To solve the replication crisis in science, couldn't we simply start from scratch?

The idea of the scientific method is sound. It takes an idea, designs a test for it and runs the test. If the idea holds up to the test, then there will be replications of the test to verify it works every time. If it doesn’t work every time, the scientific method demands that you figure out what the variables are, what might be different from test to test, is there a better test, etc. But you nailed the problem with it: scientists are no longer interested in science. They are interested in getting paid. Someone pays for their research and that someone wants a specific result. So to keep getting paid, the scientists will skew their results to get the one the person with the money wants.

What you are trying to overcome is human greed. Good luck. The people funding the research are not doing so out of altruistic desires to learn. They are hoping to get results that will allow them to get money in one way or another. The scientists know that they get paid as long as they can manipulate data the way it is desired. It all comes back to greed.

I do not understand the question.

I agree with @seawulf575 that greed is one factor.

The other factor that works against replication from scratch is cost. Depending on the field and the chemicals / equipment needed, it cost hundred of thousands, to millions of dollars to execute a controlled experiment. Think of what it costs to test cancer drugs, or weigh-loss drugs, or almost any other substance.

Should results be replicable and experiments provable? Absolutely. But that costs money. And it boils down to a balance between time, money, and efficacy.

And I disagree wholeheartedly with @elbanditoroso and @seawulf575 that greed is the major factor in the driving of science.

@luigirovatti I don’t know what context you are talking about.

But when you wrote: “Why can’t we simply start over from where we got the info wrong (on natural, as well as social sciences)?”, my first thought is to ask, “why do you suppose we know where the point is that “we got the info wrong””? And what would it mean to “simply start over”? And how would that really be “simple”?

@Zaku: The info’s wrong where the experiments are not replicable. To start over means to select a panel of scientists to start from the most basic experiments in scentific history, (for example, splitting the light with a prism, why the earth rotates around the sun, don’t know figure it out), to go upwards, in order to build gradually our information from what we know for sure is working. The idea in itself is simple, but it’s a big idea, not easy to carry out, never said that. Still, it’s better than what’s happening right now, don’t you think?

@Zaku, @Caravanfan: Read the following to understand what I’m talking about: https://en.m.wikipedia.org/wiki/Replication_crisis

Wait WAIT *WAIT if I close my eyes @luigirovatti you don’t exist ? ? ?

If I don’t see you and I can replicate that and have anyone do that . . . ?

Who funded a scientific study has some bearing on the results. This is well documented. Pressure to publish does also. The “replication crisis” is mostly in psychology and medicine. I’ll let you draw your own conclusions. There is no replication problem in the hard sciences.

@luigirovatti “Still, it’s better than what’s happening right now, don’t you think?”
– Well. what’s happening right as I read this, is I think an example of something similar, which is the tendency to look at a conclusion from one perspective, and to over-apply it too broadly. Also, to apply binary judgements to things where more refined and nuanced replies would be much more appropriate.

Applied with so much vagueness, only an at-least-equally vague answer seems possible.

So: Yes, and no. It depends on specifics.

Sure, it’s great to try to go back to first principles, or no principles, and try to investigate from there. Even the thought exercise of trying to even understand what all of one’s assumptions and so on, is an immense task, not to mention really letting go of all of those.

On the other hand, there are many very useful things to continue to act as if they’re correct.

I tend to assume that much of economics and sociology and political “science” is dubious at best, and would be better if it were thrown out and/or reconsidered from unconventional perspectives. Physical sciences, not so much. But even there, there’s room for questioning.

Fortunately, it’s not necessary to do only one or the other.

In fact, continuing questioning of ideas long thought correct, is supposed to be a part of scientific inquiry, and often isn’t done for a variety of reasons (some innocent and/or ignorant, and some very much not).

@Zaku: It doesn’t take a genius to realize what i suggested would be more practicable in physics, where this crisis doesn’t exist, than in social science, precisely because it’s all a bit more theoretical.