Pages

30 March 2019

How to Read a Scientific Paper in 5 Steps

by Ashley Hamer

In a perfect world, we'd all have access to the same information, the training to make the right judgments about it, and the wisdom to know how it should be implemented. But the fact is, we don't. Scientific research is often blocked behind paywalls, and even when it's freely available, few people have the ability or patience to wade through its scientific jargon. That's why websites like this one exist: to wade through it for you and put it into language that's easy to understand. But if you really want to be a responsible consumer of information, sometimes you've got to wade through it yourself. And for that, we've got this handy guide. Here's how you can read a scientific paper and actually understand what it means.

First Things First: How to Get Access

That thing we said about paywalls is definitely no joke: More than three-quarters of scholarly papers on the web are only available if you're either affiliated with a university or another institution that can afford a subscription, or you can afford to buy them (usually at about the price of a tank of gas). But there are online tools out there that can help: Google Scholarsearch results will usually indicate if there's a PDF of the paper freely (and legally) available, and Chrome extensions like Unpaywall and Open Access Button will automatically search the web for a free and legal version. There's also the old-fashioned way: Just contact the authors and ask. As Canadian researcher Dr. Holly Witteman noted on Twitter, "If you just email us to ask for our papers, we are allowed to send them to you for free, and we will be genuinely delighted to do so."

Step 1: Skim and Take Notes

Most scientific articles follow the structure known as IMRAD, which stands for Introduction, Methods, Results, and Discussion. Sometimes they go by different names and sometimes there are extra sections, but this is the rough outline that the majority follow. Check out where each of these sections are located and get a feel for what each one is trying to accomplish. If you see terms you don't understand, write them down so you can look them up later. Pro tip: Most acronyms are usually defined the first time they're used in a paper, so a quick control + F search to find that first mention can be helpful.

Check the publish date so you know how fresh this research is, and give the journal title a sniff test — journals like Science and Nature are easy to judge as reputable, but if you've never heard of the publication before, make sure it's not one of the many fake or predatory journals out there. It helps to check the journal's impact factor.

Step 2: Re-read, But Don't Start at the Beginning

At the very top of the paper (and in most cases, even before you get through the paywall), there's an abstract, which summarizes the entire paper. One of the most common mistakes people make with reading scientific papers is to read the abstract and move on with their life. Not only is this a hard-to-read section that gives you a small fraction of the information you're looking for, but it's also biased: It says what the authors think is important. You're reading this article to judge that for yourself, aren't you?

Instead, start with the introduction. The introduction gives the lay of the land, so to speak, and explains what previous researchers have done in this particular area of study. That's important for a layperson who's just dropping in: "Why are they testing X with Z when Y seems like the more obvious route?" "Well, because some researchers in 1998 tried testing X with Y and it didn't work out too well." The introduction is also the most clearly written part of most papers. Cherish it while you can.

Step 3: Check Out the Discussion

All this skipping around might feel like heresy if this were a traditional piece of literature, but scientific papers aren't designed to tell you the dramatic story of their discoveries; they're just designed so that other scientists can repeat the experiment. That means that you're well within your rights to skip ahead and see how the experiment turned out. The discussion section will usually give you a pretty clear picture of the paper's findings, but you'll also be left with a few questions. That's what you want. Write them down, then see if you can answer them in the next step.

Step 4: Get into the Nitty-Gritty

Here's where you read through the results and perhaps the methods section, look up any terms you don't understand, and take a close look at any numbers or measurements used. Here are a few to be mindful of:

Sample Size

You can usually find this in the methods section, either as a phrase like "A total of 100 participants" or after the letter "n," as in "n=100." A larger sample size means it's less likely that the results were just a fluke. Don't dismiss an experiment for having a small sample size, however: Depending on the area of study and the statistical methods used, experiments using smaller sample sizes can be perfectly legitimate. A quick comparison with other similar studies can help here.

The Word "Signficiant"

Like many terms, this one means something totally different in science than it does in casual conversation. If you or I said a result was "significant," we'd mean something like "important" or "noteworthy"; if a scientist writes it, they mean that the result was statistically unlikely to have arisen by random chance — that is, it's probably for real. Statistical significance is usually measured with a "p-value," or a number that measures just how real the result probably was. If p equals less than 0.05 or 0.01 (depending on the field of study) it's considered statistically significant.

Standard Deviation and Confidence Intervals

Nothing in science is said with 100 percent certainty, and a confidence interval(CI) is a measure of just how uncertain things are. If a scientist wanted to take data about a sample of a population and extrapolate it to say things about the whole population, they'd probably be mostly right — with a sliver of doubt. The CI is a frequently misunderstood measure (even many scientists don't understand it), but suffice it to say this number just tells you how accurate their statistical methods were.

To determine the confidence interval, you need the standard deviation, sometimes written as "SD." This number is a little easier to understand: it's just a measure of how spread out the data is. You can think of it like an "average of the average." If the average age of a Curiosity reader is 30, for example, that could include a lot of people who are 29 and 31, or it could include a lot of people in grade school and nursing homes. The standard deviation tells you how tightly the data points are clustered around the average, or how narrow the bell curve is. (If you're curious, 30 percent of Curiosity readers are younger than 25 and the majority are between 25 and 44. Where do you fall?)

Step 5: Check Your Understanding

Once you're done finding the answers to any questions you had and looking up any terms that were unclear, do a final review of what you learned and what you still don't understand. Here's a handy list of questions to ask yourselffrom biology professors Mary Purugganan, Ph.D. and Jan Hewitt, Ph.D. at Rice University:

What specific problem does this research address? Why is it important?
Is the method used a good one? The best one?
What are the specific findings? Am I able to summarize them in one or two sentences?
Are the findings supported by persuasive evidence?
Is there an alternative interpretation of the data that the author did not address?
How are the findings unique/new/unusual or supportive of other work in the field?
How do these results relate to the work I'm interested in? To other work I've read about?
What are some of the specific applications of the ideas presented here? What are some further experiments that would answer remaining questions?

No comments:

Post a Comment