I wrote this blog post as an assist for those who are reading scientific literature for the first time. Unlike books, papers can be a little daunting to find and digest. So, I wrote this guide among a few others, and I plan on doing a few more and some example reviews.

Let me know if y'all find this useful, or if there is something that isn't clear. I've posted a separate blog post on where to find papers and on peer review, so consider this just another step in finding and analyzing real scientific data.

Page with formatting:
http://primaltribes.com/content/primal-science-how-read-scientific-literature

Primal Science: How to Read Scientific Literature

Literature, it's is not like other science writing!

    For some, "science literature" doesn't come out as a self-defining word. Let me explain what it is by comparing it to other science writing:

1) In a lab or field, raw data is collected. This is the most basic level of all research, and is (often unintelligibly) recorded in lab notebooks, field journals, and computer files. It's the closest to the source but isn't published and is largely impossible to read, since the researcher wrote it for himself 99% of the time.

2) This raw data is written into an intelligible form that is readable to the scientific community (scientists, professors, bureaucrats in the big funding agencies), and is submitted to journals for archiving. It is published in a journal only if it passes the rigors of peer review. This what we mean by scientific literature, and it is the best accessible content from a scientific standpoint.

3) These papers are then collected and written about as a whole in book-like "literature reviews." Sometimes these reviews are also published in journals, and in other cases they are written as actual books. This content tends to be pretty good in terms of real science, and does the hard work of finding papers.

4) Alternatively, if the content of a paper is "sexy," the lab will write up a press release and put it on the internet.

5) Content-hungry magazines (Scientific American) and news agencies (Discovery) bypass the good content (2 and 3), and go directly for the easy-read press releases. They take the "sexy" content, dress it up, airbrush it, and then use a little creative license to make the content sell. You can guess what I'm about to say about its reliability.

    If you haven't read scientific writing before, it can be a bit overwhelming. Unlike articles in Scientific American or in your local paper, scientific literature has to pass peer review, and as such has to be written so that it is scientifically bulletproof. Because of this, the relatively small claims made in each paper can often be buried in several pages of background literature and data. Fortunately, there is a way to cut through the verbiage and find the meat you really want.
 
A section-by-section look at a paper

    Each field does its research papers a little differently, with different names. I'm a chemical engineer by field, and ChemE/Chemistry journals tend to use a abstract-intro-methods-results-discussion-conclusions format. Meanwhile, the paper I'm reviewing out of the American Journal of Clinical Nutrition uses a abstract-intro-methods/subjects-results-discussion format. It's a small difference overall, but to me it looks quite different. In general though, you will get a format that tends to look like this:
 
Abstract

    The abstract is a summary of the entire paper, meant to fit into 1/4 to 1/3 of a page. Read this first to see if the paper in front of you is really relevant to what you are looking for.

Introduction/Background/Objective

    This opening section of the paper describes the problem that needs to be solved, and all work that has been performed in previous research. It's acutally not where you start reading in my opinion, but it is a good section for finding other papers as this section tends to have a lot of citations of good papers.

Experimental Methods

    This is the "how to" section, and if you're a researcher, it's how your benchmark your work against the rest of the field. If you are just looking to determine the paper's quality, this is a great area to check whether the researcher did experiments that would generate meaningful results (example: did they separate grass-fed and grain-fed meat?).

Results

    This is where most of the meat is. Data organized into an intelligible form, which is what you really wanted.

Discussion

    The discussion section is where the paper's author interprets the results. Poor scientific logic would be to just accept the claims here; double-check and critique everything here.

Conclusions

     At the end of it all, the author re-hashes his discussion. Usually, it's the conclusions that are cited in news articles, and they are meaningless if the experimental and discussion sections were not good.

Finding the "meat" in the paper

    When I look through a stack of downloaded papers, the way I decide which ones to keep is to hit the abstract, methods, pictures in results, and conclusions. This is because I'm mainly looking for whether or not the paper has something interesting I can apply to my own research. Since most Primals and Paleo eaters aren't looking to do research so much, I would recommend reading the abstract and results first, and then fact-checking by next reading the methods and discussion section.

    A good rule of thumb is that the meat is in the results. This is where the raw data is made intelligible, and it's what the rest of the paper is really based on. And, if you just want a quick "does X cause Y and to what extent?" this section of the paper will tell you what you want to know. After this, the discussion section gives you a few ideas on how to interpret the data, though it's important to fact-check the author as bias is always a problem in science writing (and it's not always intentional).

    Even though the meat is in the results and discussion, there is a lot of value in the other sections. The methods section tells you if the research was done right or if there's a hole in the work. The introduction section is good for citations. By citations, I mean that previous good work in the field is the basis for the paper you're reading now, and in the introduction the author cites that good work. Since most papers use a number system, it's easy to find good papers by following the intro section to the footnotes. If you are trying to pick apart a vegan paper, it's also a good way to start figuring out where the bad science is coming from.
 
Literature reviews at PrimalTribes

    If you would like to suggest a paper to our research team, or contribute to the literature research that's taking place here at PrimalTribes, please be sure to check out The Research Team.
 
Until next time, have fun tackling some papers, and stay Primal!