This way, people will be reluctant to cite too many useless papers for their friends, and will be reluctant to depend on unreliable ones. Work with good methodology would be more popular to cite or rely on.

Maybe peer review could even become optional this way. If other researchers trust your work enough to risk their own papers by citing it, that would act as a peer approval in the long term. Actual peer review would only be needed to give an immediate indication of the quality.

We would first need to solve the versioning problem where there's no way to update a paper when mistakes are discovered or simply to improve it.

This is an interesting idea, but I don't know what real-world problem it would actually solve. As someone who reads a lot of scientific papers, I can't think of a single one whose correctness is actually dependent on the correctness of the results of a paper it cites, as though the paper in question is "buggy" because it "calls another paper as a subroutine" and that subroutine is buggy. Could you give an example or two of this?

I could see this making sense in mathematics since mathematical proofs really do "call other papers as subroutines" (theorems), but not so much in science.

What field are you reading where each paper is independent of all the others? As the most obvious example, in biomed pretty much every paper says something like "as done previously[ref x, y]" in the methods section. It is an extremely annoying, but very common, practice (usually refs x and y contain somewhat different methods descriptions and/or also contain the "as done previously" phrase). There are many other ways the papers depend on each other too ("we used these cells because they don't express this protein[ref z]", etc)

Basically, to me, an empirical result arrived at using the scientific method is, well, an empirical result arrived at using the scientific method. And it seems to me that the only thing that would invalidate such a result is if it were found that the researchers did not follow the protocol they described in their paper. The fact that the researchers might have merely cited another paper, or used the method described in another paper, in which the researchers were found to have "cheated" in this way, doesn't seem to invalidate the paper in question. It's possible that the researchers in question could have happened to also cheat, but I don't see any reason to assume it's any more likely because it happened with a paper they cited (unless of course they're the same researchers building on their own previous work).

    > > whose correctness is actually
    > > dependent on the correctness of the
    > > results of *a* paper it cites

    > where each paper is independent of
    > *all* the others

Really, the information should be abstracted away from the paper, a la freebase or wikidata, and there should be strict rules (e.g. the protocols stackoverflow enforces) such that no cycles can occur, such that propositions can be resolved to axioms, and such that it is practical to compute scores in a finite period of time with reasonable resource constraints...

See also gitxiv.org (at attempt at literate research for computer science and physics papers that can be paired with code). It doesn't address the problem described, but if statements of a paper are codified (e.g. curry howard isomorphism) in a strict way, it makes it easier for "linked statements" to have a quantifiable value.

However, some of the papers tend to discuss the results in the light of the current empirical results AND the earlier empirical results ("from this result, we now see that A → B, and from the earlier results we already know that B → C, so there's some evidence that A → C"); this of course biases the discussion and may lead to wrong conclusions.

By the way, the largest problems I see in the papers are not directly "wrong" results, but like the "Chinese whispers" or "broken telephone" game. The paper A makes some assumptions and uses certain test protocol, measurements etc, and then paper B refers the result but misrepresents it slightly, which may cause overgeneralisation. A concrete example: paper A tests the learning of some phoneme with a training period and before-after test and publishes the result. Then the paper B later references the paper A mentioning that A has shown that the learners "learn" the phoneme. However, represented this way – "unqualified" – this is an overgeneralisation: the students haven't actually shown to have learned the phoneme in any other way than using the test. Indeed – other studies have shown that using certain kind of tests are generally very poorly correlated with what people usually mean by "having learnt" something – being able to use it in spontaneous communication.

There's also a trickier problem in your simple binary flag: papers can have incorrect proofs of correct statements.

I'd love to see a Semantic Web approach where you don't cite papers but results in papers. So while a result in a paper might be wrong, you don't have to throw out all the results in all the papers citing this paper.

Also, citeseerx and crossref are your friends.

Pubmed central similarly has the ability to return much of this citation data programatically (for certain scientific subdomains)

If you're passionate about prototyping such a service, let me know and we can get you an invite to the Archive Labs slack (we're a non-profit, volunteer-run community which incubates open access and public good projects). Literate, reproducible research and knowledge + citation graphs are two of our emphases.

Good luck

You might also like metacademy.org (one of my favourite projects -- it's a package manager for knowledge built on DAGs)

> We would first need to solve the versioning problem where there's no way to update a paper when mistakes are discovered or simply to improve it.

You can certainly update what a DOI points to.

Apparently we've identified the problem enough to say "there's a problem". We can say "look at all the ways you can manipulate this analysis to achieve the desired outcome". We can say "gosh, science is harder than we thought". But it seems we're still far from a convincing solution.

The fact that statistical analysis is so liable to manipulation seems to call the entire thing into question. In the article they take comfort from the fact that many of the labs analyzing the red card/race data arrived at similar conclusions. One would assume this is because they made similar choices in the analysis. But what says those were the right sorts of choices? Could it not simply be that the labs shared the same biases and errors, making that outcome more common? Is a proper analysis really determined by (essentially) democratic vote? If that's what we've arrived at, it gives me less rather than more confidence in the robustness of the scientific process.

It feels like something fundamental has to be reimagined. It's difficult to prove things about the world---but maybe it's actually near-impossible? Or maybe we need to get real about the cost of actually demonstrating anything reliably. Instead of individual labs running one-off experiments it becomes researchers collaborating openly to build the perfect experiments, which are then run by many different labs, then analyzed collectively in the open for strengths and weaknesses, then reformulated, sent out again, and so on iteratively until at the end of years of research one little bit of probably-truth drips out the bottom of the system.

But that bit would be something we could build on.

Scientists always hesitate to talk about "proof," since we hold our beliefs on a tentative basis. However, some scientific theories have been corroborated to such a degree that they can be regarded as proven for all practical purposes, within broad domains. These include classical mechanics, relativity, gravitation, quantum mechanics, thermodynamics, electromagnetism, and Darwinian evolution. This level of "proof" is trusted on a daily basis in regular lab work, e.g., if an electrical circuit violates Ohm's Law, you don't throw out Ohm's Law. You look for a bad solder joint.

Granted, science is presently dominated by the life sciences. I'm sure the physical sciences have their own problems -- we shouldn't be overconfident -- but they may be qualitatively different problems. In my view, a journalist shouldn't generalize about "science" without showing that they've considered whether the generalization is justified.

Disclosure: My background is in physics, and my day job is in scientific instrumentation. It's a business where people are obsessive about replication. When an instrument goes out the door, it's been put through a stringent set of tests that could be described as replications, and the results of these tests are compared across different designs, models, batches of components, and even competing brands, all around the world.

While the system is highly flawed now, I would argue that collaborating to build perfect experiments carried out by multiple labs would not be a good idea. You point out if people made the same sorts of choices then perhaps everyone just has the same biases and errors. But if all the people on one experiment collaborate, then they'll definitely all have the same biases. Instead, independent verification in which the nitty-gritty details are abstracted away for a more general methods section would allow for more alternate takes at approaching a problem.

The trouble is everyone wants to be the special snowflake discovering the special effect in their own special experiment, so now we have too much messy unreliable and inaccessible data to be of much use in the real world.

Science itself can never be broken, but when people are cheating the system that makes it effective to actually figure out what is true and what isn't, for personal gain, then that system is broken.

Unconstructive criticism is worse than non-constructive mathematical proofs, as those tend to further actual results later.

It's WAY too hard to be a scientist. The kind of salary and other sacrifices scientists are asked to make are unfair and surely deter many to leave science.

Science is the only thing moving everything forward and if the funding strategy is look for irrational people who will work insanely hard for next-to-nothing salaries...it doesn't sound like a great strategy.

The average person is losing faith in science because of high profile failures, and the distrust is only increasing.

I fear Nasim Talib is right.

https://medium.com/incerto/the-intellectual-yet-idiot-13211e...

Science has a huge PR problem in that it is FAR less funded than the people with money and power whom it sometimes comes into conflict with.

That shouldn't be possible! Science is supposed to be the best possible epistemological methodology, and here it is being beaten in "success rate of determining true from false" by guessing. What's immensely frustrating is that it's not a question of whether we're just not smart enough to tell true from false, we clearly have the power (since the guesses were often right) but we're not using it. Whatever "truth compass" the guessers were using should be part of the scientific process somehow. That's what is "broken".

The problem is that without this expert panel making these calls, both studies got published, were given the same credibility, and got consumed by non-experts. Peer review and the publishing process are supposed to weed out the obviously bad studies, but they're not doing so.

Individual papers are not a source of truth, they're bits of cumulative evidence hopefully leading to the truth.

Who would you rather use it: those who sell to develop a responsible framework for using it or those who seek to abuse it?

This tech is only going to become more accurate and more widespread over time. Ignoring it isn't an option. We've got little to no idea how to responsibly use human emotion as an input and decades worth of collective experience on how to exploit it.

The potential for abuse is high, which means extreme transparency is needed, at the very least.

Operating at an emotional level requires a whole new set of metrics and a fundamental shift in how we do anything. Better to figure it out now than to wait until the tech becomes ubiquitous. Those looking to exploit the tech most definitely won't wait for that day. A few people I've spoken to about it were eager to explore the tech purely for the sake of improving sales of various things. I changed the subject with them after realizing they had little regard for human life in this context.

People concerned about this need to band together to find ways to productively use it and defend against its misuse.

I've been working on such a framework, if anyone's interested in exploring it with me.

This is too important to simply dismiss out of fear. Science needs more courage, not less.

Moreover, make sure to verify the hypothesis that mental state is important to scientific reproducibility.

The validity is up to the researcher to provide & will need to be done in as emotionally "safe" way/place as possible if we don't want to risk training scientists to deny their emotions to the point of separating their biological responses from their emotions. That'd be a disaster.

This tech introduces all kinds of dangerous potential effects, so anyone working on or with it will need to take extra care.

[1] read: FUNDING

Let's distinguish "science" -- the scientific method, the general advancement of human understanding over the centuries, etc. -- from "pork": institutionalized government funding, the establishment pursuing that, and all of the mundane, bureaucratic processes that ensue, and then the resulting hype, pettiness, recriminations, and sacrificing of ideals that it inculcates.

There is nothing wrong with science, although it may be harder these days to recognize it.

Pork, on the other hand, is approaching a singularity.

That machine- is broken, leaking and actually in parts moving contrary to the scientific interests of humanity.

The quality issue of science itself, could easily be remedied by replacing citations with partial repeatable experimental coverage in citations this way also ending inflation.

Hypothesis driven experiment.

Drive it into the brains of everyone who might enter the scientific profession, and then, when someone is caught with an experiment driven hypothesis, we don't have to speculate whether or not it was fraud, because they will have known better, by simple virtue of having their credentials, and we can then safely revoke those credentials.

Yes they are

Sadly, what people call science these days has nothing to do with the scientific method, it's just a bunch of idiots doing correlation and thinking it's causation.