Maybe I didnt express myself clearly enough but my last post already answers your questions, namely (a) what if it's not 100% and you dont agree to the probability, or (b) what if the chicken is invisible and thus its existence is unfalsifiable. I predicted the linear response so I added that last paragraph, but wanted to keep the example simple. The same argument applies to your alterations of it.
Honestly I feel like you're not really trying to think. You're instead selecting something you at face value disagree with and fire a blank there without regard to framing that thing in the context of this.conversation, yourself, everything you know and understand, incorporating it in your.mental framework and subjevtive reality and evaluating its validity there, and indeed without thinking about it, to see if something interesting happens in response, which is evident in how little connection with your reality or thoughts or personality your post has, cause thinking costs you energy. Your posts lack depth and mental clarity. Just assume I know 1,000 times.more than you currently assume, you dont need to be so accommodating, I can handle much more than this and I honestly think you could too if you tried to think harder.
So let me reiterate my teleporting chicken argument against your pure constructivist view devoid of objective reality and regard for logical positivism.
If you didnt agree that observing the chicken renders its existence 100% probable, it would still necessarily mean that you are more confident the chicken doesnt exist because if it didnt exist you're more confident it (or.something like it) doesnt show up than vice versa. If the chicken was invisible then its existence would be unfalsifiable and the probability of any unfalsifiable claim being true is vanishingly small compared to any falsifiable claim. So neither of these objections change anything except relative likelihoods and thus .or are they a defeater to the argument.
The argument is also not changed by priors, only the relative likelihoods are. The argument is universal as long as you agree that observing the chocken incresses your confidence.
In summary, if you agree that everyone telling you that you're arrogant increases your confidence that you actually are arrogant, you necessarily also agree that nobody telling you that you are arrogant invreases your confidence that you are not arrogant. When you say that people could be nice and thus they are being biased you're not questioning the validity of that statement you're merely asking: "by how much?"
By how much our confidence is increased. Perhaps not by much, like not observing thr chicken in your bedroom didnt raise your confidence much, you need to do different nature of tests to be sure. However, once you've exhausted the checklist, the chances that you're wrong is honestly nil at that point, even if therr are a few tests that suck balls as you say.
This is where pure constructivism fails, because no matter what, we do have a way to bootstrap ourselves into an objective reality even when our own reality is subjective.
However, you dont have enough background to engage with the argument, which is sort of where I feel this disconnect with people who can't converse at the level I want to converse. I can now spend a lot of time on something that to me feels like elementary calculus, and donate my thoughts further starting from the fundamentals in a one sided fashion (which is what I usually do), or I could skip to another topic so we don't need to be stuck here in what feels like a pervasive lack of thought. (Or maybe I need to start posting voice clips)
I'd opt for the latter, because your question regarding how to device tests is a good one and honestly more interesting. I already got something interesting to think about with your media thing prompt, which was the first thing I hadnt thought of so thank you for that.
So here's the response to your second question.
I appreciate and of course know psychological bias and how when devicing psychological experiments we need to "fool" people into thinking that they're doing a different test, in order not to have the test be affected by our own input. As I said, I've spent time thinking about this, which is why I'd like us to go a bit more into depth instead of dealing with these linear suggestions that most people will have thought of. I even apply the principle in my everyday interactions. The unfortunate thing is that I can already predict what you will say about the following explanations through linear extrapolation, which is why perhaps I should stop being so rigorous in my thought and instead just wing it and smoke weed and accept, like I originally suggested, that you don¢t care about logic or being wrong or about your discussions being limited to surface level stuff.. But alas I keep giving people the benefit of the doubt, but it's alll an argument from utility rather than being an epistemologically motivated stance. That's why we all believe in God, too, don't we? I honestly think you can do better.
Let me expand; there are in general at least three ways to account for bias: minimization, informed debiasing, and averaging the bias.
The first one is to, as you say, minimize bias. Then when you look at the results you can go, see, here's the result, tada, boom.
The second way to do it is to (1) account for the bias or to (2) be conservative about it. So in your example, even if the large majority of people were being polite, then if you understand roughly how the bias works you can weight the results of those who said you're an arrogant piece of @!*# by a larger factor so you debias the results. Or you could go with a conservative estimate; if literally the ENTIRE WORLD says you're not arrogant, then let's face it that's extraordinary even for some of the least arrogant people on Earth. So I'd personally put my money on a person, who's told that by literally the entire Earth that they are not arrogant, on not being arrogant, irrespective of any biases.
However, sometimes you don't even know if your test is biased, so neither minimizing nor debiasing works, although oftentimes being conservative works.
Thus, the third way is to average over many different uncorrelated measurements to roughly average out the bias. That's devicing different types of tests and trying out all of them, which was among one of my suggestions. That way, you reduce the effect of the bias by the root of the number of experiments, assuming equal weight to every experiment, and a bit less in reality because the biases are never entirely uncorrelated, but the correlation itself in the broader sense is not strongly metacorrelated, astonishingly; you need to be incredibly unlucky for those tests to be all metacorrelated. The funny thing is that this works the same way for social sciences as it does science experiments, and it works EVEN if you don't know if your tests are biased. Personally, I find that amazing.
In reality, you should employ everything.
So yes, all in all, my.original suggestion is valid even in the context of constructivism. I had considered all of this in the two days of the decade when I've had this problem, so we're still just scratching the surface. How the above applies to God is the start of another set of existential issues..
And yes, I can mostly predict what you will say next, so let me instead ask a question that I don't know the answer to: