4. Review the evidence
Reviewing the literature involves the following 7 steps:
- Removing duplicates
When you export references from databases, you will likely end up with duplicates. You can use EndNote to remove duplicate references. Do this before you start screening.
- Screening titles and abstracts
This is the process of identifying studies from your search for inclusion in the review. Refer back to the research protocol for inclusion and exclusion criteria.
Begin by screening titles and abstract to remove studies that don’t meet your inclusion criteria (these are stated in your protocol and are related to your question parameters, eg population group, geographical region, setting, outcomes of interest; relevant study types; language; timeframes).
A team of at least two or three reviewers is important for the screening process to reduce the risk of selection errors and bias.
- Collating the full texts
Articles that pass the initial title/ abstract screening process will need to be obtained in full text. You can use the Bookmarklet Tool to help access full texts. Interlibrary Loan Requests can be made for articles Deakin doesn’t hold.
- Screening full texts
Use your inclusion and exclusion criteria to review the full text articles.
- Evaluating studies
Critical appraisal of studies is important to assess the quality of evidence being synthesised in the review. In Business, this is often based on the ranking of the journal that an article is published in. It is also a good idea to appraise studies for their individual quality. The JBI Institute has a suite of critical appraisal tools to help assess different study types for bias, including qualitative studies and text and opinion. JBI tools can be applied to study designs used in Business research.
- Extracting data
Data relevant to the review question is extracted from included studies and often presented in a summary table. Covidence can be used to assist with data extraction.
-
Synthesising the data
Key findings are synthesised from the data. There are different types of syntheses depending on the included study types, eg qualitative, quantitative or integrative synthesis. For more information read the chapter: Collecting and combining study characteristics (pp. 197-210)
Tools to try
- EndNote – Use to export searches to your citation library and create groups to screen citations against inclusion and exclusion criteria. Also use to identify and remove duplicates
- Covidence - Can assist with the screening of citations, abstracts and full text studies for systematic and other complex reviews. It also supports risk of bias assessment, data extraction and the exporting of data and reference information. Covidence can save time as it allows for multiple reviewers to view the search results for the review and assign inclusion and exclusion criteria to the studies.