You are, for reasons unfathomable without resorting to uncharitable speculation, leaving out what this 'alteration' amounts to in practice. It doesn't increase accuracy of the perception.
Where did I even suggest that sighted evaluation "increased the accuracy of perception?" (In terms of the sound).
(Hint: nowhere).
In fact, you can get a person to report that the same device sounds different, if you make them think (via sight) that the test actually involves two devices. Such difference is of course entirely imaginary. That's how far 'bias' can go.
THIS IS WHY SENSORY PSYCHOLOGY -- whether its focus is the study of hearing, taste, smell, touch -- RELIES ON 'BLIND' PROTOCOLS.
It is why Toole, Olive, etc use them. When they point out that sight alters people's perception of sound, they are not positing benefits of sighted evaluation. They are warning against its deficits.
So please explain what good it would do to reintroduce the perceptual noise that DBT protocols are designed to filter out?
Have you, I suppose, skipped all my other posts in this thread? That seems the most reasonable explanation as to why you think you are telling me things you think I don't know, as if I haven't already written about those issues in this very thread!
The point of blind testing:
https://www.audiosciencereview.com/...-of-sighted-listening-tests.15114/post-477584
I'll give it one more whirl: when you want to strictly investigate the sound of a speaker on a listener's perception (or preference), you of course should control for the variable of sighted biases. We agree?
But why?
It's because, as I and almost everyone else on this forum have pointed out before, and now you have re-iterated: Sighted biases can influence perception. We agree?
Therefore: there is not only the phenomenon of how a speaker's sound (only the sound) is perceived. There's also a phenomenon as to how a speaker's sound is perceived COMBINED with variables involved in sighted bias (how it looks, it's design, or how much it costs, various expectation effects).
Right? Otherwise there would be no need to reduce the influence via blind tests when you are ONLY testing the effects of the sound.
Therefore: the effects of sighted bias - the way they change perception of the sound! - can be of interest as well, something to study, and something to consider. (Just as I change the perception of sound in my job).
And, as we've gone through on this forum before, there is an issue inherent in the blind testing studies: they can predict what people are most likely to prefer under the blind conditions of the test, and hence narrow down the variable with respect to just the sound. But life and human brains being messy things, the predictive power for listener preferences or user satisfaction once OUTSIDE the lab becomes much more shaky, due to the conditions under which most people ACTUALLY tend to purchase and listen to their speakers.
This is what I meant in pointing out the blind testing is great for reducing the influence of such variables in the lab, but if in practical normal-life scenarios those sighted variables (influences) are normally commingled with the sound, and this actually alters the perception of the sound (most people buy speakers in sighted auditions, and listen under sighted conditions) that's a *real* change in perception. (Just like an optical illusion is a real perception).
In principle the exact effects of the looks of a speaker could be studied for it's effect on listener perception, just as changes in sound caused by different speaker designs is correlated to listener perception/preferences. Toole et al, as far as I'm aware, have done *some* tests on this, such as the ones I referenced. But that seemed mostly to simply establish the unreliability of sighted tests, helping to justify the program of using blind tests. But the bulk of the research from then had to do with identifying specific speaker designs in regards to only the influence of their sound, not their looks (which is why all the blind test research). I'm not aware of any similarly involved systematic correlations between how a speaker looks (or other bias factors) as to how it alters specific perceptions of the sound. If there were, you'd have a scientific standard for looks as rigorous as the one Harman Kardon has produced for the sound. (Though, perhaps HK has actually done just this work, leading to how their speakers look visually. I'd be curious to know).
Another consequence of the fact that bias effects actually CHANGE the perception of sound is this: It could be completely reasonable to make a purchase based on the speaker you prefer in sighted testing. And that is EVEN IF you would make a different decision if based only on blind testing. So take two speakers that are pretty similar in performance, and under blind conditions I slightly prefer speaker A over B. But in sighted conditions, I perceive speaker B as sounding somewhat better than A. Why? Well, maybe speaker B just looks like a million bucks compared to speaker A. Speaker A is plain looking while speaker B has a gorgeous high quality wood finish, cool drivers, a neato design. Whatever aesthetic issues are causing my change in perception, the FACT IS I'm experiencing a change in perception. And I may want to simply avail myself of that phenomenon. "I may choose speaker A in blind tests, but under the conditions I'll be using this speaker, I perceive the better overall experience listening to speaker B." That is a completely rational way to make a decision. It's advised, made knowledgeably, with a clear understanding of one's goal and which action better fulfills the goal.
So, this is one reason I raise the issue of how blind testing is certainly immensely valuable for weeding out variables to understand and predict the contribution of *just the sound* to listener preferences. But once you are outside the lab those variables exist to either learn more about, or if one wants, to exploit in how one makes a purchase.