Systematic literature reviews

1. Formulate your research question

• Try to formulate a clear, well-defined research question of appropriate scope and which can be answered, not only by yes or no. 
• To be able to perform a literature review, you need to consider a subject area a sufficiently large number of original research studies. Therefore, you should do a test search before you formulate and delimit your research question. 
• Searching for existing reviews on your topic, can also help you to develop your research question and identify gaps.
• One way to structure your research question is to break it down into different parts. In other words, you need to identify the key concepts of your research question.

2. Define inclusion and exclusion criteria

• Clearly state the criteria you will use to determine whether or not a study will be included in your review.
• Examples of common inclusion and exclusion criteria are date, geographic location and study design. Study design refers to methods and procedures used to collect and analyze data.

3. Find search terms and create search blocks

• Test searching is a good way to identify the terminology of the subject area and find search terms.
• Reading key articles can help you gather search terms for your topic.
• Generate key words by investigating the words that occur in the title and abstract of relevant articles.
• A good way of achieving a structured final search query is to arrange your search terms into search blocks; these blocks should arise from the key concepts of your research question. For example, if you are interested in research about how performance is related to sustainable finance, the key concepts of the research question is "sustainable finance" and "performance" 

4. Make search strings and perform the search in a structured way

• Identify databases in which you intend to do your search. We recommend searching in specialized databases. See list of databases you have access to via the BI library. These are databases with good coverage of research related to BI disciplines. Note that Oria often returns a lot of records (not all necessarily relevant), while a article database such as Web of Science allows a more precise search. Contact the library for guidance if you have questions regarding which databases to search. Remember to register the number of search results your search has produced in each database you have searched (see below).

• The use of AND, OR and NOT, called Boolean logic, helps you to combine your search words.
• Examples of search strings: 
  • "Sustainable finance" OR "Green finance" - returns result where either "Sustainable finance" or "Green finance" is used.

• ("Sustainable finance" OR "Green finance") AND performance* - returns results where either "Sustainable finance" or "Green finance" and Performance* is used.

• Wildcards help you include alternative spellings or word forms. A wildcard replaces character(s) of a word to enable searching for variations in spelling. When used to search with word stems, called truncation, where you leave the ending of a word open. The most frequently used wildcard is *. Others are?, #, $.
  • Example of truncation is strateg* returns strategy, strategies, strategic etc.

  • The dollar sign ($) can be useful for finding both the British and American spellings of the same word. For example, Behavi$r finds behavior and behaviour.

Journal databases provided by BI 

5. Run your search

• Run your searches in the databases that you've identified as relevant to your topic.
• Make sure you document your search strings. This ensures transparency and replicability. 
• Collect ALL of the retrieved records from each search into a reference manager, such as Zotero.
• Remove duplicates by using a reference manager, so that you only have a set of unique articles in your selection.

• Database search log

• Zotero 

6.  Select studies based on your inclusion-criteria 

• Start with a title/abstract screening to remove studies that are clearly not related to your topic.
• Use your inclusion/exclusion criteria to screen the full-text of the studies.
• If you are more than one person doing the review, it can be smart to import your selection of articles from Zotero into a tool such as like Rayyan, where two reviewers can screen all studies independently.
• Read all the articles you choose to include in your study. 

Rayyan 

7. Analyze the included studies

• Analyze your included studies in light of your research question. Present your findings, including a summary of how the included studies conclude on the research topic as well as identified needs for further research. Differences and similarities in for instance methodology and population may also be relevant.  

• You may consider using text mining and visualization tools to analyze and present trends in your included studies.

• NVivo can be a useful tool in your analysis. It is a software that allows you to code text and articles. You can for instance code your articles by methodology, number of citations or sub-topics. With NVivo you can also visualize key information about your articles, such as number of articles per year in the period you study and information based on codes you have created. Read more about how to use NVivo in literature reviews here

NVivo 

8. Present results and assess the quality of evidence

• Clearly present your findings, including a description of your search strategy, such that it is transparent what you have done. By describing your methodology, your review can easily be updated in the future with new research findings.
• Provide recommendations for future directions for research to fill existing gaps in knowledge or to strengthen the body of evidence.

• PRISMA Flow Diagram
• PRISMA Checklist