Literature Review, pt. 2


Uploaded by Wesleyan on 17.06.2010

Transcript:
So what do you do in terms of the steps in the literature review process? I’m sure
that you all say that well, you go to the library; you go to the indexes, and you start
searching your brains out. No.
Before you start your literature review, your first step is to create an EndNote library.
I’m going to show you how to do that. This week — maybe today, maybe right after class
— you’re going to run to a lab and do this. That would be great. You’re going
to go to “start programs, EndNote.” It’s one of the ones that they pulled out. Then
you go to “EndNote program,” with that little brown and black square which you’ll
become very familiar with. You click on it, and you’re going to come to a screen like
this.
You have three options: learn about EndNote, create a new library or open an existing library.
The first time that you come in here you’re going to click on “create a new library.”
The next time you’re going to open an existing library. It’s as simple as that. I’m going
to create a new library. This is what you guys are going to do. I clicked on “create
a new library,” and basically it went to My Documents on this computer. You guys don’t
want to save anything on the computer you’re on. Where do you want to save it?
Q: The P drive.
[Dierker]: The P drive, okay. The other thing that I want to note and it’s not at the
right drive, and so I’m going to have to get to the right drive. The second thing is
it’s called “my EndNote library.” I don’t really like that title, and so I probably
will change the title.
I’m going to go to P drive, “QAC201 and students,” and then you guys would go into
your section and find your own folder. My folder is not in a section. It’s right here,
and so I’m going to put it there. I don’t like “EndNote library.” I’m going to
decide that I’m going to rename it. I’m going to call it — my main topic is nicotine
dependence, and so that’s what I want to call it and I say okay. You’re all going
to do that like in the next 24 hours.
Now what I have is an empty library. For now that’s absolutely all that I need, so I’m
going to close it. That’s all I needed to do. I now have a library and it now exists
on my P drive. Now when I go and do my literature review and I want to download the articles
that I’m going to get and read, it’s all ready. I should mention, too, Kendall Hobbs
is here for us. He’s a reference librarian who’s going to help us if we forget anything,
or if he needs to add anything.
I want to hear what kinds of indexes you guys have searched before. We’re going to learn
about Science Citation Index and PsycINFO, just because all of the datasets that we have
pretty much hit everything that you need.
So here’s the first rule of thumb for this particular class that I want to teach you.
I am looking at the association between two things, and so I don’t need to mess around
with keywords. Keywords are pretty general. They’ll get you into things. We want literally
to find articles — I want to — that look at the association between smoking and nicotine
dependence. That’s it. I want to look at the association between those two things.
I want those things in my title. I don’t want it in my abstract or maybe somewhere
in the keywords or whatever. You’re going to waste a lot of your time. You want it in
your title, and so I’m going to say “nicotine dependence and smoking.” All right. Go to
“title” and then search. Now, I’ll tell you how to limit your search in a moment,
but actually it comes up with a certain number of hits that I think are a reasonable number
of hits and so I want to mention it to you.
I’ve come up with 167 hits. Actually, one other thing that I want to do — how do I
get it to go to 50 per screen? Is that down here? Oh, here it is. Okay, I don’t want
to see ten per page. I want to see the maximum per page which is 50. That’s a good idea
to start with.
Now, I’m going to start reading the titles. Yes, you want to read abstracts, but you can
rule out most of the articles that you don’t need to read from reading just the title.
I would say that it would be worth your time to read 167 titles. It’s not going to take
you that long. As I read through the titles, I learn a lot about what I don’t want which
is very helpful. You want to know about what you don’t want, and I learned about what
I do want as well.
A lot of what I’ve found — remember I just wanted articles that looked at the association
between smoking and nicotine dependence. As you see, the first one is genetic epidemiology
of smoking behavior and nicotine dependence. I’m not doing genetic epidemiology, and
I’m not doing genetic work and I don’t have any genetic data so it’s a pretty sure
fire guess that I do not want that article.
This is stuff like failure of hospital employees to comply with smoke-free policies. I’m
not dealing with policy; I’m not dealing with cessation efforts. If you recall, my
interest is actually kids who are just experimenting with smoking so smoking cessation is way out
of the range of what I’m doing. Maternal smoking and pregnancy, that’s not what I’m
interested in and so it’s very easy to read the titles and know that’s not where you’re
going.
Now, I move down here and finally find one at number eleven here. What our title is,
is linking quantity and frequency profiles of cigarette smoking to the presence of nicotine
dependence. That’s exactly it; I want to know how quantity and frequency link to nicotine
dependence. This is a definite hit.
The next thing that I probably want to do — well, it’s not the next thing. I’m
going to click there, actually. Without going to find it, how do you get to the abstract?
Just click it?
[HobbsOB ]: You click the title.
[SLIDE 33]
[Dierker]: Okay, click the title. Maybe that’s the next thing that I’ll do. Okay, there
it is. Like that title is definitely a sure shot, and so I may or may not read that abstract
at that moment. I may wait for the article until I read the abstract, but a lot of titles
you’ll say, “That might be what I’m looking for, but I’m not sure.”
Now, you want to move down to read the abstracts. You don’t want to read every abstract. It’s
only if you have a question about whether or not you want that paper. At this point,
do you want to read the abstract and determine whether this might be something that is good
background for the question that you’re asking. So that was eleven. I definitely want
that. That’s fabulous. Let me click it again because I went away from it. I don’t understand
why they do that. That happens to me all the time. Once I check it and I go to another
screen, it put a check below it but it doesn’t treat it as checked anymore. I mean, it treats
it there because I can see it. Any thoughts — ?
[Hobbs]: It’s in a marked record file, so it’s been marked and when you get to the
next step it has an effect on which copy you use.
[Dierker]: So I’m going to skip through this and go quickly, but let me tell you things
as I read the titles that I knew I wanted to rule out that may help you know what to
rule out. One, I’m not doing genetic work; I don’t want any of those papers. Two, just
because it had both of my titles in the title — both of my topics in the title — oftentimes
you can tell that they weren't looking at the association or the link or the relationship
between the two things, but they were just looking at both things. They used the word
and a lot. And is not what you want. You really want words in the title that suggest that
they looked at the association between the things, so and doesn’t work.
Another thing is that a lot of it was about smoking risk factors. That wasn’t what I
was interested in. I said cessation intervention. There were a couple of studies that I thought
may be linked, but it was really about very heavy smokers. If I have enough in the experimental
smoker group, that is really what I want to focus on. I’m going to go through and click
some more references.
Number 36 was a good one. “The role of psychiatric disorders in the relationship between cigarette
smoking and nicotine dependence.” This is a good example of how doing a literature review
can make you think what other variables you may want to look at. What other variables
may be part of this relationship that you may have access to in your dataset, so this
one does what I want it to do. It looks at the association between smoking and nicotine
dependence, and it looks at how psychiatric disorders may explain part of that association.
Number 40. “Adolescent smoking trajectories in nicotine dependence.” Trajectories of
smoking or just changes or patterns in smoking, so they’re looking at the association. That
looks like a good one.
And then another one on this page would be — that was three. That’s it for that
page. The next thing that I want to do before I move to another page is I found three references.
I want to download those into the EndNote library now and not later.
You want to select selected records on this page. That’s correct. Don’t change that.
Author, title, source plus abstract — that’s correct. Don’t change that. Now you come
over here. Don’t go here: “save EndNote to EndNote Web.” You don’t have EndNote
Web. You want to go here: “save to EndNote Ref Man or Pro Site.” That’s what you
want to click on, so that’s what I’m going to click on. Now, you wait a moment. Then
this comes up. It says to open with — is this going to work? I think that it will.
Look what it did. It opened my EndNote. Isn’t that wonderful?
Now, I have three references in my EndNote library. You’re saying, “Well, that’s
not all the information.” Well, if you double click on those, you see everything so it’s
got the abstract; it’s got the title, and it’s got everything that you need for your
citation. You don’t need to worry about it anymore. You have it, you’ve captured
it and it’s yours. Look, it even has the preview down here. I have the old version.
This is even nicer.
Okay, now we’re done with that. We’re going to hit the “back” button. Now, we’re
going to go forward. I want to see the next 50 references. I have a couple here that are
good too — 52, 54. Those look absolutely perfect, and 71.
This one looks like it’s looking at the association and it talks about gender. Now,
I’m getting more information about my topic and more interesting ways that I might want
to expand my question.
[SLIDE 38]
The reason I’m sort of focused on this is because, I mean, I do the same thing. I had
to train myself to do this class, because I have some bad habits that I don’t want
to teach you guys. My bad habit is what I do is I find the right paper. I’m so excited
and I click “find it”; I get the paper, and then I go and get another paper and another
paper. Then at the end of my session oh, I didn’t download anything to EndNote. Then
I waste my time.
Right after I show you how to download these, then I’ll show you how to go back to checked
articles and how to find them. That was a really good question. You have not downloaded
your article. You have only downloaded the citation.
Anyway, I’ve picked three more from this page which is the end of this search. I do
the same thing; I’m going to put it in EndNote. Now, there’s something important that I’m
going to show you. Look, my three came. Where did my other three go? It overwrote them!
No, it didn’t overwrite them. What it does when it downloads it, it wants to show you
what it downloaded so that you know you downloaded the right thing. So if you downloaded ten
and it comes up with five, you know that something went wrong. This helps you to know whether
you got what you wanted.
As your library becomes bigger, it’s very difficult if it just sort of threw it into
the mix and showed you everything. How would you know whether it did it right or where
it put it? I used to hate this, but now I see the method to their madness. Anyway, you
go to “references, show all references.” Now I have all six.
[Hobbs]: Yes, the newest version. (The tab. Where is the tab?) Yes, it’s just under
your cursor there.
[Dierker]: Oh, I had my cursor on it. “All references.” I just don’t want you to
get worried. People always get worried when they use it and it downloaded and they think
it overwrote it. Especially, when you’ve spent like time and you have fifty references
and you download three and it only shows three. It’s a horrendous feeling. Don’t worry
about it.
Now, I’m doing great. I have six references that are absolutely on target with what I’m
interested in looking at. It’s time to read them. It’s not the time to go and find a
million other things, because this is a procrastination technique that you guys use. I know you. You
say, “Oh, this is easy and this is kind of fun. I like to do fun things and reading
articles sounds kind of stressful, so I’m just going to keep looking for articles and
looking for articles.” You’re not allowed to do that. So if you have six articles, you’ve
got to get in there, get them and read them. That would be good.
Okay, next thing. Now, I’m going to go back and do exactly — what is your first name?
Rachel. Exactly what Rachel pointed out. It’s lovely to have citations, but let’s get
the actual article
Anyway, moving forward you would click “find it.” You guys know how to do that, and then
you wait a while. Great news. Full text article. I can read it today — right now, this minute.
I’m going to click on “go.” I’m going to wait a moment. Now, you can print the article
because this is the whole article. I’m a big fan of PDFs; I really like the way that
the article looks in the journal. Now, you guys are younger than me. You may be just
so comfortable with the digital formats that you don’t care, but I really like that sort
of print. I don’t read it online either. I print it; I sit down with it and I do it
the old-fashioned way.
The next thing, now I’m going to go back and actually I’m going to go back one page
because I want to show you what happens when it isn’t online because sometimes it isn’t.
Anyway, so I’m picking number eleven because I want to find it and start reading it. Oh,
darn. No full text available.
You guys all know what to do? Click on “IOL, my username, my password.” Then look, it
has all the details. This is amazing. This has all the details of the article that I
want, and I just click “submit request.” Isn’t that great?
Anyway, but what I want you to be careful of is this: clicking on IOL doesn’t mean
that a robot goes and finds your article. It means that a human being goes and finds
your article and it takes a lot of time. You guys cannot be wasteful. There are a lot of
you, and there are not a lot of resources and people over at the library.
I would rather you get the really perfect articles. Some of them will actually be ones
that you can get right away. Read those; the ones that you can get right away before you
make decisions about what you’re going to do through IOL
There are a couple of other things that I wanted to mention. Now, yes, once you get
into this, there may be ways that you want to limit your search. You may actually search
something and end up with 500 titles or maybe more, or you may find that there is a lot
of old literature but you don’t really want to focus on the old literature. You want to
get the newer stuff first before you move back and look at the older stuff.
Ways to limit are things like the last five years. Now, you may actually be doing a topic
that people haven’t recently been very interested in, but maybe they were interested in it from
let’s say 1995 to 2001. It doesn’t mean that you can’t go back further, but you
want to get the newest information available on your topic and so you want to start with
the pretty recent stuff.
Some other things. One good thing here is that I would limit by the article. You don’t
want meeting abstracts; you don’t want review articles; you don’t want letters. I would
actually click “refine” and now my 167 goes down to 122, so I was reading a bunch
of things that I didn’t want to read because they were abstracts. You want journal articles.
I suggest that you focus in that way as well.
[Hobbs]: Now, one thing to mention about the review article is that if you’re trying
to do a quick search here, you can do major things. You can ignore those, but a good recent
review article if you’re doing a big thorough search, that can be somebody who’s done
a lot of the work for you of collecting a bunch of citations. So if you’re doing more
extensive things, the review articles can be very helpful.
[Dierker]: The review articles are good, again, for the background information. You don’t
want to cite them specifically, but for background that would be a good place to start.
Here we go. Web of Science. This is back to my main search page, right? Here are some
other ways here that you can subset to particular years here. Let’s say that you’re doing
something that’s clearly a very social science topic, you can actually get rid of all of
these checks and just do the social science. You can do social science and science. With
smoking I would do both, because there could be some in the health sciences and in the
social sciences, so that would be important. You don’t want conference proceedings, so
I do suggest that you subset to that really originally.
Okay, so a couple of other ideas for limiting your review. The past five years. You may
want to limit it by age. If I’m interested in smokers, but I really don’t want smokers
who are 85 — and particularly in PsycINFO to limit by age groups and things like that
— the Science Citation Index isn’t good for more limiting than I’ve just shown you.
Species restriction, the type of research methodology — some of these are easier to
do through PsycINFO than they are through the Science Citation Index. If you find that
in the Science Citation Index you’re just finding lots and lots of stuff and there are
other ways that you think you want to really narrow it down, you might want to move over
to PsycINFO — unless you’re doing caterpillars and then I don’t think you should. You wouldn’t
find what you need.


[SLIDE 53]
Now I’m going to have to go back, but I want to show you my slide because it’s important.
Okay, this is the most important slide today. How do you follow the citation trail? You
do not hang out in Science Citation Index for the next two weeks. You’re going to
spend a few hours in Science Citation Index in the next several days, and then you’re
going to find articles that are really right on target with the kinds of things you think
you’re looking at. You’re going to read them, and you know what you’re going to
find? In their literature reviews and in their introductions they’re going to cite other
really perfect right-on-the-mark kinds of articles.
Once you get a few perfect on-target articles from these indexes, now you’re going to
really utilize the reference sections of the journal articles that you’re reading. Obviously,
now you go in and you search on the whole title. You can narrow it down to the specific
article that you’re interested in getting. That’s the second step.
Hanging out in the index, it just keeps you too off track. Obviously, if somebody is writing
a paper exactly on your topic, they’re going to reference — if it’s a good paper and
it should be — they’re going to reference all of the right on-target papers that have
come before you, and so that’s the next step on your trail. It’s not to hang out
in the indexes forever or to think that you need 50 journal articles from the indexes.
I would say that from the indexes you want five to ten really on-target — hopefully
for some that you can actually get right now and we don’t have to wait for the library
alone; that would be a good step. And then I think from the five to ten that you get
from the indexes, you want to start really finding the absolutely perfect articles, and
then working those reference sections as well.
The next step is really an important one, too. So the next thing you do, because sometimes
you’ll find a perfect article — it’s the article and you find one, right? Basically,
in their introduction they’re saying that there hasn’t been much that’s come before.
This is basically the first article that’s looked at this. You’re going you’ve got
to be kidding me. I need more literature!
Well, maybe that article was published let’s say in 2002. There could be more literature,
but obviously you wouldn’t find that in the reference section. You actually have to
track forward. I don’t know how long we’ve been able to do this, but two years ago is
when I figured out that we could. That is that you can actually search on the title,
so let’s see here. I’m going to get rid of this. I wish that I had my title with me
now. I don’t know if that will find it, but anyway, title. Okay, yes, good. I found
it.
I found the article that I said that this is a great article and it’s exactly the
question I’m asking, and I want to track forward. If you come down here, it says that
the times it was cited after it was published is 25 times. So I click on 25, and do you
know what I find? I find 25 articles that cited that particular paper.
Now, some of them won’t be appropriate. Just because they cited a paper on the association
between smoking and nicotine dependence doesn’t mean, in fact, that it’s going to be looking
at the association between smoking and nicotine dependence. But boy, I’ll tell you, there’s
a higher probability in this group of 25 than in that group of 167 that you’re going to
find articles that are appropriate. That’s how you follow the trail.
We have one semester to do an entire research project. You cannot muck around in the research.
You’ve got to get in; get reading and get moving. I would say, Christian, I’m ready
to commit. If you find five to ten articles on your topic within the indexes and then
you start working the avenues which are to trace back through their reference sections
and forward, you don’t need to go past 25 articles — 20 to 25.
Now, some of you won’t get that far. Some of you will have 15, but please don’t have
15 because you’re lazy. Have 15 because there are only 15 articles that are really,
really appropriate to what you’re looking at. Now, you are going to have a few misfires
where you get an article where you thought it was going to be on topic, but it’s really
not what you thought it was or whatever. There are going to be some of those.
What I’m saying is that 25 articles with five misfires where you really don’t want
that article, but you read it anyway would get you down to about 20. That’s about the
ballpark that you want to be in, but the whole point is not to go and find a random 20 articles
that kind of, sort of or might have gotten to your topic. It’s to find the best 20
articles that absolutely narrow down what you’re doing.