Give researchers a lifetime word limit | Hacker News Reader

Give researchers a lifetime word limit

nature.com

52 pointswidforss46 comments

> Lifetime limits would create a natural incentive to do research that matters. Researchers would have to ask themselves, “Is this project I’m pursuing worth the words it will cost me?”

Well, researchers already have a natural time limit: everyone will die/retire. Sensible researchers ask themselves, "Is this project worth the lifetime it will cost me?" -- I don't see why a word limit would cause a substantial change in the answers to either question.

The lifetime word limit should not apply to papers publishing negative results.

You may all congratulate me (financially) for solving the replication crisis.

Hah, people would find a way to hide some novel findings in an otherwise "negative result" paper and then extensively cite it in all future papers.

The lifetime word limit is a too rigid constraint. There should be a possibility to acquire (or lose) the points you can later use for publishing your results. Then the "price" of publishing negative results could be lower than other results.

Earn more words via replicating. Start with enough words for one paper.

It's amazing how much hand wringing is done about the symptoms of problems in academia (especially life sciences) without anyone actually addressing the root causes. The reason all of these problems exist is because the space is over populated, ruthlessly competitive, and under funded. It's really that simple.

Create many more professorship positions for post docs, raise the NIH, NSF, and DOD research budgets to where they should be, and all of these problems will go away. Trying to solve these problem by throttling the number of papers people can publish is putting a bandaid on a gunshot wound.

"I firmly believe that this problem cannot be solved by more government money. If federal support for basic research were to be doubled (as many are calling for), the result would merely be to tack on a few more years of exponential expansion before we'd find ourselves in exactly the same situation again."

David Goodstein, "The Big Crunch" http://www.its.caltech.edu/~dg/crunch_art.html

I had not read this before and that's a valid point, thanks for sharing.

I think I would say two things to this. First, theoretically we could attack the problem from both sides. We could constrain supply in the way that med school prevents an oversupply of physicians while also increasing the total amount of resources available to those practicing science.

More practically, I tend to believe that basic and translational science is a huge boon to global prosperity and also an economic engine for the United States. Ergo, if we run into the same problem down the road again, just keep raising the budget and make more positions. We have a long way to go before we exhaust the benefits of additional research, and the money we spend on research is a drop in the bucket compared to what we spend on other less useful ventures.

Med school limits are a way for doctors to seize wealth from patients. The industry is reacting to their greed my moving work to nurses and PAs.

The supply constraints med schools impose most certainly are not to prevent "oversupply" of doctors. Depending on who you ask, there are two main reasons they exist. One view is that there's a finite amount of doctors willing to work at academic hospitals, and thus there's a limited amount of new medical students that can be adequately trained. Another view is that medical schools collude to form a guild to better separate the American populace from their money, by ensuring that doctors are always scarce relative to the demand for their services. This allows doctors to demand a dramatically higher compensation for their efforts than in other countries, despite being no better here than elsewhere.

In a normal economy the only thing regulating supply, is demand.

The kind of organization that imposes supply constraints to sarcastically prevent "oversupply" are called cartels. You know, like a drug cartel?

True to the definition of the word, but you have a wrong example. "Drug cartels" are anything but, and the only actual cartel in the drug trade is the United States government.

To be fair, it's a doctors declared goal, to reduce the need for themselves. From a system stand point, down regulation in moderately quiet times promotes people to be well healthy by their own. It's symmetric if its not just doctors monotonically maximizing their income, but also patients and doctors competing with each other and one another for prosperity. So the doctors who earn most, should in principle dampen their own optimal payout.

I agree, more money might contribute to more research and may or may not be a useful expenditure but I don't think it will solve the problems mentioned. It seems that limiting the number of PhD/post-doc postions such that they have a high likelyhood of ataining academic positions is the correct course. Although, I'm less familiar with the medical education system in the US/Canada it seems like this is roughly the model they follow.

We don't need industry taking advantage of the glut of PhD grads, they could be responsible for training and not take academics who quit because they can't find suitable jobs after graduation.

I smell some unsubstantiated opinions. Raise government research budgets to "where they should be" sounds like a good recipe for wasteful spending.

I might be on board with "govt matches private contributions."

It seems psychology is over funded, considering how much garbage is constantly spewed out.

I don't get it. The problem: A lot of lousy research gets published in order to game the academic system. Your solution: More funding. Do you really think those guys who are already gaming the system will not make a grab at the extra funding?

I think this problem has to be tackled at the level of the journals. Journals should not reward "interesting" or catchy results with publication. Instead, experiments should be accepted for publication before they are performed, and then published regardless of result. The criterion for evaluation should be methodological rigor and proven replicability, not potential for journalistic headlines. Change the incentives and you change the outcome. If scientists are rewarded for rigor (and they currently are NOT), they will bring rigor to their experiments. If they are rewarded for cutting corners, corners they will cut.If they don't they themselves will be cut from the profession as their CVs lag behind their less scrupulous peers.

Please no. Papers are sufficiently terse, jargon-filled and hard to follow as it is

Further, in some fields (biomedical) research takes so long, and top-tier journals are so selective, that a lot of work just never gets published. Slowing that process even further would wreak havoc on already way-too-long graduate careers. A really pernicious effect of some bosses who 'only publish top-tier-quality work' is many students get stuck with years-worth of negative results, and do not get to either move on or publish their work.

And that is because of the journals having limits.

I've had papers where we carefully wrote a long explanation of exactly how some feature was implemented. If papers were websites, just hide it behind a link, don't read it if you don't care.

Instead we were told to take it out as it wasted space.

This is a terrible idea. Here is a better idea:

- Change tenure criteria to something closer to “number of students * average quality of dissertation”

- Expect fewer (0-2) but higher-quality publications from grad students applying for tenure track positions.

Wouldn't that give people an incentive to have lots of students, thus making the PhD glut worse?

I guess it depends on how the two factors are weighted.

I think it might be equally, if not, more important to graduate only a suitable number of students/postdocs so that the majority of those that want academic positions will find those openings available.

I'm not sure how to balance these issues.

Fair point. This is a real issue in most non-engineering fields.

Maybe add a “% that get a job” factor.

Also, in STEM fields.

There’s definitely a glut in mathematics and science PhD’s. Universities keep cutting tenure track positions and cranking out more students per professor. It doesn’t hit technology and engineering as badly, since there is a steady supply of industry jobs.

Some people make the jump from science and mathematics to industry, but it is harder.

I think it doesn't seem as bad in tech/engineering due to the availability of industry jobs, at least you aren't unemployed. But in my opinion as a science/engineering PhD grad many of these positions don't require PhD they could just train undergrad engineers and be fine. The companies are taking advantage of the glut of PhD grads to get cheaply trained highly educated labor without incurring the cost to train them themselves.

This would be great because it would shake the foundations of the principal investigator model, which I think it's doing a lot of harm to research in Medicine and Biology.

Reminds me of some Warren Buffet advice on investing: Imagine you could only make 5 investments in your life. Would <your latest idea> be one of them?

The article jokingly mentions that papers function as a "pubcoin" allowing you to buy various benefits.

What if it each researcher started out with a limited amount of a cryptocurrency, "pubcoin", which must be spent to be published? More would be awarded based on success (citations?).

There is a Pareto distribution in the number of papers people write, as well as the number of citations papers get.

The really good papers often have few citations at first, then ramp up exponentially about a decade later. These papers advance the field 5+ years, and it takes most researchers about that long after the fact to notice massive shifts. This even happens in “fast moving” fields, like CS, where the joke about science progressing one funeral at a time still applies.

That lag is enough to end a career.

Superficial, but well marketed papers(republished in spammy ways) tend to have lots of citations at first, then die off, flash-in-the-pan style. The system already over-rewards those papers.

The rare researchers that can produce really good papers and also market them get turing awards, etc.

For every paper you publish, you must repeal two.

Restrict researchers to communicate using only the word "novel".

If I can only use "novel" how do I talk about how "robust" my work is!?

On the subject of short papers: http://fermatslibrary.com/s/shortest-paper-ever-published-in...

I got a better idea. Force the researcher to create and submit a little mini documentary on the subject along with the paper.

That's actually already done to some extent. See New Journal of Physics for example.

Would math and/or source code count?

This is a dupe but the first couple of attempts got nowhere.

I left my views here:

https://news.ycombinator.com/item?id=15495448

Noam Chomsky.

Yes.

Is your answer an example of the possible outcome of the proposal? Shorter papers with less content (but more frequently)?

Yes.

More research is needed.

Reply on news.ycombinator.com