Research analytics

Analytics is big business. Organizations are increasingly turning to the analysis of transactional and other data to inform decision-making. Websites are refined based on analysis of traffic. Companies may use analytics to optimize inventory. And so on.
I was struck by the parallel with business analytics when reading some brochures about Elsevier’s SciVal. These were part of a pack distributed at the Emtacl10 conference in Trondheim recently. They include some quotes:

We need reliable support to help us make smarter, defendable decisions about allocating internal funding. (Vice Chancellor for Research)

Decision making will always be subjective, but we are trying to move to a more evidence-based model. (Development manager)

The material describes the tool in terms which would not look out of place, with appropriate substitutions, in advertising for a business analytics tool:

Academic Executives need accurate research performance insight to make appropriate funding decisions and develop strategic blueprints that will lead their institutions to valuable breakthroughs.

Also in the pack was some material about the use of Scopus for bibliometrics. In this case, the material speaks about ‘research performance measurement’:

Research Performance Measurement (RPM), also known as bibliometrics, is the discipline of measuring the performance of a researcher, a collection of selected articles, a journal or an institute. ….. Quantitative metrics are gaining prominence because they are easy to benchmark, objective and globally comparative; all important in global RPM.

Of course, in this area Elsevier is competing with Thomson Reuters’ Web of Science and related products. Indeed, Thomson Reuters has a Research evaluation section on its website.
One of the drivers of such measurement is the strong interest in recent years in university ranking. And this is an area also in which Thomson Reuters has become very active, with its Global institutional profiles project and its partnership with the Times Higher to produce their widely consulted university rankings.
Of course, this interest in research analytics and metrics is not without opposition. However, it is interesting to see Elsevier and Thomson Reuters very clearly identify this emphasis within their product offerings, and the location of bibliometrics within the broader area of research analytics and metrics services to university customers.

2 thoughts on “Research analytics”

  1. As institutions get better information about the research output from their academics based on repository submissions, it strikes me that it’s possible to use the number of citations made within those papers to provide an indicator about the journals that local academics are referring to.
    That is, rather than focusing on metrics that try to a PageRank like thing across as many journals as possible, also review journals in light of local rankings, based on the journals your local academics evidently do read, as based on the articles they cite at the first order. (I know, I know, it may be that academics pick up one article via a second, and a third, that they don’t cite. But then, they also pick up references from colleagues and sources the library has no knowledge about…)
    So for example, given a list of articles in a local repository, look at them and the articles they cite, and run something like the HITS algorithm over them to get an idea of the local importance of journals based on local published output.

Leave a Reply