What's the correct way to analyze research articles?

[allcle4r]

Many of you have probably read research articles for the labs you work for. I'm hoping to do research for a lab next semester and the PI gave me two older papers to look at.

- What's the right way to "take notes" on a research paper? I've got a paper open and a Word document open. What are the sections of that document, so to speak?
- When trying to demonstrate that you understand a paper, what key points should you address/summarize? Similarly, how do you know which types of questions, or extensions on the hypotheses of the paper, are relevant?

Thanks!

---

Members don't see this ad.

 

[mcloaf]

Many of you have probably read research articles for the labs you work for. I'm hoping to do research for a lab next semester and the PI gave me two older papers to look at.

- What's the right way to "take notes" on a research paper? I've got a paper open and a Word document open. What are the sections of that document, so to speak?
- When trying to demonstrate that you understand a paper, what key points should you address/summarize? Similarly, how do you know which types of questions, or extensions on the hypotheses of the paper, are relevant?

Thanks!

I don't even know how to approach these questions... 😕

You read it, look up stuff you don't know, figure out what's going on...and understand it. Basically figure out their hypothesis and what they did to confirm/disprove that idea and why they think you should care about it. Picking out what's relevant for further study would be hard to do without background in the subject and a knowledge of the literature.

 

[Drrrrrr. Celty]

Read the abstract, figures, and conclusion. The majority of the information is condensed in the abstract and figures, conclusion tells you significant or not.

 

[Musty Goose]

Usually I just read abstract, last part of the intro and discussion. And whatever I don't understand I look it up.

 

[Bangersandmash]

I don't know what you think research articles are like, but they are like reading any textbook or novel for that matter, albeit with different subject material. I personally don't take notes on articles I read (that's what the abstract is for) but you can take notes on an article like you would a textbook.

A basic manuscript will be as follows:

Abstract
Intro (basic intro on the topic discussed, followed by in depth info on the projects specifics, followed by a hypothesis and what it hopes to achieve)
Methods (subsections vary depending on the kind of research. Can include data analysis, a description of the patient population, types of tests used etc)
Results (usually a statistical reporting. Figures/charts should show important results, usually summed up nicely in words)
Discussion (What it means, confirmation or rejection of hypothesis, limitations on research, further direction)

Essentially, if you don't know anything about what you're reading, you probably won't understand much from methods (methods is the least important). Intro will give you good background, but it still might be too specific for your knowledge base. Results are obviously important, but if you don't understand what they're trying to find or you don't know how to read the results, it can be tricky. Look for an explanation in words or easily explained charts. I think the discussion is most important. If you can't glean your own explanations from the rest of the article, this will guide you in determining if they confirmed their hypothesis, where they can go from here and flaws in the study.

Of course, if you're knowledgable about the subject, the abstract will do just fine to get the gist.

PM me if you have anymore questions. I've been in a clinical lab for 2 years that is largely research review and statistical analysis.

 

[vayntraubinator]

enjoy:

http://www.bmj.com/about-bmj/resources-readers/publications/how-read-paper

 

[911 Turbo]

Many of you have probably read research articles for the labs you work for. I'm hoping to do research for a lab next semester and the PI gave me two older papers to look at.

- What's the right way to "take notes" on a research paper? I've got a paper open and a Word document open. What are the sections of that document, so to speak?
- When trying to demonstrate that you understand a paper, what key points should you address/summarize? Similarly, how do you know which types of questions, or extensions on the hypotheses of the paper, are relevant?

Thanks!

As a new undergraduate researcher you aren't going to understand 80% of whats in that paper. Just read the abstract and methods part, and you will be set. 👍

 

[lobo.solo]

Read the abstract, figures, and conclusion. The majority of the information is condensed in the abstract and figures, conclusion tells you significant or not.

This^. Read abstract and conclusion if you want more details about a section read that section.

 

[Drrrrrr. Celty]

As a new undergraduate researcher you aren't going to understand 80% of whats in that paper. Just read the abstract and methods part, and you will be set. 👍

Methods section generally is the most useless part of a paper as it is all standardized. And frankly you're not going to understand half the papers that are written unless you've taken graduate level courses in that subject. Ex. I've always thought I was good at electrophysiology, read a paper on cardiac electrophysiology and frankly had no idea why they did anything they did, why any of it was significant, or how it related to their premise. I mean how does AP duration have anything to do with abnormal heart beat.....

 

[red7]

If I was completely new to a field, depending on how much time I had, I would hit the following sections (ranked from #1 Always to #3 If there is time):

1. Definitely go over results/figures (and also discussion/conclusion if you aren't really sure of the significance of the results).

2. If you're just starting in the lab, then you probably don't know a whole lot about the specific topic you're studying. Normally, I wouldn't read the intro/background, but when you're new it can help provide some context for the study and point you towards relevant reference papers which might give a nice overview of the current knowledge in the field (check over the titles of papers they reference in the intro/background section to see if anything is a broad review paper).

3. Materials/methods is a very important section, but right now you're not really in a position to evaluate whether or not a study has been conducted properly. If you have enough time, try to get an idea of different assays/experiments that are used in this field of research. For now, don't worry so much about super specific details - even if you're conducting your own experiments, you'll be given your own protocol. Just try to get the broad strokes of the scientific principles behind different assays.

And ask lots of questions. If you're in a lab with helpful students, they should be happy to give you some pointers. Obviously, you'll figure out pretty quickly who is a willing guide and who needs to be left alone. Good luck!

Edit: Oh and read the abstract. Always read the abstract

 

[911 Turbo]

Methods section generally is the most useless part of a paper as it is all standardized. And frankly you're not going to understand half the papers that are written unless you've taken graduate level courses in that subject. Ex. I've always thought I was good at electrophysiology, read a paper on cardiac electrophysiology and frankly had no idea why they did anything they did, why any of it was significant, or how it related to their premise. I mean how does AP duration have anything to do with abnormal heart beat.....

Lol?

OP will be familiar with lab techniques. My PI told me to pay most attention to methods section. Just because you didn't understand what was going on behind the scenes didn't mean that you didn't know how to physically do the experiment, did it?

 

[IslandStyle808]

Lol?

OP will be familiar with lab techniques. My PI told me to pay most attention to methods section. Just because you didn't understand what was going on behind the scenes didn't mean that you didn't know how to physically do the experiment, did it?

👍

Good job you deserve a cookie.

Plus the material and methods sections will give you an idea if the experiments were legit or bogus. However, first and foremost understand the abstract and then delve into the details. You will be looking up a lot of stuff OP but that is all apart of research.

 

[Narmerguy]

👍

Good job you deserve a cookie.

Plus the material and methods sections will give you an idea if the experiments were legit or bogus. However, first and foremost understand the abstract and then delve into the details. You will be looking up a lot of stuff OP but that is all apart of research.

Eh, I would disagree. Few undergraduates are in a position to make this distinction and with significantly less research time to devote than grad students/postdocs/professors, it's not really high yield for most papers. I would only read materials/methods if you understand the field very well and are extremely interested in determining whether the experimental design is airtight. Most times, especially as an undergrad, you're best served just assuming that peer-review insured a reasonable experimental design. As you get more experienced in research fields you'll get to learn and appreciate the "hot" areas under debate, the common pitfalls & technicalities that support or invalidate the use of specific techniques, and the prevailing wisdom & culture in the discipline with respect to how much rigor and specificity is needed with certain kinds of data/phenomena.

Long story short, I would say when I read a paper I try to make sure I can answer a few questions:
1) What is the problem that this paper feels it is addressing?
2) What is the "big picture" for this problem/why is it relevant? (when a paper says its trying to elucidate mechanisms of shutting off cell replication signalling, why would someone want to do this? Probably cancer related...)
3) What were the results this paper came up with (try to think of 3 distinct results that are key)
4) What did the paper list as caveats/problems in interpreting their results.
5) What might a future study need to address?

I just whipped this up, generally the nature of the paper will make some questions irrelevant while making others that I haven't listed here relevant. Make your own questions that consistently work for you. Usually the abstract/intro/figures/discussion should be sufficient to answer these questions. I personally think the intro and discussion are great starting points when you're new to research, because even the figures can be hard to interpret without experience.

When I take notes, these are often the points I try to make for myself. I've been doing research a bit longer now so I now include many more questions that are relevant for my situation, but as a burgeoning research student, if you can identify those sort of questions about a paper and answer them, you've done a pretty good job.

 

[45408]

Lol?

OP will be familiar with lab techniques. My PI told me to pay most attention to methods section. Just because you didn't understand what was going on behind the scenes didn't mean that you didn't know how to physically do the experiment, did it?

Reading the methods is only useful if you've got a pretty good handle on the topic. I don't remember much of anything about bench techniques, so all those details would be lost on me. I'm doing a clinical research project though, and I always read the methods of other similar papers to see if they did it the right way or if they did something sneaky that would have a significant impact on their results (like excluding patients that you would want to include).

Eh, I would disagree. Few undergraduates are in a position to make this distinction and with significantly less research time to devote than grad students/postdocs/professors, it's not really high yield for most papers. I would only read materials/methods if you understand the field very well and are extremely interested in determining whether the experimental design is airtight. Most times, especially as an undergrad, you're best served just assuming that peer-review insured a reasonable experimental design. As you get more experienced in research fields you'll get to learn and appreciate the "hot" areas under debate, the common pitfalls & technicalities that support or invalidate the use of specific techniques, and the prevailing wisdom & culture in the discipline with respect to how much rigor and specificity is needed with certain kinds of data/phenomena.

Agreed. 👍 Until you have that background, which takes a while to acquire, you're not going to be able to read between the lines, which is really what matters.