How to read a scientific paper

[Cindy]

I just thought it may be beneficial to some people to have a look at how to read a scientific paper.

I am a professional scientist, author of 70 peer reviewed papers, taught about 30 Honours students and 15 PhD students.

Scientific papers are how scientists communicate and is an essential part of science - if it isn't published it wasn't done, and if it wasn't peer reviewed it is meaningless.

But not all papers are equal. So how do you decide if a paper is 'good' or not? How do you read a paper? How do you decide if the information is important or mundane?

It isn't easy and does need an open mind and a method.

That said it is very important that we, as a community, benefit from what scientists, medical providers etc say about topics that are important to us.

We need the ability to decide if a study is relevant or - just noise.

Scientists contradict each other regularly, that is part of the scientific method.

A hypothesis is an idea to be tested and it should be tested rigorously.

Opinion is just that; what someone thinks.  Proof is different; it is an opinion that has been tested and shown to remain steady under the tested conditions. That proof is no longer valid when conditions have changed.

So how do we read a paper? How do we decide if the information is useful?

This is an article from Huffington Post. I decided to post this rather than information I give my students because one, I think it is good, and second, it is independent of my personal opinion.

Don't be afraid of reading the scientific literature, but please do so with awareness.

Thank You

Cindy

http://www.huffingtonpost.com/jennifer-raff/how-to-read-and-understand-a-scientific-paper_b_5501628.html

[helen2010]

Cindy

Many thanks for this.  I confess that I am guilty of going straight to the abstract and not forming my own view of the research method, results and interpretation.

Aisla

[Tysilio]

Cindy, thanks for posting this. I was a scientist in a former life, and it's sooo hard to get this across to students, much less laypeople.

I wish I thought that more people were willing to spend the time to do this right and had the attention span for it. But if they read this, at least they'll know what they might be missing, and that reading only the abstract isn't enough to form an opinion.

The author fails to mention one thing I consider critical: Once you're clear about what's being investigated, look at the source(s) of funding for the research. That's a huge indicator of whether the authors have an agenda -- especially now that so much research is funded by private entities, whether corporate or non-profit.

She also doesn't mention that even in peer-reviewed journals, 90% of what's published is crap.

[suzifrommd]

Thanks for posting this, Cindy. You've got me trying to figure out how I can fit it into my classes.

Critical thinking skills are the hardest skills (IMO) to teach. That's really what we're talking about. Not believing everything you read, but instead reading PAST what you read to how the conclusion is formed and, based on that, deciding what level of validity and attention it deserves.

I do disagree with one thing the article says. I think reading a report of a scientific finding in the press or media requires even more scrutiny than a scientific paper. Because you're not just judging the merits of the topic but also the reporter/writer's interpretation of it.

Nearly all articles in the press or online I read reviewing scientific papers or announcing findings of a study stand up embarrassingly poorly to critical examination.

[Eris]

Maybe look at some pointers from Dr Ben Goldacre, he does a lot of work examining "bad science" XD

[Jenna Marie]

Thank you for this. I used to work as a librarian trying to teach exactly this ability to college students, so now I admit it's a pet peeve whenever someone thinks a flood of random studies is the same thing as peer-reviewed robust *evidence.*

[Cindy]

OK

A follow on.

There are a few professionals on the site, besides myself, whose daily jobs involve interpretation of the scientific literature, who have volunteered to help me in this (coerced may be more accurate a term!)

If I post or link to a paper or two are there people interested in having a go at reading them, interpreting them and critiquing them? We would then comment on what you have done and give feedback.

I won't bother if there is no interest as it involves some work, but at the same time it is an opportunity to learn something from experts that to be honest would normally cost considerable tuition fees.

But maybe people are happy to just accept what is in print on the face of it?

[V M]

Is it okay if a guitar playing mad scientist writes a peculiar song about it?    Sorry Sis, I'm terrible... Weee Hee Hee

*Runs back to hide under her rock*

[Cindy]

One of the best scientist I know often plays a song at the end of his talks.

It is a great way of summarising work.

You play Slash on 'Sweet child of mine'  and I'll do my Axel impression!

Love you Sis, particularly when you are bad 

[Rachel]

Thanks Cindy,

I am guilty of reading abstracts. When I read the whole paper I usually do not employ critical thinking. My purpose most of the time is to gain one or two insights I did not have prior. Sometimes I get a lot for the time invested.

Many papers charge to read the entire paper which is a real turn-off.

[Dee Marshall]

I had a whole class on this in grad school. Best course in the degree. I'm rusty, but I'm in.

[luna nyan]

There are a few professionals on the site, besides myself, whose daily jobs involve interpretation of the scientific literature, who have volunteered to help me in this (coerced may be more accurate a term!)

Hahah!  You punted me back into an undergrad! 

If I post or link to a paper or two are there people interested in having a go at reading them, interpreting them and critiquing them? We would then comment on what you have done and give feedback.

I'll help as I can.  I have to say, I tend to get bombarded by studies made by commercial interests, so it'd be good practice for me anyway.

Edit: thread bookmarked.

[Taka]

i love reading papers!

though i mostly read abstracts, since the times when i've needed to read papers have been when there were papers i had to write for some course. read abstract, dismiss read abstract, dismiss. and out of the few papers that seem to have anything relevant to say, they still usually don't get into what i'm looking for.

i really hate it when someone links an article, and it turns out i can only read the abstract. it's not like i can trust any of that, though it can help me find out if i might have any interest at all in reading the entire paper.

the super course they teach at ntnu (where i studied), is in writing a scientific article. that course really teaches how to write things right, and find good sources. nobody could ever come out of it without learning how much is needed for an article to hold water. i didn't finish it though. mostly because i hate writing, and i hate deadlines even more. but knowing how to write really makes the good grades easier to get.

There are a few professionals on the site, besides myself, whose daily jobs involve interpretation of the scientific literature, who have volunteered to help me in this (coerced may be more accurate a term!)

If I post or link to a paper or two are there people interested in having a go at reading them, interpreting them and critiquing them? We would then comment on what you have done and give feedback.

would this be beneficial to you in any way?
i'd love to try, but i've never tried interpreting and critiquing papers before, i think. closest i've gotten is presenting a few papers on different linguistics topics to my class. i should probably get some more training before i start aiming for that master's degree. i do want an a at least on average, after all.

[luna nyan]

I've finally had time to read the link. It's a nice little intro to reading papers.  I wish someone had told me this when I was an undergrad - I had to figure all that out for myself! 

I have to say most of the articles I read are not immediately relevant to me - a lot of persnickety research that one hopes will eventually be useful...

[barbie]

Cindy,

Thanks for the link. I shared it in facebook.

I should add "That is not what I typically do." I always read the abstract first.

barbie~~