How do scientific research articles get published? How do we know their results aren’t faked? What exactly are scientific journals and how do researchers get revenue from publishing their research work?

140 views
0

How do scientific research articles get published? How do we know their results aren’t faked? What exactly are scientific journals and how do researchers get revenue from publishing their research work?

In: Other

When an article gets submitted they have a lot of people look at it and similar data in order to publish it. It generally takes months. And a lot of research is funded either at a government or private level.

The articles are submitted to journals or conferences for publication. They are then ‘peer reviewed’.

The peer review process is largely a check on relevance and basic sanity. The reviewers want to ensure the work is meaningful and not simply a rehash of previous work (although the standards for this these days are incredibly low). They also want to ensure the author is using plausible methods within the field.

In terms of data faking, we don’t know. We simply trust that the researcher was honest. Unless the work is very significant or has real-world consequences, no one is likely to replicate it. However, almost any work with far-reaching consequences *will* be replicated and they’ll discover that the conclusions are inaccurate.

When you look at people who have been caught in academic fraud, it’s almost always because they’ve established a pattern over years of publishing too-good-to-be-true work that someone finally figures out *isn’t* true.

In terms of getting revenue from their research work, researchers normally don’t. Your published works are a way of building your CV. Your CV is what gets you hired in academia and certain industry jobs. If you don’t publish anything, you’re unlikely to be considered for these jobs.

Researchers do earn money from patents, but you don’t need to publish to secure a patent – and, in many cases, it would be prudent to not publish your results if you’re planning to monetize them in this fashion.

There’s a board, the IRB, that you have to submit your hypothesis and plan to before you even start. You have to get approval from the board on your tactics for gaining data, make sure it’s ethical, fair, and unbiased. And you have to have a *ton of supporting information. If you don’t get approval you can’t submit your research. You can’t just bring a document to them and ask for it to be published. Depending on the type of research, sometimes you can get government funding. Some companies have their own research labs and will pay people to carry out experiments. For example I know hospitals always do “Evidence-based practice”. Where research teams look at certain practices and research the efficacy of it, instead of just saying “well we’ve always done it like this”.

Most people in Research don’t have any motivation to fake or sway results, they’re after the truth. That being said, I’m sure if you really wanted to you could. There are some ways to prevent that, in Blind, Double Blind, and unbiased research (where the researcher isn’t tied to the company or another entity in any way).

Let’s start with the last question.

Generally some Master’s (in some countries) and all PhD students in scientific fields get paid either with a grant they got or a grant they were offered from a prof. Where does the grant come from? Either the students or the prof or the group submit a project proposal to some authority or government body or some funding agency, they explain the problem in the field, how they intend to fix it, and how this would impact the world. If the agency is convinced, they give them the money: it’s usually from tax payers money, by can also be from pharma companies or charities or whatever (conflict of interest might arise). The money then covers everything from consumables to instruments to wages of people you hire for the project and salary or stipend of the main student. Post doctoral fellows are similar, but they get paid typically from their own grant (they bring their own project when they join the lab, and they already have an approved funding, they just need a lab to work), but post docs can also join a lab that already has a grant (usually super rich labs in top universities). Professors get money from the university itself and they also typically have side jobs as some consultant or the CEO of a spin off/start up bio tech company and so on.

So publishing per se doesn’t pay you (in fact you often end up paying for it). But publishing good science shows your employers that you’re good and productive, which makes them hire you and pay you (also for grant agencies to give you money).

Let’s get to the first questions. How’s quality controlled? Well.. It goes like this: you write up a paper, with figures and tables showing the data. You don’t show everything, that would be way too long, you skip out on the controls and validation experiments and pick out representative stuff and so on. Now when the paper is received it depends on the journal, but usually it goes through a preliminary assessment where they decide is this at all fit to be in the journal in terms of field? Is the quality standard at least in the ball park of our journal? Are the authors sketchy or from some weird place you never even heard of? Etc.

Once this step is done. Then there is I think the editor choosing whether he recommends this for publishing in the journal. This is just based on quality, significance, reputation, vision, etc. If yes, then a group of experts who are usually professors are chosen from around the world who are good at the field of that paper and pretty up to date. Sometimes the profs are the go to profs for the journal, sometimes they’re new and so on. Then these professors read the manuscript and criticize the living shit out of it. They look at the quality of the data, the validity of the data, the credibility of the methods, the sensibility of the whole approach, the interpretation of the results, does this fall in line with previous literature? Is it completely conflicting with everything we know? If so, is there a possible reason or just straight out bs? Are the controls properly done? And so on. They really scrutinize the paper and dig it to the bone.

Then it’s a dynamic process. After the first review round, the profs sent their recommendation and opinion to the editor. The editor, based on the. Comments, decides hmm.. Okay this is worth it. So he sends the author that they have preliminary acceptance if they respond to the comments. The comments can be questions for clarification, asking for the controls that aren’t shown, asking for more experiments to make sure the observed effect is real and reproducible, ask for the actual raw untouched data (this is usually done to studies showing revolutionary data, they want to make sure it’s not made up, and trust me some papers before have been found to be fake this way, they can even analyze the noise in the system to see if it’s random or there’s a pattern). They can also ask for changes to text, writing style, etc… Etc. Then it’s a cycle.. Back and forth between the authors and the reviewers, until the reviewers are satisfied and say okay. We’re good. Then it gets published in the next available issue.

Of course.. This is high end journal stuff. Like nature, science, cell, etc and also stuff like PNAS, endocrinology, etc. But crappy journals from some random country with an impact factor of 0.2 are usually extremely careless and don’t give two shits if your data is true.

Once a paper gets published in a good journal, that doesn’t mean it’s fact. Some fake data still makes its way through. And if it’s bad enough it usually gets retracted later. Like that Wakefield crap. But sometimes the model used in the paper or the protocol followed or something is actually suboptimal and renders inaccurate data or something specific to their set up but cannot be reproduced. This is usually found by later papers. I mean all science builds on itself. Most of what you do in a paper depends on protocols by other papers. And you use the protocol of another paper and it doesn’t give you their result because you have a different model, and then others find themselves in the same shoes, well eventually that paper loses its value and gets cited very little. Science is all about reproducibility. Even if you succeed in scamming a journal, it’s impossible to scam the scientific community and they find out in a few years (very easy to scam the public though.. Too damn easy).

**How do scientific research articles get published?**

A given journal has a panel of scientists that perform a peer review on submitted papers. An established scientist in a relevant field reads of the paper looking for flaws and trying to establish whether it is interesting enough to be published.

**How do we know their results aren’t faked?**

If a paper presents interesting and novel results, two things will happen. First, the paper will be scrutinized for errors and anomalies, faking data in a way that will pass deep statistical analysis is actually quite difficult. Next, if there are no obvious flaws, other scientists will try to replicate the experiment to get the same results. That will quickly expose any shenanigans, many a career has been ruined by simply overstating results, much less faking them.

**What exactly are scientific journals?**

A journal will typically be associated with a scientific organization, like a university or a professional group, although some are independent. They invite scientists in their field to submit papers, provide peer reviewers to ensure quality, and publish results they consider interesting. The more prestigious the journal, the higher the quality of the submissions, the more accomplished the peer reviewers, and the higher the cost. Some run thousands of dollars per year.

**How do researchers get revenue from publishing their research work?**

They don’t, not directly. Publishing does three things for the researcher, it ensures they get credit for the discovery, it improves their reputation (and job prospects) within the field, and it justifies their existence to their employers, particularly those in academia.

Generally they’re published in journals with a peer-review process. Typically when you submit an article to a journal they’ll give it a look over to make sure it fits the scope of the journal and if it does they’ll send it out to some other academic in the field to check the methodology, literature review, analysis, writing etc.. The reviewers don’t get paid, it’s part of the ‘service’ obligation that’s part of most academic jobs. There’s always the risk a reviewer will do a half-arsed job, but there will be at least one other reviewer, so if you do that there’s a good chance you’ll get found out and do damage to your reputation.

The reviewers usually don’t check that the data is real because they don’t have access to the necessary information. It’s up to universities/research institutions to ensure academic integrity in relation to the falsification of data. At that level there *are* people who can check up and usually there’s more than one person involved in the research project. Anyone caught falsifying data’s reputation will be ruined, they’ll certainly be fired and probably be unable to find work in the field again. Sometimes it happens, but generally there is very little motivation to falsify data and doing so involves taking a career-ending risk.