Close
Close
This guide © 2024 by UCL - Library Skills is licensed under CC BY-NC-SA 4.0 
Studies are identified by searching for studies of likely relevance to the review, followed by an assessment of relevance through sifting the search results in a structured way. This stage is guided by the review question and criteria on what is included and excluded from the review. (See Formulating a research question). These criteria could relate to specific populations, settings, date limits, geographical areas, types of interventions, or something else.
1. Searching for studies
Research evidence may be reported in many different places. References and links to these reports are identified from a variety of places.
A strategy needs to be developed for looking in these different places in order to find the studies that meet the review's inclusion criteria. What is the most effective search strategy will depend on the research question (see Sources of studies). However, a large proportion of research is published in journals and there are many databases in different topic areas that provide access to such journal papers. Database searching using Boolean searches is often undertaken as part of systematic searching supplemented by other sources (see Searching methods).
2) Sifting (or screening) studies
A systematic sift or screening of the search results is undertaken against the specific inclusion and exclusion criteria. The sifting is likely to be undertaken on the titles and abstracts of the research citations initially, followed by a further sift on the full text. If the sifting takes place without reference to the inclusion and exclusion criteria then there are risks that a) relevant studies are excluded, or b) irrelevant studies are included. The sifting is typically checked with at least one other researcher to ensure consistency and reliability of the process. The written report of the systematic review contains details of the number of citations sifted, those that met the inclusion criteria, and reasons for exclusion of full texts. Increasingly machine learning is used to prioritise or semi-automate the sifting process, particularly for large reviews. Screening can be undertaken using a variety of tools (see Software for systematic reviews).
There are many bibliographic databases that have been curated to cover specific topic areas or fields of research. These often contain journal papers, and sometimes other types of research outputs. Systematic reviews generally search at least two topic-based bibliographic databases, and often many more, depending on the topic area. They will also search other types of resources.
Consider your topic and where the likely research will be. Check the library Subject Guides and Databases for the subject areas that the topic covers to identify relevant resources. Also consider that some resources may focus on a perspective (age, geographical area), or a study design (qualitative, trials, systematic reviews), or on types of output (working papers, conference proceedings, theses). Some of the textbooks and guidance documents on systematic reviews contain lists of useful resources by subject area, and published systematic reviews describe the sources searched.
See also Searching databases.
There are a range of other potential sources of research studies, including:
Searching these sources involves a variety of techniques, such as: browsing web pages, keyword searching or 'related item searching'. It is not always possible to search for the literature in a structured, systematic way.
