This page explores some strategies you can use to ensure you are getting highly relevant, thorough search results. These strategies will work, with minor adjustments, in many library databases. If a strategy can only be used in some search tools, that will be noted.
The majority of search tools make use of controlled vocabulary: pre-set terms applied to items in that database, which describe what that thing is about. These are often called "subject headings" or "subjects" and function like tags to link items on a similar topic together, regardless of the words used in the title or abstract. In most search tools, clicking on a controlled vocabulary term will bring up a list of all items that are labeled with that term.
The tricky part about successfully utilizing controlled vocabulary is that you have to know exactly which words were used in the controlled vocabulary term. This can be quite different from the language we use everyday!
Different tools may also make use of completely different sets of controlled vocabulary. For example, a general purpose or multidisciplinary database is likely to use different, typically less technical, terms compared to an engineering database.
To find out what controlled vocabulary a particular tool uses for your topic, look for the thesaurus. The thesaurus is a list, often searchable, of all the controlled vocabulary terms used by that search tool. By using the official terms from the thesaurus, you are communicating with the computer in its own language (so to speak) and it will have a better chance of giving you relevant search results.
Finding the thesaurus varies depending on which database or index you're using, and not all of them will have a thesaurus. Look for a link near the top or near the search box that says "Thesaurus", or check the Advanced Search page. These are the places you'll most often find a thesaurus link.
Even Google Scholar provides advanced search options, if you know where to look. Exactly what is meant by "advanced search" differs from database to database, but generally these are tools that enable you to be much more specific about what and how you are searching, as well as to apply useful limits to your search results.
Some common features of advanced search interfaces are:
Many search tools now also provide these options as filters after a search has been run. Advanced search can help you retrieve the most relevant results the first time around by pre-limiting just to what you are interested in.
Additionally, many subject-specific databases have extra advanced search features designed to make searching easier for researchers in that field. For example, the chemistry database SciFinder allows searches by chemical structure drawing and includes other tools that are highly specific to chemistry and materials research.
Citation chaining is a search strategy that leverages articles you have already found in order to discover other relevant articles.
A citation chain shows the relationship between a specific article (Smith 2002) and the research that cites, or is cited by that article. Research doesn't take place in a vacuum–researchers build on and expand on previous knowledge. This knowledge is interconnected through a citation chain like the one shown below:
As you can see from the diagram, Smith (2002) cited some previous research, and also has been cited by others who are building on that work. Notice that the relationship isn't strictly linear, it's an interconnected web of research.
Citation Chain Diagram from Undergraduate Science Librarian's Diagram of the Citation Chain.
Hidden in the articles you've found so far is a wealth of information! The bibliography is a roadmap that shows who our author (Smith 2002) thinks is important in the development of their research. By looking at the references in a bibliography, you can explore the context in which Smith's research takes place and the research that Smith was building upon. Finding articles in a bibliography is a quick way to find more research on your topic. You can then check the bibliographies of those articles and (perhaps) find more relevant research.
To follow up on items from a bibliography, you'll need to get copies of the articles. You can search for them using a relevant subject-specific database, or a multidisciplinary one like Google Scholar. (Pro tip: Google Scholar is often the fastest way to get to a known item, such as something cited in a bibliography.)
You'll also want to follow those citations forward in time using cited reference searching.
Cited reference searching is the ability to search the list of references (or footnotes) found in journal articles, books, dissertations, websites, etc. It is based on the premise that you have a scholarly work in-hand that you really like and you want to see who else has used that work in their research–ergo, have included it in their list of references (or footnotes). Typically, cited reference searching involves looking for works by a particular author or for a specific item.
What you're doing with a cited reference search is following an article forward in time and answering the question: "Who has cited this article?" This is an important component of a thorough literature search. Looking at the authors who have used an article can give you valuable information about how that research has been used by other researchers.
Not every database can do cited reference searching, and this strategy is not guaranteed to find all literature that cites a particular item, but it will quickly generate a list of newer, potentially useful articles.
Some modern search tools now also include analytics features to help researchers make sense of their search results. These tools can help you:
Understanding what was captured by your search results—and brainstorming about what might have been missed—can help you make sure that you're catching all of the relevant literature. For example, a sharp decline in the number of articles published on a particular topic could indicate that there was a decline in the amount of research being done (or published) in that area, but it could also be an indication that there was a change in terminology and a new search including more modern keywords or controlled vocabulary might turn up additional articles.
See the links below for some databases with analytics features: