Properly cited scientific papers increase your credibility and provide evidence for your arguments.
Poor sources of evidence
Many writers use as evidence links to sites that have poor credibility. These include:
Websites with an ideological bend. These have the worse credibility of all because they are obviously pandering a set of beliefs. If you link to them, you identify yourself as a member of that ideological bubble. For readers who accept those beliefs, your article would be preaching to the choir. For the rest, it would generate distrust.
Blogs. You would be basing your argument on somebody who has just another opinion, and maybe even less credibility than you have.
Magazines and newspapers. These have various levels of credibility depending on their reputation. However, almost all of them can be placed somewhere in the ideological spectrum, which means that they are biased. Even if the writer makes an effort to be intellectually honest, they still will be filtering information according to their level of education and the amount of research that they did. Unfortunately, most journalists have a poor understanding of science. Of course, if you want to document an event that happened in the past, you should cite newspapers reporting it.
Science news and university press releases. These should be avoided like the plague because they have become a major corrupting influence in scientific culture. Major universities have created their own propaganda machines to tout the achievements of their professors in news releases, sometimes to position them ahead in battles for patents and awards. Needless to say, this is not conductive to honest reporting and balanced scientific discussions. It is also promoting a culture in which scientists become just another type of celebrities, like the sport stars that have long been used to prop up the prestige of universities.
I love Wikipedia and I often use hyperlinks to its pages. I do that here. However, I will discuss its use in another article.
What you should do is to go directly to the source of scientific knowledge: the scientific papers. You should do so in a careful and balanced way. However, this is hard to do for non-scientist.
In this article, I will explain a few basic things about scientific papers that you need to know to cite them properly. I will also tell you how to use PubMed to find the best papers for your writing.
The different sciences
There are many scientific disciplines: physics, astronomy, geology, biology, biochemistry, medicine, pharmacology, neuroscience, etc. To cite a paper in one science to support claims in another science would be a major blunder.
The humanities - philosophy, history, economy, literature, law, art, race studies, feminist studies, etc. - are considered different from science. Some of them, like history, are based on facts. Others, like philosophy, are based on reason and speculation. It has been argued that science and humanities belong to two cultures, often at odd with each other. In any case, you should not cite a paper in the humanities as scientific evidence.
In this article, I explain how to find papers in the general area of biomedicine, which covers chemistry, pharmacology, biology, molecular biology, medicine, neuroscience and psychology. Papers about physics, geology and other sciences are not listed in PubMed and have to be searched in other repositories.
Types of scientific papers
There are many types of scientific papers, that are used for different purposes. You need to be aware of this when you cite them and discuss them.
Here are the main types of scientific papers:
Basic research is done using instruments to examine the properties of matter and living tissue. They range from the telescopes that study the immensely big to the electron microscopes that examine the immensely small. But they also include many other techniques: mass spectroscopy, DNA sequencing, gene modification, cell cultures, electrophysiology, etc. In physiology, neuroscience, cellular biology and molecular biology, these type of experiments are called in vitro, which means in the glass - the glass of test tubes and Petri dishes.
Animal research is done in live animals (in vivo) or in tissue extracted from animals and kept alive (ex vivo). Examples of the latter are primary cell cultures and tissue slices. Most biomedical, pharmacology and neuroscience research is done on animals - mostly mice and rats, but some essential research needs to be done on monkeys. However, animal rights activists have spread the lie that animals are not necessary for scientific research. Because the same activists stalk and terrorize scientists and their families, universities and other research institutions hide the fact that they use animals, which has the unwanted consequence of giving the impression that most research is not done in animals. In fact, in the last decades, genetic modification of animals has blurred the line between basic and animal research.
Clinical research is done in humans, often to test new medication, devices and medical procedures. Most commonly, it consist on giving sick people and healthy controls a drug or a placebo. Other times, experiments are done in people. For example, using electrophysiology, positron emission tomography (PET), functional magnetic resonance imaging (fMRI) or trasncranial magnetic stimulation (TMS). Clinical research also includes epidemiology: the science that studies the spread and causes of diseases.
Surveys consist of asking people a bunch of questions and analyzing their answers using statistics. They are used in psychology, diet research and sex research.
Reviews are papers that collect, examine and discuss the papers on a particular topic. Although they are not original research, they play a fundamental role in summarizing the state of the research on that topic and in building scientific consensus. Citing reviews is your best bet when you want to prove that there is a growing scientific consensus on a topic, or when you want to cite a bunch of research at the same time. However, you should also be aware that some reviews can be biased to promote a certain scientific view and to ignore another. A lot of scientists that write reviews include a lot of their own papers in them (self-citation). This is not improper per se - scientists usually write reviews about something that they know well precisely because they have been working on it - but it can be abused.
Meta-analyses also summarize a lot of previous work, usually on clinical research. However, unlike reviews, they do so in a systematic way by pulling the data from previous studies and doing a statistical analysis of all those data together. This lead to strong conclusions.
When you cite a paper, it is a good idea to mention what type of paper it is.
What is peer-review?
You may have heard that, for evidence to be reliable, it had to be published in peer-reviewed papers. But, what is peer-review? How do you know that a paper has been peer-reviewed?
When modern science was being created in the 19th century, a tradition was established that scientific papers should be evaluated by other scientists before being published. These scientists are called reviewers and should be anonymous and unbiased either in favor or against the authors of the paper.
These days, when a paper is submitted to a journal, an editor assigns it to three reviewers, who remain anonymous and do not know who the two other reviewers are. They decide whether to accept the paper, require modifications, or reject it. The second and third options are the most likely outcomes. The editor then makes a final decision by balancing the opinions of the reviewers. Sometimes, a paper goes to several review cycles before being accepted.
Almost all papers deposited in PubMed are peer-reviewed, except letters to the editor. If a paper was published in a reputable scientific journal, it is safe to assume that it was peer-reviewed.
Peer-review is also used to score grant proposals to obtain the government grants that fund most of the research done these days. In the USA, these grant are administered by the National Institutes of Health (NIH), the Department of Veteran Affairs (VA), the Department of Defense and the National Science Foundation.
In my scientific career, I published 65 peer-reviewed papers and obtained half a dozen grants from the NIH and the VA. On the other side of the equation, I reviewed many papers and grants proposals to the NIH and the VA.
Although peer-review has been criticized, it is by far the best way for gate-keeping and quality control for science to remain in the hands of scientists, instead of being handed over to corporations and politicians who would quickly corrupt science.
What parts of a paper should you read?
Most writers don’t read the papers that they cite. This can backfire when a clever reader uses the paper that you cite to counter your own argument, which is embarrassing and undermines your credibility.
However, reading a scientific paper is very hard and requires a specialized education in the scientific field of the paper. Unless you have a Ph.D. or similar degree in that subject, you should approach the paper knowing that you are in way over your head. Even if you are a scientist, reading a paper can take you two or three hours. So, what do you do?
You should read the abstract, which is short summary of the paper. You should also skim the paper looking for the information that relates to your writing.
Yes, I know… The full text of many papers is behind a paywall. I will address how to deal with that in another article.
There are parts of the paper that are written in less specialized language accessible to most people. These are:
The abstract. This is a summary of the paper in 250 words or less - a limit imposed by most journals. Even if the paper is behind a paywall, its abstract should be accessible in PubMed and other repositories. Most scientists only read the abstract of a paper, unless they are looking for detailed information. The problem is that abstracts condense a lot of information to get under the 250 word limit, so they are hard to read. They also use specialized language because that’s the only way to condense information. Therefore, an abstract is hard to read for non-specialists.
The introduction is the text following the Abstract, without a heading. It presents the state of the research on this topic and the goals of the paper. Because it uses less specialized language, it tends to be easier to understand. It tells you what the paper is about and what is the previous knowledge on its subject.
The figures are often hard to understand, but sometimes they are the best way to appreciate the results and how solid they are. Some figures are beautiful, veritable works of art. But don’t grab figures and put them in your article. You need to request permission to do that, not from the authors but from the journal, which usually owns the copyright.
The discussion is the last part of the paper, after the results and before the list of references. Here, the authors go to town explaining what they found and why it is important. Like the introduction, it tends to be in a more colloquial language.
The reference list can be used to find other papers on the same topic. However, you would be moving backwards in time when you do that, so you will miss the most recent research.
How to judge the quality of a scientific paper
Every scientist knows that there is a world of a difference in the quality of papers. Publishing in journals like Nature, Science, Cell or PNAS is a lifetime achievement. On the other end of the spectrum, there are a bunch of new journals that will beg you to publish in them (I get half a dozen emails every day) - if you pay their publication fees, of course! They are just a way to get your research money.
Therefore, to cite papers, you need to be aware of the hierarchy of journals. It’s not the same to cite a journal in Nature as a paper in the Journal of Irreproducible Results - which actually exists!.
However, there is a way to quality of a journal: look at its impact factor. This is a metric that was created in the 60s and 70s by the Institute for Scientific Information (ISI) based on how often papers in that journal are cited. Journals with impact factors higher than 5 are exceptionally good. Run-of-the mill journals have impact factors between 3 and 5. Some good journals that publish a lot of papers can have impact factors below 3, but below an impact factor of 1 we are definitely in low-quality territory.
Using the impact factor to determine the quality of journals and papers has many problems. Nevertheless, it is useful for those unfamiliar with science.
How to use PubMed to search for papers
PubMed is a free, searchable repository of all reputable scientific papers on areas of biomedicine published in the entire world. It is run by the National Library of Medicine, which is part of the NIH and funded by the USA government.
Since a lot of biomedical research done in the United States, and even abroad, is funded by the USA government, it is now mandatory that scientists deposit papers created with that funding in PubMed. Therefore, you can find the full text of many papers in PubMed. If the full version is not accessible there, at least you can find the citation and the abstract.
Here is a simple way to do a PubMed search to find papers on a given topic:
Follow this link to PubMed https://pubmed.ncbi.nlm.nih.gov/.
You will see a big white rectangle with Search to the right of it. However, I do not recommend that you do your search here, because this will give you an imprecise search with too many results. For example, searching the word choking gave me 43,679 papers. Way too many to even scan their titles.
Instead, click on Advanced under the search bar. This takes you to a more complex search page, which is actually better for our purposes.
In the box All Fields, scroll down to Title or Title/Abstract. This will restrict your search to paper containing your keywords in the title, or in the title or the abstract, respectively. Enter your keyword in the bar to the right and press ADD, then Search. For choking, I found 509 references with Title and 2,434 references with Title/Abstract.
These are still too many, so you need to narrow down your search. For example, if you are interested in the practice of choking during sex, you can use the AND function to restrict your search to papers that have the keywords choking and sex. The search (choking[Title/Abstract]) AND (sex[Title/Abstract]) gave me only 106 results.
You can now scroll through the titles and check those that may be interesting in the big square boxes next to the title.
At the top left of the list of results, you will find options to save your checked references in a file, have them emailed to you, or send them to the Clipboard, My Bibliography, Collections or a citation manager. I normally send them to the clipboard, where I can delete those that I don’t want and add more citations from new searches. When I am done, I send the citations to my citation manager.
I will address the use of a citation manager in another article.
Intellectual honesty and credibility
Now, it is up to you to use this information.
Anybody who loves and respects science should have intellectual honesty:
“Intellectual honesty is a personal commitment to search for the truth by examining the evidence and thinking rationally, to tell the truth, and to act according to the truth.” Hermes Solenzol in Sex, Science & Spirit.
I wish I could tell you that being intellectually honest and building your credibility would get you more readers, but I am afraid that this would not be true. You will find yourself arguing with unethical fools who will use all kinds of dirty tricks to protect their beliefs.
However, in the end, self-respect is one of the most valuable things that you will achieve with your writing.