Proposing hypotheses in the scientific method : better approaches?

So here’s what bugs me about the scientific method. In the most recent results in AI, the best outcomes were obtained from artificial systems that don’t start with pre-existing human supplied assumptions.

To me, the idea of “proposing a hypothesis” seems like an injection of dirty and statistically likely to be false information into a process.

Suppose you are investigating an animal that is in a box. You can only indirectly measure parameters about that animal.

You determine eventually that it’s a large land animal. Someone has felt a tiny patch of skin but they don’t know where it was on the animal.

Scientist 1 : “My hypothesis is that it’s an elephant!”

Scientist 2 : “My hypothesis is that it’s a hippopotamus!”

And you have a senseless ego match and breathless headlines get published that you found there’s an elephant in the box when it will later turn out to have been a tiger.

The alternate approach I am suggesting is that each measurement is a constraint on the probability function of what could be in the box. Each subsequent measurement you make narrows the set of what it could be.

So a scientific paper could be “based on the evidence, the likely remaining possibilities are an elephant, a tiger, a hippopotamus, or a very short giraffe in that box. This set of experiments is designed to reduce the set of possibilities to 2 by determining if the enclosed animal has fur…”

I thought of this when hearing about “dark matter”, which the evidence does not in any remote way suggest that the phenomenon that causes the observations is in fact a form of matter. Right now it could be a large set of things. Advancing the hypothesis that the culprit is a form of matter with never before seen properties is just hubris.

You don’t seem to understand how a hypothesis works. Or the scientific method.

Those are not hypotheses. Those are conclusions. A hypothesis comes before the experiment, not after. In your example, the researchers would start with their hypothesis and then conduct an experiment to see if the results are consistent with the hypothesis. If one of them thinks it is an elephant, then they will design an experiment specifically trying to see if it is an elephant. Perhaps, they will try to feel for a trunk. If they feel a trunk, they will report that in their findings. Another scientist might come along and build off that previous study by trying to feel for large tusks. The absence of tusks doesn’t prove that it isn’t an elephant, but the presence of large tusks and a trunk are pretty good evidence that it is an elephant. Regardless, they still haven’t proven anything, they’ve just got a running theory right now. It will take many more experiments, all showing results consistent with an elephant, before anyone accepts that there is indeed an elephant in there.

Scientific papers already kind of do that.

A hypothesis doesn’t get “advanced”. It gets tested. The only way to prove a hypothesis wrong is to test it.

What do you mean by “advancing the hypothesis”? A hypothesis is not a belief. It’s not something the researcher thinks is true. Very often, the researcher is trying very hard to disprove the hypothesis because that would be a more interesting and groundbreaking result.

For example, some physicists try hard to measure the effects of General Relativity. So General Relativity (or its predictions) is the hypothesis. They would be ecstatic if they managed to disprove it, because that would lead to new theories to replace Relativity.

A much better approach would be for you to try to develop some basic understanding of what the scientific method is, including what a hypothesis involves, before you go proposing changes to a methodology that has produced enormous advances over hundreds of years.

Just a thought.

Learning what a probability function is would also be helpful, as would the ability to come up with examples in which the noise made by a hungry animal isn’t being completely disregarded.

A hypothesis can be binary, but it can also be probabilistic. Someone saying “based on all available photographs, people in Navadad’s family in generations before Nava’s own always have light hair and eyes, therefore so will everybody in Nava’s generation” would have been ignoring both the genetic input of my generation’s other-side parents and the existence of my great-grandmother Honoria, she of the black hair and eyes; they wouldn’t even be wrong in their prediction, but also in their premises. Someone saying “based on all available photographs, people in Navadad’s family in generations before Nava’s own almost always have light hair and eyes, therefore it is highly likely that the amount of people with light hair and eyes in Nava’s own generation will be higher than that in the general Spanish population” would have been correct, but he would have been correct partly because he didn’t ignore previous evidence, partly because his understanding of genetics was better, and partly because he was expressing something which is probabilistic by its very nature (genetics) in appropriately-probabilistic terms.
OTOH, someone hypothesizing that members of the Nava family and in Nava’s generation will be human… yep, that one would have been right. No need to invoke probabilities there. We may be more or less weird, but we all count as human.

Ideally all experiments are based on testing a hypothesis. I forget who said it, but just gathering data without a hypothesis is stamp-collecting. Of course, you do start gathering data in order to formulate a hypothesis, but then you start. Even the assumption that it is a lion or elephant or hippo or giraffe is still a hypothesis and suggests which observations will be useful to test it or to narrow it down.

My dissertation was mostly stamp-collecting. A large part of chemistry is “hey look at this cool new stuff I can do!”

That’s how creating hypotheses work. Well, sort of, the data gathering that led to that specific hypothesis seems suspect.

“Dark matter” is a label, not an exhaustive description of a hypothesis. Underneath the “Dark matter” label there are many actual hypotheses, based on the evidence.

And when testing these hypotheses, which do all suggest dark matter is matter, one seeks, among other things, tests which would show that it isn’t matter.

Now, assuming you know something all the astrophysics I’ve read as a layman didn’t inform me of, what is this large set of things that could be causing the phenomena “dark matter” is hypothesized to explain, and that fit the observational data so far?

There are many proposed. And all can curvefit to the data. Such as ‘gravity scales in a way that explains it’ to ‘space itself has another property on large scales that explains it’ and so on. It’s a set of possibilities. Yes, if you actually see filaments through a telescope that is an observation that massively reduces the probability that it is not matter.

Doesn’t make the scientists rooting for the ‘matter’ camp right, just lucky, though.

I don’t care either way I am just wondering if the hundreds of years old scientific method is perhaps suboptimal.

You obviously don’t understand the scientific method, or the nature of the multiple observations that have helped with the understanding of dark matter. I merely picked one.

Their importance is that in combination they refute the “gravity scales in a way that explains it” and “space has a property on larger scales that explains it”-type hypotheses.

Saying “you don’t understand something” without evidence is advancing a hypothesis without support. Same to Colibri above.

As for the latter paragraph, I said that. Reread my post.

Relax guys, he doesn’t understand AI or derp learning either.

What a Freudian slip!

The evidence is that you proposed a hypothetical about the scientific method that bears very little resemblance to the scientific method.

Further evidence is that you continue to talk about “advancing a hypothesis without support,” which is meaningless. You design an experiment based on a hypothesis, and then conduct that experiment to gather data that either supports or disproves your hypothesis. In other words, the hypothesis comes before support.

Your posts are the evidence.