If you find that you are making some of these mistakes, please seek support from your supervisor and/or make a 121 appointment with an academic librarian.
Common searching mistakes when conducting a systematic review
| Failure to scope the topic at the start of your dissertation |
- check if there is an existing current systematic review
- e.g. search in Pubmed - restrict to systematic reviews as well as Prospero & Cochrane
- make sure there are enough research papers relevant to your topic to conduct a systematic review before you submit your protocol/proposal
|
| Poor Searching |
- thinking the more databases you search the better
- not making your search manageable across each databases e.g. too many hits
- Searching Pubmed and medline (Medline is part of Pubmed)
- choosing the wrong database for your subject
- spelling errors in your search terms
- incorrect use of AND/OR/NOT & truncation/proximity/phrase searching/
- translating searches across different databases - may need to adjust free text terms to make searches manageable e.g. Scopus or Web of Science
- not using limiters e.g. years, language across all databases
- for some databases - removal of animal studies can be beneficial
|
| difference between conducting and reporting systematic review guidance |
- Understand the difference between the different techniques /methods for conducting & reporting systematic reviews
- Use Cochrane or CRD guidance for conducting review methods and PRISMA guidelines/checklist for reporting your systematic review.
|
| describing not justifying searches in methodology section |
- don't just state what databases searched, underpin with evidence
- e.g. Systematic Review searches require a sensitive approach to identify as many relevant studies (Higgins et al., 2011) Both published and unpublished sources will be searched in order to minimise publication bias (Booth et al., 2016). The following databases will be searched - CINAHL, Medline and Scopus.
|
| Reporting the search strategy incorrectly |
- No search date included of when database searches were conducted
- No details on databases platforms/date coverage
- just writing keywords only - no full search strategy included in dissertation
- platforms vs databases - know the difference between the 2
- platform = a company that provides access to a number of databases across the same interface e.g. Ebsco
- database = a searchable organised/item information containing journal articles etc. e.g. CINAHL or Medline
- incorrect spelling
- No information on how search results will be managed (use of tools, deduplication)
- Reporting search results in methodology section
|
| Poor reporting of Prisma flowchart |
- Numbers don't add up
- don't split over 2 pages
- use template, don't create your own flowchart
- no context to your Prisma in your dissertation - explain what PRISMA flowchart is in the text of your dissertation
- Understanding the screening part of the Prisma flowchart
- 1. Title of article screening
- 2. Abstract of article screening
- 3. Full text of article screening
|
| No evaluation of search |
- no discussion of strengths and limitations of search
- potential bias issue due to search and impact on the validity of the review findings
|
