Thing 16: Bibliometrics

This Thing is about Bibliometrics and their value in the research community.

Bibliometrics are numerical indicators which are used for two purposes. First, to give an indication of the academic impact of a specific piece of research - i.e. the affect of the research upon the scientific community. Second, to give a measure of the academic output of that research - this is linked to productivity (University of Surrey - 23 Things 2019).

There are several types of metrics that can be used:

1. Impact at the point of publication - this tends to focus on the journal in which the research was published.
2. Impact post publication - here the influence of the research output is considered. Usually 2 years after publication.  
3. Impact from enabling knowledge transfer - this comes from linking dissimilar areas of research, where-by the research publication acts as a knowledge bridge'.
4. Impact through collaboration - this arises when the research is completed with others, perhaps between authors or institutions. 

However, when comparing bibliometrics, the indicators must be normalised. This is done by comparing like-for-like research. Un-normalised indicators include the number of citations, h-indices or Journal Impact Factors. For example, older papers will have had more time to gather citations, while newer papers will not. Additionally a review-type paper will almost certainly have more citations than an experimental paper focusing on a specific subject, however the latter may have a greater impact on the research community.  

Below, some papers used in my research are evaluated using bibliometrics found using Web of Science (WoS), Scopus (Sc), and Google Scholar (GS). The papers are: 

• Belmonte, H. M. S., Mulheron, M. & Smith, P. A., 2007. Weibull analysis, extrapolations and implications for condition assessment of cast iron watermains. Fatigue & Fracture of Engineering Materials & Structures, 30(N/a), pp. 964-990.

• Turgoose, S., 1982. Post-excavation changes in iron antiquities. Studies in Conservation, 27(N/a), pp. 97-101

• Velichko, A., Holzapfel, C. & Mücklich, F., 2007. 3D Characterisation of Graphite Morphologies in Cast Iron. Advanced Engineering Materials, 9(1-2), pp. 39-45.

Only two of the three papers could be found in the three data bases mentioned earlier, and the citations for each paper vary slightly. It is also seen that the oldest paper is receives the most citations, and in all cases, Google Scholar (GS) yields the highest number of citations - this is summarised below: 

• Belmonte et al (2007) - 11 (WoS), 12 (Sc), 17 (GS)

• Turgoose (1982) - N/A (WoS), 85 (Sc), 160 (GS)

• Velichko et al (2007) - 37 (WoS), 41 (Sc), 56 (GS)

There also exists a new feild of Biblimetric data - Altmetrics. These give a measure of how infroamtion is used in differenct media services such as Google, Facebook, Twitter, ect. A rather interesting tool which summarises this data is the Altmetric donut. 

Thanks for reading this blog post :) I've managed to find a picture that rather well sums up my trepidation towards bibliometrics...

Image from Upsplash, by Nariman Mesharrafa