Building a Search Strategy
- Create a reference set of 5-10 articles you would like to include in your review.
- List possible key concepts.
- Try to come up with preliminary search terms for each concept.
- Compare your 5-10 articles with the titles and abstracts that show up.
- Combine your key concepts into one search query then assess the scope of your search query by looking at the number of results and their relevancy.
- If you are not seeing what you are expecting, you may need to adjust the scope of your search or your search terms
- If you see what you are expecting, spend more time developing relevant search terms.
Adapted from modules created by Janneke Staaks from the University of Amsterdam
https://osf.io/ga2du
Subject Headings versus Keywords
Subject headings are keywords that are added to articles within a database. Subject heading serve as an aid in the retrieval of results by tagging articles on a given topic with the same heading, regardless of the exact words that are used.
Databases like APA PsycINFO, CINAHL, Embase, and Medline via PubMed have their own list of subject headings. Please note that a subject heading may vary between databases.
A well known kind of subject heading is a MeSH term. MeSH is an abbreviation for Medical Subject Heading from the database, Medline via PubMed. MeSH headings enable PubMed to establish a persistent term to stand for a concept with many synonyms. Some articles also may have no abstract and unless concepts are in the title, we can only retrieve them by MeSH searching.
Please be aware that not all databases will use the same subject headings. For example, CINAHL uses Medication Compliance while PubMed uses Medication Adherence.
Keyword searches are keywords that are not included within the database's subject headings. It is recommended to search particular fields such as titles, abstracts, and author-supplied keywords, especially if the articles are not assigned subject headings. The most common source of keywords may come from author-supplied keywords from a journal article.
Using subject headings only for your searches will not pick up articles recently added to databases because the articles may not have subject headings yet. It may also take years for a new concept to be added to a database as a subject heading.
Adapted from modules created by Janneke Staaks from the University of Amsterdam
https://osf.io/dfz2k
Combining Search Terms Together
Searching with AND combines terms. It gives you results that satisfy both search terms (only where the two overlap).
Some databases use + for AND. Searching with OR “means more”! It gives you everything available for each term, including when they overlap.
Some databases use | for OR. Searching with NOT excludes an entire term, including any overlap between terms. Some databases use – for NOT.
Searching Phrases and Variations of a Word
Databases will allow you to use symbols such as quotation marks and stars to search for specific phrases or variations of words. The symbols that you can use may vary in each database.
Phrase Searching
Quotation marks keep search terms together
"global medicine"; "community health worker"
Use thoughtfully & cautiously... "diabetes prevention" will not find "prevention of diabetes"
You can also try using dashes between your search terms.
systematic-review, integrative-review
Also be aware that truncation & other wild cards (*,?$, etc.) may not always work if used in a phrase
Parentheses
Used to group search terms together.
(mood or depression) AND exercise
(youth or teenager OR adolescent)
Truncation and Wild Cards !*%$
Searches variations of a word
nurs* will search nurse, nurses, nursing, etc.
colo#r will search color or colour
Please note that truncation turns off automatic term mapping in PubMed and the process that includes the MeSH term and any specific terms indented under that term in the MeSH hierarchy. For example, heart attack* will not map to the MeSH term Myocardial Infarction or include any of the more specific terms, e.g., Myocardial Stunning; Shock, Cardiogenic. PubMed does not also allow wild cards.
Proximity Searching
Proximity searching is an advanced search technique used to search for words that are near or within a certain number of words from each other. In most databases, you can add a number to the proximity operator to indicate how close you want the words from each other. If you decide to do proximity searching, I recommend starting with having the words within or near 3 words from each other. Please note that PubMed does not allow proximity searching.
APA PsycINFO
PRE Finds the words that are within a certain number of words in the order you entered them.
health PRE/3 intervention
NEAR Finds the words if they are with a certain number of words in any order.
online NEAR/3 consult
CINAHL
Wn Stands for within. Finds the words that are within a certain number of words in the order you entered them.
mobile W3 application
Nn Stands for near. Finds the words if they are with a certain number of words in any order
fall N3 reduction
Embase
NEXT/n Use for terms that are within 'n' words of each other in the order you entered them.
mobile NEXT/3 device
NEAR/n Use for terms that are within 'n' words of each other in any order.
fall NEAR/3 prevention
MEDLINE via Ovid
ADJ finds two terms next to each other in the specified order
death adj education
ADJn Finds two terms next to each other in any order
safety adj3 culture