The Literature Review

A systematic review is an evaluation of the literature on a topic that uses an explicit methodology in order to answer a specific research question. They were originated in the Health Sciences fields and initially developed by Cochrane, an organization that organizes medical research findings. While mostly used originally for clinical medical questions, the systematic review process has expanded into other fields. Here are some of the elements of a Systematic Review:

• Uses a protocol to guide searching and methods
• Follows reporting guidelines like PRISMA
• Clearly tracks and reports on the search tactics used
• Evaluates for bias within the review process
• If using statistical analysis of data within the studies reviewed, becomes a Meta-Analysis

N.B. **A Systematic Review with a goal of being published can take between 9-12 months and sometimes more, and requires multiple authors with different areas of expertise. If you are doing a systematic review as a class assignment or an individual project, this will differ**

Use the Cochrane Handbook for Systematic Reviews of Interventions and and the JBI Manual for Evidence Synthesis to learn more about each of these steps. Read below for a summary of each step and resources associated. Image from the Systematic Review Research Guide at the University of Alabama at Birmingham.

Formulate your Question

This step includes a Preliminary or Exploratory search: casually looking in databases for existing research on this topic, checking to see if there may be too many or too little relevant results, and determining your formal research question through this process. At this point, you will also want to determine your research team and work with the other team members in formulating your question (unless you are doing a systematic review as a class assignment or solo project). Check the box on this page describing the PICO system for help designing a question.

Develop Your Protocol

A protocol is a written explanation of the process you will take in your systematic review. This will often include what databases you will search, what your search strategy will be, your inclusion and exclusion criteria, and other necessary details. If you plan to publish your systematic review in a journal, often registering your protocol prior will be a requirement. Protocols are most often published in Prospero in the Health Care and Social Care fields. If you are not in these fields or do not plan on publishing, it can still be useful to develop a protocol for accountability.

Conduct Search

Using your workflow from your protocol, begin your search by using your search strings and filters decided from your preliminary search. Be meticulous about tracking your result numbers, databases, and where you are storing this information. See the section of this guide on searching for information about constructing a search using Boolean Operators, Punctuation, and Filters.

Select Studies

The process of selecting studies for inclusion is called screening. Most systematic reviews include a Title/Abstract screening and a Full-Text Screening. The top resource used in this process is called Covidence. You can create an account in Covidence through Columbia University. You select studies by determining if they fit within your inclusion/exclusion criteria that you have developed in your protocol.

Appraise Studies

Typical systematic reviews include an assessment of the risk of bias or quality assessment. the Cochrane handbook details this process using their tool called RoB 2 in Chapter 8. Covidence also has a built in Assessment of the risk of bias tool.

Extract Data

Covidence has a data extraction tool that can help in this process. the Cochrane handbook also details procedures for data extraction from your results in Chapter 5.

Analyze/Interpret Results

Bring it all together in your review by providing interpretation of your results. Utilizing the PRISMA reporting checklist is essential here.

A well constructed research question will translate into a robust systematic search. Different disciplines have different ways of expressing the essential elements of a research question. Below are three common question frameworks, with PICO being the most popular.

PICO

A PICO Question is often used in Systematic Reviews to clarify the research interest and generate search terms. Determining your PICO elements and breaking them down into search terms can help narrow your results. PICO stands for: Population (or Patient, Problem) - the group or phenomenon you are researching, Intervention - the mechanism for change you are attempting to apply to the population, Comparison (or Control) - An alternative, or what you are measuring your intervention against, and Outcome - what intended effect or consequence the intervention will cause.

Picking Key Words

1. Use your research questions to pick out essential keywords: see the Venn Diagram on joining topics into a research focus.
2. Find related terms and synonyms for each concept in your research question. for example:

art museums 

• Cultural heritage

• Museums

• Galleries

Constructing Searches

Most search engines, including Educat+ and CLIO, use standard search language and symbols to convey searches:

1. Boolean operators connect your terms: AND connects different topics together, OR connects related terms within each topic, NOT removes that search term from results. See the example below.

2. Quotations: using quotation marks allows you to search for an exact phrase i.e. "Teachers College" instead of 'Teachers' and 'College'.

3. Parentheses: help group sets of terms to delineate relationships within the search.

4. Question mark: a question mark in a search has a similar function to an asterisk, but for only one letter i.e. 'wom?n' will capture woman and women.

5. Asterisks: adding an asterisk at the end of a word will capture related words with the same root or base, i.e. 'philosoph*' will capture philosophy, philosophical, philosopher etc.

Example Search

(“art museum*” OR museu* OR galler*) AND (hiring OR recruit*) AND (divers* OR DEAI OR DEI) AND (manage* OR executiv* OR leader*)

Lincoln School Of Teachers College. Summer Demonstration. Recreation Program. (1939). Historical Photographs of Teachers College. Courtesy of Gottesman Libraries.