| Kai Petersen, Robert Feldt, Shahid Mujtaba, Michael Mattsson
Systematic mapping studies in software engineering are a method to classify and structure a research area by analyzing the frequency of publications in different categories. This approach helps determine the coverage of a research field and can be used to answer specific research questions by combining different facets of the classification scheme. The paper describes the process of conducting a systematic mapping study, compares it with systematic reviews, and provides guidelines for both.
The systematic mapping process involves defining research questions, conducting a search for relevant papers, screening papers for inclusion and exclusion, keywording abstracts to develop a classification scheme, and data extraction and mapping of studies. The classification scheme is developed by identifying keywords and concepts from abstracts and combining them to form categories. The study also discusses the use of visualization techniques, such as bubble plots, to represent the frequency of publications in different categories.
Systematic reviews, on the other hand, focus on analyzing existing primary studies to determine the state of evidence in a specific area. They often involve more detailed analysis and use methods like meta-analysis. The paper compares systematic maps and reviews, highlighting differences in goals, breadth, depth, and validity. Systematic maps are less detailed but can provide a broader overview of a research area, while systematic reviews focus on in-depth analysis of specific topics.
The paper also discusses the importance of using both methods complementarily. Systematic maps can be used first to get an overview of a research area, followed by systematic reviews to investigate specific topics in more detail. The paper provides guidelines for conducting systematic maps and reviews, emphasizing the importance of visualizing results and using structured abstracts to improve the clarity and completeness of research findings.Systematic mapping studies in software engineering are a method to classify and structure a research area by analyzing the frequency of publications in different categories. This approach helps determine the coverage of a research field and can be used to answer specific research questions by combining different facets of the classification scheme. The paper describes the process of conducting a systematic mapping study, compares it with systematic reviews, and provides guidelines for both.
The systematic mapping process involves defining research questions, conducting a search for relevant papers, screening papers for inclusion and exclusion, keywording abstracts to develop a classification scheme, and data extraction and mapping of studies. The classification scheme is developed by identifying keywords and concepts from abstracts and combining them to form categories. The study also discusses the use of visualization techniques, such as bubble plots, to represent the frequency of publications in different categories.
Systematic reviews, on the other hand, focus on analyzing existing primary studies to determine the state of evidence in a specific area. They often involve more detailed analysis and use methods like meta-analysis. The paper compares systematic maps and reviews, highlighting differences in goals, breadth, depth, and validity. Systematic maps are less detailed but can provide a broader overview of a research area, while systematic reviews focus on in-depth analysis of specific topics.
The paper also discusses the importance of using both methods complementarily. Systematic maps can be used first to get an overview of a research area, followed by systematic reviews to investigate specific topics in more detail. The paper provides guidelines for conducting systematic maps and reviews, emphasizing the importance of visualizing results and using structured abstracts to improve the clarity and completeness of research findings.