Skip to main content

NS-663 Biobehavioral Principles and Theories

Keyword Searching

Keyword searching is the type of searching you typically do in PubMed, CINAHL, MGH OneSearch, and many other databases. At it's most basic, it looks at the words you type into the search box and finds articles that contain those words in their titles and abstracts.

Finding the Right Keywords

Because you are looking for words that appear in titles and abstracts, it follows that you need to know which words authors are using to talk about your topic. This can present some problems when you are at the beginning of a project, because you may not know much about your topic and the way people are talking about it yet. That's why your first few searches are less about finding the perfect articles and more about reading titles and abstracts so that you can find the right keywords.

As you read through the titles and abstracts, start collecting words you think might be good search terms.

Modifying an Unsuccessful Search

If you aren't happy with the list of articles your search brings back, here are some ways you might think about changing your search.

1. Too many articles / Articles aren't on topic

  • Make your topic more specific. You could add words to describe the
    • population
    • setting
    • treatment or intervention
    • outcome
  • Make your terms more specific. For example
    • "women" becomes "women over 50"
    • "analgesic" becomes "opioid"
    • "recovery" becomes "length of stay"

2. Too few articles

  • Make your topic less specific
    • focus on the core ideas (remove unnecessary details)
    • choose less specific terms ("exercise therapy" becomes "exercise")
  • Look for different search words 
    • synonyms - brainstorm synonyms (e.g. "length of stay" = "hospital stay") or different forms of a word (e.g. therapy, therapies, therapeutic)
    • scan abstracts and subject headings to find out which words authors and databases are using to describe your topic.

Photo by Anant Nath Sharma, used with permission under a Creative Commons license