How to (seriously) read a scientific paper

[Andy]

Adam Ruben’s tongue-in-cheek column about the common difficulties and frustrations of reading a scientific paper broadly resonated among Science Careers readers. Many of you have come to us asking for more (and more serious) advice on how to make sense of the scientific literature, so we’ve asked a dozen scientists at different career stages and in a broad range of fields to tell us how they do it. Although it is clear that reading scientific papers becomes easier with experience, the stumbling blocks are real, and it is up to each scientist to identify and apply the techniques that work best for them. The responses have been edited for clarity and brevity.

How do you approach reading a paper?
I start by reading the abstract. Then, I skim the introduction and flip through the article to look at the figures. I try to identify the most prominent one or two figures, and I really make sure I understand what's going on in them. Then, I read the conclusion/summary. Only when I have done that will I go back into the technical details to clarify any questions I might have.
- Jesse Shanahan, master's candidate in astronomy at Wesleyan University in Middletown, Connecticut

https://www.sciencemag.org/careers/2016/03/how-seriously-read-scientific-paper

 

[Liessa]

Good question I don't know if you intend for this to be a discussion on this topic, but I'm happy to share what seems to work for me.

I read the abstract and leaf/scroll through the paper to see what figures and sections it has.

Then I go to the references. This may only work after you've read a good bunch of papers but I find it highly enlightening to see which definitions are used (do they mention CCC/ICC for example), and I look at the prevalence of BPS references. This gives some sense regarding where the current paper's authors are coming from. And I keep an eye out for authors/papers I know in general because that helps me have the right context in mind when I come across those references later.

Then I read the discussion and the conclusion. Usually this yields some 'why how but..' questions in my mind that I can look out for.

Then I scan/read the methods/analysis parts (depends on the subject, how well I understand it, and how much energy I have to look up every term I may still be unfamiliar with).



I'm sure there are more ways to go about and would love to hear if others have tips!

 

[Trish]

I read the abstract, then the conclusion, then decide how much of the rest I want to read. Graphs and tables are useful.

 

[JemPD]

Reading the article in the OP, it becomes clearer to me why so many scientists have fallen for the PACE (et al) nonsense.... very few of the people quoted in the article refer to reading papers with a critical eye - checking that the methodology is valid etc. It's like they're mostly coming at them with the attitude of the authors knowing more than the reader, & expecting it to automatically be valid & they want to learn it. There isn't as much mention of assessments of whether the study is likely to be any good in terms of methodology & whether the conclusions are warranted, checking whether the data do actually say what the conclusion suggests etc.

There is mention of it, but it seems much less prominent/common.

 

[Andy]

Good question I don't know if you intend for this to be a discussion on this topic, but I'm happy to share what seems to work for me.

This article is much like many others I share, I either think that it is interesting myself or that others might find it interesting, and that, broadly speaking, the subject is of relevance to the purpose of the forum in some way. So if it inspires conversation then great, if not then I'm not too worried.

 

[Arnie Pye]

On the same subject, this might be of interest :

http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.660.6210&rep=rep1&type=pdf

Title : How to read a paper - The basics of evidence-based medicine.

The link is to the second edition which is copyright 2001. It's a free 240-page PDF download.

Several later editions are available on Amazon for eye-watering prices. The latest is the 5th Edition.

Have I read it? I started it but didn't have the stamina to get very far, and now I've forgotten what I read. The book gets a fair number of good reviews (for what is essentially a textbook).