This step of the process can feel a bit like trial and error. When searching, you may need to add or remove keywords or subject headings or add filters to get good search results.
Here is an example of a research question that went through multiple refinements to find the most relevant research:
A researcher is looking for literature that addresses the question, "Which interventions reduce stigma in pregnant or postpartum women with HIV?" They start out with a PICO question:
Their first search might look like this:
| HIV OR human immunodeficiency virus | |
| AND | pregnant OR postpartum |
| AND | interventions OR strategies OR methods |
| AND | reduce stigma |
The researcher decides to add some synonyms and use truncation in the search:
| HIV OR MM "Acquired Immunodeficiency Syndrome" | |
| AND | pregnant OR postpartum |
| AND | intervention OR strateg* OR methods |
| AND | reduc* OR decrease OR minimiz* OR prevent* |
| AND | stigma |
Because they are seeing a lot of studies that talk about stigma, but are not on reducing stigma as a major outcome, the researcher decides to include a proximity search. Proximity searching is a way to search for two or more words that occur within a certain number of words from each other. In this case, the researcher wants the word "stigma" to be found three words or less away from the outcome word, so they would use the code N3 between the terms. In addition to their revised search terms, they might also use a filter to limit their search results to only articles that were published in the last five years.
| HIV OR MM "Acquired Immunodeficiency Syndrome" | |
| AND | pregnant OR postpartum |
| AND | intervention OR strateg* OR methods |
| AND | (reduc* OR decrease OR minimiz* OR prevent*) N3 stigma |
As you can see, researchers often have to go through multiple iterations of revising their search -- adding or removing keywords, using a mix of keywords and subject headings, and using filters and boolean operators to focus their search to find relevant information.
Using FiltersFilters allow you to limit the results you get in a database. You might think having a lot of results come up in your search is great, but having too many results can be overwhelming. What would you do if your search came up with over 10,000 hits? Would you have the time, patience, and attention span to scan each article to see if it is worth your time?
Your answer is probably no. And that is fine -- because that is why filters were created. Filters allow you to narrow down your search results to just those articles that are relevant to your search. There are many kinds of limits -- you can limit by type of research (RCT, systematic review, case study, etc.), by year published, by subject heading -- some databases even let you limit by the subject species (such as tests involving rats) or by gender or age (only studies done on adolescent females).
