Expectation Bias

[firedog]

45 minutes ago, FIndingit said:

"Expectation bias" is the new "Giant killer" of audio forums.

 

Giant killer of discussion.

The actual problem is that most audiophiles insist on living in a fantasy world where it doesn't exist; and especially doesn't exist in their own minds. 

[firedog]

10 hours ago, kennyb123 said:

I had been catching up on this interesting thread but then this comment caused me to take a detour.  I’ve lately been taking a greater interest in psychology as I’ve been curious about what makes people tick.

 

I have heard it repeatedly often on audio forums that we all suffer from expectation bias - and it is this bias that leads many to imaging things that aren’t there.  I’ve not spent long researching this but I’ve yet to find anything that confirms that.  
 

Expectation bias comes up most often as “researcher bias” and it pertains more to conducting research in a way that achieves an outcome that meets the researcher’s expectations. That’s different than the charge leveled at those who report having heard certain qualities from a component.  It could be argued that it’s researcher bias that leads some to discount observations from listeners that don’t conform to their expectations.
 

I will keep digging into this.  Sorry about going off topic here.  Thanks @GoldenOnefor such an interesting thread.

Nah. We loose the term loosely in audiophilia, it could also be called confirmation bias, or a couple of other things. It's also the reason people talk about double blind testing: so that the biases of both the tester and the subject being tested are removed.

It's extremely well established that such perception bias exists and influences all aspects of human perception: sight, hearing, and the other senses. To the extent that we can hear something that isn't there, just because a visual reference is telling us it is. Or see something that isn't there, just because their are visual clues that tell our brains it "should" be there. 

Audiophiles tend to arrogantly assume they are immune to it. They aren't. Ever (in any sighted/conscious testing).

[firedog]

5 minutes ago, dericchan1 said:

I thought expectation bias means a fantasy world that exist in one’s minds?

meaning they live in a fantasy world where expectation bias doesn't exist.

[firedog]

12 minutes ago, kennyb123 said:

I’m glad you mentioned this because it supports my argument.  Typically only a small percentage of those given the placebo react as if it actually did something.  Blind testing relies on this number being small as it represents the control group.  They can know a drug is effective when those actually given the medication have a reaction that exceeds that of the control group
 

If perception bias were as bad as some in audio say, double-blind testing would be useless because mostly everyone given the drug would imagine that it helped.  That double-blind testing can be counted on as being reliable is proof that expectation bias isn’t as big of a factor as some make it out to be.  

 

Sorry, Kenny this and your other posts suggest a basic misunderstading of the concepts. You've simply gotten it wrong. 

Placebo effects can be large, even in drug testing. The idea doesn't depend on it being near 100%, as you seem to think. The strength of the placebo effect also depends on the malady involved. 

 

In audio it's even more important, as unlike drug testing, we aren't curing a disease or eliminating a specific physical reaction, which is a clear outcome. In audio it can have subtle effects. So unlike drug testing, the effect is altering say, bass perception, not eliminating it. 

 

There are so many tests showing perception bias in humans that it isn't in dispute. Including audio perception. Ever looked up the McGurk effect? Try it.

Your exercise doing Google searches don't undo decades of research. 

 

 

[firedog]

8 minutes ago, firedog said:

Sorry, Kenny this and your other posts suggest a basic misunderstading of the concepts. You've simply gotten it wrong. 

Placebo effects can be large, even in drug testing. The idea doesn't depend on it being near 100%, as you seem to think. The strength of the placebo effect also depends on the malady involved. 

 

In audio it's even more important, as unlike drug testing, we aren't curing a disease or eliminating a specific physical reaction, which is a clear outcome. In audio it can have subtle effects. So unlike drug testing, the effect is altering say, bass perception, not eliminating it. 

 

There are so many tests showing perception bias in humans that it isn't in dispute. Including audio perception. Ever looked up the McGurk effect? Try it. https://www.youtube.com/watch?v=2k8fHR9jKVM

Your exercise doing Google searches don't undo decades of research. 

 

 

 

[firedog]

7 hours ago, kennyb123 said:

Great then I’m sure it would be easy for you to provide support for your claim that expectation bias impacts us all the time when we evaluate audio.  
 

The McGurk effect has no relevance to this topic.  This has nothing to do with bias.  It pertains to the interaction between hearing and vision in speech perception.  Nice that you can call out a study by name.  How about naming one that actually supports your assertions around audio.  I’m genuinely interested in seeing this and have yet to find anything myself.  

I have no interest in  spending my  time  looking for links to provide for you. In your posts I've mostly seen someone who did a quick Google search and then made broad, and incorrect conclusions based on misunderstanding. Probably misunderstanding brought about by a desire not to understand the material in a way that threatens your position. The evidence is overwhelming for perception bias in all forms of human perception. But I get that whatever evidence you are presented with, you will say it doesn't apply/isn't good enough for audio. 

 

The McGurk effect has everything to do with this: it's one example of how ideas in our brain effect what we hear. It's just a different form of perception bias. Do you think listening to music is so different that somehow our brains totally change how they work when the aural stimuli come from a speaker?

At Harmon, they found that even experienced listeners' perceptions of speakers were effected by visual clues, such seeing the size of the speaker or the brand of the speaker. Get it? Visual clues can effect audio perception.

 

A few years ago research was published showing people's  evaluation of volume: the same volume signal was perceived as having different loudness levels, depending on the graphic shown when the audio was playing. The audio was accompanied by the picture of electronics, all of which were the same except for the color of the volume knob on the device. The one with the red knob was perceived as having the loudest volume, even though there was no volume change.

 

But none of that has anything to do with audio, right?

[firedog]

7 hours ago, kennyb123 said:

t's because you know it will be difficult to prove you assertion.  You make sweeping assertions and then use logical fallacies as proof.  

No, it's because I really don't think it's worth my time. Find the links to stuff I've mentionsed 

 

7 hours ago, kennyb123 said:

I don't know about you, but I was able to adjust my thinking to be able to prioritize what I heard over what I saw.  So if there is any relevance to expectation bias, this would suggest that if I know my biases, I can adjust my perception to remove that bias.  

No. And this shows you misunderstand the whole concept. Expectation biases aren't conscious. You can't know with certainty what they are and you can't control them as you think. This is typical wrong and arrogant thinking of audiophiles that they are exempt from a basic fact of human behavior and perception.

 

7 hours ago, kennyb123 said:

7 hours ago, kennyb123 said:

Of course it does but again it doesn't prove your argument about expectation bias.  All that proves is under those specific conditions, it was possible to influence bias in the observers.  No one is disputing that manipulation is possible.  But these findings can only be applied to the home if similar conditions are present.  I often turn off the lights and close my eyes when I do serious listening comparisons just so I can better focus on what I'm doing.    I know many other audiophiles who take similar steps.  Where are the studies on how much less of a factor expectation bias when care is taken to approach the comparison fairly?  There are none.

 

Again, audiophile myths with no connection to reality. There are studies showing education about a specific bias under specific controlled circumstances can have some effect on results. But that doesn't prove your point or disprove mine. You can't control your expectation biases in your listening room as you keep claiming. Especially since you don't even know what they are.

It's simply false thinking to think that you can.  That's why  blind testing and double blind testing is used when scientific evaluations are made. They are necessary because the ability to eliminate expectation bias as you claim doesn't exist. 

[firedog]

Okay, I've decided to relent.

Here's some links:

https://www.harman.com/documents/AudioScience_0.pdf

see page 10. 

 

Quote

"Other variables were also tested, and the results indicated that, in the sighted tests, listeners substantially ignored large differences in sound quality attributable to loudspeaker position in the listening room and to program material. In other words, knowledge of the product identity was at least as important a factor in the tests as the principal acoustical factors. Incidentally, many of these listeners were very experienced and, some of them thought, able to ignore the visually-stimulated biases [7]. 

 

Here's a small list of links. Btw, hearing aids are much more extensively studied than audio (because there isn't money/research money in audio, but there is in hearing aids) but the experience is highly applicable to audiophilia. Perception bias in evaluation of identical hearing aids has been shown.

 

1. de Lange, Floris P., Micha Heilbron, and Peter Kok. "How do expectations shape perception?." Trends in cognitive sciences 22, no. 9 (2018): 764-779. (PDF)
2. Wikipedia contributors, "Schema (psychology)," Wikipedia, The Free Encyclopedia, https://en.wikipedia.org/w/index.php?title=Schema_(psychology)&oldid=925260703 (accessed November 28, 2019).
3. Taylor, S. E., & Crocker, J. (1981). Schematic bases of social information processing. In E. T. Higgins, C. A. Herman, & M. P. Zanna (Eds.), Social cognition: The Ontario Symposium on Personality and Social Psychology (pp. 89-134). Hillsdale, NJ: Erlbaum. (PDF)
4. Schwarz, Katharina A., Roland Pfister, and Christian Büchel. "Rethinking explicit expectations: connecting placebos, social cognition, and contextual perception." Trends in Cognitive Sciences 20, no. 6 (2016): 469-480. (PDF)
5. Dawes, Piers, Samantha Powell, and Kevin J. Munro. "The placebo effect and the influence of participant expectation on hearing aid trials." Ear and hearing 32, no. 6 (2011): 767-774.
6. Dawes, Piers, Rachel Hopkins, and Kevin J. Munro. "Placebo effects in hearing-aid trials are reliable." International journal of audiology 52, no. 7 (2013): 472-477.
7. Wikipedia contributors, "Nocebo," Wikipedia, The Free Encyclopedia, https://en.wikipedia.org/w/index.php?title=Nocebo&oldid=924529979 (accessed November 28, 2019).
8. Politzer, Thomas. “Vision is Our Dominant Sense.” Brainline. https://www.brainline.org/article/vision-our-dominant-sense (accessed November 28, 2019).
9. Berger, Christopher C., and H. Henrik Ehrsson. "Mental imagery induces cross-modal sensory plasticity and changes future auditory perception." Psychological science 29, no. 6 (2018): 926-935. (PDF)
10. Palusis, Kelly Lynn. "Expression and Emotion in Music: How Expression and Emotion Affect the Audience’s Perception of a Performance." (2017). (PDF)
11. Tsay, Chia-Jung. "Sight over sound in the judgment of music performance." Proceedings of the National Academy of Sciences 110, no. 36 (2013): 14580-14585. (HTML)
12. West, Richard F., Russell J. Meserve, and Keith E. Stanovich. "Cognitive sophistication does not attenuate the bias blind spot." Journal of personality and social psychology 103, no. 3 (2012): 506. (PDF)

 

There are many many more. This is just a quick sample I found. 

 

I'm still guessing you'll reject them all as somehow not relevant.

 

It's amazing how only the tiny subset of audiophiles  have superhuman powers of perception unavailable to other humans, and can eliminate expectation bias when listening to their systems. 

[firedog]

4 hours ago, GregWormald said:

"How does expectation bias affect my perception when I go into a listening comparison to learn if I can actually hear a difference rather than with an expectation?"

 

You have expectations whether you are aware of them or not.

[firedog]

11 hours ago, fas42 said:

 

I see ... although, I recall there were some misguided astronomers who complained that the Hubble space telescope suffered from blurring of the fine detail, when first launched. And some foolish engineers were convinced that this was real, and  believed that they could understand the cause of this subjective appraisal. And even more foolishly, the government spent a fortune on a tweak, which expectation bias led astronomers to believe that it resolved the blurring; that it was no longer an issue.

 

Interesting how mass delusion can spread around, in all these fields,

 

You recall wrong: https://www.cbsnews.com/news/an-ingenius-fix-for-hubbles-famously-flawed-vision/

 

They made a "contact lens" to fit over the lens to fix it.

[firedog]

10 hours ago, manisandher said:

The whole of science relies on intuition and curiousity

Sort of. It's a necessary but not sufficient condition.

People who are totally unscientific also have intuition and curiosity.

 

Science is a method for finding out repeatable objective results and producing hypotheses that get us closer to the truth about the natural world. 

That's why eliminating all bias from testing and discovering something that's repeatable is important.

Your "experiment" is interesting, but is only a first step.

But it shows why this stuff is almost never done in audio. The comments about experiment design and the need to do more are spot on. It's difficult and expensive  to do right, and the incentive really isn't there. There's little academic interest, and commercial interest is mostly against it. 

 

[firedog]

1 hour ago, botrytis said:

 

NO - a hypothesis is based on previous data, similar type of experiments, etc. It is not intuition; it many times is extrapolation.

Agreed. The scientific use of the word is different than the lay use. A scientific hypothesis is supposed to be a proposed explanation of previously known data, etc. And it needs to be something that can be tested. 

[firedog]

2 hours ago, The Computer Audiophile said:

It’s helpful to talk about specifics. Do a few examples come to mind?

Most of the cable manufacturers.

Shenyata "Research", for example.

Anyone using the word "quantum" in describing principles of their audio equipment.