As you may remember, a little while back, I asked you what your biggest struggle was.
Today, I’m going to try and tackle one of the most popular topics that were mentioned:
There’s way too much on my topic! How can I narrow it down so that it’s manageable?
First, let me preface this by saying that this is really a question you should talk about with your tutor / teacher / adviser if you have one. They might be able to stir you in a more appropriate direction or help you focus on the most important point.
That said, if you’re on your own or want to have a go at it before getting your adviser involved, here is my professional opinion, as a librarian.
From a very technical point of view, there are several tools you can use to narrow down a search that ended turning up way too many results.
Let’s go over each of them and see how you can apply their principles to help you narrow down your topic.
1. Making sure your search terms only turn up relevant results
Try this out: think of a random topic, type your keywords in the search box of your favourite database, and have a look at the first few pages of results that come up.
Are all of those results actually relevant to the topic you had in mind?
Or do you find some stray results that have nothing to do with it?
Example: I want to find documents about Sith lords in the universe of Star Wars.
I type in “sith” and do a search in my library’s database.
I end up finding interesting results actually on topic…
… as well as documents about Static Induction Thyristor (or SiTH) used in physics.
Tightening your search equation
To eliminate those stray results from your search, you’ve got several options.
The Boolean Operator NOT
You could start by using the Boolean “NOT”.
NOT will help you eliminate unwanted meanings from your results. You use it by putting behind it keywords that you want to exclude from your search.
i.e. the search equation Hypnotism NOT Magic will help you find articles about the therapeutic uses of hypnotism by eliminating articles about the use of hypnotism in magic shows.
If one of your search terms is a phrase composed of several words, you can use quotation marks.
By putting quotation marks around an expression, you make sure that the results will return the exact phrase (i.e. “Harry Potter”) rather than the words of the phrase being used in the same documents but not side by side (i.e. “Harry Smith was a potter by trade…”).
The gist of it
As you can see, my instinct as a librarian would be to try and eliminate the noise surrounding your search: we want to get rid of all the results that are not actually relevant to your topic.
This is a good test to see if it’s really your topic that’s too wide… or if it’s just your search method that was not refined enough!
If you try this out but still feel like you’re drowning under the sheer mass of the literature, then go to the next step.
2. Focusing on a smaller part of your topic
If you’ve got a truly large topic, then you will find that it is very often composed of sub-topics.
Sometimes, those sub-topics are very obvious. But if they’re not, go ahead and read the titles and abstracts of the first few pages of your search results. By combing through them carefully, you should be able to see subdivisions appear.
This is also something that your supervisor should be able to help you with quite easily.
For example, I did a simple search on “shift work” and found out that results could be divided into several broad categories: physical health impact, performance implications, mental health impact, etc.
Each category could also be divided further. For example, relating to the physical health impact of shift work, I found many articles about the workers’ sleep, the fatigue experienced, their susceptibility to various illnesses, etc.
Refining your topic
In order to focus your literature search on one of the sub-topics you’ve identified, there are two tools I would recommend for you to use.
The Boolean Operator AND
An instinctive way to narrow down a search to a sub-topic could be to simply add extra keywords to your original search and linking them to the rest with the Boolean “AND”.
What this does is that you’re asking the database to now return documents that contain both keyword1 AND keyword2.
i.e. if I want articles that are about hypnotism as used for dental surgery on children I could write: hypnotism AND “dental surgery” AND children.
You can add as many keywords linked by AND as you want, in order to narrow down your search further and further. But I would recommend to first check that each of your keywords return a reasonable amount of results (like, at the very least 500 results for a search comprising just one individual keyword – but a few millions would be better!) in order to avoid your search to return no results at all.
For example, the search “shift work” AND “psychological impact” AND nurses returns no results.
To test it out, I do one search per keyword.
A search for just “shift work” returns 5,000 results.
Another independent search on “psychological impact” returns 49 results.
A final search on nurses returns 12,000 results.
When I string all of those keywords together with the Boolean AND, the database is going to search for documents that have “shift work”, “psychological impact” and “nurses” in them at the same time.
But we know that the database only has 49 documents with the exact words “psychological impact” in them.
What are the chances that within those 49 documents, some of them will also have both the keywords “shift work” or “nurses” in them? Well, very low. And that’s why we ended up with no results for that equation.
How to repair this?
Well, now you know that the term “psychological impact” is the problem. I would try and change it for something that returns more results. Try things out in the database. Maybe a using a truncation like in the equation psycholog* AND impact would work better?
Another possibility would be to change database. Maybe the one you’re using now is too focused on another subject area and doesn’t contain any psychological documents in it? Explore your other options with the help of your librarian.
You can also work it in steps: add an extra keyword, check how many results you get. If you still have a reasonable number of results (more than you could have a look at in a reasonable amount of time, for example, more than 100 – ymmv) add another extra keyword, check your number of results, etc.
If you’re really miffed about the kind of subdivision you could use within your topic, you have also the possibility of using artificial subdivisions that have been inputted by well-meaning librarians.
I’m talking of using limiters.
The idea is that all the documents in the database have been indexed according to some specific criteria (which are going to vary depending on which database you’re using).
You can use those built-in criterias to help you narrow down your search.
Classic limiters will include:
- Subject / classification (here is a perfect list of sub-topics that span your current results — just pick one and see what happens!);
- Publication date (could you only focus on the most recent research on the topic?);
- Full-text access (if your literature review is just a stand-alone exercise, it might make sense for you to limit your search to documents that you can actually read right now. But beware, if you’re supposed to be doing a thorough search on your topic, then you can’t use this one: you’re supposed to try and find other ways to access the full-text!); …
But, most importantly, check out the specific limiters that are proposed in the specific database you’re using – they often will be the most useful in your context.
- PubMed will allow you to limit by species — could you limit your search on this medication to tests that have been done on humans only?
- PsycINFO allows you to limit your search by age and gender of the population studied — could you limit yourself to studying autism in school-aged girls only?
- ERIC can limit your search by audience — could you limit your search to school administrators only? Or counselors? Or policymakers? …
All in all, it can really pay to take a bit of time and analyse which kind of limitations your database allows for.
Those limiters will vary from database to database (and their scope will vary depending on your first pool of results), so always take some time to check them out when you start using a new database.
How to apply those techniques to your topic
Try and identify as many sub-topics and limitations as possible within your own topic.
Is there one specific aspect that you could limit yourself to? Once again, this is something that your supervisor will be able to weigh in on more effectively than a librarian!
If you’re not really sure, you can try things out and see how they go (this is a technique that is really efficient if you’re still at the beginning of your academic career and only doing this for an exercise of some sort…).
- List all of your possible sub-topics / limiters.
- Limit your search to one of them, using the tools discussed above.
- Have a look at the results you found.
- Do you end up with still way too many results? Then try and limit a little further to get to a sub-sub-topic.
- Do you end up with very few results? You’re too narrow! Try and widen your search a bit? Or try another of your sub-topics instead.
- Continue until you end up with a number of results that you feel comfortable with and a sub-topic that feels right.
By following this method, you end up tailoring your topic to the kind of results existing in the database. It might seem a bit backwards, but it’s totally fair if you’re doing this for an exercise and you weren’t completely set on a specific topic ahead of time (like you would be if you were undertaking a specific research project).
But what if you’re stuck with a very wide topic that you need to master? Let’s have a look at this dilemma in the next section.
3. What if you’re still doomed with an enormous load of results?
Divide and conquer
In those cases, you can use the “sub-topic” system shown above to try and make things more manageable.
Here, you don’t want to eliminate huge parts of your topic, but to divide and conquer.
Here is the technique I suggest:
- Do a literature search on sub-topic 1.
- Read all relevant results and summarize them.
- Write a mini-literature review for the sub-topic.
- Go to your next sub-topic and repeat.
The idea is to focus on one small part of the work at a time, analyse that part of the literature, and produce a little bit of end result.
You will probably have to re-work each mini-literature review when incorporating them together, but at least you shall have a good starting point!
Work in waves
If even your sub-topic seems insurmountably wide, then I’ve got another technique to help you go through it bit by bit… But I’ll tell you about that one next week since it’s a tad complex and this article is already super long. Stay tuned!
Have you ever encountered problems trying to narrow down a research topic? How did you manage? Tell us all about it in the comments!