at the researcher. I guess what is commonly referred to as expectation bias in this forum, is this effect: http://en.wikipedia.org/wiki/Subject-expectancy_effect



When a listener on his own is comparing things, *he* is the 'researcher' or 'observer'.


And expectation is just one bias. Take your pick of these others:


http://en.wikipedia.org/wiki/List_of_cognitive_biases

or check out the list in the TOC for Chapter 4 of this textbook  (google lets you do it)

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #58 –
Quote from: krabapple on
This is why no *experts* (and I seem to *remember* you saying you defer to experts on things) on hearing or memory would rely on your report on its own.

Deferring to the expertise of a community is not the same thing as deferring to an expert. Come to think of it, I'd be far more comfortable if a few links to academic articles providing evidence for the phenomenon under discussion had been dropped earlier in the thread. Or at least some easy to find bibliographical references.

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #59 –
Quote from: gnusmas997 on
These tests allow us to conclude that there’s no difference between the way wave and flac sounds, but we still need to explain why the audiophile reported a difference.


No, we don't need to explain this, unless we want to examine why the audiophile determines his or her PREFERENCE.

Preference is inviolate.  There is no arguing preference, and one can use what one PREFERS to use to establish PREFERENCE.

Yes, it is possible to examine why preference exists in some cases, via complex research.

But, if an individual prefers how hir stereo sounds while using red speaker wire, so be it.

The problem comes about when s/he argues that hir preference is MORE than preference.
-----
J. D. (jj) Johnston

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #60 –
Quote from: db1989 on
Quote from: gnusmas997 on
if wave really sounds better to the audiophile because of his expectations, then he is fully justified in using it over flac. In fact, he is fully justified in preaching this usage, in the hopes that other people will also share his expectations and thus hear wave differently from the rest of us.
Personally, I must disagree. If there are any valid candidates for valid measures of morality that could elevate it above pure subjectivism, truth has to be at or near the top of the list. Surely, as a scientifically/objectively based community, we can all apportion some degree of plausibility to that. Saying people can believe whatever they want is one thing: thinking it’s good for them to be able to coax others into others believing the same unverifiable claims and the untruths that generally underlie them is another. But I’m by no means a philosopher, as is probably evident, so I’ll stop here. :|



One is free to use one's preference, but I would submit that it is unethical to try to perpetuate a preference that is not solidly based on some kind of evidence.
-----
J. D. (jj) Johnston

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #61 –
I totally agree. By the same token, if you hadn’t posted that second post, I’d have posted this after your first post.
Quote from: Woodinville on
The problem comes about when s/he argues that hir preference is MORE than preference.
I would separate subjective preferences (favourite band, colour, and so on) from the assertion of differences that are contrary to all rational expectations. The former are inherently untestable. The latter are eminently subject to testability. Where I, and I think most people here, have a problem is where people refuse to prove the things they claim so emphatically.

Alternatively stated, people can have preferences, and whether subjective preferences can ever be evaluated is a topic for another kind of philosophy, but if those preferences are testable and seem specifically orthogonal to reality (as opposed to purely aesthetic, etc.), no one is under any obligation to respect them until evidence is provided.

By the same logic, and going back to the post by gnusmas997 quoted by Woodinville two posts up, I personally don’t care why people report imaginary phenomena. Maybe others are enjoying this round of semantics and nit-picking, and it’s perfectly possible that I’m just lacking in the relevant philosophical skills needed to appreciate it fully…but I would say that an unverifiable claim is an unverifiable claim. Does the person really hear the difference? Try to convince them to test it properly, but good luck separating some of them from their sentimental need to feel special. Are they lying for cynical reasons? Welcome to the world.

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #62 –
Quote from: db1989 on
Alternatively stated, people can have preferences, and whether subjective preferences can ever be evaluated is a topic for another kind of philosophy, but if those preferences are testable and seem specifically orthogonal to reality (as opposed to purely aesthetic, etc.), no one is under any obligation to respect them until evidence is provided.

All the more, most of the times in audiophile field the path to create a belief starts, dangerously I'd say, from a pseudo-scientific statement and leads to an intuitive, even if wrong, application to sound quality. So that it's not a real preference, which could be perfectly legitimate at a subjective level if formed AFTER a sensorial experience, but a sort of aprioristic knowledge which prevents the listener to form a preference, because he already know "rationally" what to like and what not to!

If you state that analog is better than digital because "the world doesn't go step by step but in a temporal continuum", if you state that 24 bits are better than 16 and show a stairstepped curve side by side to a smoother one, you create an apparent but intuitive rational expectation: any surprise if nine people out of ten, when they know what they're listening, will find better the one that their brains already know it is better?

All this has nothing to do with freedom of choice, which is the base of preference.

Proof is that the vast majority of people who hasn't been instructed before, as to say not audiophilities-aware listeners, and has had the freedom of choice, has chosen CD upon LP over the years and then lossy upon lossless for mass distribution (both legal and illegal ).
... I live by long distance.

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #63 –
That’s a very good way of describing it. I couldn’t find a way to translate what I was thinking into words very well, so you’ve helped to clarify that bit of it.

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #64 –
Quote from: Woodinville on
But, if an individual prefers how hir stereo sounds while using red speaker wire, so be it.

This leads us back to the starting point of this thread: can wire color make stereos sound differently for specific individuals? If yes, on what evidence is that based?

Medical analogy: can pill color improve subjectively reported symptoms? Based on recent findings the answer is: logically we can't conclude that pill color has no effect. What we can say is that, if it exists, the effect is going to be small and that there's, presently, no evidence supporting its existence (=it is indistinguishable from bias).

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #65 –
Please excuse me if you’ve already tried to explain this umpteen times, but if so, tell me again. How is the placebo effect any different from expectation bias in the scenario you posit? The patient is given a pill made of sugar and water; the patient expects it to be real and to have an effect, so through the mysterious magic of the placebo effect, they start to feel better.

The paper you brought instead suggests that reports about placebos can be biased. This, to me, is a valid concern, but seems to be separate from whether perception was genuinely affected. Untangling the two might well be difficult, hence conclusions such as those of the paper. I defer to others to authoritatively comment on the paper in context, but I personally don’t see how reporting bias does much to suggest that placebo doesn’t actually occur, although it’s obviously a confounding factor in need of attention. Expecting an inherently subjective being to provide reports completely free of subjectivity seems futile, so things like this will always complicate matters, but in general, we still need to pay attention to reports about perception. And anyway, if we just require objective testing, all these complications seem to magically vanish!

And lest we forget:
Quote
Despite low effects in general and the risk of bias, the authors acknowledged that large effects of placebo interventions may occur in certain situations.

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #66 –
Quote from: db1989 on
Please excuse me if you’ve already tried to explain this umpteen times, but if so, tell me again. How is the placebo effect any different from expectation bias in the scenario you posit? The patient is given a pill made of sugar and water; the patient expects it to be real and to have an effect, so through the mysterious magic of the placebo effect, they start to feel better.

This used to be the standard view on the subject, and many researchers ended their papers on placebo with suggestions about how doctors could apply this mysterious magic in their practices, so as to help patients improve at less risk and no extra expense.

The three papers I mentioned in my first post reviewed much of the evidence for this mysterious magic and arrived at a series of conclusions:
a)it doesn't affect binary outcomes (cured/not cured).
b)it doesn't affect symptoms and conditions that can be objectively measured (like blood pressure)
c)there is no evidence that it affects subjectively reported symptoms (like pain and nausea).

On the third conclusion, they say that a small effect could be found (it varies widely among symptoms, methodology used and kinds of patients). But this small effect can't be distinguished from bias. 

If you don't want to read any of three papers, here's a popular article, by one of the co-authors, in which he summarizes their findings:
The Placebo Myth
Quote
When clinical outcomes of placebos were measured on a binary scale, such as improved/not improved, we could not detect any placebo effects. Nor did we find any effect on objective outcomes measured on a continuous scale, such as blood pressure or weight loss. We did find an apparent effect on subjective continuous outcomes such as pain, but these results were weak and unreliable. We noticed, for example that the larger the study, the smaller the effect – and small studies are statistically less reliable than bigger ones, and often exaggerate the effect. The weak effect of placebos on subjective outcomes, such as pain, may also be spurious for another reason. In almost all relevant studies, the patients were randomly divided into three groups: those receiving a medically active treatment, those receiving a placebo treatment, and those receiving no treatment. But it is a well-known phenomenon in health care research that it is impossible in these kinds of studies to prevent a bias in the way patients report their subjective states.

I think the author speaks very clearly. He does come short of saying placebo effect doesn't exist. He calls it "a myth", "spurious", "rather weak even if real", and so on.

But at this point I'm using these medical analogies just to clarify what I'm getting at. Earlier in the thread it was suggested to me that the phenomenon is widely documented in audio and backed up by rigorous evidence. I'm basically just waiting for the references.

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #67 –
Well, I wish you luck with that (not sarcastically ).

To be fair, it should be noted that you keep referring to a single author/pair, and there seems to have been some controversy about the conclusions. I’m not claiming to have an understanding of all the discussion around this, so I leave it to others to make pronouncements, but we should try to retain a sense of balance by not relying on single authors/groups, ater all.

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #68 –
The forum is chalk-full of claims that one thing sounds better than another where both things are equal without any effort to demonstrate how in an objective way. Many of them will be found in the recycle bin.  There are also claims (or suggestions) of audible differences which include failed ABX tests (p >> 5%) as "evidence".

It is expected that it be proven that there isn't some grand conspiracy that all of those people are being dishonest for reasons x, y and z?

It doesn't work that way; and personally, I don't care if the horse won't drink.

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #69 –
Even though I'm a layman, if someone were to ask me for proof of the McGurk effect's existence I'd be able to provide links to some half a dozen papers available online.

Many posters have claimed that the phenomenon we are discussing is supported by evidence. I don't think it's odd to ask them to be kind of enough to either present an instance of such evidence or tell us where to find it.

On the other hand, if they will withdraw their claim that this is a well documented phenomenon and just say that they know it exists from their personal experience and anedoctal report, the whole discussion will have to be reframed. I hope it's clear by now that I'm very skeptical about personal experience. I wouldn't trust my own shadow if it didn't provide me with a DBT.

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #70 –
Quote from: db1989 on
How is the placebo effect any different from expectation bias in the scenario you posit? The patient is given a pill made of sugar and water; the patient expects it to be real and to have an effect, so through the mysterious magic of the placebo effect, they start to feel better.


In their definitions, as far as I read it: Placebo would be if the sugar pill treatment actually cures (at least to some extent). Not only does the patient feel better, or report to feel better, but - through this perception or otherwise - the sugar pill treatment helps rid the disease.

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #71 –
@OP: The forum has a search function.  You are free to use it.  I was polite in asking you to conduct your own research. At this time I will simply tell you that I'm not going to do it for you.

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #72 –
Quote from: greynol on
The forum has a search function.

It was used before I started the thread. With the exception of a result from 2003, I found nothing relevant. Maybe my search was flawed, as I was not well acquainted with the forum's terminology.

But just to be clear, I'm not questioning that individuals come here and report all the time that they can hear differences that they later fail to hear in a DBT (or, at least, they fail to prove that they can hear the differences). I know this and it's very annoying. I'm not saying that they're lying either - although I have my doubts about the integrity of some journalists. What I'm saying is that when they say that they've heard a difference they're just as wrong as the patient who claims that his pain was diminshed after eating a sugar pill.

They didn't hear any difference because there isn't any. The patients didn't feel any improvement because there wasn't any.

All that's required to establish the falsehood of the first sentence is a proper experiment. For instance: subjects consistently rated the item labelled wave more highly than the one labelled flac, even though the wave condition actually led to playing of 96kbps mp3...

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #73 –
Quote from: Porcus on
In their definitions, as far as I read it: Placebo would be if the sugar pill treatment actually cures (at least to some extent). Not only does the patient feel better, or report to feel better, but - through this perception or otherwise - the sugar pill treatment helps rid the disease.


If that's the case, translating the definition (as it is) to audio would not be possible, as it would mean asserting that placebo (or expectation bias) in the subject doing the test is actually changing whatever he's listening to. And that cannot happen AFAIK.

Medic. study—placebo tends to lack important effect—relevant to audio?

Reply #74 –
I can tell you first-hand that you can hear things through the power of suggestion that don't actually exist. When told that they do exist prior to taking a double-blind test when they don't actually exist you can hear them anyway to the point that you think you hear it in X but not in Y.

Woodinville has referred to the phenomena as steering.

Seeing that we're still nitpicking terms, the person taking a test is called the testee. The items being auditioned for testing are what are called the subjects.