I conducted a bibliometric coupling analysis on firm growth using VOSviewer, which resulted in 80 papers grouped into 4 clusters.
It is important to note that I had a total of 668 papers initially. To select the top 80 papers, I set the minimum number of document citations to 1 and used the Total Strength Link in VOSviewer.
The largest cluster comprises 28 papers. However, after reading the full text of all these papers, I found no significant similarities among them. Some papers focus on the persistence of growth, while others explore the impacts of high-growth firms or measure high-growth firms' characteristics.
I should mention that, for instance, three papers in this cluster discuss the persistence of growth, whereas I have another cluster that is primarily dedicated to this topic. I believe these papers should be part of the cluster focused on persistence of growth due to their thematic similarity, rather than being included in this larger cluster.
Now, I am seeking guidance on how to address this issue where the papers within this cluster are heterogeneous, making it challenging to identify a single common theme.
can i ingnore some of papers in each cluster that are not relevant to main theme of that cluster?
or what are alternative solutions for this problem?
I would appreciate your insights and recommendations on how to approach this situation.
To address heterogeneity in a paper cluster, ignoring irrelevant papers may lead to loss of valuable information. Instead, consider redefining the cluster based on common themes identified through manual and automated methods. Text mining and topic modeling can identify frequent words, phrases, and underlying topics. Reorganizing the papers based on themes can help understand relationships and identify main themes.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Probability
- Probability Theory
- Reliability