Read time: 5 min
Recently, I covered the literature review exercise and mentioned that some tools exist to visualize the bibliometric networks of scientific literature.
As the volume of academic literature continues to grow, conducting a comprehensive literature review becomes challenging. This is where VOSviewer comes in very handy.
VOSviewer is a user-friendly software, developed by Nees Van Eck and Ludo Waltman from CWTS, Leiden University. This software allows researchers to visualize and map scientific literature, making the process of literature review more efficient and insightful.
Here, I present briefly the essential steps to use VOSviewer effectively in a literature review from data collection to visualization. By the end of this guide, you’ll be able to conduct better data analysis and visualization.
Step 1: Data Collection
Before diving into VOSviewer, you need to collect your data. Start by identifying the academic database or search engine where you’ll source your literature. Common choices include Web of Science, PubMed, Scopus, and Google Scholar. Ensure that your search terms are well-defined to retrieve relevant literature.
1. Define Your Research Objectives
Begin by clarifying your research objectives and the scope of your literature review. What are the specific topics or keywords you want to explore? Or do you simply want to group papers based on their citation linkages? Having a clear focus will help you refine your search and gather relevant data.
2. Set Inclusion and Exclusion Criteria
Establish criteria for including or excluding papers in your review. This may include the publication date, language, document type, and more. These criteria will help you filter the data effectively.
3. Data Retrieval
Execute your search in the chosen database and retrieve the relevant academic papers. Make sure to export the data in a compatible format for VOSviewer, such as Txt or RIS. Here, I use Web of Science on one of my topics of interest: “scientific mobility”. I found 5,697 papers that seem relevant. Obviously, I don’t have time to read all of them.
Step 2: Preparing Data for VOSviewer
Now that you have your data, it’s time to prepare it for analysis in VOSviewer.
1. Data Cleaning and Formatting
Review the downloaded data for irrelevant records, or inconsistencies. Cleaning the data ensures that your analysis is based on high-quality data. For instance, there might be some issues at the record level such as authors/abstracts written on multiple lines.
VOSviewer requires specific formatting for data import. Ensure that your data is organized in a way that includes essential information such as titles, abstracts, authors, and citations. This is the required data to create the maps.
For instance, in Web of Science, make sure that you export the full record with cited references. I personally use the Tab Delimited File export from Web of Science. Note that you might need to do multiple exports (500 being the limit in Web of Science) and then merge the text files.
Step 3: Using VOSviewer
You’re now ready to work with VOSviewer. Follow these steps to create visualizations of your literature of interest:
1. Importing Data
Launch VOSviewer and choose “Create a new map”. Then, import your prepared data file. VOSviewer will analyze the data and display an initial network map of your literature.
2. Creating the Network Map
There are several types of analyses. In this guide, I cover two of them. First, the co-occurrence analysis will group papers based on related or similar keywords. Second, the citation analysis will reveal the linkages between papers.
This co-occurrence analysis shows how the different terms are related in your set of publications. The size of the nodes represent their occurrence in your papers of interest.
Moving on to the citation analysis, the following network map represents the relationships between different papers based on citations/references. Nodes represent documents, and links between nodes indicate citations. Clusters are colored with distinct colors.
You can customize your map by adjusting parameters like the minimum number of citations for a document to be included.
Don’t forget to save your maps/networks that you can re-open anytime later (step 3.1: File > Open)
3. Fine-Tuning the Visualization
VOSviewer offers various options to fine-tune your visualization. You can adjust the layout, node size, colors, and labels to make the map more informative and visually appealing.
4. Analyzing Clusters
VOSviewer automatically identifies clustersr groups o of related documents. These clusters represent thematic groups within your literature. Explore these clusters to gain insights into the major themes in your research field.
Clusters of papers Clusters of keywords
5. Exploring Citation Patterns
Use VOSviewer to analyze citation patterns. By understanding how papers cite each other, you can discover influential works and explore the evolution of ideas in your field.
Step 4: Interpretation and Insights
As you explore your VOSviewer map, pay attention to the following aspects:
1. Cluster Analysis:
Identify clusters of related papers or clusters of terms/keywords. These clusters can represent subfields, hot topics, or recurring themes in your literature. Dive deeper into each cluster to uncover specific research trends.
2. Citation Analysis
Examine the citation patterns within your map. Which papers are frequently cited? Are there works that have significantly influenced the field? Understanding these dynamics can help to identify critical papers.
3. Identifying Key Papers
Look for papers with high citation counts or centrality in the network (have a look at the size of the nods and the width of the linkages). These are often pivotal works that play a significant role in your research area.
4. Generating New Search Terms
VOSviewer can suggest new keywords and terms based on the co-occurrence of words in your literature. These suggestions can be valuable for refining your search queries and identifying related papers.
You are now equipped to use VOSviewer and enhance the quality and efficiency of your literature reviews. For more details about all the functionalities, refer to the manual guide.
My favourite things this week
1- I spent 5 days in Kuwait recently, visiting several leading research institutions and funders. I met about 400 researchers there. And it was good to reconnect with some of them this week. It’s always rich discussing so many exciting research projects but also about the challenges faced by researchers. Getting to know more about the national initiatives was also insightful.
2- The Highly Cited Researchers 2023 list is out. Each researcher selected has published multiple Highly Cited Papers™ which rank in the top 1% by citations for their field(s) and publication year in the Web of Science™ over the past 10 years. Highly Cited Researchers™ (HCR) are 1 in 1,000. And it’s always fascinating to see stories behind the HCR.
Quote of the Week
“The network is what carries the collective response to your performance.” Albert-László Barabási, The Formula: The Universal Laws of Success