Reading Time: 3 minutes
The Google Search home screen: is this where academics start their research?

Any academic study that looks at institutional repositories is welcome, so we can get a better idea of how many repositories there are, how they function in different countries, and how they are changing. However, I was surprised to see a recent article referred to in Scholarly Kitchen as a demonstration of the seeming inadequacy of repositories. The article, written by three researchers at a Spanish institution claims that papers in institutional repositories are not as visible as those same papers held elsewhere, due to a lack of investment in metadata and search engine optimization by the repositories. In other words, repositories are not doing their job well enough.

I looked at the article (Orduña-Malea et al 2024), and, on reflection, it appears to me a rather odd piece of research.

  1. It uses a commercial SEO tool, Ubersuggest, to measure traffic to Spanish institutional repositories from Google Search. SEO and keyword ranking is used for commercial sites, not for academic sites. Few academics are interested in keywords so that their articles appear at the top of search results in Google: this is the kind of kind of approach more relevant to pizza restaurants than to academic institutions. After I had loaded Ubersuggest to my PC, I then found myself receiving frequent adverts to buy it – not what I expect from an academic tool.
  2. In any case, Google Search is not what most researchers would use for academic searching. Back in 2016, Simon Inger and Tracey Gardner, in a survey cited by this article, stated that “Google Scholar is used more than Google in the academic sector”, and that is much more true today. In fact, Inger and Gardner’s research identified ten different channels to discovery, of which a general search engine was just one. But the way researchers discover content are even more numerous than that. Recommendations, reference linking and forward and backward citation linking are all widely used. Trying to assess the success of an institutional repository by looking at a very small proportion of the total discovery traffic is unlikely to produce meaningful results.
  3. Worse, the authors themselves explain that Spanish repositories only contain a small fraction of the institution’s scholarly output – which surely invalidates the research. Wikipedia estimates a total of 104,353 scholarly articles published in Spain in 2020, suggesting a total of some millions of articles overall; but the largest repository in Spain, Digital CISC of the Spanish National Research Council, contains only around 350,000 articles in total. If I were a researcher, I would not use any search engine to find content from such an incomplete resource, even though I might search for one article via its title (something the authors of this article didn’t consider). But the authors claim, perversely, that “a potential explanation for the low visibility recorded here is the general absence of repository website optimization”.
  4. The authors claim IRs have “weak web authority” due to lack of external links from reputable sites”. This is not a relevant criterion for academic articles – although it might be a common way of comparing commercial websites.
  5. Among their recommendations, the authors state that articles “should seek to include additional content that is more suited to a non-academic public”. But scholarly articles are not primarily written for a non-academic public!

Conclusion

I’m mystified why the authors carried out this research. They claim “the visibility of institutional repositories in search engine results is of particular relevance”. I don’t think that is true at all. Repositories aim to hold content that might not be freely available otherwise. They were never set up to be measured by traffic and downloads, to be compared with other websites. The way that content is found in repositories is not typically by Google Search.

Most importantly, the article worries me because it gives the idea that repositories are somehow inadequate. On the contrary, my belief is that repositories fulfil a key function, that of ensuring open science and open research, by ensuring that content is accessible to all. Scholarly articles are retrieved by many routes, of which Google Scholar is only one. Nonetheless, placing the article in the institutional repository is a key component of Open Science. Repositories perform perform that function well, despite a lack of public awareness or even, it would seem, public understanding.

Views: 303