This blogpost covers the main findings and implications of a study of Swiss higher education institutions’ findability in commercial bibliometric databases, conducted within the project “Towards Open Bibliometric Indicators” (TOBI). It underscores the relevance of organisational identifiers and language disambiguation for the findability and visibility of Swiss HEIs in bibliometric databases. It also showcases the widely varying publication counts across commercial bibliometric databases.
Switzerland has a diverse and vibrant landscape of higher education institutions (HEIs), with each institution contributing to the country’s scientific and educational record. To assess Swiss HEIs scholarly impact, it is crucial to understand the findability and coverage of Swiss HEIs within commercial bibliometric databases like Dimensions, Scopus, and Web of Science (WoS).
The project “Towards Open Bibliometric Indicators” (TOBI) takes on this task and aims to benchmark open bibliometric databases for bibliometrics in Switzerland. To have a point of comparison, it is useful to know how Switzerland is represented in established bibliometric databases. As one out of many steps, we have investigated the representation of Swiss HEIs across these commercial databases, shedding light on Swiss HEIs’ findability and on document classification within those platforms.
Findability of Institutions in Commercial Databases
Ensuring the findability of Swiss HEIs in bibliometric databases requires rich metadata, clear institutional identifiers, and adequate indexing. Our work included all 52 institutions listed by swissuniversities[1] as Accredited Swiss Higher Education Institutions in accordance with the Higher Education Funding and Coordination Act[2] (see https://www.swissuniversities.ch/themen/lehre-studium/akkreditierte-schweizer-hochschulen), focusing on their representation in Dimensions, Scopus, and WoS. Commercial bibliometric databases can be used through acquiring bulk-data or by using their specific Application Programming Interfaces (APIs). However, many research institutions might not have data science staff specifically available to focus on scientometric analyses via API access or a data dump. Under these circumstances, the web-applications offered by Dimensions, Scopus and WoS might be the first and easiest entry point for publication monitoring. Therefore, this blogpost focuses on the webapp access options of the respective databases.
Findability of Swiss HEIs
Within Dimensions 46 out of 52 Swiss institutions were easily findable with a clear identifier, whereas in WoS this was only the case for 17 institutions and in Scopus for 26 (see Figure 1 below and full list here). In particular, universities of applied sciences and universities of teacher education (German: pädagogische Hochschulen) had a disadvantage when it comes to findability in the respective databases. Many of them do not seem to be indexed at all within WoS or Scopus. The six institutions that were not found in Dimensions were:
- Pädagogische Hochschule Nordwestschweiz PHNW
- SUPSI – Dipartimento formazione e apprendimento SUPSIDFA
- Hochschulinstitut Schaffhausen HSSH
- Schweizerisches universitäres Institut für traditionelle chinesische Medizin SWISS TCMUNI
- Swiss UMEF
- Swiss Business School SBS (found in Scopus and WoS though)
Figure 1.
Within Dimensions, 46 out of 52 Swiss institutions were easily findable with a clear identifier, whereas in WoS this was only the case for 17 institutions and in Scopus for 26.
The fact that institutions were not found within a certain database does not mean that research outputs from those institutions are not included in the databases at all. It merely highlights that finding and selecting those institutions easily and evaluating their research output is mostly not possible with the webapps of Scopus and WoS, which might be problematic as the webapps are a frequent entry point for users. Admittedly, all three commercial databases have different strengths and weaknesses such as coverage of certain fields of research, or they might do better or worse in covering research institutions on a global scale or in specific countries. However, our finding for Swiss HEIs is in line with studies on the differences in coverage of commercial bibliometric databases (e.g. Harzing 2019, Visser et al. 2021), with WoS being the most selective also for journals and Dimensions being the most comprehensive of the three (Singh et al. 2021).
Organisational identifiers
Not only in terms of the number of institutions that were found in the respective database, but also regarding other aspects of findability, Dimensions has shown some advantages. Findability within Dimensions profits from identification of institutions with help of Dimensions’ internal grid-identifier (e.g. ‘grid.5801.c’) or the external, open ROR-identifier (e.g. ‘05a28rw58’ for ETH Zurich), while Scopus builds on an internal 8-digit code (e.g. 60025858 for ETH Zurich) and WoS on an organisation’s name (e.g. “OG=(ETH Zurich)”) (see also Purnell 2022: Table 1). One cannot search with the external, open ROR identifiers in the Scopus or WoS webapps, which is disadvantageous for findability. Meanwhile, what selecting “ETH Zurich” actually means in the respective databases, can differ widely. For example, the identifier “OG=(ETH Zurich)” in WoS includes a variety of child organisations and related organisations (see list of institutions included for “ETH Zurich” in Dimensions, Scopus, WoS). In essence, “ETH Zurich” means very different things in the different bibliometric databases.
“ETH Zurich” means very different things in the different bibliometric databases.
Language disambiguation in Swiss languages
Findability of organisations when using the names of institutions in different languages is particularly relevant in the Swiss context, with its four official languages (German, French, Italian, Romansh). Even though not each institution has been checked in all Swiss languages, when checking for institutions’ names in a national language and English, Dimensions performed best in terms of findability of organisations when using the names of institutions in different languages.
Other than in Scopus or Dimensions, in WoS the names of institutions in their original language often were not found at all. Instead, mostly the English name had to be used in WoS to find institutions. On the other hand, there are examples in WoS like “Università della Svizzera italiana” (en: University of Italian Switzerland), which can be found in the Italian version, or “FHNW University of Applied Sciences & Arts Northwestern Switzerland”, which can be found by using the English name or even with the abbreviation “FHNW”. Thus, there seem to be language inconsistencies that are relevant for finding institutions.
In sum, the WoS webapp language disambiguation appears as weakest and the one of Dimensions as strongest, which further improves findability of institutions within the database’s webapp. Such differences are not unexpected given of the scope of the respective database (WoS focuses almost exclusively on publications in English, whereas Dimensions is more inclusive, and not including language disambiguation on the organisation level might be problematic here). Some clear language connections between abbreviations and institutions’ names are not found in any of the three webapps. For example, searching for “Sankt Gallen” in the organisation search of Dimensions, Scopus, or WoS webapp does not reveal the “University of St Gallen”. Using for example “St Gallen” works though).
Publication Coverage
The amounts of Swiss HEI publications for different publication types differ across the commercial bibliometric databases. This is due to a variety of factors that have already been investigated with different focuses (e.g. Mongeon/Paul-Hus 2016; Singh et al. 2021), that are hard to identify in an exhaustive manner, and that cannot be presented within this blogpost in full detail. Figure 2 illustrates an exemplary comparison of the number of articles found for several Swiss HEIs, which illustrates that the differences in publication numbers can be substantial, depending on the institution.
Figure 2.
Furthermore, many scientometric analyses focus on journal articles, while in some disciplines, publication output such as monographies or book chapters have played or still play an important role. This implies that the selectiveness of commercial databases might play more of a role for some disciplines or institutions than for others. Web of Science in particular hardly includes any book publications, as Figure 3 shows.
Figure 3.
In conclusion, the investigation highlights the role of organisational identifiers and language disambiguation for the findability and visibility of Swiss HEIs in bibliometric databases. It also showcases the widely varying publication counts that commercial bibliometric databases indicate for Swiss HEIs. Only when keeping these aspects of findability and differing publication counts in mind, we can ensure robust scientometric analyses and informed decision-making in HEIs.
Stay tuned for future updates on bibliometrics, database comparisons and open bibliometric indicators!
Acknowledgement
This blogpost was partly inspired by ChatGPT 3.5, version 23 February 2024 (OpenAI 2021).
This blogpost has been written in the context of the project “Towards Open Bibliometric Indicators” (TOBI), co-funded by swissuniversities and the ETH Library.
References
Donner, P., Rimmert, C., & van Eck, N. J. (2020). Comparing institutional-level bibliometric research performance indicator values based on different affiliation disambiguation systems. Quantitative Science Studies, 1(1), 150-170.
Harzing, A. W. (2019). Two new kids on the block: How do Crossref and Dimensions compare with Google Scholar, Microsoft Academic, Scopus and the Web of Science?. Scientometrics, 120(1), 341-349.
Mongeon, P., & Paul-Hus, A. (2016). The journal coverage of Web of Science and Scopus: a comparative analysis. Scientometrics, 106, 213-228.
Singh, V. K., Singh, P., Karmakar, M., Leta, J., & Mayr, P. (2021). The journal coverage of Web of Science, Scopus and Dimensions: A comparative analysis. Scientometrics, 126, 5113-5142.
OpenAI. (2021). GPT-3.5 (ChatGPT) [Computer software]. Retrieved from https://openai.com, accessed 21.02.2024.
Purnell, P. J. (2022). The prevalence and impact of university affiliation discrepancies between four bibliographic databases—Scopus, Web of Science, Dimensions, and Microsoft Academic. Quantitative Science Studies, 3(1), 99-121.
Visser, M., Van Eck, N. J., & Waltman, L. (2021). Large-scale comparison of bibliographic data sources: Scopus, Web of Science, Dimensions, Crossref, and Microsoft Academic. Quantitative science studies, 2(1), 20-41.
Footnotes
[1] Swissuniversities is the Rectors’ Conference of the Swiss Universities, i.e. the umbrella organisation of Swiss universities (https://www.swissuniversities.ch/).
[2] See https://www.swissuniversities.ch/themen/lehre-studium/akkreditierte-schweizer-hochschulen, accessed 21.02.2024.
