A comparative study of first and all-author bibliographic coupling analysis based on Scientometrics

This paper takes 2,876 papers published in Scientometrics as the sample, and the time span of these papers is nearly ten years. We have studied the intellectual structure and development trend of Scientometrics in two periods within five years, and made an empirical exploration and comparison of two methods of author bibliographic coupling methods—first author bibliographic coupling (FABCA) and all authors bibliographic coupling (AABCA). The results show that the intellectual structure of Scientometrics in the two periods are relatively stable, but the intellectual structure of the latter period is clearer, and the correlation and integration of the research topics is stronger than that of the previous period. There is a strong correlation between AABCA and FABCA, but each has its own characteristics, and there are subtle differences between them. The similarity between them can be found by authors rank correlation analysis, the cosine similarity analysis, and discipline research topic detection. However, the factor models fitting analysis, research topics detection and evolution show that there are differences between them. AABCA is sensitive to the discovery of research hotspots, and can detect more research topics than FABCA. While FABCA has advantages over AABCA in reflecting the frontier and innovation of the discipline. It can provide more information. Therefore, FABCA and AABCA have their own characteristics and cannot be replaced. The combination of the two methods is a powerful and effective research method to explore the intellectual structure and evolution trend of the discipline.