This library guide provides:
A systematic review is much more than a literature review. It uses an explicit searching methodology to perform a comprehensive search and critical appraisal of selected primary studies with the aim of reducing bias to answer a clearly defined research question.
Cochrane reviews are considered the gold standard of systematic reviews. Watch the clip below for further explanation.
Systematic reviews include:
a clearly stated set of objectives with pre-defined eligibility criteria for studies
an explicit, reproducible search strategy and methodology
a systematic search that attempts to identify all studies that would meet the eligibility criteria
an assessment of the validity of the findings of the included studies, for example through the assessment of risk of bias
A systematic-style review follows a similar process to systematic reviews, but they may include fewer databases, use less stringent inclusion and exclusion criteria, and have a less detailed protocol. The purpose of a systematic-style review is to provide a broad overview of the available literature on a topic rather than to provide a definitive answer to a specific research question.
The first step is to determine what kind of review is right for your research question.
The review type will depend on your:
Speak with your supervisor or team leader before choosing a review type, or consult the tools below.
The video below from the University of Texas Library summarizes the different review types and is shared under a CC BY NC licence.
A scoping review aims to assess the scope of literature in a particular field or research area, identifying gaps in the literature and potential areas for future research. It is typically broader in scope and less focused than a systematic review, and is often the first step toward undertaking a full systematic review.
Munn et al. (2018) have since established an expanded definition and purpose of scoping reviews to include the following features:
Conducting a full systematic review is a more rigorous process, and you will need about two to three researchers on your team to ensure the elimination of bias in the review process.
The best place to register your scoping review protocol is the Open Science Framework (OSF). Some other platforms no longer accept the registration of these kinds of reviews, such as PROSPERO, and the JBI Systematic Review Register is only open to members of the JBI Collaboration.
Documenting results
Documenting results is important for both scoping reviews and systematic reviews. While scoping reviews and systematic reviews have different aims and methods, both involve a comprehensive search for relevant studies and the synthesis of evidence to answer a research question.
You can also consult the JBI Manual for Evidence Synthesis for guidance on conducting scoping reviews, and the University of South Australia's library guide on scoping reviews, or watch the clip below for further information.
Take a look at some examples of published systematic reviews so you know what you are working towards. Systematic reviews are conducted more commonly in fields like health and medicine, yet, increasingly, they are also being published in other disciplines such as humanities, education and sciences.
You can find systematic reviews in several key databases (use it as a search term, or limit) - here are some key sources:
Once you have formulated your research question and decided the type of review you will conduct, you will need to check whether your research question has been answered. Reviews that are currently in progress are called protocols. Protocols outline the scope and intention of the review.
You can access the registers below to check and register your review protocol:
For further advice on how to develop a protocol see the resources below:
GenAI cannot conduct a full systematic or scoping review. Human mediation is still required to validate the reliability of the evidence to minimise potential bias because GenAI can “hallucinate” or produce false references and research results.
The Preferred Reporting Guidelines Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) remains the gold standard process for publishing systematic and scoping reviews. An essential part of PRISMA is to document search strategies in discipline-specific databases to ensure transparency and the reproducibility of those results.
Despite its limitations, the combination of GenAI tools and human expertise can make stages of the systematic review process more efficient and thorough.
See the table below for some tips on how GenAI tools can assist the review process. Please note that new tools are emerging all the time and that this a rapidly changing space.
For more information on how to use GenAI tools in your research consult this page.
|
Workflow |
Tools |
Use |
Limitations |
|
Identify some MeSH or other subject headings & keywords to help develop your search strategy |
Brainstorm topics, generate lists of keywords, narrow topics, provide summaries of works |
Never an exhaustive list. Researchers are still required to search for subject headings in subject-specific databases to satisfy the requirements of PRISMA. |
|
|
Identify some seed references, seminal authors and key literature |
Highlights key research papers, seminal authors & citation networks. |
Inaccuracies in complex or niche research areas. Can “hallucinate” or produce false references. |
|
|
Identify some research gaps |
Finds connections, if they exist, between papers.
|
Usually requires an edit to ensure writing reflects an appropriate voice and academic tone. |
|
|
Summarise and synthesise large
|
|
Reads research articles, reports and book chapters in seconds and breaks them into bite-sized sections, extracts key findings |
Inaccuracies in complex or niche research areas |
|
Assist in testing the review search strategy |
|
Can test the veracity of your search strategy in the final stages of screening by detecting “extra” research not found in discipline-specific databases or published after the search period. |
Can “hallucinate” or produce false references. |