When conducting a literature review, it is imperative to brainstorm a list of keywords related to your topic. Examining the titles, abstracts, and author-provided keywords of pertinent literature is a great starting point.
Things to keep in mind:
One way to visually organize your thoughts is to create a table where each column represents one concept in your research question. For example, if your research question is...
Does social media play a role in the number of eating disorder diagnoses in college-aged women?
...then your table might look something like this:
Social media | Eating disorders | College-aged women |
---|---|---|
Bulimia (nervosa) | Young women | |
Anorexia (nervosa) | Females | |
Binge eating (disorder) | Young adults [refine by women] | |
TikTok |
Orthorexia (nervosa) |
College students [refine by women] |
Generative AI tools, such as chatbots, are actually quite helpful at this stage when it comes to brainstorming synonyms and other related terms. You can also look at author-provided keywords from benchmark articles (key papers related to your topic), databases' controlled vocabularies, or do a preliminary search and look through abstracts from relevant papers.
Generative AI tools: ChatGPT, Google Gemini (formerly Bard), Claude, Microsoft Copilot
For more information on how to incorporate AI tools into your research, check out the section on AI Tools.
Boolean searching yields more effective and precise search results. Boolean operators include AND, OR, and NOT. These are logic-based words that help search engines narrow down or broaden search results.
Using the Operators
AND
The Boolean operator AND tells a search engine that you want to find information about two (or more) search terms. For example, sustainability AND plastics. This will narrow down your search results because the search engine will only bring back results that include both search terms.
OR
The Boolean operator OR tells the search engine that you want to find information about either search term you've entered. For example, sustainability OR plastics. This will broaden your search results because the search engine will bring back any results that have either search term in them.
NOT
The Boolean operator NOT tells the search engine that you want to find information about the first search term, but nothing about the second. For example, sustainability NOT plastics. This will narrow down your research results because the search engine will bring back only resources about the first search term (sustainability), but exclude any resources that include the second search term (plastics).
Some databases offer a thesaurus, controlled vocabulary, or list of available subject headings that are assigned to each of its records, either by an indexer or by the original author. The use of controlled vocabularies is a highly effective, efficient, and deliberate way of comprehensively discovering the material within a field of study.
Web of Science's Core Collection offers a list of subject categories that are searchable by the Web of Science Categories field.
When in doubt, reach out to a librarian! We are highly skilled in effective searching and creating search strategies, from simple keyword searches to several-paragraph-long search strings.
Reach out to a Duke University Libraries librarian at asklib@duke.edu or use the chat function.
While not essential for traditional literature reviews, documenting your search can help you:
Documenting your search will help you stay organized and save time when tweaking your search strategy. This is a critical step for rigorous review papers, such as systematic reviews.
One of the easiest ways to document your search strategy is to use a table like this:
Database | Search Strategy | Results |
---|---|---|
Web of Science | (formaldehyde OR "dibutyl phthalate" OR acetone OR toluene) AND exposure AND ("beauty salon*" OR "nail salon*") |
44 results 12/18/2023 |
Rarely will you construct a search strategy that yields appropriate and comprehensive results in one try. Literature searching is an iterative process.
If you find that you're receiving too many results, try the following tips:
On the other hand, if you're not receiving enough results:
Grey (or gray) literature refers to research materials and publications that are not commercially published or widely distributed through traditional academic channels. If you are tasked with doing an intensive type of review or evidence synthesis, or you are involved in research related to policy-making, you will likely want to include searching for grey literature. This type of literature includes:
For more information on grey literature, please see our Grey Literature guide.
For more in-depth information related to grey literature searching in medicine, please visit Duke Medical Center Library's guide.
Please see the Data Sets and Collections page from our Statistical Sciences guide.