Hacker News Reader

Comment by montecarl | Hacker News Reader

I spent a couple months implementing a promising optimization technique that was published in Nature. I assumed that I was missing something or had a bug in my code because I could not reproduce any of the results. It turns out, one guy in my research group knew the author. He requested the source code for me to the program used to generate the data for publication. That program did not even come close to doing what the paper claimed due to some very serious bugs. When I brought up these issues, nobody really seemed to care much. It wouldn't be worth anyone's time to try to correct the issue or write to the journal.

View on HN

> It wouldn't be worth anyone's time to try to correct the issue or write to the journal.

I don't see how a research scientist could say this; something seems culturally wrong there. Nature is a pretty serious publication. You spent months on it, maybe someone else will. How can it not be worth your time to 'try to correct the issue or write to the journal'?

-Nature is full of bogus, like all journals, if not more. The reason for its high impact factor isn't due to consistently solid papers all around, it's due to a few very high impact, foundational papers and a crowd of minor impact/shaky ones that were sneaked in by big shots and friends of the editors. Look up retraction rates by journal and prepare to be surprised.

-I don't think you realize the step it takes to actually write to Nature and say: "Excuse me, that paper you published is wrong and you should retract it." You're liable to alienate all of the authors as well as piss off the editorial board for pointing out that they let a mistake slip through. You should be prepared to face vigorous backlash, and be completely confident in your own results if you don't want the repercussions to overwhelm you. Many communities related to specific fields are small and niche, so making enemies with one team often means that you've actually cut yourself off a good part of that community. You should be prepared for awkward moments in conferences when you meet each other, incendiary questions when giving talks and scathing "anonymous" peer-reviews for your future submitted work. It's far easier to assume good faith, give the authors the benefit of the doubt ("yeah the implementation is buggy but what software isn't? The main idea is probably sound") and move on.

-I don't know about other fields, but in mine, there's an unofficial accepted consensus on a set of very high-profile papers that happen to be either complete bunk or utterly useless. Again, we're talking about Science, Nature, etc. These papers made a lot of noise when they came out years ago, and now no one in their right mind would base their work on them. Again, you can't just knock on Science's editorial board and go "Excuse me, that hyped paper from a very big shot is useless and you shouldn't have published it", so it's just something people in the community know and whisper among themselves. I guess it may seem strange to outsiders though.

> I guess it may seem strange to outsiders though.

Of course it seems strange. The output of the scientific/academic community is often presented to people as being a source of truth and scientists do little to dissuade this. To learn that it's all rotten to the core is upsetting.

It's not rotten to the core, it's just imperfect, like all human communities are. Nepotism, conflicts of interests, human error, unconscious biases, etc. don't magically disappear just because we're in academia. As I said below: "All in all, I'd say we're doing fine. We're just not the ethereal source of truth that some people hold us to be, the very same people who, after claming that 'God is dead', are very quick to replace Him with His sillicon-based equivalent around which we would act like priests, except with lab coats in lieu of clerical garments."

It's all about trust, really. Many people choose to trust state-of-the-art medical research when they do vaccinations on their kids; some do not. You can indeed choose not to trust us as a 'source of truth', and it seems a fair chunk of the US political landscape is being led down that path; we'll see where it goes.

Most of those other systems don’t claim to be designed to prevent them intrinsically via process, though.

We never claimed to be completely free of bias, or immune to nepotism. We're human.

I think you're making it to be a bigger deal than it actually is. Most of these things are just noise. We just ignore it. If someone rises to prominence with outrageously fake results, rest assured that they will get shot down quickly by concurrents (all the incentive to publish fake results lies within highly competitive fields). If you know what you're doing, and work in a lab where people know what they're doing too, you can still do pretty good science. I would guess than in all communities, academic or not, there's a nonzero amount of bogus, over-hype, nepotism, politics and what not, and a subset of people who actually get things done and make the whole field advance. After a while, the test of time truly determines what was actually useful from what is bunk. That's how it's always been.

Brian Wansink didn't get shot down quickly at all.

I think it's one of those things you just have to learn to live with :/

And believe me I hate being defeatist, it made me a lot angrier when I was younger and I consider it to be one of the things that stopped me trying to go down the academic route so far.

While coders (who aren't just being boosters for particular technologies) have to live with the fact that the world is run on mountains of bad code, scientists of anyone engaging with academia or journal papers has to sleep at night realising how much bad science is being done and how much the publication process is skewed to publishing and politics over quality or fact checking.

I can no more fix it than I can fix all software bugs or politics in big corporations...

If the most prominent repositories of research not only do nothing about but actively resist and retaliate against substantiated challenges to published results, that sounds to me like something pretty rotten, at or near the core.

The core principles of science are in conflict with natural human behavior (which is one reason it took so long to invent science), so saying “they’re only human” is really no excuse.

It makes me wonder why more effort isn't put into finding alternatives to this fundamentally broken system.

How do you propose getting started?

Start paying people to find negative results, and you'll get negative results.

Unfortunately, you can't build a career out of doing QA for science. Nobody will fund you.

I don't think it is all rotten to the core. I think it is a human establishment trying to do something very complicated with varying results. Sometimes they get things right and sometimes they make mistakes and sometimes they are infiltrated by bad actors.

What I try to do when reading academic papers is, if there are multiple on the same topic, read them and try to understand the differences. If different papers come to the same conclusion, that should make you a lot more confident in the result. If there aren't many papers on the subject then you just need to understand that this is kind of a "best effort" thing that is more likely to be true than a random guess, but not certainly accurate.

It's not rotten to the core. It's a complex system, full of human beings, with egos and incentives and other reasons for doing things.

Different labs also do different things - mine is currently reworking a paper that was basically done because something we did might matter, and my students have developing tests as part of their workload for developing code.

It's not rotten to the core. There are good people in there, but good people are not rewarded, and advancement selects for some bad traits. Maybe that's scarier than it being rotten to the core.

Pretty interesting use of 'assume good faith' there!

> people in the community know and whisper among themselves

If no one else, tenured academics should be calling out bunk in their fields, or are failing, IMO.

> You're liable to alienate all of the authors as well as piss off the editorial board for pointing out that they let a mistake slip through.

Better rewards for uncovering bad science keeps it from devolving into a political contest. If you're not seeing the right incentives in place, you're either missing something or you're seeing an opportunity to capture some value by implementing those incentives.

> Better rewards for uncovering bad science keeps it from devolving into a political contest.

No. If you increase rewards for uncovering bad science, it will just lead to false accusations.

> If you're not seeing the right incentives in place

Some people think that the very existence of incentives is the problem here. https://hbr.org/1993/09/why-incentive-plans-cannot-work

> If you increase rewards for uncovering bad science, it will just lead to false accusations.

Isn't that what this study suggests is already happening, in as much as false accusations are just more bad science?

> Some people think that the very existence of incentives is the problem here. https://hbr.org/1993/09/why-incentive-plans-cannot-work

This article is full of opaque generalizations about human behavior that more accurately describes the baser impulses of individuals attempting to game systems they don't understand (participants in psychological experiments) and don't so much reflect the actions of self-conscious professionals with any semblance of dedication to their fields.

The role of science in society is to increase the sum of knowledge on which progress can be made. If a scientist knowingly lets false information be published he is causing a disservice to society, harming progress.

Scientific papers are made public for a reason. The correct information should not be accessible only to a small circle of people in the know. If incorrect papers are out there, those who can correct them have a duty to do so.

>If a scientist knowingly lets false information be published he is causing a disservice to society, harming progress.

Often it's not so much false as empty or useless. Nevertheless, since all people in a community know the values of individual papers, and these are the same people 'making progress' in that field, I don't see how science is so much harmed. I agree the situation is kinda ridiculous, but it's not a "sky is falling" situation either.

>If incorrect papers are out there, those who can correct them have a duty to do so.

Sure. How much are you willing to pay me for me to prioritize this 'duty' over the literal hundreds of duties I already have at my lab?

Curiosity killed the cat and all that, but what is your field? I had a look at your profile, comments and submissions (hopefully that doesn't sound creepy!) and I can't tell.

It's Saturday morning (and I got students' papers to mark) :)

>I don't see how a research scientist could say this; something seems culturally wrong there.

I can tell you haven't spent time as a grad student in a top university in the US. There is almost no incentive for a research scientist to pursue this. They're very busy and stressed out, and this will not help them in any way. Your argument that one person wasted a lot of time, and so others will be spared that pointless effort is a sound one, but you have to realize that in some (sub)disciplines, scientists view a good bulk of the research as problematic and a waste of time. Why go through the trouble for this particular case?

Also, I doubt a simple email to Nature will change much. There would likely be a somewhat lengthy process, which will suck more of your time. And to be brutally honest, the chances are higher that the code did produce those results, but the grad students/post docs have been modifying the code for their next batch of research. Even something as basic as version control is unheard of in much of scientific research.

Definitely a cultural problem, as you describe it.

> it wouldn't be worth anyone's time to...write to the journal.

I'm sorry, what? It almost sounds like everyone in your lab is scum. Hopefully YOU at least spent the 60 seconds to write an email?

It's not at all uncommon that implementations provided with papers do not actually do what the paper says it does. You're often lucky when there is an implementation at all. But most of the time it's just that running the exact same implementation under the same conditions requires setting up a very specific environment, installing specific versions of libraries, using some niche software and converting data from one byzantine format to another. Each deviation from the original paper is liable to subtly affect the results in nondeterministic ways. That's why no one really gets surprised or even cares, life is too short to call out all the bad software written in academia.

> requires setting up a very specific environment, installing specific versions of libraries, using some niche software and converting data from one byzantine format to another

Containerisation is fairly mature and simple to use. Many in other fields struggle with these exact same issues and are able to create reproducible environments just fine.

I find it amazing that those publishing don't include their implementation, all that work locked away on a rusty hdd.

In the software dev world--but outside of BSD--I would argue that containerization is extremely immature. Outside the software dev world, it is not easy to use.

Containerization is not simple to use for many users of scientific code.

Hell, I'm pretty sure in my department, the first response would be "What's that?"

Containers may be a bit difficult, but every researcher should be capable of using a VM and distributing it.

I'm excited if I can get my colleagues to use version control.

Git is way more complicated than VMware.

Use SVN just like everybody else.

Do you think that they would use a VCS that was less invasive and more transparent to their workflow, like dropbox?

I'm thinking of making a VCS that simply runs in the background and

- Automatically records every file save (effectively a git commit without a message)

- Allows adding messages through tagging (like git tag)

- Handles 'branching' just by asking you make a copy of the directory with a different name, properly understands how to diff/merge/etc copied files/directories that have since diverged.

- Has built in support for large files

Almost certainly, yes.

You obviously don't know many researchers (if any).

If a research result requires specific versions of libraries to replicate, wouldn't that completely invalidate any scientific claim being made?

Any change in precision or numerical methods that affects results surely must be well within the error margins.

I was more referring to changes in the API where the input and output suddenly have to be in different formats in the middle of a pipeline, causing a crash. What can also happen is that somehow the old format is still valid and gets processed all the same, thus yielding nonsensical results. Sometimes a lab devises their own format which no one else uses, and the specification may be updated without notice between the moment they publish and the moment you try things out. Most people have no idea about things like 'backwards compatibility', 'unit tests', 'containers', etc. Code is just a tool to them and the fact that they had to write some is annoying them in itself.

Grad students / postdocs / human lab rats aren't scum, the incentives just aren't in place to promote good behavior (such as calling other researchers out on their bullshit). If you're trying to acquire a vaunted tenure track job, you can't afford to piss off $senior_tenured_researcher_at_prestigious_institution, since $senior could blacklist you so that you won't get hired at the incredibly small set of universities out there. Sometimes things work out despite pissing off major powers (Carl Sagan technically had to "settle" for Cornell due to being denied tenure at Harvard, in no small part because of a bad recommendation letter from Harold Urey [0]), but not often.

Even if you do manage to get a tenure track job, you pretty much have to keep your head down for 7 years in order to secure your position.

And once you have tenure, you still get attacked vociferously. Look at what happened when Andrew Gelman rightly pointed out that Susan Fiske (and other social psychologists) have been abusing statistics for years. Rather than a hearty "congratulations", he was called a "methodological terrorist" and a great humdrum came about [1].

When framed against these circumstances, it should be evident that there is literally nothing to gain and everything to lose from sending out a short e-mail pointing out that someone's model doesn't work.

[0] https://www.reddit.com/r/todayilearned/comments/15m8om/til_c...

[1] https://www.businessinsider.com/susan-fiske-methodological-t...

I'm a researcher myself and I guess this is one of those "does the end justify the means?" scenarios... Out bad research and its perpetrators and science loses out on a scientist that actually wants to do good work. Or don't and then watch yourself rationalize worse decisions later on for the sake of your research, slowly becoming as corrupt as they were and realizing that a lot of your cited work could potentially be as bad (or worse) as the ones you helped get published.

I really believe we need a better way. Privately funded / bootstrapped OPEN research comes to mind as a potential solution to bring some healthy competition to this potentially corrupt system. Mathematicians are starting to do this, I think computational researchers have the potential to be next.

> Grad students / postdocs / human lab rats aren't scum, the incentives just aren't in place to promote good behavior

The question is, would additional incentives promote good behavior or just lead to more measurement dysfunction. Some people think that just giving the "right" incentives is needed, but actual research shows otherwise.

https://hbr.org/1993/09/why-incentive-plans-cannot-work

Without reading through that very long text, claiming that incentives don't influence human behavior is a wildly exotic claim.

There is near infinite evidence to the contrary. That said, constructing a system with "the right incentives" can of course be devilishly hard or even impossible.

The claim is that it does change behavior, but only temporarily and it doesn't change the culture in a positive way / doesn't motivate people. It ends up feeling like a way of manipulating. That being said, according to this article, the entire incentive system would need to be dismantled. Simply adding more incentives wouldn't necessarily produce higher quality, at least not in the long run. So essentially the process of incentivizing new amazing research for funding is the primary issue and adding incentives for pointing out issues would just be a bandaid.

This sounds like a good critique of naive incentive schemes.

I don't think there is any doubt that humans follow incentives.

But working out what the core incentive problems are, and actually changing them might be both (1) intellectually difficult, and (2) challenge some sacred beliefs and strong power structures, thus making it practically impossible.

The HBR article's discussion of incentives is not really quite what I was thinking of when I wrote my comment. Specifically, the article you cite refers to the well-known phenomenon of how introducing extrinsic rewards via positive reinforcement is counterproductive in the long run. I've often noticed this form of "incentive" / reward being offered in the gamification of open science, such as via the Mozilla Open Science Badges [0], which in my opinion are a waste of time, effort, and money that do little to address systemic problems with scientific publishing.

With regard to the issue of grad students being unwilling to come forward and report mistakes, incentives wouldn't be added, but rather positive punishment [1] would be removed, which would then allow rewards for intrinsically motivated [2] actions.

[0] https://cos.io/our-services/open-science-badges/

[1] https://en.wikipedia.org/wiki/Punishment_(psychology)

[2] https://msu.edu/~dwong/StudentWorkArchive/CEP900F01-RIP/Webb...

Reply on news.ycombinator.com