Use the Library’s search table template to help you plan your search strategy.
A search statement is a combination of your search terms (free text and controlled vocabulary) boolean operators, and truncation symbols that will be entered into the search boxes of library databases. You will have used the library's search strategy template to identify all the terms you will use and reviewed our guide to adapting your search for each database. You can see the search statements or search strings used for a number of databases in section 4.
Once you've identified your main concepts, the next step is to convert each concept into a list of relevant 'free text terms' and 'controlled vocabulary' to use in your search of each of your chosen databases.
You will use both the free text terms and controlled vocabulary (where included in databases) in your search. Using both, will ensure you retrieve as many relevant results as possible in your search. For databases that do not have subject headings, simply search using the free text terms you identified.
Free text Terms
Free text terms or sometimes called 'keywords' are terms that you may identify as being relevant to use in your search from your own research and also from reviewing articles and studies related to your research.
Controlled Vocabulary
Controlled vocabulary are a set of approved 'standardised' words or phrases that describe concepts and ensure uniformity across database records (articles). Also known as know as subject headings they are used in a number of key databases e.g. Medline/Embase//Cinahl/PsycInfo.
Subject headings are applied manually and it can take some time to apply them hence it’s important to also search a database using free text terms. The advantage of using controlled vocabulary/subject headings is that your search will automatically search for all relevant synonyms for your concept. For example, the concept 'cancer' can be described in a number of ways, articles may talk about cancers or tumours or neoplasms. Using free text terms you need to include all possible terms (synonyms). If however you search using the controlled vocabulary term (in Medline it's Neoplasms), it will retrieve ALL articles about cancers irrespective of the word/term used. It will automatically find articles using all synonyms. See example below.
Note that:
Boolean Operators are used to connect terms so that you can search for multiple terms at once.
| AND |
Use AND to combine multiple concepts |
diet AND exercise |
| OR | Use OR when you have similar words to describe a concept or topic At least one of the words connected with OR should be found in each result |
exercise OR physical activity |
| NOT |
Use NOT to exclude results containing terms |
animals NOT humans |
Search operators such as wildcards, truncation and parenthesis can help you expand/limit your search.
| Quotation Marks/ Inverted Commas | " " |
Use for exact phrase searching, terms must appear exactly "cancer prevention" |
| Truncation | * |
Searches for words that begin the same but end differently comput* will find results that mention the terms |
| Wildcards | # or ? |
Searches for alternative spellings, replaces zero or one a single letter behavio#r |
NOTE: Search operators may vary depending on the database you are using. Check the help section of your chosen database for more information.
Database Limit Options
Articles indexed in databases will include fields such as language, publication type and date of publication etc which means you will have the option to limit your search using such fields once you have conducted your search. Not all articles will be 'indexed' in detail immediately and therefore if you were to apply such limits, you could miss out on important studies/papers. Also it is generally not advised to limit by language or date when conducting systematic reviews as key studies could be missed and also limiting to English language can be considered as 'language bias'. Both Cochrane and the Campbell Collaboration advise not to restrict the searches by language. Restricting by date would have to be justified in terms of excluding earlier studies.
Limiting to 'human' only studies
The best way to do this, is not to use human filters in the databases, again as new studies are added to the databases, full indexing will follow and is not always done immediately. It is advised instead to identify animal only studies first and then exclude those from the search results (double negative). Here's a simple example to show how you can do this in PubMed. Note in this instance, the excluded studies may actually be relevant! So it is always advisable to be cautious when using limiters.
