Finding statistics - beware!
The next pages will cover how to find statistics. We find statistics by searching for our publications in databases or platforms. However, you should always check the data that you find. There are two main issues that can affect data quality, which you should be aware of:
Database coverage
Several different data sources / platforms are available for finding statistics - and results may differ between them. This is because they do not all have the same coverage; thus, a particular database may not contain all of your publications and/or all publications that cite you.
Example: You have authored 12 publications (10 articles, 2 books). Searching for yourself in Web of Science (WoS), you find only 10 publications, cited 25 times. In Google Scholar (GS) you find 11 of your publications, cited 35 times. Why? For the publications, it is likely that one of your books is indexed in GS, but not WoS. The second book was not found in either database. For the citations, 10 sources that cite you are not indexed in WoS but are in GS; these may be from journals or publication types which are not indexed in WoS (e.g. theses, reports).
Articles tend to be well-covered by many data sources/platforms, so you have a choice in where to find statistics. On the other hand, books, reports, datasets, software etc. are more likely to lack coverage or only be covered on specialist platforms. Suggestions for which database/platform to use for various publication types are included on the relevant pages later in this module.
Generally, two of the most-used databases are Web of Science or Google Scholar. Web of Science (WoS) focuses on journals, so statistics from WoS are most reliable for fields where most publications are articles in English; for example, the natural sciences, medicine, and some social sciences. If WoS covers your field adequately, it is generally seen as a more reliable source. However, WoS will lack coverage when publishing is dominated by books, preprints, non-English publications, or new/niche/local journals; for example the some fields in the humanities. For these therefore, statistics in WoS will likely be underestimates (or can't be found at all); the publications will be better covered in Google Scholar. You can check with your subject librarian for advice about coverage of your field. For a recent comparison of coverage between databases, see Martín-Martín et al. 2018.
Author ambiguity
If you use your name to do a quick search for your publications/citations, the resulting data can often be incorrect - either mixed up with someone else's, or incomplete. This is due to author ambiguity (when authors have similar names, or have been inconsistent in how they write their name/affiliation). Some databases, such as Web of Science, put considerable work into identifying authors and linking them to IDs to try and avoid this problem.
Example: You have published some articles at UiB using your full name (Kari Nordmann), and some using K. Nordmann. Later, after a name & job change, you publish using your new name Kari Danskemann at UiO. Meanwhile, a researcher called Kjersti Nordmann is publishing, also using K. Nordmann. You will have to be careful when searching to find all of your works, and to separate them from Kjersti's. This could be avoided by creating an author ID.
If you plan to use Web of Science, please do a thorough author search using name variants, or preferably your ORCID, to ensure you have your correct publication set - see the image below for how to do this (choose the "Researcher" search tab, and then choose whether you are searching by "Name" og "Author identifier"). Finding your correct data is easier if you maintain a profile on Web of Science (see Module 1 for how to do this).