How to read scientific papers

Academic literature is not like other literature. It is not exciting. It is not amusing. It is not suspenseful. If I had to pick one word to describe it, it would be &quot;viscous.&quot; It’s like literary molasses: hard to move through, easy to get stuck in, and generally unpleasant in large doses. The following strategies can help you get through even the densest papers. These ideas are taken from my own experience, as well as from an excellent compilation of tips and tricks from scientists around the world, published in Science magazine.

It’s a skill, not a talent.
I can pretty much guarantee that the first academic paper you read will be a stressful experience. You don’t remember, but so was your first time walking. But the more you do it, the easier it becomes. Most professors and grad students try to read at least one paper every week, just to stay in practice.

Go in with a purpose.
Don’t read a paper just because your adviser told you to. Think about what he/she expects you to learn from it. What kinds of information should you expect to find? How will this paper help your research? How does it relate to things you already know?

Use the buddy system.
Arrange a time with one of your lab-mates to sit down and parse through a difficult paper. Explain the concepts to each other. You may be surprised how much more you will both get out of it.

Get a good reference manager.
A good reference manager is like EasyBib (minus the ads) plus Adobe Acrobat (minus the price tag). It should allow you, for free, to keep all your articles in a single place, synced with the cloud, and has built-in citation exporting, annotating, and organizational tools. Mendeley and Zotero are both great options.

Use the abstract as a filter.
Pretty much all articles are worth someone’s time. The purpose of an abstract is to help you know whether the article is worth your time. Go through it carefully before you move on to the article itself.

Start with the pictures.
Sure, the figures in an article probably won’t make much sense without the context of the article itself. But for most articles, diving straight into the text will be just as confusing. That’s the whole point of this strategy. If your first 10 minutes of reading is going to be confusing anyway, you might as well be confused while looking at something interesting. You’ll be less likely to give up or fall asleep in any case.

Take notes directly on the article.
You can circle, highlight, draw lines to connect concepts, plus your notes will always be readily available if you go back to reread it. There are many ways to do this, so find one that you like. Touchscreen pens work great if you like to hand-write your notes. Most PDF readers will let you attach comments/notes to specific locations on a document.

Familiarize yourself with the notation.
It has been said that physicists would rather use someone else’s toothbrush than use their notation. This, perhaps, explains why there are more notation systems in physics than there are stars in the observable universe. Understanding what notation is being used in a paper is critical to understanding what’s being said.

Know the variables before you tackle the equations.
This is similar to the last point. Write down each variable that the author introduces and try to explain what it represents in your own words. Draw pictures with the variables labeled. The equations won’t make any sense if you don’t know what the variables represent.

Keep the big picture in mind.
It’s easy to get lost in the multi-line equations, run-on sentences, and invented formalisms that are all too common in academic literature. If you find yourself stuck on a single point, step back and ask yourself how that point fits into the context of the article as a whole. This can help frame difficult (or poorly explained) concepts in a way that makes them easier to understand.

You don’t always have to read every word.
I’ve literally referenced hundreds, if not thousands, of pages of academic literature, code documentation, and Wikipedia articles in this paper. I would have to read all day, every day, for months to get through it all. You and I don’t have that kind of time. If a paper only has one paragraph that’s relevant to your research, then just read that paragraph.

Occasionally pause to remind yourself that you are, in fact, intelligent.
In a fantasticly relatable editorial, Adam Ruben says that &quot;nothing makes you feel stupid quite like reading a scientific journal article.&quot; If you find yourself getting discouraged, take a second to remember that you’re reading some of the most difficult literature the human race has ever come up with. Good for you!