Running your search
Now that you've collected relevant subject headings and keywords, and have applied appropriate search techniques for your chosen database, it's time to build your concepts and run your search following the line by line method.
Check out the video (11:00) below which demonstrates line by line searching using the Search History tool in EBSCO databases.
Building and running a line by line search
To support your learning we have written step-by-step instructions which reflect the above video. You can refer to the Example MEDLINE Search (DOC, 29KB) as you follow the steps.
Step 1: Enter your keywords with field codes
Apply field codes to your keywords, copy and paste each line into the database, click ‘search’. In EBSCOHost databases, keyword search lines in the title and abstract fields will be formatted using the field code TI for Title and AB for Abstract:
TI "physical* activ*" OR AB "physical* activ*"
Repeat this formatting for each of your keywords:
TI sport* OR AB sport*
TI jogging OR AB jogging
Explanation: Limiting your keyword search to specific fields e.g. Title and Abstract, is an effective way to narrow your search and yield more precise results aligned to your topic.
Note
Databases in other platforms may use different field codes to denote their search fields. You can follow the same process above of checking the field code in the search history and formatting your search lines with the field codes to efficiently build your searches.
Step 2: Add relevant subject headings
Add the subject headings you have collected on individual search lines, using the database's subject heading tool to add them to your search history.
You can get a good idea of relevant subject headings by:
- Checking what subject headings have been assigned to your gold set articles.
- Searching your concept by keyword and seeing what subject headings appear in the most relevant search results.
Step 3: Combine your concept lines with OR
Combine the keyword and subject heading lines with OR to finish building your first concept.
It is important to add both keywords and subject headings your concepts to create a robust and comprehensive search. By using OR to combine your concept lines, you are telling the database that 'any/all of these keywords and subject headings represent my concept'. This helps minimise the likelihood of missing important literature.
The below image demonstrates a concept which has followed the line by line search method. S1, S2 and S3 are is a keyword search lines, and S4 is a subject heading search line. S5 combines lines S1-S4 with OR, which finishes building the concept.
From the search results in the Search History above one can see that the keywords are retrieving over 12,500 results, while the subject heading yields nearly 8,000 results. Combining them in line 5 results in over 16,000 results. Neither the keywords nor the subject headings alone can retrieve all the relevant results which is why both are combined.
Step 4: Repeat steps 1-3 with each of your concepts
Step 5: Join all concepts in the final line of the search
The final line of a systematic search combines each concept set into one results using the AND operator. By using AND, you are telling the database that each concept must be represented in the final results. For example, if a search strategy uses AND to combine the concepts of physical activity, adolescents and Australia, then all articles in the search results must represent each concept in either the title, abstract or subject heading.
Step 6: Apply search filters
If you will be applying any filters to your search, only do so on the final line.
Two different types of filters exist:
- In-built database limiters let you limit your results based on selected criteria. For example, some databases may have limiters for age demographics or language of text. These vary between databases, so use with caution.
- Search filters are not covered in this guide - see the Systematic Review Toolkit for more information
Note
Note: When searching in Embase, it is important to set the source to only search across Embase, excluding duplicated content from other databases.